diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,50542 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 7212,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0,
+      "eval_loss": 2.829824447631836,
+      "eval_runtime": 408.1353,
+      "eval_samples_per_second": 100.489,
+      "eval_steps_per_second": 1.571,
+      "step": 0
+    },
+    {
+      "epoch": 0.00027731558513588466,
+      "grad_norm": 46.86890411376953,
+      "learning_rate": 0.0,
+      "loss": 2.4417,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005546311702717693,
+      "grad_norm": 69.00373077392578,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 2.3396,
+      "step": 2
+    },
+    {
+      "epoch": 0.0008319467554076539,
+      "grad_norm": 53.255943298339844,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 2.3484,
+      "step": 3
+    },
+    {
+      "epoch": 0.0011092623405435386,
+      "grad_norm": 21.859601974487305,
+      "learning_rate": 7.5e-07,
+      "loss": 2.3949,
+      "step": 4
+    },
+    {
+      "epoch": 0.0013865779256794233,
+      "grad_norm": 18.474227905273438,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.3011,
+      "step": 5
+    },
+    {
+      "epoch": 0.0016638935108153079,
+      "grad_norm": 18.92083740234375,
+      "learning_rate": 1.25e-06,
+      "loss": 2.1921,
+      "step": 6
+    },
+    {
+      "epoch": 0.0019412090959511925,
+      "grad_norm": 17.222856521606445,
+      "learning_rate": 1.5e-06,
+      "loss": 2.2153,
+      "step": 7
+    },
+    {
+      "epoch": 0.0022185246810870773,
+      "grad_norm": 31.592514038085938,
+      "learning_rate": 1.7500000000000002e-06,
+      "loss": 2.0413,
+      "step": 8
+    },
+    {
+      "epoch": 0.0024958402662229617,
+      "grad_norm": 17.17295265197754,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.0861,
+      "step": 9
+    },
+    {
+      "epoch": 0.0027731558513588465,
+      "grad_norm": 17.179834365844727,
+      "learning_rate": 2.25e-06,
+      "loss": 2.0175,
+      "step": 10
+    },
+    {
+      "epoch": 0.003050471436494731,
+      "grad_norm": 17.518646240234375,
+      "learning_rate": 2.5e-06,
+      "loss": 1.9135,
+      "step": 11
+    },
+    {
+      "epoch": 0.0033277870216306157,
+      "grad_norm": 14.855606079101562,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 1.7351,
+      "step": 12
+    },
+    {
+      "epoch": 0.0036051026067665,
+      "grad_norm": 13.514114379882812,
+      "learning_rate": 3e-06,
+      "loss": 1.5465,
+      "step": 13
+    },
+    {
+      "epoch": 0.003882418191902385,
+      "grad_norm": 12.117414474487305,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 1.4642,
+      "step": 14
+    },
+    {
+      "epoch": 0.004159733777038269,
+      "grad_norm": 10.935081481933594,
+      "learning_rate": 3.5000000000000004e-06,
+      "loss": 1.3851,
+      "step": 15
+    },
+    {
+      "epoch": 0.004437049362174155,
+      "grad_norm": 7.275962829589844,
+      "learning_rate": 3.75e-06,
+      "loss": 1.1289,
+      "step": 16
+    },
+    {
+      "epoch": 0.004714364947310039,
+      "grad_norm": 6.399021148681641,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.1277,
+      "step": 17
+    },
+    {
+      "epoch": 0.004991680532445923,
+      "grad_norm": 6.132956027984619,
+      "learning_rate": 4.250000000000001e-06,
+      "loss": 1.1483,
+      "step": 18
+    },
+    {
+      "epoch": 0.005268996117581808,
+      "grad_norm": 5.525564670562744,
+      "learning_rate": 4.5e-06,
+      "loss": 1.0578,
+      "step": 19
+    },
+    {
+      "epoch": 0.005546311702717693,
+      "grad_norm": 5.441694259643555,
+      "learning_rate": 4.75e-06,
+      "loss": 1.0647,
+      "step": 20
+    },
+    {
+      "epoch": 0.005823627287853577,
+      "grad_norm": 5.160792827606201,
+      "learning_rate": 5e-06,
+      "loss": 0.9961,
+      "step": 21
+    },
+    {
+      "epoch": 0.006100942872989462,
+      "grad_norm": 5.569485664367676,
+      "learning_rate": 5.25e-06,
+      "loss": 1.0063,
+      "step": 22
+    },
+    {
+      "epoch": 0.006378258458125347,
+      "grad_norm": 4.869104385375977,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.9848,
+      "step": 23
+    },
+    {
+      "epoch": 0.0066555740432612314,
+      "grad_norm": 3.172858238220215,
+      "learning_rate": 5.750000000000001e-06,
+      "loss": 0.9317,
+      "step": 24
+    },
+    {
+      "epoch": 0.006932889628397116,
+      "grad_norm": 2.5935134887695312,
+      "learning_rate": 6e-06,
+      "loss": 0.9086,
+      "step": 25
+    },
+    {
+      "epoch": 0.007210205213533,
+      "grad_norm": 1.7386329174041748,
+      "learning_rate": 6.25e-06,
+      "loss": 0.8976,
+      "step": 26
+    },
+    {
+      "epoch": 0.0074875207986688855,
+      "grad_norm": 1.3860479593276978,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.8668,
+      "step": 27
+    },
+    {
+      "epoch": 0.00776483638380477,
+      "grad_norm": 1.3284790515899658,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 0.8438,
+      "step": 28
+    },
+    {
+      "epoch": 0.008042151968940654,
+      "grad_norm": 1.208060383796692,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.8601,
+      "step": 29
+    },
+    {
+      "epoch": 0.008319467554076539,
+      "grad_norm": 0.99210125207901,
+      "learning_rate": 7.25e-06,
+      "loss": 0.836,
+      "step": 30
+    },
+    {
+      "epoch": 0.008596783139212423,
+      "grad_norm": 0.7937288284301758,
+      "learning_rate": 7.5e-06,
+      "loss": 0.829,
+      "step": 31
+    },
+    {
+      "epoch": 0.00887409872434831,
+      "grad_norm": 0.706200361251831,
+      "learning_rate": 7.75e-06,
+      "loss": 0.8006,
+      "step": 32
+    },
+    {
+      "epoch": 0.009151414309484194,
+      "grad_norm": 0.9658659100532532,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.7989,
+      "step": 33
+    },
+    {
+      "epoch": 0.009428729894620078,
+      "grad_norm": 1.174869418144226,
+      "learning_rate": 8.25e-06,
+      "loss": 0.8159,
+      "step": 34
+    },
+    {
+      "epoch": 0.009706045479755962,
+      "grad_norm": 0.5839990973472595,
+      "learning_rate": 8.500000000000002e-06,
+      "loss": 0.7837,
+      "step": 35
+    },
+    {
+      "epoch": 0.009983361064891847,
+      "grad_norm": 0.6130610704421997,
+      "learning_rate": 8.75e-06,
+      "loss": 0.7799,
+      "step": 36
+    },
+    {
+      "epoch": 0.010260676650027731,
+      "grad_norm": 0.7108742594718933,
+      "learning_rate": 9e-06,
+      "loss": 0.7611,
+      "step": 37
+    },
+    {
+      "epoch": 0.010537992235163616,
+      "grad_norm": 0.7302682995796204,
+      "learning_rate": 9.25e-06,
+      "loss": 0.7942,
+      "step": 38
+    },
+    {
+      "epoch": 0.010815307820299502,
+      "grad_norm": 0.5843620896339417,
+      "learning_rate": 9.5e-06,
+      "loss": 0.765,
+      "step": 39
+    },
+    {
+      "epoch": 0.011092623405435386,
+      "grad_norm": 0.41768163442611694,
+      "learning_rate": 9.750000000000002e-06,
+      "loss": 0.7112,
+      "step": 40
+    },
+    {
+      "epoch": 0.01136993899057127,
+      "grad_norm": 0.5103988647460938,
+      "learning_rate": 1e-05,
+      "loss": 0.762,
+      "step": 41
+    },
+    {
+      "epoch": 0.011647254575707155,
+      "grad_norm": 0.47128552198410034,
+      "learning_rate": 1.025e-05,
+      "loss": 0.7509,
+      "step": 42
+    },
+    {
+      "epoch": 0.01192457016084304,
+      "grad_norm": 1.0201480388641357,
+      "learning_rate": 1.05e-05,
+      "loss": 0.7507,
+      "step": 43
+    },
+    {
+      "epoch": 0.012201885745978924,
+      "grad_norm": 0.3908264935016632,
+      "learning_rate": 1.075e-05,
+      "loss": 0.7629,
+      "step": 44
+    },
+    {
+      "epoch": 0.012479201331114808,
+      "grad_norm": 0.4154920279979706,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.7531,
+      "step": 45
+    },
+    {
+      "epoch": 0.012756516916250694,
+      "grad_norm": 0.4213290512561798,
+      "learning_rate": 1.125e-05,
+      "loss": 0.7472,
+      "step": 46
+    },
+    {
+      "epoch": 0.013033832501386578,
+      "grad_norm": 0.6245641112327576,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 0.7404,
+      "step": 47
+    },
+    {
+      "epoch": 0.013311148086522463,
+      "grad_norm": 0.44496941566467285,
+      "learning_rate": 1.175e-05,
+      "loss": 0.7427,
+      "step": 48
+    },
+    {
+      "epoch": 0.013588463671658347,
+      "grad_norm": 0.4155629575252533,
+      "learning_rate": 1.2e-05,
+      "loss": 0.7147,
+      "step": 49
+    },
+    {
+      "epoch": 0.013865779256794232,
+      "grad_norm": 0.37920621037483215,
+      "learning_rate": 1.225e-05,
+      "loss": 0.7426,
+      "step": 50
+    },
+    {
+      "epoch": 0.014143094841930116,
+      "grad_norm": 0.3893055319786072,
+      "learning_rate": 1.25e-05,
+      "loss": 0.7176,
+      "step": 51
+    },
+    {
+      "epoch": 0.014420410427066,
+      "grad_norm": 0.3363882005214691,
+      "learning_rate": 1.2750000000000002e-05,
+      "loss": 0.7063,
+      "step": 52
+    },
+    {
+      "epoch": 0.014697726012201887,
+      "grad_norm": 0.4444830119609833,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.6945,
+      "step": 53
+    },
+    {
+      "epoch": 0.014975041597337771,
+      "grad_norm": 0.3413512706756592,
+      "learning_rate": 1.3250000000000002e-05,
+      "loss": 0.7288,
+      "step": 54
+    },
+    {
+      "epoch": 0.015252357182473655,
+      "grad_norm": 0.4114389717578888,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 0.7575,
+      "step": 55
+    },
+    {
+      "epoch": 0.01552967276760954,
+      "grad_norm": 0.36049914360046387,
+      "learning_rate": 1.3750000000000002e-05,
+      "loss": 0.7217,
+      "step": 56
+    },
+    {
+      "epoch": 0.015806988352745424,
+      "grad_norm": 0.41267284750938416,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.7166,
+      "step": 57
+    },
+    {
+      "epoch": 0.01608430393788131,
+      "grad_norm": 0.4639422297477722,
+      "learning_rate": 1.4249999999999999e-05,
+      "loss": 0.7069,
+      "step": 58
+    },
+    {
+      "epoch": 0.016361619523017193,
+      "grad_norm": 0.36772483587265015,
+      "learning_rate": 1.45e-05,
+      "loss": 0.7247,
+      "step": 59
+    },
+    {
+      "epoch": 0.016638935108153077,
+      "grad_norm": 0.3546575903892517,
+      "learning_rate": 1.475e-05,
+      "loss": 0.7128,
+      "step": 60
+    },
+    {
+      "epoch": 0.01691625069328896,
+      "grad_norm": 0.31919416785240173,
+      "learning_rate": 1.5e-05,
+      "loss": 0.7207,
+      "step": 61
+    },
+    {
+      "epoch": 0.017193566278424846,
+      "grad_norm": 0.3498699367046356,
+      "learning_rate": 1.525e-05,
+      "loss": 0.7065,
+      "step": 62
+    },
+    {
+      "epoch": 0.01747088186356073,
+      "grad_norm": 0.35648590326309204,
+      "learning_rate": 1.55e-05,
+      "loss": 0.7146,
+      "step": 63
+    },
+    {
+      "epoch": 0.01774819744869662,
+      "grad_norm": 0.30697041749954224,
+      "learning_rate": 1.575e-05,
+      "loss": 0.6805,
+      "step": 64
+    },
+    {
+      "epoch": 0.018025513033832503,
+      "grad_norm": 0.5759001970291138,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.7267,
+      "step": 65
+    },
+    {
+      "epoch": 0.018302828618968387,
+      "grad_norm": 0.263336718082428,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 0.6936,
+      "step": 66
+    },
+    {
+      "epoch": 0.01858014420410427,
+      "grad_norm": 0.2977915108203888,
+      "learning_rate": 1.65e-05,
+      "loss": 0.6685,
+      "step": 67
+    },
+    {
+      "epoch": 0.018857459789240156,
+      "grad_norm": 0.3028334081172943,
+      "learning_rate": 1.675e-05,
+      "loss": 0.7109,
+      "step": 68
+    },
+    {
+      "epoch": 0.01913477537437604,
+      "grad_norm": 0.3265489935874939,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.7023,
+      "step": 69
+    },
+    {
+      "epoch": 0.019412090959511925,
+      "grad_norm": 0.2899531126022339,
+      "learning_rate": 1.725e-05,
+      "loss": 0.6985,
+      "step": 70
+    },
+    {
+      "epoch": 0.01968940654464781,
+      "grad_norm": 0.29272034764289856,
+      "learning_rate": 1.75e-05,
+      "loss": 0.7125,
+      "step": 71
+    },
+    {
+      "epoch": 0.019966722129783693,
+      "grad_norm": 0.3114602863788605,
+      "learning_rate": 1.775e-05,
+      "loss": 0.6836,
+      "step": 72
+    },
+    {
+      "epoch": 0.020244037714919578,
+      "grad_norm": 0.28768229484558105,
+      "learning_rate": 1.8e-05,
+      "loss": 0.6809,
+      "step": 73
+    },
+    {
+      "epoch": 0.020521353300055462,
+      "grad_norm": 0.270345002412796,
+      "learning_rate": 1.825e-05,
+      "loss": 0.6776,
+      "step": 74
+    },
+    {
+      "epoch": 0.020798668885191347,
+      "grad_norm": 0.2635841369628906,
+      "learning_rate": 1.85e-05,
+      "loss": 0.6645,
+      "step": 75
+    },
+    {
+      "epoch": 0.02107598447032723,
+      "grad_norm": 0.3204723000526428,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.6631,
+      "step": 76
+    },
+    {
+      "epoch": 0.021353300055463115,
+      "grad_norm": 0.27179455757141113,
+      "learning_rate": 1.9e-05,
+      "loss": 0.693,
+      "step": 77
+    },
+    {
+      "epoch": 0.021630615640599003,
+      "grad_norm": 0.26558464765548706,
+      "learning_rate": 1.925e-05,
+      "loss": 0.6715,
+      "step": 78
+    },
+    {
+      "epoch": 0.021907931225734888,
+      "grad_norm": 0.3682558834552765,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 0.6826,
+      "step": 79
+    },
+    {
+      "epoch": 0.022185246810870772,
+      "grad_norm": 0.281429648399353,
+      "learning_rate": 1.9750000000000002e-05,
+      "loss": 0.6599,
+      "step": 80
+    },
+    {
+      "epoch": 0.022462562396006656,
+      "grad_norm": 0.28537672758102417,
+      "learning_rate": 2e-05,
+      "loss": 0.6695,
+      "step": 81
+    },
+    {
+      "epoch": 0.02273987798114254,
+      "grad_norm": 0.274913489818573,
+      "learning_rate": 2.025e-05,
+      "loss": 0.678,
+      "step": 82
+    },
+    {
+      "epoch": 0.023017193566278425,
+      "grad_norm": 0.28847208619117737,
+      "learning_rate": 2.05e-05,
+      "loss": 0.669,
+      "step": 83
+    },
+    {
+      "epoch": 0.02329450915141431,
+      "grad_norm": 0.30678853392601013,
+      "learning_rate": 2.075e-05,
+      "loss": 0.6647,
+      "step": 84
+    },
+    {
+      "epoch": 0.023571824736550194,
+      "grad_norm": 0.28266021609306335,
+      "learning_rate": 2.1e-05,
+      "loss": 0.671,
+      "step": 85
+    },
+    {
+      "epoch": 0.02384914032168608,
+      "grad_norm": 0.2712315320968628,
+      "learning_rate": 2.125e-05,
+      "loss": 0.6579,
+      "step": 86
+    },
+    {
+      "epoch": 0.024126455906821963,
+      "grad_norm": 0.30666086077690125,
+      "learning_rate": 2.15e-05,
+      "loss": 0.6598,
+      "step": 87
+    },
+    {
+      "epoch": 0.024403771491957847,
+      "grad_norm": 0.257932186126709,
+      "learning_rate": 2.175e-05,
+      "loss": 0.6348,
+      "step": 88
+    },
+    {
+      "epoch": 0.02468108707709373,
+      "grad_norm": 0.3133629560470581,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.6611,
+      "step": 89
+    },
+    {
+      "epoch": 0.024958402662229616,
+      "grad_norm": 0.27258774638175964,
+      "learning_rate": 2.2250000000000002e-05,
+      "loss": 0.6736,
+      "step": 90
+    },
+    {
+      "epoch": 0.0252357182473655,
+      "grad_norm": 0.3201597034931183,
+      "learning_rate": 2.25e-05,
+      "loss": 0.6918,
+      "step": 91
+    },
+    {
+      "epoch": 0.025513033832501388,
+      "grad_norm": 0.26909735798835754,
+      "learning_rate": 2.275e-05,
+      "loss": 0.6738,
+      "step": 92
+    },
+    {
+      "epoch": 0.025790349417637273,
+      "grad_norm": 0.40945449471473694,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.6615,
+      "step": 93
+    },
+    {
+      "epoch": 0.026067665002773157,
+      "grad_norm": 0.3059796392917633,
+      "learning_rate": 2.3250000000000003e-05,
+      "loss": 0.6677,
+      "step": 94
+    },
+    {
+      "epoch": 0.02634498058790904,
+      "grad_norm": 0.2737233638763428,
+      "learning_rate": 2.35e-05,
+      "loss": 0.6436,
+      "step": 95
+    },
+    {
+      "epoch": 0.026622296173044926,
+      "grad_norm": 0.3231774866580963,
+      "learning_rate": 2.375e-05,
+      "loss": 0.6735,
+      "step": 96
+    },
+    {
+      "epoch": 0.02689961175818081,
+      "grad_norm": 0.28404903411865234,
+      "learning_rate": 2.4e-05,
+      "loss": 0.6688,
+      "step": 97
+    },
+    {
+      "epoch": 0.027176927343316695,
+      "grad_norm": 0.3392276465892792,
+      "learning_rate": 2.425e-05,
+      "loss": 0.684,
+      "step": 98
+    },
+    {
+      "epoch": 0.02745424292845258,
+      "grad_norm": 0.3021562099456787,
+      "learning_rate": 2.45e-05,
+      "loss": 0.6886,
+      "step": 99
+    },
+    {
+      "epoch": 0.027731558513588463,
+      "grad_norm": 0.2843964099884033,
+      "learning_rate": 2.4750000000000002e-05,
+      "loss": 0.6358,
+      "step": 100
+    },
+    {
+      "epoch": 0.028008874098724348,
+      "grad_norm": 0.3120342195034027,
+      "learning_rate": 2.5e-05,
+      "loss": 0.669,
+      "step": 101
+    },
+    {
+      "epoch": 0.028286189683860232,
+      "grad_norm": 0.2752966284751892,
+      "learning_rate": 2.499999878045941e-05,
+      "loss": 0.6231,
+      "step": 102
+    },
+    {
+      "epoch": 0.028563505268996116,
+      "grad_norm": 0.28501492738723755,
+      "learning_rate": 2.4999995121837877e-05,
+      "loss": 0.6205,
+      "step": 103
+    },
+    {
+      "epoch": 0.028840820854132,
+      "grad_norm": 0.2797880172729492,
+      "learning_rate": 2.4999989024136113e-05,
+      "loss": 0.6697,
+      "step": 104
+    },
+    {
+      "epoch": 0.029118136439267885,
+      "grad_norm": 0.2929855287075043,
+      "learning_rate": 2.4999980487355314e-05,
+      "loss": 0.642,
+      "step": 105
+    },
+    {
+      "epoch": 0.029395452024403773,
+      "grad_norm": 0.264911025762558,
+      "learning_rate": 2.4999969511497135e-05,
+      "loss": 0.6575,
+      "step": 106
+    },
+    {
+      "epoch": 0.029672767609539658,
+      "grad_norm": 0.2797522246837616,
+      "learning_rate": 2.4999956096563725e-05,
+      "loss": 0.6566,
+      "step": 107
+    },
+    {
+      "epoch": 0.029950083194675542,
+      "grad_norm": 0.24777652323246002,
+      "learning_rate": 2.49999402425577e-05,
+      "loss": 0.6288,
+      "step": 108
+    },
+    {
+      "epoch": 0.030227398779811426,
+      "grad_norm": 0.34936287999153137,
+      "learning_rate": 2.4999921949482157e-05,
+      "loss": 0.6506,
+      "step": 109
+    },
+    {
+      "epoch": 0.03050471436494731,
+      "grad_norm": 0.30599579215049744,
+      "learning_rate": 2.499990121734066e-05,
+      "loss": 0.6554,
+      "step": 110
+    },
+    {
+      "epoch": 0.030782029950083195,
+      "grad_norm": 0.29751792550086975,
+      "learning_rate": 2.499987804613726e-05,
+      "loss": 0.6698,
+      "step": 111
+    },
+    {
+      "epoch": 0.03105934553521908,
+      "grad_norm": 0.2642778754234314,
+      "learning_rate": 2.4999852435876473e-05,
+      "loss": 0.6337,
+      "step": 112
+    },
+    {
+      "epoch": 0.031336661120354964,
+      "grad_norm": 0.2584931552410126,
+      "learning_rate": 2.49998243865633e-05,
+      "loss": 0.661,
+      "step": 113
+    },
+    {
+      "epoch": 0.03161397670549085,
+      "grad_norm": 0.266797810792923,
+      "learning_rate": 2.4999793898203212e-05,
+      "loss": 0.6368,
+      "step": 114
+    },
+    {
+      "epoch": 0.03189129229062673,
+      "grad_norm": 0.35552042722702026,
+      "learning_rate": 2.4999760970802155e-05,
+      "loss": 0.6364,
+      "step": 115
+    },
+    {
+      "epoch": 0.03216860787576262,
+      "grad_norm": 0.28450194001197815,
+      "learning_rate": 2.4999725604366562e-05,
+      "loss": 0.661,
+      "step": 116
+    },
+    {
+      "epoch": 0.0324459234608985,
+      "grad_norm": 0.3352636694908142,
+      "learning_rate": 2.4999687798903327e-05,
+      "loss": 0.6439,
+      "step": 117
+    },
+    {
+      "epoch": 0.032723239046034386,
+      "grad_norm": 0.2475953847169876,
+      "learning_rate": 2.499964755441983e-05,
+      "loss": 0.6344,
+      "step": 118
+    },
+    {
+      "epoch": 0.03300055463117027,
+      "grad_norm": 0.30431386828422546,
+      "learning_rate": 2.4999604870923926e-05,
+      "loss": 0.6459,
+      "step": 119
+    },
+    {
+      "epoch": 0.033277870216306155,
+      "grad_norm": 0.2653152644634247,
+      "learning_rate": 2.499955974842394e-05,
+      "loss": 0.6425,
+      "step": 120
+    },
+    {
+      "epoch": 0.03355518580144204,
+      "grad_norm": 0.29490575194358826,
+      "learning_rate": 2.4999512186928675e-05,
+      "loss": 0.6427,
+      "step": 121
+    },
+    {
+      "epoch": 0.03383250138657792,
+      "grad_norm": 0.2630308270454407,
+      "learning_rate": 2.4999462186447415e-05,
+      "loss": 0.6597,
+      "step": 122
+    },
+    {
+      "epoch": 0.03410981697171381,
+      "grad_norm": 0.26787513494491577,
+      "learning_rate": 2.4999409746989914e-05,
+      "loss": 0.6622,
+      "step": 123
+    },
+    {
+      "epoch": 0.03438713255684969,
+      "grad_norm": 0.25667890906333923,
+      "learning_rate": 2.499935486856641e-05,
+      "loss": 0.6335,
+      "step": 124
+    },
+    {
+      "epoch": 0.03466444814198558,
+      "grad_norm": 0.26751402020454407,
+      "learning_rate": 2.4999297551187603e-05,
+      "loss": 0.6358,
+      "step": 125
+    },
+    {
+      "epoch": 0.03494176372712146,
+      "grad_norm": 0.2815951108932495,
+      "learning_rate": 2.4999237794864683e-05,
+      "loss": 0.6615,
+      "step": 126
+    },
+    {
+      "epoch": 0.03521907931225735,
+      "grad_norm": 0.2573346793651581,
+      "learning_rate": 2.499917559960931e-05,
+      "loss": 0.6463,
+      "step": 127
+    },
+    {
+      "epoch": 0.03549639489739324,
+      "grad_norm": 0.26202693581581116,
+      "learning_rate": 2.4999110965433615e-05,
+      "loss": 0.6436,
+      "step": 128
+    },
+    {
+      "epoch": 0.03577371048252912,
+      "grad_norm": 0.267046719789505,
+      "learning_rate": 2.4999043892350213e-05,
+      "loss": 0.6433,
+      "step": 129
+    },
+    {
+      "epoch": 0.036051026067665005,
+      "grad_norm": 0.2713761329650879,
+      "learning_rate": 2.499897438037219e-05,
+      "loss": 0.6314,
+      "step": 130
+    },
+    {
+      "epoch": 0.03632834165280089,
+      "grad_norm": 0.2704955041408539,
+      "learning_rate": 2.4998902429513115e-05,
+      "loss": 0.6471,
+      "step": 131
+    },
+    {
+      "epoch": 0.036605657237936774,
+      "grad_norm": 0.25811654329299927,
+      "learning_rate": 2.4998828039787027e-05,
+      "loss": 0.6346,
+      "step": 132
+    },
+    {
+      "epoch": 0.03688297282307266,
+      "grad_norm": 0.2768125832080841,
+      "learning_rate": 2.4998751211208432e-05,
+      "loss": 0.6327,
+      "step": 133
+    },
+    {
+      "epoch": 0.03716028840820854,
+      "grad_norm": 0.29026105999946594,
+      "learning_rate": 2.499867194379233e-05,
+      "loss": 0.6632,
+      "step": 134
+    },
+    {
+      "epoch": 0.03743760399334443,
+      "grad_norm": 0.26648250222206116,
+      "learning_rate": 2.4998590237554182e-05,
+      "loss": 0.6414,
+      "step": 135
+    },
+    {
+      "epoch": 0.03771491957848031,
+      "grad_norm": 0.2578074336051941,
+      "learning_rate": 2.4998506092509938e-05,
+      "loss": 0.6459,
+      "step": 136
+    },
+    {
+      "epoch": 0.037992235163616196,
+      "grad_norm": 0.2555679678916931,
+      "learning_rate": 2.4998419508676014e-05,
+      "loss": 0.6561,
+      "step": 137
+    },
+    {
+      "epoch": 0.03826955074875208,
+      "grad_norm": 0.25471994280815125,
+      "learning_rate": 2.4998330486069304e-05,
+      "loss": 0.6616,
+      "step": 138
+    },
+    {
+      "epoch": 0.038546866333887965,
+      "grad_norm": 0.2434554398059845,
+      "learning_rate": 2.4998239024707183e-05,
+      "loss": 0.6423,
+      "step": 139
+    },
+    {
+      "epoch": 0.03882418191902385,
+      "grad_norm": 0.23697395622730255,
+      "learning_rate": 2.4998145124607485e-05,
+      "loss": 0.629,
+      "step": 140
+    },
+    {
+      "epoch": 0.039101497504159734,
+      "grad_norm": 0.2652537226676941,
+      "learning_rate": 2.4998048785788547e-05,
+      "loss": 0.6558,
+      "step": 141
+    },
+    {
+      "epoch": 0.03937881308929562,
+      "grad_norm": 0.2602185606956482,
+      "learning_rate": 2.499795000826916e-05,
+      "loss": 0.6427,
+      "step": 142
+    },
+    {
+      "epoch": 0.0396561286744315,
+      "grad_norm": 0.23875969648361206,
+      "learning_rate": 2.49978487920686e-05,
+      "loss": 0.6097,
+      "step": 143
+    },
+    {
+      "epoch": 0.03993344425956739,
+      "grad_norm": 0.2549594044685364,
+      "learning_rate": 2.4997745137206618e-05,
+      "loss": 0.6477,
+      "step": 144
+    },
+    {
+      "epoch": 0.04021075984470327,
+      "grad_norm": 0.2528778910636902,
+      "learning_rate": 2.4997639043703437e-05,
+      "loss": 0.6028,
+      "step": 145
+    },
+    {
+      "epoch": 0.040488075429839156,
+      "grad_norm": 0.252888023853302,
+      "learning_rate": 2.499753051157976e-05,
+      "loss": 0.6547,
+      "step": 146
+    },
+    {
+      "epoch": 0.04076539101497504,
+      "grad_norm": 0.289661705493927,
+      "learning_rate": 2.4997419540856762e-05,
+      "loss": 0.6604,
+      "step": 147
+    },
+    {
+      "epoch": 0.041042706600110924,
+      "grad_norm": 0.27772676944732666,
+      "learning_rate": 2.49973061315561e-05,
+      "loss": 0.6321,
+      "step": 148
+    },
+    {
+      "epoch": 0.04132002218524681,
+      "grad_norm": 0.29435357451438904,
+      "learning_rate": 2.4997190283699904e-05,
+      "loss": 0.6539,
+      "step": 149
+    },
+    {
+      "epoch": 0.04159733777038269,
+      "grad_norm": 0.2796315848827362,
+      "learning_rate": 2.4997071997310774e-05,
+      "loss": 0.6816,
+      "step": 150
+    },
+    {
+      "epoch": 0.04187465335551858,
+      "grad_norm": 0.2854909598827362,
+      "learning_rate": 2.4996951272411794e-05,
+      "loss": 0.621,
+      "step": 151
+    },
+    {
+      "epoch": 0.04215196894065446,
+      "grad_norm": 0.3513517677783966,
+      "learning_rate": 2.499682810902652e-05,
+      "loss": 0.6462,
+      "step": 152
+    },
+    {
+      "epoch": 0.042429284525790346,
+      "grad_norm": 0.2492416799068451,
+      "learning_rate": 2.4996702507178988e-05,
+      "loss": 0.6455,
+      "step": 153
+    },
+    {
+      "epoch": 0.04270660011092623,
+      "grad_norm": 0.2352532148361206,
+      "learning_rate": 2.49965744668937e-05,
+      "loss": 0.6168,
+      "step": 154
+    },
+    {
+      "epoch": 0.04298391569606212,
+      "grad_norm": 0.25850751996040344,
+      "learning_rate": 2.4996443988195644e-05,
+      "loss": 0.6452,
+      "step": 155
+    },
+    {
+      "epoch": 0.04326123128119801,
+      "grad_norm": 0.23972827196121216,
+      "learning_rate": 2.499631107111028e-05,
+      "loss": 0.6447,
+      "step": 156
+    },
+    {
+      "epoch": 0.04353854686633389,
+      "grad_norm": 0.24847468733787537,
+      "learning_rate": 2.499617571566354e-05,
+      "loss": 0.6386,
+      "step": 157
+    },
+    {
+      "epoch": 0.043815862451469775,
+      "grad_norm": 0.25739696621894836,
+      "learning_rate": 2.4996037921881837e-05,
+      "loss": 0.6417,
+      "step": 158
+    },
+    {
+      "epoch": 0.04409317803660566,
+      "grad_norm": 0.23640736937522888,
+      "learning_rate": 2.4995897689792062e-05,
+      "loss": 0.6451,
+      "step": 159
+    },
+    {
+      "epoch": 0.044370493621741544,
+      "grad_norm": 0.25362861156463623,
+      "learning_rate": 2.4995755019421577e-05,
+      "loss": 0.6525,
+      "step": 160
+    },
+    {
+      "epoch": 0.04464780920687743,
+      "grad_norm": 0.2607216536998749,
+      "learning_rate": 2.4995609910798214e-05,
+      "loss": 0.6276,
+      "step": 161
+    },
+    {
+      "epoch": 0.04492512479201331,
+      "grad_norm": 0.2426438182592392,
+      "learning_rate": 2.4995462363950295e-05,
+      "loss": 0.6375,
+      "step": 162
+    },
+    {
+      "epoch": 0.0452024403771492,
+      "grad_norm": 0.35536178946495056,
+      "learning_rate": 2.499531237890661e-05,
+      "loss": 0.6502,
+      "step": 163
+    },
+    {
+      "epoch": 0.04547975596228508,
+      "grad_norm": 0.2616370618343353,
+      "learning_rate": 2.4995159955696417e-05,
+      "loss": 0.6422,
+      "step": 164
+    },
+    {
+      "epoch": 0.045757071547420966,
+      "grad_norm": 0.2493521124124527,
+      "learning_rate": 2.4995005094349473e-05,
+      "loss": 0.6314,
+      "step": 165
+    },
+    {
+      "epoch": 0.04603438713255685,
+      "grad_norm": 0.25228554010391235,
+      "learning_rate": 2.4994847794895977e-05,
+      "loss": 0.6154,
+      "step": 166
+    },
+    {
+      "epoch": 0.046311702717692735,
+      "grad_norm": 0.24656261503696442,
+      "learning_rate": 2.4994688057366635e-05,
+      "loss": 0.6241,
+      "step": 167
+    },
+    {
+      "epoch": 0.04658901830282862,
+      "grad_norm": 0.27083900570869446,
+      "learning_rate": 2.4994525881792612e-05,
+      "loss": 0.627,
+      "step": 168
+    },
+    {
+      "epoch": 0.046866333887964504,
+      "grad_norm": 0.24816080927848816,
+      "learning_rate": 2.499436126820555e-05,
+      "loss": 0.6237,
+      "step": 169
+    },
+    {
+      "epoch": 0.04714364947310039,
+      "grad_norm": 0.2589535415172577,
+      "learning_rate": 2.499419421663758e-05,
+      "loss": 0.6139,
+      "step": 170
+    },
+    {
+      "epoch": 0.04742096505823627,
+      "grad_norm": 0.2533140182495117,
+      "learning_rate": 2.499402472712129e-05,
+      "loss": 0.6533,
+      "step": 171
+    },
+    {
+      "epoch": 0.04769828064337216,
+      "grad_norm": 0.23976951837539673,
+      "learning_rate": 2.499385279968975e-05,
+      "loss": 0.645,
+      "step": 172
+    },
+    {
+      "epoch": 0.04797559622850804,
+      "grad_norm": 0.24593569338321686,
+      "learning_rate": 2.4993678434376507e-05,
+      "loss": 0.5958,
+      "step": 173
+    },
+    {
+      "epoch": 0.048252911813643926,
+      "grad_norm": 0.283794641494751,
+      "learning_rate": 2.4993501631215593e-05,
+      "loss": 0.6565,
+      "step": 174
+    },
+    {
+      "epoch": 0.04853022739877981,
+      "grad_norm": 0.23555535078048706,
+      "learning_rate": 2.4993322390241496e-05,
+      "loss": 0.6077,
+      "step": 175
+    },
+    {
+      "epoch": 0.048807542983915694,
+      "grad_norm": 0.24843829870224,
+      "learning_rate": 2.4993140711489203e-05,
+      "loss": 0.6001,
+      "step": 176
+    },
+    {
+      "epoch": 0.04908485856905158,
+      "grad_norm": 0.24111098051071167,
+      "learning_rate": 2.4992956594994156e-05,
+      "loss": 0.6445,
+      "step": 177
+    },
+    {
+      "epoch": 0.04936217415418746,
+      "grad_norm": 0.2340569943189621,
+      "learning_rate": 2.499277004079228e-05,
+      "loss": 0.6174,
+      "step": 178
+    },
+    {
+      "epoch": 0.04963948973932335,
+      "grad_norm": 0.26766470074653625,
+      "learning_rate": 2.499258104891998e-05,
+      "loss": 0.6167,
+      "step": 179
+    },
+    {
+      "epoch": 0.04991680532445923,
+      "grad_norm": 0.2567934989929199,
+      "learning_rate": 2.499238961941413e-05,
+      "loss": 0.6394,
+      "step": 180
+    },
+    {
+      "epoch": 0.050194120909595116,
+      "grad_norm": 0.27323460578918457,
+      "learning_rate": 2.4992195752312093e-05,
+      "loss": 0.6337,
+      "step": 181
+    },
+    {
+      "epoch": 0.050471436494731,
+      "grad_norm": 0.3052992522716522,
+      "learning_rate": 2.4991999447651686e-05,
+      "loss": 0.6389,
+      "step": 182
+    },
+    {
+      "epoch": 0.050748752079866885,
+      "grad_norm": 0.24447570741176605,
+      "learning_rate": 2.4991800705471218e-05,
+      "loss": 0.6165,
+      "step": 183
+    },
+    {
+      "epoch": 0.051026067665002776,
+      "grad_norm": 0.319871187210083,
+      "learning_rate": 2.499159952580947e-05,
+      "loss": 0.5816,
+      "step": 184
+    },
+    {
+      "epoch": 0.05130338325013866,
+      "grad_norm": 0.2554628551006317,
+      "learning_rate": 2.4991395908705693e-05,
+      "loss": 0.646,
+      "step": 185
+    },
+    {
+      "epoch": 0.051580698835274545,
+      "grad_norm": 0.2671261727809906,
+      "learning_rate": 2.499118985419962e-05,
+      "loss": 0.6592,
+      "step": 186
+    },
+    {
+      "epoch": 0.05185801442041043,
+      "grad_norm": 0.2524307370185852,
+      "learning_rate": 2.4990981362331462e-05,
+      "loss": 0.6178,
+      "step": 187
+    },
+    {
+      "epoch": 0.052135330005546314,
+      "grad_norm": 0.2809012532234192,
+      "learning_rate": 2.4990770433141898e-05,
+      "loss": 0.6234,
+      "step": 188
+    },
+    {
+      "epoch": 0.0524126455906822,
+      "grad_norm": 0.23483595252037048,
+      "learning_rate": 2.499055706667208e-05,
+      "loss": 0.6209,
+      "step": 189
+    },
+    {
+      "epoch": 0.05268996117581808,
+      "grad_norm": 0.2775763273239136,
+      "learning_rate": 2.4990341262963654e-05,
+      "loss": 0.618,
+      "step": 190
+    },
+    {
+      "epoch": 0.05296727676095397,
+      "grad_norm": 0.27081090211868286,
+      "learning_rate": 2.499012302205872e-05,
+      "loss": 0.6377,
+      "step": 191
+    },
+    {
+      "epoch": 0.05324459234608985,
+      "grad_norm": 0.24989834427833557,
+      "learning_rate": 2.4989902343999865e-05,
+      "loss": 0.6179,
+      "step": 192
+    },
+    {
+      "epoch": 0.053521907931225736,
+      "grad_norm": 0.27669793367385864,
+      "learning_rate": 2.498967922883015e-05,
+      "loss": 0.6298,
+      "step": 193
+    },
+    {
+      "epoch": 0.05379922351636162,
+      "grad_norm": 0.2672564387321472,
+      "learning_rate": 2.4989453676593106e-05,
+      "loss": 0.6536,
+      "step": 194
+    },
+    {
+      "epoch": 0.054076539101497505,
+      "grad_norm": 0.24099047482013702,
+      "learning_rate": 2.4989225687332752e-05,
+      "loss": 0.61,
+      "step": 195
+    },
+    {
+      "epoch": 0.05435385468663339,
+      "grad_norm": 0.235582634806633,
+      "learning_rate": 2.4988995261093566e-05,
+      "loss": 0.654,
+      "step": 196
+    },
+    {
+      "epoch": 0.054631170271769273,
+      "grad_norm": 0.2795652747154236,
+      "learning_rate": 2.4988762397920517e-05,
+      "loss": 0.6224,
+      "step": 197
+    },
+    {
+      "epoch": 0.05490848585690516,
+      "grad_norm": 0.22800379991531372,
+      "learning_rate": 2.4988527097859045e-05,
+      "loss": 0.6186,
+      "step": 198
+    },
+    {
+      "epoch": 0.05518580144204104,
+      "grad_norm": 0.24810528755187988,
+      "learning_rate": 2.4988289360955053e-05,
+      "loss": 0.6286,
+      "step": 199
+    },
+    {
+      "epoch": 0.05546311702717693,
+      "grad_norm": 0.21688294410705566,
+      "learning_rate": 2.4988049187254935e-05,
+      "loss": 0.598,
+      "step": 200
+    },
+    {
+      "epoch": 0.05574043261231281,
+      "grad_norm": 0.23709554970264435,
+      "learning_rate": 2.4987806576805562e-05,
+      "loss": 0.6598,
+      "step": 201
+    },
+    {
+      "epoch": 0.056017748197448695,
+      "grad_norm": 0.24982847273349762,
+      "learning_rate": 2.4987561529654263e-05,
+      "loss": 0.6342,
+      "step": 202
+    },
+    {
+      "epoch": 0.05629506378258458,
+      "grad_norm": 0.2215258628129959,
+      "learning_rate": 2.498731404584886e-05,
+      "loss": 0.6159,
+      "step": 203
+    },
+    {
+      "epoch": 0.056572379367720464,
+      "grad_norm": 0.25177863240242004,
+      "learning_rate": 2.4987064125437643e-05,
+      "loss": 0.6289,
+      "step": 204
+    },
+    {
+      "epoch": 0.05684969495285635,
+      "grad_norm": 0.24716275930404663,
+      "learning_rate": 2.498681176846937e-05,
+      "loss": 0.628,
+      "step": 205
+    },
+    {
+      "epoch": 0.05712701053799223,
+      "grad_norm": 0.2888506054878235,
+      "learning_rate": 2.49865569749933e-05,
+      "loss": 0.6216,
+      "step": 206
+    },
+    {
+      "epoch": 0.05740432612312812,
+      "grad_norm": 0.24658267199993134,
+      "learning_rate": 2.4986299745059127e-05,
+      "loss": 0.6132,
+      "step": 207
+    },
+    {
+      "epoch": 0.057681641708264,
+      "grad_norm": 0.24297240376472473,
+      "learning_rate": 2.4986040078717063e-05,
+      "loss": 0.6201,
+      "step": 208
+    },
+    {
+      "epoch": 0.057958957293399886,
+      "grad_norm": 0.2425074279308319,
+      "learning_rate": 2.4985777976017767e-05,
+      "loss": 0.5997,
+      "step": 209
+    },
+    {
+      "epoch": 0.05823627287853577,
+      "grad_norm": 0.25336262583732605,
+      "learning_rate": 2.498551343701238e-05,
+      "loss": 0.6489,
+      "step": 210
+    },
+    {
+      "epoch": 0.058513588463671655,
+      "grad_norm": 0.23498830199241638,
+      "learning_rate": 2.498524646175253e-05,
+      "loss": 0.6341,
+      "step": 211
+    },
+    {
+      "epoch": 0.058790904048807546,
+      "grad_norm": 0.24754488468170166,
+      "learning_rate": 2.49849770502903e-05,
+      "loss": 0.6538,
+      "step": 212
+    },
+    {
+      "epoch": 0.05906821963394343,
+      "grad_norm": 0.24818097054958344,
+      "learning_rate": 2.4984705202678266e-05,
+      "loss": 0.6098,
+      "step": 213
+    },
+    {
+      "epoch": 0.059345535219079315,
+      "grad_norm": 0.22981123626232147,
+      "learning_rate": 2.498443091896947e-05,
+      "loss": 0.6072,
+      "step": 214
+    },
+    {
+      "epoch": 0.0596228508042152,
+      "grad_norm": 0.2612292766571045,
+      "learning_rate": 2.4984154199217434e-05,
+      "loss": 0.626,
+      "step": 215
+    },
+    {
+      "epoch": 0.059900166389351084,
+      "grad_norm": 0.26491644978523254,
+      "learning_rate": 2.4983875043476153e-05,
+      "loss": 0.6495,
+      "step": 216
+    },
+    {
+      "epoch": 0.06017748197448697,
+      "grad_norm": 0.22399267554283142,
+      "learning_rate": 2.4983593451800096e-05,
+      "loss": 0.6341,
+      "step": 217
+    },
+    {
+      "epoch": 0.06045479755962285,
+      "grad_norm": 0.2318061739206314,
+      "learning_rate": 2.498330942424421e-05,
+      "loss": 0.5787,
+      "step": 218
+    },
+    {
+      "epoch": 0.06073211314475874,
+      "grad_norm": 0.2700578272342682,
+      "learning_rate": 2.498302296086392e-05,
+      "loss": 0.6314,
+      "step": 219
+    },
+    {
+      "epoch": 0.06100942872989462,
+      "grad_norm": 0.22675910592079163,
+      "learning_rate": 2.4982734061715112e-05,
+      "loss": 0.5714,
+      "step": 220
+    },
+    {
+      "epoch": 0.061286744315030506,
+      "grad_norm": 0.2522087097167969,
+      "learning_rate": 2.4982442726854173e-05,
+      "loss": 0.6053,
+      "step": 221
+    },
+    {
+      "epoch": 0.06156405990016639,
+      "grad_norm": 0.22665978968143463,
+      "learning_rate": 2.4982148956337935e-05,
+      "loss": 0.6156,
+      "step": 222
+    },
+    {
+      "epoch": 0.061841375485302275,
+      "grad_norm": 0.24832209944725037,
+      "learning_rate": 2.4981852750223726e-05,
+      "loss": 0.6406,
+      "step": 223
+    },
+    {
+      "epoch": 0.06211869107043816,
+      "grad_norm": 0.2526067793369293,
+      "learning_rate": 2.498155410856935e-05,
+      "loss": 0.6312,
+      "step": 224
+    },
+    {
+      "epoch": 0.06239600665557404,
+      "grad_norm": 0.24257095158100128,
+      "learning_rate": 2.4981253031433076e-05,
+      "loss": 0.6456,
+      "step": 225
+    },
+    {
+      "epoch": 0.06267332224070993,
+      "grad_norm": 0.25497207045555115,
+      "learning_rate": 2.4980949518873648e-05,
+      "loss": 0.6047,
+      "step": 226
+    },
+    {
+      "epoch": 0.06295063782584581,
+      "grad_norm": 0.23874424397945404,
+      "learning_rate": 2.498064357095029e-05,
+      "loss": 0.614,
+      "step": 227
+    },
+    {
+      "epoch": 0.0632279534109817,
+      "grad_norm": 0.24398040771484375,
+      "learning_rate": 2.498033518772271e-05,
+      "loss": 0.6212,
+      "step": 228
+    },
+    {
+      "epoch": 0.06350526899611758,
+      "grad_norm": 0.27126333117485046,
+      "learning_rate": 2.498002436925107e-05,
+      "loss": 0.6051,
+      "step": 229
+    },
+    {
+      "epoch": 0.06378258458125347,
+      "grad_norm": 0.22852414846420288,
+      "learning_rate": 2.497971111559602e-05,
+      "loss": 0.6133,
+      "step": 230
+    },
+    {
+      "epoch": 0.06405990016638935,
+      "grad_norm": 0.22752775251865387,
+      "learning_rate": 2.4979395426818696e-05,
+      "loss": 0.5893,
+      "step": 231
+    },
+    {
+      "epoch": 0.06433721575152523,
+      "grad_norm": 0.27361559867858887,
+      "learning_rate": 2.4979077302980683e-05,
+      "loss": 0.6431,
+      "step": 232
+    },
+    {
+      "epoch": 0.06461453133666112,
+      "grad_norm": 0.3071225881576538,
+      "learning_rate": 2.497875674414406e-05,
+      "loss": 0.617,
+      "step": 233
+    },
+    {
+      "epoch": 0.064891846921797,
+      "grad_norm": 0.3025614023208618,
+      "learning_rate": 2.4978433750371382e-05,
+      "loss": 0.6294,
+      "step": 234
+    },
+    {
+      "epoch": 0.06516916250693289,
+      "grad_norm": 0.21824614703655243,
+      "learning_rate": 2.4978108321725667e-05,
+      "loss": 0.6189,
+      "step": 235
+    },
+    {
+      "epoch": 0.06544647809206877,
+      "grad_norm": 0.23781217634677887,
+      "learning_rate": 2.497778045827042e-05,
+      "loss": 0.6115,
+      "step": 236
+    },
+    {
+      "epoch": 0.06572379367720466,
+      "grad_norm": 0.3532952666282654,
+      "learning_rate": 2.497745016006961e-05,
+      "loss": 0.6091,
+      "step": 237
+    },
+    {
+      "epoch": 0.06600110926234054,
+      "grad_norm": 0.26994985342025757,
+      "learning_rate": 2.4977117427187692e-05,
+      "loss": 0.6078,
+      "step": 238
+    },
+    {
+      "epoch": 0.06627842484747642,
+      "grad_norm": 0.2395590841770172,
+      "learning_rate": 2.4976782259689587e-05,
+      "loss": 0.6437,
+      "step": 239
+    },
+    {
+      "epoch": 0.06655574043261231,
+      "grad_norm": 0.2347521334886551,
+      "learning_rate": 2.49764446576407e-05,
+      "loss": 0.6015,
+      "step": 240
+    },
+    {
+      "epoch": 0.0668330560177482,
+      "grad_norm": 0.2603704333305359,
+      "learning_rate": 2.49761046211069e-05,
+      "loss": 0.6441,
+      "step": 241
+    },
+    {
+      "epoch": 0.06711037160288408,
+      "grad_norm": 0.21548427641391754,
+      "learning_rate": 2.4975762150154542e-05,
+      "loss": 0.6059,
+      "step": 242
+    },
+    {
+      "epoch": 0.06738768718801996,
+      "grad_norm": 0.2270384132862091,
+      "learning_rate": 2.497541724485045e-05,
+      "loss": 0.6148,
+      "step": 243
+    },
+    {
+      "epoch": 0.06766500277315585,
+      "grad_norm": 0.23716577887535095,
+      "learning_rate": 2.497506990526192e-05,
+      "loss": 0.6135,
+      "step": 244
+    },
+    {
+      "epoch": 0.06794231835829173,
+      "grad_norm": 0.24414962530136108,
+      "learning_rate": 2.4974720131456736e-05,
+      "loss": 0.6363,
+      "step": 245
+    },
+    {
+      "epoch": 0.06821963394342762,
+      "grad_norm": 0.24200287461280823,
+      "learning_rate": 2.497436792350314e-05,
+      "loss": 0.6278,
+      "step": 246
+    },
+    {
+      "epoch": 0.0684969495285635,
+      "grad_norm": 0.25273025035858154,
+      "learning_rate": 2.497401328146986e-05,
+      "loss": 0.6012,
+      "step": 247
+    },
+    {
+      "epoch": 0.06877426511369938,
+      "grad_norm": 0.2444678694009781,
+      "learning_rate": 2.4973656205426094e-05,
+      "loss": 0.6218,
+      "step": 248
+    },
+    {
+      "epoch": 0.06905158069883527,
+      "grad_norm": 0.25649040937423706,
+      "learning_rate": 2.4973296695441523e-05,
+      "loss": 0.6678,
+      "step": 249
+    },
+    {
+      "epoch": 0.06932889628397115,
+      "grad_norm": 0.2729082703590393,
+      "learning_rate": 2.4972934751586292e-05,
+      "loss": 0.6018,
+      "step": 250
+    },
+    {
+      "epoch": 0.06960621186910704,
+      "grad_norm": 0.23950397968292236,
+      "learning_rate": 2.4972570373931026e-05,
+      "loss": 0.6342,
+      "step": 251
+    },
+    {
+      "epoch": 0.06988352745424292,
+      "grad_norm": 0.27791211009025574,
+      "learning_rate": 2.4972203562546825e-05,
+      "loss": 0.5948,
+      "step": 252
+    },
+    {
+      "epoch": 0.0701608430393788,
+      "grad_norm": 0.24157491326332092,
+      "learning_rate": 2.4971834317505266e-05,
+      "loss": 0.6346,
+      "step": 253
+    },
+    {
+      "epoch": 0.0704381586245147,
+      "grad_norm": 0.23233160376548767,
+      "learning_rate": 2.4971462638878394e-05,
+      "loss": 0.6023,
+      "step": 254
+    },
+    {
+      "epoch": 0.07071547420965059,
+      "grad_norm": 0.22705158591270447,
+      "learning_rate": 2.4971088526738737e-05,
+      "loss": 0.6314,
+      "step": 255
+    },
+    {
+      "epoch": 0.07099278979478647,
+      "grad_norm": 0.2364932894706726,
+      "learning_rate": 2.4970711981159294e-05,
+      "loss": 0.6239,
+      "step": 256
+    },
+    {
+      "epoch": 0.07127010537992236,
+      "grad_norm": 0.2369173765182495,
+      "learning_rate": 2.4970333002213535e-05,
+      "loss": 0.6056,
+      "step": 257
+    },
+    {
+      "epoch": 0.07154742096505824,
+      "grad_norm": 0.23944415152072906,
+      "learning_rate": 2.4969951589975415e-05,
+      "loss": 0.6188,
+      "step": 258
+    },
+    {
+      "epoch": 0.07182473655019413,
+      "grad_norm": 0.24431641399860382,
+      "learning_rate": 2.4969567744519357e-05,
+      "loss": 0.6393,
+      "step": 259
+    },
+    {
+      "epoch": 0.07210205213533001,
+      "grad_norm": 0.23058977723121643,
+      "learning_rate": 2.4969181465920254e-05,
+      "loss": 0.623,
+      "step": 260
+    },
+    {
+      "epoch": 0.0723793677204659,
+      "grad_norm": 0.2396584004163742,
+      "learning_rate": 2.4968792754253483e-05,
+      "loss": 0.6085,
+      "step": 261
+    },
+    {
+      "epoch": 0.07265668330560178,
+      "grad_norm": 0.28117048740386963,
+      "learning_rate": 2.496840160959489e-05,
+      "loss": 0.612,
+      "step": 262
+    },
+    {
+      "epoch": 0.07293399889073766,
+      "grad_norm": 0.22773273289203644,
+      "learning_rate": 2.49680080320208e-05,
+      "loss": 0.6215,
+      "step": 263
+    },
+    {
+      "epoch": 0.07321131447587355,
+      "grad_norm": 0.25395745038986206,
+      "learning_rate": 2.496761202160801e-05,
+      "loss": 0.6197,
+      "step": 264
+    },
+    {
+      "epoch": 0.07348863006100943,
+      "grad_norm": 0.2748405337333679,
+      "learning_rate": 2.496721357843379e-05,
+      "loss": 0.5983,
+      "step": 265
+    },
+    {
+      "epoch": 0.07376594564614532,
+      "grad_norm": 0.37056779861450195,
+      "learning_rate": 2.496681270257589e-05,
+      "loss": 0.5974,
+      "step": 266
+    },
+    {
+      "epoch": 0.0740432612312812,
+      "grad_norm": 0.2506537139415741,
+      "learning_rate": 2.4966409394112528e-05,
+      "loss": 0.6279,
+      "step": 267
+    },
+    {
+      "epoch": 0.07432057681641709,
+      "grad_norm": 0.3248004615306854,
+      "learning_rate": 2.4966003653122406e-05,
+      "loss": 0.5968,
+      "step": 268
+    },
+    {
+      "epoch": 0.07459789240155297,
+      "grad_norm": 0.27633965015411377,
+      "learning_rate": 2.4965595479684685e-05,
+      "loss": 0.6207,
+      "step": 269
+    },
+    {
+      "epoch": 0.07487520798668885,
+      "grad_norm": 0.2533873915672302,
+      "learning_rate": 2.4965184873879015e-05,
+      "loss": 0.6428,
+      "step": 270
+    },
+    {
+      "epoch": 0.07515252357182474,
+      "grad_norm": 0.24860017001628876,
+      "learning_rate": 2.496477183578552e-05,
+      "loss": 0.5717,
+      "step": 271
+    },
+    {
+      "epoch": 0.07542983915696062,
+      "grad_norm": 0.2541423738002777,
+      "learning_rate": 2.4964356365484797e-05,
+      "loss": 0.6331,
+      "step": 272
+    },
+    {
+      "epoch": 0.07570715474209651,
+      "grad_norm": 0.21915870904922485,
+      "learning_rate": 2.4963938463057907e-05,
+      "loss": 0.6167,
+      "step": 273
+    },
+    {
+      "epoch": 0.07598447032723239,
+      "grad_norm": 0.21635891497135162,
+      "learning_rate": 2.4963518128586393e-05,
+      "loss": 0.602,
+      "step": 274
+    },
+    {
+      "epoch": 0.07626178591236828,
+      "grad_norm": 0.23918819427490234,
+      "learning_rate": 2.4963095362152282e-05,
+      "loss": 0.5869,
+      "step": 275
+    },
+    {
+      "epoch": 0.07653910149750416,
+      "grad_norm": 0.22294320166110992,
+      "learning_rate": 2.496267016383806e-05,
+      "loss": 0.6256,
+      "step": 276
+    },
+    {
+      "epoch": 0.07681641708264005,
+      "grad_norm": 0.2167348712682724,
+      "learning_rate": 2.49622425337267e-05,
+      "loss": 0.5936,
+      "step": 277
+    },
+    {
+      "epoch": 0.07709373266777593,
+      "grad_norm": 0.21333451569080353,
+      "learning_rate": 2.496181247190164e-05,
+      "loss": 0.6221,
+      "step": 278
+    },
+    {
+      "epoch": 0.07737104825291181,
+      "grad_norm": 0.22917450964450836,
+      "learning_rate": 2.4961379978446793e-05,
+      "loss": 0.6132,
+      "step": 279
+    },
+    {
+      "epoch": 0.0776483638380477,
+      "grad_norm": 0.228413388133049,
+      "learning_rate": 2.496094505344656e-05,
+      "loss": 0.5953,
+      "step": 280
+    },
+    {
+      "epoch": 0.07792567942318358,
+      "grad_norm": 0.20908214151859283,
+      "learning_rate": 2.4960507696985796e-05,
+      "loss": 0.6081,
+      "step": 281
+    },
+    {
+      "epoch": 0.07820299500831947,
+      "grad_norm": 0.2270585596561432,
+      "learning_rate": 2.4960067909149846e-05,
+      "loss": 0.5915,
+      "step": 282
+    },
+    {
+      "epoch": 0.07848031059345535,
+      "grad_norm": 0.23176094889640808,
+      "learning_rate": 2.4959625690024524e-05,
+      "loss": 0.6126,
+      "step": 283
+    },
+    {
+      "epoch": 0.07875762617859124,
+      "grad_norm": 0.24496515095233917,
+      "learning_rate": 2.495918103969612e-05,
+      "loss": 0.6245,
+      "step": 284
+    },
+    {
+      "epoch": 0.07903494176372712,
+      "grad_norm": 0.21696555614471436,
+      "learning_rate": 2.4958733958251394e-05,
+      "loss": 0.6104,
+      "step": 285
+    },
+    {
+      "epoch": 0.079312257348863,
+      "grad_norm": 0.22292231023311615,
+      "learning_rate": 2.4958284445777584e-05,
+      "loss": 0.6164,
+      "step": 286
+    },
+    {
+      "epoch": 0.07958957293399889,
+      "grad_norm": 0.24104639887809753,
+      "learning_rate": 2.4957832502362404e-05,
+      "loss": 0.6002,
+      "step": 287
+    },
+    {
+      "epoch": 0.07986688851913477,
+      "grad_norm": 0.22866299748420715,
+      "learning_rate": 2.495737812809404e-05,
+      "loss": 0.6143,
+      "step": 288
+    },
+    {
+      "epoch": 0.08014420410427066,
+      "grad_norm": 0.23108075559139252,
+      "learning_rate": 2.495692132306115e-05,
+      "loss": 0.6591,
+      "step": 289
+    },
+    {
+      "epoch": 0.08042151968940654,
+      "grad_norm": 0.24755387008190155,
+      "learning_rate": 2.4956462087352868e-05,
+      "loss": 0.5883,
+      "step": 290
+    },
+    {
+      "epoch": 0.08069883527454243,
+      "grad_norm": 0.23270832002162933,
+      "learning_rate": 2.4956000421058807e-05,
+      "loss": 0.5727,
+      "step": 291
+    },
+    {
+      "epoch": 0.08097615085967831,
+      "grad_norm": 0.23588238656520844,
+      "learning_rate": 2.4955536324269048e-05,
+      "loss": 0.6178,
+      "step": 292
+    },
+    {
+      "epoch": 0.0812534664448142,
+      "grad_norm": 0.26772478222846985,
+      "learning_rate": 2.4955069797074147e-05,
+      "loss": 0.6214,
+      "step": 293
+    },
+    {
+      "epoch": 0.08153078202995008,
+      "grad_norm": 0.22470323741436005,
+      "learning_rate": 2.495460083956514e-05,
+      "loss": 0.61,
+      "step": 294
+    },
+    {
+      "epoch": 0.08180809761508596,
+      "grad_norm": 0.22901985049247742,
+      "learning_rate": 2.495412945183353e-05,
+      "loss": 0.5965,
+      "step": 295
+    },
+    {
+      "epoch": 0.08208541320022185,
+      "grad_norm": 0.23927484452724457,
+      "learning_rate": 2.49536556339713e-05,
+      "loss": 0.6229,
+      "step": 296
+    },
+    {
+      "epoch": 0.08236272878535773,
+      "grad_norm": 0.2505173087120056,
+      "learning_rate": 2.49531793860709e-05,
+      "loss": 0.5974,
+      "step": 297
+    },
+    {
+      "epoch": 0.08264004437049362,
+      "grad_norm": 0.7846159934997559,
+      "learning_rate": 2.4952700708225263e-05,
+      "loss": 0.6097,
+      "step": 298
+    },
+    {
+      "epoch": 0.0829173599556295,
+      "grad_norm": 0.23284678161144257,
+      "learning_rate": 2.4952219600527786e-05,
+      "loss": 0.6161,
+      "step": 299
+    },
+    {
+      "epoch": 0.08319467554076539,
+      "grad_norm": 0.22659331560134888,
+      "learning_rate": 2.4951736063072356e-05,
+      "loss": 0.5917,
+      "step": 300
+    },
+    {
+      "epoch": 0.08347199112590127,
+      "grad_norm": 0.24401767551898956,
+      "learning_rate": 2.4951250095953315e-05,
+      "loss": 0.6163,
+      "step": 301
+    },
+    {
+      "epoch": 0.08374930671103716,
+      "grad_norm": 0.23994800448417664,
+      "learning_rate": 2.4950761699265487e-05,
+      "loss": 0.6035,
+      "step": 302
+    },
+    {
+      "epoch": 0.08402662229617304,
+      "grad_norm": 0.293527752161026,
+      "learning_rate": 2.495027087310418e-05,
+      "loss": 0.6148,
+      "step": 303
+    },
+    {
+      "epoch": 0.08430393788130892,
+      "grad_norm": 0.2797812819480896,
+      "learning_rate": 2.4949777617565156e-05,
+      "loss": 0.6249,
+      "step": 304
+    },
+    {
+      "epoch": 0.08458125346644481,
+      "grad_norm": 0.2422715574502945,
+      "learning_rate": 2.4949281932744672e-05,
+      "loss": 0.6064,
+      "step": 305
+    },
+    {
+      "epoch": 0.08485856905158069,
+      "grad_norm": 0.2489105761051178,
+      "learning_rate": 2.4948783818739446e-05,
+      "loss": 0.6176,
+      "step": 306
+    },
+    {
+      "epoch": 0.08513588463671658,
+      "grad_norm": 0.23189565539360046,
+      "learning_rate": 2.4948283275646672e-05,
+      "loss": 0.6172,
+      "step": 307
+    },
+    {
+      "epoch": 0.08541320022185246,
+      "grad_norm": 0.21257632970809937,
+      "learning_rate": 2.4947780303564015e-05,
+      "loss": 0.6132,
+      "step": 308
+    },
+    {
+      "epoch": 0.08569051580698835,
+      "grad_norm": 0.23266074061393738,
+      "learning_rate": 2.4947274902589628e-05,
+      "loss": 0.6001,
+      "step": 309
+    },
+    {
+      "epoch": 0.08596783139212424,
+      "grad_norm": 0.21587929129600525,
+      "learning_rate": 2.4946767072822126e-05,
+      "loss": 0.6381,
+      "step": 310
+    },
+    {
+      "epoch": 0.08624514697726013,
+      "grad_norm": 0.23052595555782318,
+      "learning_rate": 2.4946256814360594e-05,
+      "loss": 0.6643,
+      "step": 311
+    },
+    {
+      "epoch": 0.08652246256239601,
+      "grad_norm": 0.2822146713733673,
+      "learning_rate": 2.4945744127304598e-05,
+      "loss": 0.6331,
+      "step": 312
+    },
+    {
+      "epoch": 0.0867997781475319,
+      "grad_norm": 0.22692646086215973,
+      "learning_rate": 2.4945229011754184e-05,
+      "loss": 0.6126,
+      "step": 313
+    },
+    {
+      "epoch": 0.08707709373266778,
+      "grad_norm": 0.2250347137451172,
+      "learning_rate": 2.4944711467809855e-05,
+      "loss": 0.6308,
+      "step": 314
+    },
+    {
+      "epoch": 0.08735440931780367,
+      "grad_norm": 0.21644283831119537,
+      "learning_rate": 2.4944191495572604e-05,
+      "loss": 0.587,
+      "step": 315
+    },
+    {
+      "epoch": 0.08763172490293955,
+      "grad_norm": 0.22959665954113007,
+      "learning_rate": 2.494366909514389e-05,
+      "loss": 0.6138,
+      "step": 316
+    },
+    {
+      "epoch": 0.08790904048807544,
+      "grad_norm": 0.24681390821933746,
+      "learning_rate": 2.4943144266625645e-05,
+      "loss": 0.6309,
+      "step": 317
+    },
+    {
+      "epoch": 0.08818635607321132,
+      "grad_norm": 0.22859139740467072,
+      "learning_rate": 2.4942617010120282e-05,
+      "loss": 0.5937,
+      "step": 318
+    },
+    {
+      "epoch": 0.0884636716583472,
+      "grad_norm": 0.20714016258716583,
+      "learning_rate": 2.4942087325730678e-05,
+      "loss": 0.5925,
+      "step": 319
+    },
+    {
+      "epoch": 0.08874098724348309,
+      "grad_norm": 0.2056405246257782,
+      "learning_rate": 2.494155521356019e-05,
+      "loss": 0.5922,
+      "step": 320
+    },
+    {
+      "epoch": 0.08901830282861897,
+      "grad_norm": 0.22429367899894714,
+      "learning_rate": 2.4941020673712644e-05,
+      "loss": 0.6141,
+      "step": 321
+    },
+    {
+      "epoch": 0.08929561841375486,
+      "grad_norm": 0.2454768568277359,
+      "learning_rate": 2.494048370629235e-05,
+      "loss": 0.6221,
+      "step": 322
+    },
+    {
+      "epoch": 0.08957293399889074,
+      "grad_norm": 0.21887235343456268,
+      "learning_rate": 2.493994431140408e-05,
+      "loss": 0.6249,
+      "step": 323
+    },
+    {
+      "epoch": 0.08985024958402663,
+      "grad_norm": 0.23439091444015503,
+      "learning_rate": 2.493940248915308e-05,
+      "loss": 0.6145,
+      "step": 324
+    },
+    {
+      "epoch": 0.09012756516916251,
+      "grad_norm": 0.21770575642585754,
+      "learning_rate": 2.4938858239645087e-05,
+      "loss": 0.6123,
+      "step": 325
+    },
+    {
+      "epoch": 0.0904048807542984,
+      "grad_norm": 0.24734006822109222,
+      "learning_rate": 2.4938311562986284e-05,
+      "loss": 0.6223,
+      "step": 326
+    },
+    {
+      "epoch": 0.09068219633943428,
+      "grad_norm": 0.22917009890079498,
+      "learning_rate": 2.4937762459283348e-05,
+      "loss": 0.6041,
+      "step": 327
+    },
+    {
+      "epoch": 0.09095951192457016,
+      "grad_norm": 0.22535157203674316,
+      "learning_rate": 2.4937210928643423e-05,
+      "loss": 0.6449,
+      "step": 328
+    },
+    {
+      "epoch": 0.09123682750970605,
+      "grad_norm": 0.21863703429698944,
+      "learning_rate": 2.4936656971174134e-05,
+      "loss": 0.6144,
+      "step": 329
+    },
+    {
+      "epoch": 0.09151414309484193,
+      "grad_norm": 0.24071593582630157,
+      "learning_rate": 2.4936100586983563e-05,
+      "loss": 0.6391,
+      "step": 330
+    },
+    {
+      "epoch": 0.09179145867997782,
+      "grad_norm": 0.21045182645320892,
+      "learning_rate": 2.4935541776180275e-05,
+      "loss": 0.613,
+      "step": 331
+    },
+    {
+      "epoch": 0.0920687742651137,
+      "grad_norm": 0.250699520111084,
+      "learning_rate": 2.493498053887332e-05,
+      "loss": 0.6155,
+      "step": 332
+    },
+    {
+      "epoch": 0.09234608985024959,
+      "grad_norm": 0.22076334059238434,
+      "learning_rate": 2.4934416875172202e-05,
+      "loss": 0.6184,
+      "step": 333
+    },
+    {
+      "epoch": 0.09262340543538547,
+      "grad_norm": 0.22932595014572144,
+      "learning_rate": 2.4933850785186906e-05,
+      "loss": 0.6234,
+      "step": 334
+    },
+    {
+      "epoch": 0.09290072102052135,
+      "grad_norm": 0.2126377820968628,
+      "learning_rate": 2.4933282269027898e-05,
+      "loss": 0.5768,
+      "step": 335
+    },
+    {
+      "epoch": 0.09317803660565724,
+      "grad_norm": 0.21872107684612274,
+      "learning_rate": 2.49327113268061e-05,
+      "loss": 0.5847,
+      "step": 336
+    },
+    {
+      "epoch": 0.09345535219079312,
+      "grad_norm": 0.22751103341579437,
+      "learning_rate": 2.4932137958632922e-05,
+      "loss": 0.6241,
+      "step": 337
+    },
+    {
+      "epoch": 0.09373266777592901,
+      "grad_norm": 0.24364197254180908,
+      "learning_rate": 2.493156216462025e-05,
+      "loss": 0.5956,
+      "step": 338
+    },
+    {
+      "epoch": 0.09400998336106489,
+      "grad_norm": 0.2077159285545349,
+      "learning_rate": 2.493098394488043e-05,
+      "loss": 0.5911,
+      "step": 339
+    },
+    {
+      "epoch": 0.09428729894620078,
+      "grad_norm": 0.24238905310630798,
+      "learning_rate": 2.4930403299526292e-05,
+      "loss": 0.629,
+      "step": 340
+    },
+    {
+      "epoch": 0.09456461453133666,
+      "grad_norm": 0.22944410145282745,
+      "learning_rate": 2.492982022867113e-05,
+      "loss": 0.596,
+      "step": 341
+    },
+    {
+      "epoch": 0.09484193011647254,
+      "grad_norm": 0.22006259858608246,
+      "learning_rate": 2.492923473242872e-05,
+      "loss": 0.581,
+      "step": 342
+    },
+    {
+      "epoch": 0.09511924570160843,
+      "grad_norm": 0.2297179251909256,
+      "learning_rate": 2.4928646810913307e-05,
+      "loss": 0.6107,
+      "step": 343
+    },
+    {
+      "epoch": 0.09539656128674431,
+      "grad_norm": 0.21393971145153046,
+      "learning_rate": 2.4928056464239614e-05,
+      "loss": 0.5773,
+      "step": 344
+    },
+    {
+      "epoch": 0.0956738768718802,
+      "grad_norm": 0.23898737132549286,
+      "learning_rate": 2.4927463692522825e-05,
+      "loss": 0.6119,
+      "step": 345
+    },
+    {
+      "epoch": 0.09595119245701608,
+      "grad_norm": 0.22290126979351044,
+      "learning_rate": 2.4926868495878613e-05,
+      "loss": 0.5721,
+      "step": 346
+    },
+    {
+      "epoch": 0.09622850804215197,
+      "grad_norm": 0.23102609813213348,
+      "learning_rate": 2.4926270874423113e-05,
+      "loss": 0.5735,
+      "step": 347
+    },
+    {
+      "epoch": 0.09650582362728785,
+      "grad_norm": 0.22276602685451508,
+      "learning_rate": 2.4925670828272935e-05,
+      "loss": 0.5799,
+      "step": 348
+    },
+    {
+      "epoch": 0.09678313921242374,
+      "grad_norm": 0.229088693857193,
+      "learning_rate": 2.492506835754517e-05,
+      "loss": 0.6191,
+      "step": 349
+    },
+    {
+      "epoch": 0.09706045479755962,
+      "grad_norm": 0.22365529835224152,
+      "learning_rate": 2.4924463462357373e-05,
+      "loss": 0.5932,
+      "step": 350
+    },
+    {
+      "epoch": 0.0973377703826955,
+      "grad_norm": 0.21552829444408417,
+      "learning_rate": 2.492385614282757e-05,
+      "loss": 0.6091,
+      "step": 351
+    },
+    {
+      "epoch": 0.09761508596783139,
+      "grad_norm": 0.2322327196598053,
+      "learning_rate": 2.4923246399074272e-05,
+      "loss": 0.6216,
+      "step": 352
+    },
+    {
+      "epoch": 0.09789240155296727,
+      "grad_norm": 0.27425798773765564,
+      "learning_rate": 2.4922634231216458e-05,
+      "loss": 0.5915,
+      "step": 353
+    },
+    {
+      "epoch": 0.09816971713810316,
+      "grad_norm": 0.21547527611255646,
+      "learning_rate": 2.492201963937357e-05,
+      "loss": 0.6003,
+      "step": 354
+    },
+    {
+      "epoch": 0.09844703272323904,
+      "grad_norm": 0.24001803994178772,
+      "learning_rate": 2.4921402623665535e-05,
+      "loss": 0.5879,
+      "step": 355
+    },
+    {
+      "epoch": 0.09872434830837493,
+      "grad_norm": 0.2128361016511917,
+      "learning_rate": 2.492078318421275e-05,
+      "loss": 0.6192,
+      "step": 356
+    },
+    {
+      "epoch": 0.09900166389351081,
+      "grad_norm": 0.23891720175743103,
+      "learning_rate": 2.492016132113608e-05,
+      "loss": 0.6237,
+      "step": 357
+    },
+    {
+      "epoch": 0.0992789794786467,
+      "grad_norm": 0.2530137300491333,
+      "learning_rate": 2.4919537034556876e-05,
+      "loss": 0.5975,
+      "step": 358
+    },
+    {
+      "epoch": 0.09955629506378258,
+      "grad_norm": 0.22864577174186707,
+      "learning_rate": 2.4918910324596944e-05,
+      "loss": 0.6085,
+      "step": 359
+    },
+    {
+      "epoch": 0.09983361064891846,
+      "grad_norm": 0.21646267175674438,
+      "learning_rate": 2.4918281191378573e-05,
+      "loss": 0.5734,
+      "step": 360
+    },
+    {
+      "epoch": 0.10011092623405435,
+      "grad_norm": 0.21921531856060028,
+      "learning_rate": 2.491764963502453e-05,
+      "loss": 0.6003,
+      "step": 361
+    },
+    {
+      "epoch": 0.10038824181919023,
+      "grad_norm": 0.22741259634494781,
+      "learning_rate": 2.491701565565804e-05,
+      "loss": 0.59,
+      "step": 362
+    },
+    {
+      "epoch": 0.10066555740432612,
+      "grad_norm": 0.2382003366947174,
+      "learning_rate": 2.4916379253402815e-05,
+      "loss": 0.6021,
+      "step": 363
+    },
+    {
+      "epoch": 0.100942872989462,
+      "grad_norm": 0.20885150134563446,
+      "learning_rate": 2.4915740428383032e-05,
+      "loss": 0.5973,
+      "step": 364
+    },
+    {
+      "epoch": 0.10122018857459789,
+      "grad_norm": 0.1941784769296646,
+      "learning_rate": 2.491509918072334e-05,
+      "loss": 0.591,
+      "step": 365
+    },
+    {
+      "epoch": 0.10149750415973377,
+      "grad_norm": 0.21724678575992584,
+      "learning_rate": 2.491445551054887e-05,
+      "loss": 0.6212,
+      "step": 366
+    },
+    {
+      "epoch": 0.10177481974486967,
+      "grad_norm": 0.29596778750419617,
+      "learning_rate": 2.4913809417985213e-05,
+      "loss": 0.6241,
+      "step": 367
+    },
+    {
+      "epoch": 0.10205213533000555,
+      "grad_norm": 0.23365665972232819,
+      "learning_rate": 2.4913160903158443e-05,
+      "loss": 0.6243,
+      "step": 368
+    },
+    {
+      "epoch": 0.10232945091514144,
+      "grad_norm": 0.23263859748840332,
+      "learning_rate": 2.4912509966195098e-05,
+      "loss": 0.5946,
+      "step": 369
+    },
+    {
+      "epoch": 0.10260676650027732,
+      "grad_norm": 0.22337226569652557,
+      "learning_rate": 2.4911856607222196e-05,
+      "loss": 0.6287,
+      "step": 370
+    },
+    {
+      "epoch": 0.1028840820854132,
+      "grad_norm": 0.23225417733192444,
+      "learning_rate": 2.491120082636722e-05,
+      "loss": 0.5927,
+      "step": 371
+    },
+    {
+      "epoch": 0.10316139767054909,
+      "grad_norm": 0.22292552888393402,
+      "learning_rate": 2.4910542623758142e-05,
+      "loss": 0.6208,
+      "step": 372
+    },
+    {
+      "epoch": 0.10343871325568497,
+      "grad_norm": 0.21180188655853271,
+      "learning_rate": 2.4909881999523382e-05,
+      "loss": 0.5652,
+      "step": 373
+    },
+    {
+      "epoch": 0.10371602884082086,
+      "grad_norm": 0.2395281195640564,
+      "learning_rate": 2.4909218953791853e-05,
+      "loss": 0.5922,
+      "step": 374
+    },
+    {
+      "epoch": 0.10399334442595674,
+      "grad_norm": 0.2313883900642395,
+      "learning_rate": 2.4908553486692926e-05,
+      "loss": 0.6083,
+      "step": 375
+    },
+    {
+      "epoch": 0.10427066001109263,
+      "grad_norm": 0.21677231788635254,
+      "learning_rate": 2.4907885598356456e-05,
+      "loss": 0.6115,
+      "step": 376
+    },
+    {
+      "epoch": 0.10454797559622851,
+      "grad_norm": 0.21811628341674805,
+      "learning_rate": 2.4907215288912766e-05,
+      "loss": 0.5815,
+      "step": 377
+    },
+    {
+      "epoch": 0.1048252911813644,
+      "grad_norm": 0.22422359883785248,
+      "learning_rate": 2.4906542558492652e-05,
+      "loss": 0.6161,
+      "step": 378
+    },
+    {
+      "epoch": 0.10510260676650028,
+      "grad_norm": 0.2190743088722229,
+      "learning_rate": 2.4905867407227377e-05,
+      "loss": 0.5554,
+      "step": 379
+    },
+    {
+      "epoch": 0.10537992235163617,
+      "grad_norm": 0.25590968132019043,
+      "learning_rate": 2.490518983524869e-05,
+      "loss": 0.5856,
+      "step": 380
+    },
+    {
+      "epoch": 0.10565723793677205,
+      "grad_norm": 0.26324909925460815,
+      "learning_rate": 2.490450984268879e-05,
+      "loss": 0.6057,
+      "step": 381
+    },
+    {
+      "epoch": 0.10593455352190793,
+      "grad_norm": 0.2394174039363861,
+      "learning_rate": 2.490382742968037e-05,
+      "loss": 0.6045,
+      "step": 382
+    },
+    {
+      "epoch": 0.10621186910704382,
+      "grad_norm": 0.23230458796024323,
+      "learning_rate": 2.4903142596356586e-05,
+      "loss": 0.6188,
+      "step": 383
+    },
+    {
+      "epoch": 0.1064891846921797,
+      "grad_norm": 0.21763205528259277,
+      "learning_rate": 2.4902455342851067e-05,
+      "loss": 0.5626,
+      "step": 384
+    },
+    {
+      "epoch": 0.10676650027731559,
+      "grad_norm": 0.469051718711853,
+      "learning_rate": 2.490176566929791e-05,
+      "loss": 0.5909,
+      "step": 385
+    },
+    {
+      "epoch": 0.10704381586245147,
+      "grad_norm": 0.24806742370128632,
+      "learning_rate": 2.4901073575831697e-05,
+      "loss": 0.6215,
+      "step": 386
+    },
+    {
+      "epoch": 0.10732113144758736,
+      "grad_norm": 0.22851231694221497,
+      "learning_rate": 2.4900379062587463e-05,
+      "loss": 0.593,
+      "step": 387
+    },
+    {
+      "epoch": 0.10759844703272324,
+      "grad_norm": 0.24515169858932495,
+      "learning_rate": 2.489968212970074e-05,
+      "loss": 0.6036,
+      "step": 388
+    },
+    {
+      "epoch": 0.10787576261785913,
+      "grad_norm": 0.24662603437900543,
+      "learning_rate": 2.4898982777307506e-05,
+      "loss": 0.6153,
+      "step": 389
+    },
+    {
+      "epoch": 0.10815307820299501,
+      "grad_norm": 0.2459113895893097,
+      "learning_rate": 2.4898281005544227e-05,
+      "loss": 0.5771,
+      "step": 390
+    },
+    {
+      "epoch": 0.1084303937881309,
+      "grad_norm": 0.23075874149799347,
+      "learning_rate": 2.489757681454784e-05,
+      "loss": 0.6297,
+      "step": 391
+    },
+    {
+      "epoch": 0.10870770937326678,
+      "grad_norm": 0.24344393610954285,
+      "learning_rate": 2.4896870204455746e-05,
+      "loss": 0.5993,
+      "step": 392
+    },
+    {
+      "epoch": 0.10898502495840266,
+      "grad_norm": 0.2444470226764679,
+      "learning_rate": 2.4896161175405826e-05,
+      "loss": 0.6159,
+      "step": 393
+    },
+    {
+      "epoch": 0.10926234054353855,
+      "grad_norm": 0.24199549853801727,
+      "learning_rate": 2.4895449727536435e-05,
+      "loss": 0.6177,
+      "step": 394
+    },
+    {
+      "epoch": 0.10953965612867443,
+      "grad_norm": 0.20678602159023285,
+      "learning_rate": 2.4894735860986385e-05,
+      "loss": 0.5894,
+      "step": 395
+    },
+    {
+      "epoch": 0.10981697171381032,
+      "grad_norm": 0.25881609320640564,
+      "learning_rate": 2.489401957589498e-05,
+      "loss": 0.631,
+      "step": 396
+    },
+    {
+      "epoch": 0.1100942872989462,
+      "grad_norm": 0.2568078637123108,
+      "learning_rate": 2.489330087240198e-05,
+      "loss": 0.5902,
+      "step": 397
+    },
+    {
+      "epoch": 0.11037160288408208,
+      "grad_norm": 0.2495458871126175,
+      "learning_rate": 2.489257975064763e-05,
+      "loss": 0.6141,
+      "step": 398
+    },
+    {
+      "epoch": 0.11064891846921797,
+      "grad_norm": 0.6050971150398254,
+      "learning_rate": 2.489185621077263e-05,
+      "loss": 0.597,
+      "step": 399
+    },
+    {
+      "epoch": 0.11092623405435385,
+      "grad_norm": 0.22337263822555542,
+      "learning_rate": 2.489113025291817e-05,
+      "loss": 0.58,
+      "step": 400
+    },
+    {
+      "epoch": 0.11120354963948974,
+      "grad_norm": 0.20583049952983856,
+      "learning_rate": 2.4890401877225898e-05,
+      "loss": 0.5751,
+      "step": 401
+    },
+    {
+      "epoch": 0.11148086522462562,
+      "grad_norm": 0.2487124800682068,
+      "learning_rate": 2.488967108383795e-05,
+      "loss": 0.6009,
+      "step": 402
+    },
+    {
+      "epoch": 0.1117581808097615,
+      "grad_norm": 0.24986512959003448,
+      "learning_rate": 2.4888937872896908e-05,
+      "loss": 0.6203,
+      "step": 403
+    },
+    {
+      "epoch": 0.11203549639489739,
+      "grad_norm": 0.28655165433883667,
+      "learning_rate": 2.488820224454585e-05,
+      "loss": 0.6037,
+      "step": 404
+    },
+    {
+      "epoch": 0.11231281198003328,
+      "grad_norm": 0.24651272594928741,
+      "learning_rate": 2.4887464198928317e-05,
+      "loss": 0.5853,
+      "step": 405
+    },
+    {
+      "epoch": 0.11259012756516916,
+      "grad_norm": 0.1938582807779312,
+      "learning_rate": 2.4886723736188318e-05,
+      "loss": 0.5888,
+      "step": 406
+    },
+    {
+      "epoch": 0.11286744315030504,
+      "grad_norm": 0.22223535180091858,
+      "learning_rate": 2.4885980856470338e-05,
+      "loss": 0.627,
+      "step": 407
+    },
+    {
+      "epoch": 0.11314475873544093,
+      "grad_norm": 0.24378454685211182,
+      "learning_rate": 2.4885235559919328e-05,
+      "loss": 0.5827,
+      "step": 408
+    },
+    {
+      "epoch": 0.11342207432057681,
+      "grad_norm": 0.2019236534833908,
+      "learning_rate": 2.4884487846680727e-05,
+      "loss": 0.5976,
+      "step": 409
+    },
+    {
+      "epoch": 0.1136993899057127,
+      "grad_norm": 0.21661922335624695,
+      "learning_rate": 2.4883737716900424e-05,
+      "loss": 0.6013,
+      "step": 410
+    },
+    {
+      "epoch": 0.11397670549084858,
+      "grad_norm": 0.26957792043685913,
+      "learning_rate": 2.4882985170724787e-05,
+      "loss": 0.63,
+      "step": 411
+    },
+    {
+      "epoch": 0.11425402107598447,
+      "grad_norm": 0.21899108588695526,
+      "learning_rate": 2.4882230208300668e-05,
+      "loss": 0.5935,
+      "step": 412
+    },
+    {
+      "epoch": 0.11453133666112035,
+      "grad_norm": 0.2505897879600525,
+      "learning_rate": 2.488147282977537e-05,
+      "loss": 0.5689,
+      "step": 413
+    },
+    {
+      "epoch": 0.11480865224625623,
+      "grad_norm": 0.20966675877571106,
+      "learning_rate": 2.4880713035296686e-05,
+      "loss": 0.5893,
+      "step": 414
+    },
+    {
+      "epoch": 0.11508596783139212,
+      "grad_norm": 0.26599064469337463,
+      "learning_rate": 2.4879950825012864e-05,
+      "loss": 0.5912,
+      "step": 415
+    },
+    {
+      "epoch": 0.115363283416528,
+      "grad_norm": 0.22095918655395508,
+      "learning_rate": 2.487918619907264e-05,
+      "loss": 0.6068,
+      "step": 416
+    },
+    {
+      "epoch": 0.11564059900166389,
+      "grad_norm": 0.20822377502918243,
+      "learning_rate": 2.4878419157625206e-05,
+      "loss": 0.5783,
+      "step": 417
+    },
+    {
+      "epoch": 0.11591791458679977,
+      "grad_norm": 0.20983396470546722,
+      "learning_rate": 2.4877649700820232e-05,
+      "loss": 0.6258,
+      "step": 418
+    },
+    {
+      "epoch": 0.11619523017193566,
+      "grad_norm": 0.2288864701986313,
+      "learning_rate": 2.4876877828807864e-05,
+      "loss": 0.6196,
+      "step": 419
+    },
+    {
+      "epoch": 0.11647254575707154,
+      "grad_norm": 0.20762163400650024,
+      "learning_rate": 2.4876103541738714e-05,
+      "loss": 0.5674,
+      "step": 420
+    },
+    {
+      "epoch": 0.11674986134220743,
+      "grad_norm": 0.2152256816625595,
+      "learning_rate": 2.4875326839763863e-05,
+      "loss": 0.5681,
+      "step": 421
+    },
+    {
+      "epoch": 0.11702717692734331,
+      "grad_norm": 0.25224751234054565,
+      "learning_rate": 2.4874547723034865e-05,
+      "loss": 0.5948,
+      "step": 422
+    },
+    {
+      "epoch": 0.11730449251247921,
+      "grad_norm": 0.21316662430763245,
+      "learning_rate": 2.4873766191703752e-05,
+      "loss": 0.5757,
+      "step": 423
+    },
+    {
+      "epoch": 0.11758180809761509,
+      "grad_norm": 0.20757247507572174,
+      "learning_rate": 2.4872982245923014e-05,
+      "loss": 0.5903,
+      "step": 424
+    },
+    {
+      "epoch": 0.11785912368275098,
+      "grad_norm": 0.23846663534641266,
+      "learning_rate": 2.487219588584563e-05,
+      "loss": 0.5735,
+      "step": 425
+    },
+    {
+      "epoch": 0.11813643926788686,
+      "grad_norm": 0.21389099955558777,
+      "learning_rate": 2.4871407111625027e-05,
+      "loss": 0.5998,
+      "step": 426
+    },
+    {
+      "epoch": 0.11841375485302275,
+      "grad_norm": 0.21840502321720123,
+      "learning_rate": 2.487061592341513e-05,
+      "loss": 0.5854,
+      "step": 427
+    },
+    {
+      "epoch": 0.11869107043815863,
+      "grad_norm": 0.23358672857284546,
+      "learning_rate": 2.4869822321370308e-05,
+      "loss": 0.6212,
+      "step": 428
+    },
+    {
+      "epoch": 0.11896838602329451,
+      "grad_norm": 0.24467387795448303,
+      "learning_rate": 2.4869026305645418e-05,
+      "loss": 0.5937,
+      "step": 429
+    },
+    {
+      "epoch": 0.1192457016084304,
+      "grad_norm": 0.24679329991340637,
+      "learning_rate": 2.486822787639579e-05,
+      "loss": 0.6027,
+      "step": 430
+    },
+    {
+      "epoch": 0.11952301719356628,
+      "grad_norm": 0.22588002681732178,
+      "learning_rate": 2.4867427033777206e-05,
+      "loss": 0.5707,
+      "step": 431
+    },
+    {
+      "epoch": 0.11980033277870217,
+      "grad_norm": 0.20728443562984467,
+      "learning_rate": 2.486662377794594e-05,
+      "loss": 0.5857,
+      "step": 432
+    },
+    {
+      "epoch": 0.12007764836383805,
+      "grad_norm": 0.2292574942111969,
+      "learning_rate": 2.4865818109058732e-05,
+      "loss": 0.6288,
+      "step": 433
+    },
+    {
+      "epoch": 0.12035496394897394,
+      "grad_norm": 0.22358085215091705,
+      "learning_rate": 2.4865010027272784e-05,
+      "loss": 0.6043,
+      "step": 434
+    },
+    {
+      "epoch": 0.12063227953410982,
+      "grad_norm": 0.21650134027004242,
+      "learning_rate": 2.4864199532745776e-05,
+      "loss": 0.5772,
+      "step": 435
+    },
+    {
+      "epoch": 0.1209095951192457,
+      "grad_norm": 0.21783700585365295,
+      "learning_rate": 2.486338662563585e-05,
+      "loss": 0.608,
+      "step": 436
+    },
+    {
+      "epoch": 0.12118691070438159,
+      "grad_norm": 0.2252453863620758,
+      "learning_rate": 2.4862571306101633e-05,
+      "loss": 0.5783,
+      "step": 437
+    },
+    {
+      "epoch": 0.12146422628951747,
+      "grad_norm": 0.22224466502666473,
+      "learning_rate": 2.4861753574302217e-05,
+      "loss": 0.5823,
+      "step": 438
+    },
+    {
+      "epoch": 0.12174154187465336,
+      "grad_norm": 0.24375957250595093,
+      "learning_rate": 2.486093343039716e-05,
+      "loss": 0.5872,
+      "step": 439
+    },
+    {
+      "epoch": 0.12201885745978924,
+      "grad_norm": 0.20903299748897552,
+      "learning_rate": 2.4860110874546495e-05,
+      "loss": 0.6237,
+      "step": 440
+    },
+    {
+      "epoch": 0.12229617304492513,
+      "grad_norm": 0.23007185757160187,
+      "learning_rate": 2.485928590691072e-05,
+      "loss": 0.6188,
+      "step": 441
+    },
+    {
+      "epoch": 0.12257348863006101,
+      "grad_norm": 0.23085376620292664,
+      "learning_rate": 2.4858458527650814e-05,
+      "loss": 0.5693,
+      "step": 442
+    },
+    {
+      "epoch": 0.1228508042151969,
+      "grad_norm": 0.2241743952035904,
+      "learning_rate": 2.485762873692822e-05,
+      "loss": 0.6294,
+      "step": 443
+    },
+    {
+      "epoch": 0.12312811980033278,
+      "grad_norm": 0.20904746651649475,
+      "learning_rate": 2.4856796534904845e-05,
+      "loss": 0.6301,
+      "step": 444
+    },
+    {
+      "epoch": 0.12340543538546866,
+      "grad_norm": 0.6742352843284607,
+      "learning_rate": 2.4855961921743083e-05,
+      "loss": 0.5524,
+      "step": 445
+    },
+    {
+      "epoch": 0.12368275097060455,
+      "grad_norm": 0.20682546496391296,
+      "learning_rate": 2.4855124897605782e-05,
+      "loss": 0.5907,
+      "step": 446
+    },
+    {
+      "epoch": 0.12396006655574043,
+      "grad_norm": 0.2383589744567871,
+      "learning_rate": 2.485428546265627e-05,
+      "loss": 0.5865,
+      "step": 447
+    },
+    {
+      "epoch": 0.12423738214087632,
+      "grad_norm": 0.2051754891872406,
+      "learning_rate": 2.4853443617058348e-05,
+      "loss": 0.6112,
+      "step": 448
+    },
+    {
+      "epoch": 0.1245146977260122,
+      "grad_norm": 0.2156454175710678,
+      "learning_rate": 2.4852599360976274e-05,
+      "loss": 0.5913,
+      "step": 449
+    },
+    {
+      "epoch": 0.12479201331114809,
+      "grad_norm": 0.22987020015716553,
+      "learning_rate": 2.485175269457479e-05,
+      "loss": 0.5873,
+      "step": 450
+    },
+    {
+      "epoch": 0.12506932889628397,
+      "grad_norm": 0.20809032022953033,
+      "learning_rate": 2.4850903618019102e-05,
+      "loss": 0.582,
+      "step": 451
+    },
+    {
+      "epoch": 0.12534664448141986,
+      "grad_norm": 0.2254360467195511,
+      "learning_rate": 2.485005213147489e-05,
+      "loss": 0.5998,
+      "step": 452
+    },
+    {
+      "epoch": 0.12562396006655574,
+      "grad_norm": 0.214163139462471,
+      "learning_rate": 2.4849198235108296e-05,
+      "loss": 0.5884,
+      "step": 453
+    },
+    {
+      "epoch": 0.12590127565169162,
+      "grad_norm": 0.21463198959827423,
+      "learning_rate": 2.484834192908594e-05,
+      "loss": 0.606,
+      "step": 454
+    },
+    {
+      "epoch": 0.1261785912368275,
+      "grad_norm": 0.20102332532405853,
+      "learning_rate": 2.4847483213574908e-05,
+      "loss": 0.6012,
+      "step": 455
+    },
+    {
+      "epoch": 0.1264559068219634,
+      "grad_norm": 0.19328515231609344,
+      "learning_rate": 2.4846622088742765e-05,
+      "loss": 0.5749,
+      "step": 456
+    },
+    {
+      "epoch": 0.12673322240709928,
+      "grad_norm": 0.20251993834972382,
+      "learning_rate": 2.484575855475753e-05,
+      "loss": 0.6121,
+      "step": 457
+    },
+    {
+      "epoch": 0.12701053799223516,
+      "grad_norm": 0.21547801792621613,
+      "learning_rate": 2.484489261178771e-05,
+      "loss": 0.6019,
+      "step": 458
+    },
+    {
+      "epoch": 0.12728785357737105,
+      "grad_norm": 0.21968044340610504,
+      "learning_rate": 2.4844024260002276e-05,
+      "loss": 0.5863,
+      "step": 459
+    },
+    {
+      "epoch": 0.12756516916250693,
+      "grad_norm": 0.21164929866790771,
+      "learning_rate": 2.4843153499570648e-05,
+      "loss": 0.5995,
+      "step": 460
+    },
+    {
+      "epoch": 0.12784248474764282,
+      "grad_norm": 0.2152341902256012,
+      "learning_rate": 2.4842280330662753e-05,
+      "loss": 0.6374,
+      "step": 461
+    },
+    {
+      "epoch": 0.1281198003327787,
+      "grad_norm": 0.19914227724075317,
+      "learning_rate": 2.4841404753448963e-05,
+      "loss": 0.5919,
+      "step": 462
+    },
+    {
+      "epoch": 0.12839711591791458,
+      "grad_norm": 0.2268274873495102,
+      "learning_rate": 2.4840526768100124e-05,
+      "loss": 0.5913,
+      "step": 463
+    },
+    {
+      "epoch": 0.12867443150305047,
+      "grad_norm": 0.21451812982559204,
+      "learning_rate": 2.483964637478756e-05,
+      "loss": 0.6146,
+      "step": 464
+    },
+    {
+      "epoch": 0.12895174708818635,
+      "grad_norm": 0.1978655308485031,
+      "learning_rate": 2.483876357368305e-05,
+      "loss": 0.5938,
+      "step": 465
+    },
+    {
+      "epoch": 0.12922906267332224,
+      "grad_norm": 0.20545656979084015,
+      "learning_rate": 2.4837878364958865e-05,
+      "loss": 0.6172,
+      "step": 466
+    },
+    {
+      "epoch": 0.12950637825845812,
+      "grad_norm": 0.21529193222522736,
+      "learning_rate": 2.483699074878772e-05,
+      "loss": 0.5794,
+      "step": 467
+    },
+    {
+      "epoch": 0.129783693843594,
+      "grad_norm": 0.2971234917640686,
+      "learning_rate": 2.4836100725342818e-05,
+      "loss": 0.6166,
+      "step": 468
+    },
+    {
+      "epoch": 0.1300610094287299,
+      "grad_norm": 0.1968923807144165,
+      "learning_rate": 2.4835208294797824e-05,
+      "loss": 0.5898,
+      "step": 469
+    },
+    {
+      "epoch": 0.13033832501386577,
+      "grad_norm": 0.2248852252960205,
+      "learning_rate": 2.483431345732688e-05,
+      "loss": 0.5984,
+      "step": 470
+    },
+    {
+      "epoch": 0.13061564059900166,
+      "grad_norm": 0.21942903101444244,
+      "learning_rate": 2.4833416213104588e-05,
+      "loss": 0.5984,
+      "step": 471
+    },
+    {
+      "epoch": 0.13089295618413754,
+      "grad_norm": 0.22266723215579987,
+      "learning_rate": 2.4832516562306024e-05,
+      "loss": 0.5858,
+      "step": 472
+    },
+    {
+      "epoch": 0.13117027176927343,
+      "grad_norm": 0.21460357308387756,
+      "learning_rate": 2.483161450510674e-05,
+      "loss": 0.5763,
+      "step": 473
+    },
+    {
+      "epoch": 0.1314475873544093,
+      "grad_norm": 0.333474725484848,
+      "learning_rate": 2.4830710041682735e-05,
+      "loss": 0.6024,
+      "step": 474
+    },
+    {
+      "epoch": 0.1317249029395452,
+      "grad_norm": 0.1983480155467987,
+      "learning_rate": 2.4829803172210515e-05,
+      "loss": 0.5898,
+      "step": 475
+    },
+    {
+      "epoch": 0.13200221852468108,
+      "grad_norm": 0.2835070788860321,
+      "learning_rate": 2.482889389686702e-05,
+      "loss": 0.571,
+      "step": 476
+    },
+    {
+      "epoch": 0.13227953410981697,
+      "grad_norm": 0.2176080197095871,
+      "learning_rate": 2.4827982215829674e-05,
+      "loss": 0.5875,
+      "step": 477
+    },
+    {
+      "epoch": 0.13255684969495285,
+      "grad_norm": 0.2436138391494751,
+      "learning_rate": 2.482706812927638e-05,
+      "loss": 0.5965,
+      "step": 478
+    },
+    {
+      "epoch": 0.13283416528008873,
+      "grad_norm": 0.21060815453529358,
+      "learning_rate": 2.4826151637385495e-05,
+      "loss": 0.5881,
+      "step": 479
+    },
+    {
+      "epoch": 0.13311148086522462,
+      "grad_norm": 0.49135246872901917,
+      "learning_rate": 2.4825232740335847e-05,
+      "loss": 0.5742,
+      "step": 480
+    },
+    {
+      "epoch": 0.1333887964503605,
+      "grad_norm": 0.20535485446453094,
+      "learning_rate": 2.4824311438306742e-05,
+      "loss": 0.5877,
+      "step": 481
+    },
+    {
+      "epoch": 0.1336661120354964,
+      "grad_norm": 0.20854201912879944,
+      "learning_rate": 2.482338773147795e-05,
+      "loss": 0.6065,
+      "step": 482
+    },
+    {
+      "epoch": 0.13394342762063227,
+      "grad_norm": 0.20914287865161896,
+      "learning_rate": 2.4822461620029708e-05,
+      "loss": 0.5919,
+      "step": 483
+    },
+    {
+      "epoch": 0.13422074320576816,
+      "grad_norm": 0.20028036832809448,
+      "learning_rate": 2.4821533104142724e-05,
+      "loss": 0.5707,
+      "step": 484
+    },
+    {
+      "epoch": 0.13449805879090404,
+      "grad_norm": 0.22616969048976898,
+      "learning_rate": 2.4820602183998185e-05,
+      "loss": 0.5896,
+      "step": 485
+    },
+    {
+      "epoch": 0.13477537437603992,
+      "grad_norm": 0.2049257457256317,
+      "learning_rate": 2.4819668859777728e-05,
+      "loss": 0.5693,
+      "step": 486
+    },
+    {
+      "epoch": 0.1350526899611758,
+      "grad_norm": 0.21746453642845154,
+      "learning_rate": 2.4818733131663473e-05,
+      "loss": 0.6177,
+      "step": 487
+    },
+    {
+      "epoch": 0.1353300055463117,
+      "grad_norm": 0.20084752142429352,
+      "learning_rate": 2.4817794999838004e-05,
+      "loss": 0.5871,
+      "step": 488
+    },
+    {
+      "epoch": 0.13560732113144758,
+      "grad_norm": 0.2062511444091797,
+      "learning_rate": 2.4816854464484378e-05,
+      "loss": 0.5975,
+      "step": 489
+    },
+    {
+      "epoch": 0.13588463671658346,
+      "grad_norm": 0.2201562523841858,
+      "learning_rate": 2.4815911525786118e-05,
+      "loss": 0.5683,
+      "step": 490
+    },
+    {
+      "epoch": 0.13616195230171935,
+      "grad_norm": 0.22616079449653625,
+      "learning_rate": 2.4814966183927213e-05,
+      "loss": 0.6306,
+      "step": 491
+    },
+    {
+      "epoch": 0.13643926788685523,
+      "grad_norm": 0.21003180742263794,
+      "learning_rate": 2.4814018439092128e-05,
+      "loss": 0.6064,
+      "step": 492
+    },
+    {
+      "epoch": 0.13671658347199112,
+      "grad_norm": 0.2046622782945633,
+      "learning_rate": 2.481306829146579e-05,
+      "loss": 0.6107,
+      "step": 493
+    },
+    {
+      "epoch": 0.136993899057127,
+      "grad_norm": 0.2102370411157608,
+      "learning_rate": 2.4812115741233606e-05,
+      "loss": 0.596,
+      "step": 494
+    },
+    {
+      "epoch": 0.13727121464226288,
+      "grad_norm": 0.20774902403354645,
+      "learning_rate": 2.4811160788581434e-05,
+      "loss": 0.6111,
+      "step": 495
+    },
+    {
+      "epoch": 0.13754853022739877,
+      "grad_norm": 0.20868700742721558,
+      "learning_rate": 2.481020343369561e-05,
+      "loss": 0.604,
+      "step": 496
+    },
+    {
+      "epoch": 0.13782584581253465,
+      "grad_norm": 0.20590144395828247,
+      "learning_rate": 2.4809243676762947e-05,
+      "loss": 0.606,
+      "step": 497
+    },
+    {
+      "epoch": 0.13810316139767054,
+      "grad_norm": 0.2019280344247818,
+      "learning_rate": 2.4808281517970716e-05,
+      "loss": 0.6034,
+      "step": 498
+    },
+    {
+      "epoch": 0.13838047698280642,
+      "grad_norm": 0.22689440846443176,
+      "learning_rate": 2.4807316957506656e-05,
+      "loss": 0.5715,
+      "step": 499
+    },
+    {
+      "epoch": 0.1386577925679423,
+      "grad_norm": 0.2134653776884079,
+      "learning_rate": 2.4806349995558986e-05,
+      "loss": 0.6184,
+      "step": 500
+    },
+    {
+      "epoch": 0.1389351081530782,
+      "grad_norm": 0.20334339141845703,
+      "learning_rate": 2.4805380632316377e-05,
+      "loss": 0.5804,
+      "step": 501
+    },
+    {
+      "epoch": 0.13921242373821408,
+      "grad_norm": 0.20713390409946442,
+      "learning_rate": 2.4804408867967984e-05,
+      "loss": 0.5898,
+      "step": 502
+    },
+    {
+      "epoch": 0.13948973932334996,
+      "grad_norm": 0.21584905683994293,
+      "learning_rate": 2.4803434702703422e-05,
+      "loss": 0.5957,
+      "step": 503
+    },
+    {
+      "epoch": 0.13976705490848584,
+      "grad_norm": 0.21197180449962616,
+      "learning_rate": 2.4802458136712775e-05,
+      "loss": 0.5981,
+      "step": 504
+    },
+    {
+      "epoch": 0.14004437049362173,
+      "grad_norm": 0.19864031672477722,
+      "learning_rate": 2.4801479170186597e-05,
+      "loss": 0.6027,
+      "step": 505
+    },
+    {
+      "epoch": 0.1403216860787576,
+      "grad_norm": 0.21110500395298004,
+      "learning_rate": 2.4800497803315913e-05,
+      "loss": 0.5882,
+      "step": 506
+    },
+    {
+      "epoch": 0.1405990016638935,
+      "grad_norm": 0.20834285020828247,
+      "learning_rate": 2.4799514036292215e-05,
+      "loss": 0.5935,
+      "step": 507
+    },
+    {
+      "epoch": 0.1408763172490294,
+      "grad_norm": 0.22122903168201447,
+      "learning_rate": 2.4798527869307454e-05,
+      "loss": 0.6011,
+      "step": 508
+    },
+    {
+      "epoch": 0.1411536328341653,
+      "grad_norm": 0.21510954201221466,
+      "learning_rate": 2.4797539302554064e-05,
+      "loss": 0.6266,
+      "step": 509
+    },
+    {
+      "epoch": 0.14143094841930118,
+      "grad_norm": 0.20589859783649445,
+      "learning_rate": 2.479654833622494e-05,
+      "loss": 0.5858,
+      "step": 510
+    },
+    {
+      "epoch": 0.14170826400443706,
+      "grad_norm": 0.20928624272346497,
+      "learning_rate": 2.4795554970513445e-05,
+      "loss": 0.6006,
+      "step": 511
+    },
+    {
+      "epoch": 0.14198557958957295,
+      "grad_norm": 0.2174837589263916,
+      "learning_rate": 2.4794559205613412e-05,
+      "loss": 0.5792,
+      "step": 512
+    },
+    {
+      "epoch": 0.14226289517470883,
+      "grad_norm": 0.20877033472061157,
+      "learning_rate": 2.4793561041719137e-05,
+      "loss": 0.5662,
+      "step": 513
+    },
+    {
+      "epoch": 0.14254021075984472,
+      "grad_norm": 0.240639790892601,
+      "learning_rate": 2.479256047902539e-05,
+      "loss": 0.5824,
+      "step": 514
+    },
+    {
+      "epoch": 0.1428175263449806,
+      "grad_norm": 0.21567635238170624,
+      "learning_rate": 2.479155751772741e-05,
+      "loss": 0.5833,
+      "step": 515
+    },
+    {
+      "epoch": 0.14309484193011648,
+      "grad_norm": 0.2284121960401535,
+      "learning_rate": 2.4790552158020896e-05,
+      "loss": 0.6057,
+      "step": 516
+    },
+    {
+      "epoch": 0.14337215751525237,
+      "grad_norm": 0.19480617344379425,
+      "learning_rate": 2.478954440010203e-05,
+      "loss": 0.5972,
+      "step": 517
+    },
+    {
+      "epoch": 0.14364947310038825,
+      "grad_norm": 0.20838883519172668,
+      "learning_rate": 2.4788534244167443e-05,
+      "loss": 0.6373,
+      "step": 518
+    },
+    {
+      "epoch": 0.14392678868552414,
+      "grad_norm": 0.21365465223789215,
+      "learning_rate": 2.4787521690414245e-05,
+      "loss": 0.5796,
+      "step": 519
+    },
+    {
+      "epoch": 0.14420410427066002,
+      "grad_norm": 2.2805471420288086,
+      "learning_rate": 2.4786506739040018e-05,
+      "loss": 0.5915,
+      "step": 520
+    },
+    {
+      "epoch": 0.1444814198557959,
+      "grad_norm": 0.34635624289512634,
+      "learning_rate": 2.47854893902428e-05,
+      "loss": 0.6325,
+      "step": 521
+    },
+    {
+      "epoch": 0.1447587354409318,
+      "grad_norm": 0.39266762137413025,
+      "learning_rate": 2.47844696442211e-05,
+      "loss": 0.5756,
+      "step": 522
+    },
+    {
+      "epoch": 0.14503605102606767,
+      "grad_norm": 0.31766456365585327,
+      "learning_rate": 2.4783447501173907e-05,
+      "loss": 0.5703,
+      "step": 523
+    },
+    {
+      "epoch": 0.14531336661120356,
+      "grad_norm": 0.24752533435821533,
+      "learning_rate": 2.478242296130066e-05,
+      "loss": 0.5878,
+      "step": 524
+    },
+    {
+      "epoch": 0.14559068219633944,
+      "grad_norm": 0.24595655500888824,
+      "learning_rate": 2.4781396024801272e-05,
+      "loss": 0.5819,
+      "step": 525
+    },
+    {
+      "epoch": 0.14586799778147533,
+      "grad_norm": 0.2457636296749115,
+      "learning_rate": 2.478036669187614e-05,
+      "loss": 0.599,
+      "step": 526
+    },
+    {
+      "epoch": 0.1461453133666112,
+      "grad_norm": 0.244289368391037,
+      "learning_rate": 2.4779334962726096e-05,
+      "loss": 0.5922,
+      "step": 527
+    },
+    {
+      "epoch": 0.1464226289517471,
+      "grad_norm": 0.23528233170509338,
+      "learning_rate": 2.477830083755247e-05,
+      "loss": 0.6032,
+      "step": 528
+    },
+    {
+      "epoch": 0.14669994453688298,
+      "grad_norm": 0.2198038101196289,
+      "learning_rate": 2.477726431655704e-05,
+      "loss": 0.5954,
+      "step": 529
+    },
+    {
+      "epoch": 0.14697726012201887,
+      "grad_norm": 0.23673711717128754,
+      "learning_rate": 2.4776225399942066e-05,
+      "loss": 0.5938,
+      "step": 530
+    },
+    {
+      "epoch": 0.14725457570715475,
+      "grad_norm": 0.2085774540901184,
+      "learning_rate": 2.4775184087910262e-05,
+      "loss": 0.5856,
+      "step": 531
+    },
+    {
+      "epoch": 0.14753189129229063,
+      "grad_norm": 0.21415582299232483,
+      "learning_rate": 2.4774140380664816e-05,
+      "loss": 0.5751,
+      "step": 532
+    },
+    {
+      "epoch": 0.14780920687742652,
+      "grad_norm": 0.2082296758890152,
+      "learning_rate": 2.4773094278409388e-05,
+      "loss": 0.5573,
+      "step": 533
+    },
+    {
+      "epoch": 0.1480865224625624,
+      "grad_norm": 0.20202411711215973,
+      "learning_rate": 2.4772045781348093e-05,
+      "loss": 0.5883,
+      "step": 534
+    },
+    {
+      "epoch": 0.1483638380476983,
+      "grad_norm": 0.20766015350818634,
+      "learning_rate": 2.477099488968553e-05,
+      "loss": 0.6066,
+      "step": 535
+    },
+    {
+      "epoch": 0.14864115363283417,
+      "grad_norm": 0.2137647122144699,
+      "learning_rate": 2.4769941603626744e-05,
+      "loss": 0.597,
+      "step": 536
+    },
+    {
+      "epoch": 0.14891846921797006,
+      "grad_norm": 0.23699134588241577,
+      "learning_rate": 2.4768885923377265e-05,
+      "loss": 0.587,
+      "step": 537
+    },
+    {
+      "epoch": 0.14919578480310594,
+      "grad_norm": 0.21466752886772156,
+      "learning_rate": 2.4767827849143087e-05,
+      "loss": 0.5725,
+      "step": 538
+    },
+    {
+      "epoch": 0.14947310038824183,
+      "grad_norm": 0.20940807461738586,
+      "learning_rate": 2.476676738113067e-05,
+      "loss": 0.5807,
+      "step": 539
+    },
+    {
+      "epoch": 0.1497504159733777,
+      "grad_norm": 0.22769619524478912,
+      "learning_rate": 2.476570451954693e-05,
+      "loss": 0.6089,
+      "step": 540
+    },
+    {
+      "epoch": 0.1500277315585136,
+      "grad_norm": 0.20399393141269684,
+      "learning_rate": 2.4764639264599266e-05,
+      "loss": 0.5705,
+      "step": 541
+    },
+    {
+      "epoch": 0.15030504714364948,
+      "grad_norm": 0.2241872102022171,
+      "learning_rate": 2.4763571616495535e-05,
+      "loss": 0.5731,
+      "step": 542
+    },
+    {
+      "epoch": 0.15058236272878536,
+      "grad_norm": 0.20283614099025726,
+      "learning_rate": 2.4762501575444062e-05,
+      "loss": 0.6051,
+      "step": 543
+    },
+    {
+      "epoch": 0.15085967831392125,
+      "grad_norm": 0.2145642638206482,
+      "learning_rate": 2.4761429141653646e-05,
+      "loss": 0.6069,
+      "step": 544
+    },
+    {
+      "epoch": 0.15113699389905713,
+      "grad_norm": 0.2139946073293686,
+      "learning_rate": 2.4760354315333546e-05,
+      "loss": 0.6055,
+      "step": 545
+    },
+    {
+      "epoch": 0.15141430948419302,
+      "grad_norm": 0.22807584702968597,
+      "learning_rate": 2.4759277096693486e-05,
+      "loss": 0.5945,
+      "step": 546
+    },
+    {
+      "epoch": 0.1516916250693289,
+      "grad_norm": 0.2132754623889923,
+      "learning_rate": 2.4758197485943657e-05,
+      "loss": 0.5975,
+      "step": 547
+    },
+    {
+      "epoch": 0.15196894065446478,
+      "grad_norm": 0.2016879767179489,
+      "learning_rate": 2.4757115483294724e-05,
+      "loss": 0.5863,
+      "step": 548
+    },
+    {
+      "epoch": 0.15224625623960067,
+      "grad_norm": 0.227370485663414,
+      "learning_rate": 2.475603108895782e-05,
+      "loss": 0.583,
+      "step": 549
+    },
+    {
+      "epoch": 0.15252357182473655,
+      "grad_norm": 0.22234570980072021,
+      "learning_rate": 2.475494430314453e-05,
+      "loss": 0.5962,
+      "step": 550
+    },
+    {
+      "epoch": 0.15280088740987244,
+      "grad_norm": 0.20360559225082397,
+      "learning_rate": 2.4753855126066916e-05,
+      "loss": 0.587,
+      "step": 551
+    },
+    {
+      "epoch": 0.15307820299500832,
+      "grad_norm": 0.23359502851963043,
+      "learning_rate": 2.475276355793751e-05,
+      "loss": 0.5967,
+      "step": 552
+    },
+    {
+      "epoch": 0.1533555185801442,
+      "grad_norm": 0.23216257989406586,
+      "learning_rate": 2.47516695989693e-05,
+      "loss": 0.5845,
+      "step": 553
+    },
+    {
+      "epoch": 0.1536328341652801,
+      "grad_norm": 0.21213343739509583,
+      "learning_rate": 2.475057324937575e-05,
+      "loss": 0.5821,
+      "step": 554
+    },
+    {
+      "epoch": 0.15391014975041598,
+      "grad_norm": 0.21203738451004028,
+      "learning_rate": 2.4749474509370784e-05,
+      "loss": 0.5792,
+      "step": 555
+    },
+    {
+      "epoch": 0.15418746533555186,
+      "grad_norm": 0.21234023571014404,
+      "learning_rate": 2.4748373379168805e-05,
+      "loss": 0.5985,
+      "step": 556
+    },
+    {
+      "epoch": 0.15446478092068774,
+      "grad_norm": 0.20847538113594055,
+      "learning_rate": 2.4747269858984658e-05,
+      "loss": 0.595,
+      "step": 557
+    },
+    {
+      "epoch": 0.15474209650582363,
+      "grad_norm": 0.20475518703460693,
+      "learning_rate": 2.474616394903368e-05,
+      "loss": 0.5821,
+      "step": 558
+    },
+    {
+      "epoch": 0.1550194120909595,
+      "grad_norm": 0.211504727602005,
+      "learning_rate": 2.474505564953166e-05,
+      "loss": 0.572,
+      "step": 559
+    },
+    {
+      "epoch": 0.1552967276760954,
+      "grad_norm": 0.21250484883785248,
+      "learning_rate": 2.4743944960694854e-05,
+      "loss": 0.5748,
+      "step": 560
+    },
+    {
+      "epoch": 0.15557404326123128,
+      "grad_norm": 0.2148432582616806,
+      "learning_rate": 2.4742831882739988e-05,
+      "loss": 0.5881,
+      "step": 561
+    },
+    {
+      "epoch": 0.15585135884636717,
+      "grad_norm": 0.19098572432994843,
+      "learning_rate": 2.4741716415884257e-05,
+      "loss": 0.5989,
+      "step": 562
+    },
+    {
+      "epoch": 0.15612867443150305,
+      "grad_norm": 0.20260894298553467,
+      "learning_rate": 2.474059856034531e-05,
+      "loss": 0.567,
+      "step": 563
+    },
+    {
+      "epoch": 0.15640599001663893,
+      "grad_norm": 0.21840746700763702,
+      "learning_rate": 2.4739478316341282e-05,
+      "loss": 0.6054,
+      "step": 564
+    },
+    {
+      "epoch": 0.15668330560177482,
+      "grad_norm": 0.2050980031490326,
+      "learning_rate": 2.473835568409075e-05,
+      "loss": 0.5842,
+      "step": 565
+    },
+    {
+      "epoch": 0.1569606211869107,
+      "grad_norm": 0.20163971185684204,
+      "learning_rate": 2.473723066381278e-05,
+      "loss": 0.5823,
+      "step": 566
+    },
+    {
+      "epoch": 0.1572379367720466,
+      "grad_norm": 0.2088451236486435,
+      "learning_rate": 2.473610325572689e-05,
+      "loss": 0.5995,
+      "step": 567
+    },
+    {
+      "epoch": 0.15751525235718247,
+      "grad_norm": 0.20921272039413452,
+      "learning_rate": 2.4734973460053056e-05,
+      "loss": 0.585,
+      "step": 568
+    },
+    {
+      "epoch": 0.15779256794231836,
+      "grad_norm": 0.22330057621002197,
+      "learning_rate": 2.473384127701175e-05,
+      "loss": 0.5888,
+      "step": 569
+    },
+    {
+      "epoch": 0.15806988352745424,
+      "grad_norm": 0.2152683287858963,
+      "learning_rate": 2.4732706706823876e-05,
+      "loss": 0.5942,
+      "step": 570
+    },
+    {
+      "epoch": 0.15834719911259013,
+      "grad_norm": 0.20223170518875122,
+      "learning_rate": 2.4731569749710824e-05,
+      "loss": 0.5781,
+      "step": 571
+    },
+    {
+      "epoch": 0.158624514697726,
+      "grad_norm": 0.20824022591114044,
+      "learning_rate": 2.4730430405894446e-05,
+      "loss": 0.6404,
+      "step": 572
+    },
+    {
+      "epoch": 0.1589018302828619,
+      "grad_norm": 0.19907240569591522,
+      "learning_rate": 2.4729288675597058e-05,
+      "loss": 0.5983,
+      "step": 573
+    },
+    {
+      "epoch": 0.15917914586799778,
+      "grad_norm": 0.20674046874046326,
+      "learning_rate": 2.472814455904144e-05,
+      "loss": 0.5595,
+      "step": 574
+    },
+    {
+      "epoch": 0.15945646145313366,
+      "grad_norm": 0.19486385583877563,
+      "learning_rate": 2.4726998056450833e-05,
+      "loss": 0.5783,
+      "step": 575
+    },
+    {
+      "epoch": 0.15973377703826955,
+      "grad_norm": 0.2102123498916626,
+      "learning_rate": 2.4725849168048965e-05,
+      "loss": 0.5809,
+      "step": 576
+    },
+    {
+      "epoch": 0.16001109262340543,
+      "grad_norm": 0.21006052196025848,
+      "learning_rate": 2.4724697894060005e-05,
+      "loss": 0.5882,
+      "step": 577
+    },
+    {
+      "epoch": 0.16028840820854132,
+      "grad_norm": 0.22287555038928986,
+      "learning_rate": 2.47235442347086e-05,
+      "loss": 0.6012,
+      "step": 578
+    },
+    {
+      "epoch": 0.1605657237936772,
+      "grad_norm": 0.20599472522735596,
+      "learning_rate": 2.4722388190219852e-05,
+      "loss": 0.5971,
+      "step": 579
+    },
+    {
+      "epoch": 0.16084303937881309,
+      "grad_norm": 0.21176591515541077,
+      "learning_rate": 2.4721229760819348e-05,
+      "loss": 0.5954,
+      "step": 580
+    },
+    {
+      "epoch": 0.16112035496394897,
+      "grad_norm": 0.24732773005962372,
+      "learning_rate": 2.4720068946733123e-05,
+      "loss": 0.5818,
+      "step": 581
+    },
+    {
+      "epoch": 0.16139767054908485,
+      "grad_norm": 0.20434054732322693,
+      "learning_rate": 2.4718905748187677e-05,
+      "loss": 0.5745,
+      "step": 582
+    },
+    {
+      "epoch": 0.16167498613422074,
+      "grad_norm": 0.20684310793876648,
+      "learning_rate": 2.4717740165409988e-05,
+      "loss": 0.5663,
+      "step": 583
+    },
+    {
+      "epoch": 0.16195230171935662,
+      "grad_norm": 0.2029474377632141,
+      "learning_rate": 2.471657219862749e-05,
+      "loss": 0.5855,
+      "step": 584
+    },
+    {
+      "epoch": 0.1622296173044925,
+      "grad_norm": 0.2033785730600357,
+      "learning_rate": 2.4715401848068086e-05,
+      "loss": 0.6119,
+      "step": 585
+    },
+    {
+      "epoch": 0.1625069328896284,
+      "grad_norm": 0.21371322870254517,
+      "learning_rate": 2.4714229113960135e-05,
+      "loss": 0.6022,
+      "step": 586
+    },
+    {
+      "epoch": 0.16278424847476428,
+      "grad_norm": 0.20918406546115875,
+      "learning_rate": 2.4713053996532477e-05,
+      "loss": 0.569,
+      "step": 587
+    },
+    {
+      "epoch": 0.16306156405990016,
+      "grad_norm": 0.2060522437095642,
+      "learning_rate": 2.4711876496014407e-05,
+      "loss": 0.5982,
+      "step": 588
+    },
+    {
+      "epoch": 0.16333887964503604,
+      "grad_norm": 0.20782527327537537,
+      "learning_rate": 2.4710696612635688e-05,
+      "loss": 0.6015,
+      "step": 589
+    },
+    {
+      "epoch": 0.16361619523017193,
+      "grad_norm": 0.20826764404773712,
+      "learning_rate": 2.4709514346626536e-05,
+      "loss": 0.6094,
+      "step": 590
+    },
+    {
+      "epoch": 0.1638935108153078,
+      "grad_norm": 0.20720824599266052,
+      "learning_rate": 2.4708329698217652e-05,
+      "loss": 0.6054,
+      "step": 591
+    },
+    {
+      "epoch": 0.1641708264004437,
+      "grad_norm": 0.19394385814666748,
+      "learning_rate": 2.4707142667640193e-05,
+      "loss": 0.5812,
+      "step": 592
+    },
+    {
+      "epoch": 0.16444814198557958,
+      "grad_norm": 0.2022271454334259,
+      "learning_rate": 2.4705953255125777e-05,
+      "loss": 0.6084,
+      "step": 593
+    },
+    {
+      "epoch": 0.16472545757071547,
+      "grad_norm": 0.21304059028625488,
+      "learning_rate": 2.4704761460906488e-05,
+      "loss": 0.5673,
+      "step": 594
+    },
+    {
+      "epoch": 0.16500277315585135,
+      "grad_norm": 0.20137831568717957,
+      "learning_rate": 2.470356728521488e-05,
+      "loss": 0.5945,
+      "step": 595
+    },
+    {
+      "epoch": 0.16528008874098724,
+      "grad_norm": 0.20188415050506592,
+      "learning_rate": 2.470237072828397e-05,
+      "loss": 0.5849,
+      "step": 596
+    },
+    {
+      "epoch": 0.16555740432612312,
+      "grad_norm": 0.206806018948555,
+      "learning_rate": 2.4701171790347233e-05,
+      "loss": 0.5863,
+      "step": 597
+    },
+    {
+      "epoch": 0.165834719911259,
+      "grad_norm": 0.2093089371919632,
+      "learning_rate": 2.4699970471638613e-05,
+      "loss": 0.601,
+      "step": 598
+    },
+    {
+      "epoch": 0.1661120354963949,
+      "grad_norm": 0.19595085084438324,
+      "learning_rate": 2.4698766772392524e-05,
+      "loss": 0.5993,
+      "step": 599
+    },
+    {
+      "epoch": 0.16638935108153077,
+      "grad_norm": 0.20450963079929352,
+      "learning_rate": 2.469756069284384e-05,
+      "loss": 0.5875,
+      "step": 600
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.18902625143527985,
+      "learning_rate": 2.4696352233227894e-05,
+      "loss": 0.5943,
+      "step": 601
+    },
+    {
+      "epoch": 0.16694398225180254,
+      "grad_norm": 0.19143831729888916,
+      "learning_rate": 2.469514139378049e-05,
+      "loss": 0.5925,
+      "step": 602
+    },
+    {
+      "epoch": 0.16722129783693843,
+      "grad_norm": 0.20280803740024567,
+      "learning_rate": 2.46939281747379e-05,
+      "loss": 0.596,
+      "step": 603
+    },
+    {
+      "epoch": 0.1674986134220743,
+      "grad_norm": 0.20762760937213898,
+      "learning_rate": 2.4692712576336848e-05,
+      "loss": 0.5951,
+      "step": 604
+    },
+    {
+      "epoch": 0.1677759290072102,
+      "grad_norm": 0.209476500749588,
+      "learning_rate": 2.4691494598814536e-05,
+      "loss": 0.5988,
+      "step": 605
+    },
+    {
+      "epoch": 0.16805324459234608,
+      "grad_norm": 0.23190903663635254,
+      "learning_rate": 2.4690274242408617e-05,
+      "loss": 0.5928,
+      "step": 606
+    },
+    {
+      "epoch": 0.16833056017748196,
+      "grad_norm": 0.20941099524497986,
+      "learning_rate": 2.4689051507357218e-05,
+      "loss": 0.6001,
+      "step": 607
+    },
+    {
+      "epoch": 0.16860787576261785,
+      "grad_norm": 0.20067985355854034,
+      "learning_rate": 2.468782639389893e-05,
+      "loss": 0.572,
+      "step": 608
+    },
+    {
+      "epoch": 0.16888519134775373,
+      "grad_norm": 0.20099548995494843,
+      "learning_rate": 2.4686598902272793e-05,
+      "loss": 0.5603,
+      "step": 609
+    },
+    {
+      "epoch": 0.16916250693288962,
+      "grad_norm": 0.20298384130001068,
+      "learning_rate": 2.4685369032718343e-05,
+      "loss": 0.5657,
+      "step": 610
+    },
+    {
+      "epoch": 0.1694398225180255,
+      "grad_norm": 0.19231431186199188,
+      "learning_rate": 2.4684136785475544e-05,
+      "loss": 0.5628,
+      "step": 611
+    },
+    {
+      "epoch": 0.16971713810316139,
+      "grad_norm": 0.20296776294708252,
+      "learning_rate": 2.468290216078485e-05,
+      "loss": 0.6011,
+      "step": 612
+    },
+    {
+      "epoch": 0.16999445368829727,
+      "grad_norm": 0.19989420473575592,
+      "learning_rate": 2.468166515888716e-05,
+      "loss": 0.5876,
+      "step": 613
+    },
+    {
+      "epoch": 0.17027176927343315,
+      "grad_norm": 0.19636170566082,
+      "learning_rate": 2.4680425780023852e-05,
+      "loss": 0.5852,
+      "step": 614
+    },
+    {
+      "epoch": 0.17054908485856904,
+      "grad_norm": 0.21776345372200012,
+      "learning_rate": 2.4679184024436757e-05,
+      "loss": 0.5988,
+      "step": 615
+    },
+    {
+      "epoch": 0.17082640044370492,
+      "grad_norm": 0.2017216682434082,
+      "learning_rate": 2.4677939892368183e-05,
+      "loss": 0.6135,
+      "step": 616
+    },
+    {
+      "epoch": 0.1711037160288408,
+      "grad_norm": 0.218113973736763,
+      "learning_rate": 2.4676693384060884e-05,
+      "loss": 0.5727,
+      "step": 617
+    },
+    {
+      "epoch": 0.1713810316139767,
+      "grad_norm": 0.2070799022912979,
+      "learning_rate": 2.4675444499758093e-05,
+      "loss": 0.6229,
+      "step": 618
+    },
+    {
+      "epoch": 0.17165834719911258,
+      "grad_norm": 0.211971253156662,
+      "learning_rate": 2.4674193239703496e-05,
+      "loss": 0.5909,
+      "step": 619
+    },
+    {
+      "epoch": 0.1719356627842485,
+      "grad_norm": 0.21103401482105255,
+      "learning_rate": 2.4672939604141248e-05,
+      "loss": 0.5805,
+      "step": 620
+    },
+    {
+      "epoch": 0.17221297836938437,
+      "grad_norm": 0.22362381219863892,
+      "learning_rate": 2.467168359331597e-05,
+      "loss": 0.6006,
+      "step": 621
+    },
+    {
+      "epoch": 0.17249029395452026,
+      "grad_norm": 0.1989423632621765,
+      "learning_rate": 2.4670425207472737e-05,
+      "loss": 0.5895,
+      "step": 622
+    },
+    {
+      "epoch": 0.17276760953965614,
+      "grad_norm": 0.20032650232315063,
+      "learning_rate": 2.46691644468571e-05,
+      "loss": 0.5856,
+      "step": 623
+    },
+    {
+      "epoch": 0.17304492512479203,
+      "grad_norm": 0.21420548856258392,
+      "learning_rate": 2.466790131171506e-05,
+      "loss": 0.5835,
+      "step": 624
+    },
+    {
+      "epoch": 0.1733222407099279,
+      "grad_norm": 0.2181633710861206,
+      "learning_rate": 2.466663580229309e-05,
+      "loss": 0.5608,
+      "step": 625
+    },
+    {
+      "epoch": 0.1735995562950638,
+      "grad_norm": 0.19224753975868225,
+      "learning_rate": 2.4665367918838135e-05,
+      "loss": 0.5826,
+      "step": 626
+    },
+    {
+      "epoch": 0.17387687188019968,
+      "grad_norm": 0.20331954956054688,
+      "learning_rate": 2.4664097661597576e-05,
+      "loss": 0.5948,
+      "step": 627
+    },
+    {
+      "epoch": 0.17415418746533556,
+      "grad_norm": 0.20249582827091217,
+      "learning_rate": 2.4662825030819282e-05,
+      "loss": 0.5894,
+      "step": 628
+    },
+    {
+      "epoch": 0.17443150305047145,
+      "grad_norm": 0.19642974436283112,
+      "learning_rate": 2.466155002675158e-05,
+      "loss": 0.5938,
+      "step": 629
+    },
+    {
+      "epoch": 0.17470881863560733,
+      "grad_norm": 0.20837126672267914,
+      "learning_rate": 2.466027264964325e-05,
+      "loss": 0.589,
+      "step": 630
+    },
+    {
+      "epoch": 0.17498613422074322,
+      "grad_norm": 0.1986762434244156,
+      "learning_rate": 2.465899289974355e-05,
+      "loss": 0.5677,
+      "step": 631
+    },
+    {
+      "epoch": 0.1752634498058791,
+      "grad_norm": 0.47144341468811035,
+      "learning_rate": 2.4657710777302183e-05,
+      "loss": 0.6075,
+      "step": 632
+    },
+    {
+      "epoch": 0.17554076539101499,
+      "grad_norm": 0.2005637288093567,
+      "learning_rate": 2.465642628256934e-05,
+      "loss": 0.5863,
+      "step": 633
+    },
+    {
+      "epoch": 0.17581808097615087,
+      "grad_norm": 0.2160159796476364,
+      "learning_rate": 2.465513941579564e-05,
+      "loss": 0.5661,
+      "step": 634
+    },
+    {
+      "epoch": 0.17609539656128675,
+      "grad_norm": 0.22309495508670807,
+      "learning_rate": 2.4653850177232203e-05,
+      "loss": 0.6029,
+      "step": 635
+    },
+    {
+      "epoch": 0.17637271214642264,
+      "grad_norm": 0.20880426466464996,
+      "learning_rate": 2.4652558567130585e-05,
+      "loss": 0.6039,
+      "step": 636
+    },
+    {
+      "epoch": 0.17665002773155852,
+      "grad_norm": 0.20371706783771515,
+      "learning_rate": 2.4651264585742813e-05,
+      "loss": 0.5974,
+      "step": 637
+    },
+    {
+      "epoch": 0.1769273433166944,
+      "grad_norm": 0.19604718685150146,
+      "learning_rate": 2.464996823332138e-05,
+      "loss": 0.6012,
+      "step": 638
+    },
+    {
+      "epoch": 0.1772046589018303,
+      "grad_norm": 0.2000109851360321,
+      "learning_rate": 2.4648669510119235e-05,
+      "loss": 0.6038,
+      "step": 639
+    },
+    {
+      "epoch": 0.17748197448696618,
+      "grad_norm": 0.19588269293308258,
+      "learning_rate": 2.46473684163898e-05,
+      "loss": 0.5949,
+      "step": 640
+    },
+    {
+      "epoch": 0.17775929007210206,
+      "grad_norm": 0.2040427029132843,
+      "learning_rate": 2.4646064952386945e-05,
+      "loss": 0.5616,
+      "step": 641
+    },
+    {
+      "epoch": 0.17803660565723795,
+      "grad_norm": 0.2059299200773239,
+      "learning_rate": 2.4644759118365014e-05,
+      "loss": 0.5785,
+      "step": 642
+    },
+    {
+      "epoch": 0.17831392124237383,
+      "grad_norm": 0.1959904134273529,
+      "learning_rate": 2.464345091457881e-05,
+      "loss": 0.5691,
+      "step": 643
+    },
+    {
+      "epoch": 0.17859123682750971,
+      "grad_norm": 0.20045484602451324,
+      "learning_rate": 2.46421403412836e-05,
+      "loss": 0.5831,
+      "step": 644
+    },
+    {
+      "epoch": 0.1788685524126456,
+      "grad_norm": 0.18542559444904327,
+      "learning_rate": 2.4640827398735105e-05,
+      "loss": 0.5666,
+      "step": 645
+    },
+    {
+      "epoch": 0.17914586799778148,
+      "grad_norm": 0.29157590866088867,
+      "learning_rate": 2.463951208718952e-05,
+      "loss": 0.5841,
+      "step": 646
+    },
+    {
+      "epoch": 0.17942318358291737,
+      "grad_norm": 0.20582380890846252,
+      "learning_rate": 2.46381944069035e-05,
+      "loss": 0.618,
+      "step": 647
+    },
+    {
+      "epoch": 0.17970049916805325,
+      "grad_norm": 0.20922648906707764,
+      "learning_rate": 2.4636874358134153e-05,
+      "loss": 0.5831,
+      "step": 648
+    },
+    {
+      "epoch": 0.17997781475318914,
+      "grad_norm": 0.19867978990077972,
+      "learning_rate": 2.463555194113906e-05,
+      "loss": 0.5734,
+      "step": 649
+    },
+    {
+      "epoch": 0.18025513033832502,
+      "grad_norm": 0.19810400903224945,
+      "learning_rate": 2.463422715617626e-05,
+      "loss": 0.6086,
+      "step": 650
+    },
+    {
+      "epoch": 0.1805324459234609,
+      "grad_norm": 0.22697949409484863,
+      "learning_rate": 2.4632900003504246e-05,
+      "loss": 0.5942,
+      "step": 651
+    },
+    {
+      "epoch": 0.1808097615085968,
+      "grad_norm": 0.21418651938438416,
+      "learning_rate": 2.4631570483381992e-05,
+      "loss": 0.5793,
+      "step": 652
+    },
+    {
+      "epoch": 0.18108707709373267,
+      "grad_norm": 0.22276844084262848,
+      "learning_rate": 2.4630238596068914e-05,
+      "loss": 0.5998,
+      "step": 653
+    },
+    {
+      "epoch": 0.18136439267886856,
+      "grad_norm": 0.237818643450737,
+      "learning_rate": 2.4628904341824898e-05,
+      "loss": 0.5787,
+      "step": 654
+    },
+    {
+      "epoch": 0.18164170826400444,
+      "grad_norm": 0.20757392048835754,
+      "learning_rate": 2.46275677209103e-05,
+      "loss": 0.5603,
+      "step": 655
+    },
+    {
+      "epoch": 0.18191902384914033,
+      "grad_norm": 0.20871873199939728,
+      "learning_rate": 2.4626228733585926e-05,
+      "loss": 0.5689,
+      "step": 656
+    },
+    {
+      "epoch": 0.1821963394342762,
+      "grad_norm": 0.2344467043876648,
+      "learning_rate": 2.4624887380113048e-05,
+      "loss": 0.5887,
+      "step": 657
+    },
+    {
+      "epoch": 0.1824736550194121,
+      "grad_norm": 0.20889438688755035,
+      "learning_rate": 2.4623543660753397e-05,
+      "loss": 0.5699,
+      "step": 658
+    },
+    {
+      "epoch": 0.18275097060454798,
+      "grad_norm": 0.23723891377449036,
+      "learning_rate": 2.4622197575769173e-05,
+      "loss": 0.5691,
+      "step": 659
+    },
+    {
+      "epoch": 0.18302828618968386,
+      "grad_norm": 0.20685730874538422,
+      "learning_rate": 2.462084912542303e-05,
+      "loss": 0.585,
+      "step": 660
+    },
+    {
+      "epoch": 0.18330560177481975,
+      "grad_norm": 0.18916456401348114,
+      "learning_rate": 2.4619498309978085e-05,
+      "loss": 0.5785,
+      "step": 661
+    },
+    {
+      "epoch": 0.18358291735995563,
+      "grad_norm": 0.19421158730983734,
+      "learning_rate": 2.4618145129697916e-05,
+      "loss": 0.5742,
+      "step": 662
+    },
+    {
+      "epoch": 0.18386023294509152,
+      "grad_norm": 0.19799606502056122,
+      "learning_rate": 2.4616789584846575e-05,
+      "loss": 0.5642,
+      "step": 663
+    },
+    {
+      "epoch": 0.1841375485302274,
+      "grad_norm": 0.20754088461399078,
+      "learning_rate": 2.4615431675688556e-05,
+      "loss": 0.5793,
+      "step": 664
+    },
+    {
+      "epoch": 0.1844148641153633,
+      "grad_norm": 0.20479615032672882,
+      "learning_rate": 2.4614071402488822e-05,
+      "loss": 0.6009,
+      "step": 665
+    },
+    {
+      "epoch": 0.18469217970049917,
+      "grad_norm": 0.20695596933364868,
+      "learning_rate": 2.4612708765512803e-05,
+      "loss": 0.607,
+      "step": 666
+    },
+    {
+      "epoch": 0.18496949528563505,
+      "grad_norm": 0.21166419982910156,
+      "learning_rate": 2.4611343765026385e-05,
+      "loss": 0.5889,
+      "step": 667
+    },
+    {
+      "epoch": 0.18524681087077094,
+      "grad_norm": 0.19035880267620087,
+      "learning_rate": 2.4609976401295914e-05,
+      "loss": 0.5596,
+      "step": 668
+    },
+    {
+      "epoch": 0.18552412645590682,
+      "grad_norm": 0.2189519852399826,
+      "learning_rate": 2.4608606674588196e-05,
+      "loss": 0.595,
+      "step": 669
+    },
+    {
+      "epoch": 0.1858014420410427,
+      "grad_norm": 0.20535226166248322,
+      "learning_rate": 2.4607234585170506e-05,
+      "loss": 0.5785,
+      "step": 670
+    },
+    {
+      "epoch": 0.1860787576261786,
+      "grad_norm": 0.20723526179790497,
+      "learning_rate": 2.4605860133310577e-05,
+      "loss": 0.6205,
+      "step": 671
+    },
+    {
+      "epoch": 0.18635607321131448,
+      "grad_norm": 0.22765034437179565,
+      "learning_rate": 2.4604483319276596e-05,
+      "loss": 0.5739,
+      "step": 672
+    },
+    {
+      "epoch": 0.18663338879645036,
+      "grad_norm": 0.19783975183963776,
+      "learning_rate": 2.4603104143337212e-05,
+      "loss": 0.6001,
+      "step": 673
+    },
+    {
+      "epoch": 0.18691070438158625,
+      "grad_norm": 0.21098697185516357,
+      "learning_rate": 2.4601722605761547e-05,
+      "loss": 0.5636,
+      "step": 674
+    },
+    {
+      "epoch": 0.18718801996672213,
+      "grad_norm": 0.20571714639663696,
+      "learning_rate": 2.4600338706819175e-05,
+      "loss": 0.6031,
+      "step": 675
+    },
+    {
+      "epoch": 0.18746533555185801,
+      "grad_norm": 0.2101418673992157,
+      "learning_rate": 2.4598952446780127e-05,
+      "loss": 0.5854,
+      "step": 676
+    },
+    {
+      "epoch": 0.1877426511369939,
+      "grad_norm": 0.20562447607517242,
+      "learning_rate": 2.45975638259149e-05,
+      "loss": 0.6007,
+      "step": 677
+    },
+    {
+      "epoch": 0.18801996672212978,
+      "grad_norm": 0.20215946435928345,
+      "learning_rate": 2.4596172844494454e-05,
+      "loss": 0.601,
+      "step": 678
+    },
+    {
+      "epoch": 0.18829728230726567,
+      "grad_norm": 0.2149016410112381,
+      "learning_rate": 2.45947795027902e-05,
+      "loss": 0.5965,
+      "step": 679
+    },
+    {
+      "epoch": 0.18857459789240155,
+      "grad_norm": 0.1985412836074829,
+      "learning_rate": 2.4593383801074025e-05,
+      "loss": 0.563,
+      "step": 680
+    },
+    {
+      "epoch": 0.18885191347753744,
+      "grad_norm": 0.20387370884418488,
+      "learning_rate": 2.459198573961826e-05,
+      "loss": 0.5816,
+      "step": 681
+    },
+    {
+      "epoch": 0.18912922906267332,
+      "grad_norm": 0.19687046110630035,
+      "learning_rate": 2.4590585318695703e-05,
+      "loss": 0.5761,
+      "step": 682
+    },
+    {
+      "epoch": 0.1894065446478092,
+      "grad_norm": 0.2053958773612976,
+      "learning_rate": 2.458918253857962e-05,
+      "loss": 0.5931,
+      "step": 683
+    },
+    {
+      "epoch": 0.1896838602329451,
+      "grad_norm": 0.20455330610275269,
+      "learning_rate": 2.4587777399543726e-05,
+      "loss": 0.5739,
+      "step": 684
+    },
+    {
+      "epoch": 0.18996117581808097,
+      "grad_norm": 0.2084437757730484,
+      "learning_rate": 2.4586369901862204e-05,
+      "loss": 0.5659,
+      "step": 685
+    },
+    {
+      "epoch": 0.19023849140321686,
+      "grad_norm": 0.20842646062374115,
+      "learning_rate": 2.4584960045809686e-05,
+      "loss": 0.5863,
+      "step": 686
+    },
+    {
+      "epoch": 0.19051580698835274,
+      "grad_norm": 0.19156675040721893,
+      "learning_rate": 2.4583547831661283e-05,
+      "loss": 0.5738,
+      "step": 687
+    },
+    {
+      "epoch": 0.19079312257348863,
+      "grad_norm": 0.19893507659435272,
+      "learning_rate": 2.4582133259692546e-05,
+      "loss": 0.5739,
+      "step": 688
+    },
+    {
+      "epoch": 0.1910704381586245,
+      "grad_norm": 0.20070527493953705,
+      "learning_rate": 2.4580716330179505e-05,
+      "loss": 0.5703,
+      "step": 689
+    },
+    {
+      "epoch": 0.1913477537437604,
+      "grad_norm": 0.20454630255699158,
+      "learning_rate": 2.4579297043398636e-05,
+      "loss": 0.5735,
+      "step": 690
+    },
+    {
+      "epoch": 0.19162506932889628,
+      "grad_norm": 0.21206796169281006,
+      "learning_rate": 2.4577875399626877e-05,
+      "loss": 0.5852,
+      "step": 691
+    },
+    {
+      "epoch": 0.19190238491403216,
+      "grad_norm": 0.20151346921920776,
+      "learning_rate": 2.4576451399141627e-05,
+      "loss": 0.6033,
+      "step": 692
+    },
+    {
+      "epoch": 0.19217970049916805,
+      "grad_norm": 0.1981053203344345,
+      "learning_rate": 2.457502504222075e-05,
+      "loss": 0.5874,
+      "step": 693
+    },
+    {
+      "epoch": 0.19245701608430393,
+      "grad_norm": 0.197993203997612,
+      "learning_rate": 2.457359632914257e-05,
+      "loss": 0.5918,
+      "step": 694
+    },
+    {
+      "epoch": 0.19273433166943982,
+      "grad_norm": 0.2022867351770401,
+      "learning_rate": 2.4572165260185857e-05,
+      "loss": 0.5786,
+      "step": 695
+    },
+    {
+      "epoch": 0.1930116472545757,
+      "grad_norm": 0.20423941314220428,
+      "learning_rate": 2.457073183562986e-05,
+      "loss": 0.5781,
+      "step": 696
+    },
+    {
+      "epoch": 0.1932889628397116,
+      "grad_norm": 0.2390977442264557,
+      "learning_rate": 2.4569296055754275e-05,
+      "loss": 0.5727,
+      "step": 697
+    },
+    {
+      "epoch": 0.19356627842484747,
+      "grad_norm": 0.2169281542301178,
+      "learning_rate": 2.4567857920839256e-05,
+      "loss": 0.5638,
+      "step": 698
+    },
+    {
+      "epoch": 0.19384359400998336,
+      "grad_norm": 0.18818534910678864,
+      "learning_rate": 2.4566417431165427e-05,
+      "loss": 0.5722,
+      "step": 699
+    },
+    {
+      "epoch": 0.19412090959511924,
+      "grad_norm": 0.1905328780412674,
+      "learning_rate": 2.456497458701386e-05,
+      "loss": 0.5632,
+      "step": 700
+    },
+    {
+      "epoch": 0.19439822518025512,
+      "grad_norm": 0.19498996436595917,
+      "learning_rate": 2.45635293886661e-05,
+      "loss": 0.5953,
+      "step": 701
+    },
+    {
+      "epoch": 0.194675540765391,
+      "grad_norm": 0.19959990680217743,
+      "learning_rate": 2.456208183640414e-05,
+      "loss": 0.583,
+      "step": 702
+    },
+    {
+      "epoch": 0.1949528563505269,
+      "grad_norm": 0.2034797966480255,
+      "learning_rate": 2.456063193051043e-05,
+      "loss": 0.5883,
+      "step": 703
+    },
+    {
+      "epoch": 0.19523017193566278,
+      "grad_norm": 0.20221780240535736,
+      "learning_rate": 2.455917967126789e-05,
+      "loss": 0.5928,
+      "step": 704
+    },
+    {
+      "epoch": 0.19550748752079866,
+      "grad_norm": 0.21273115277290344,
+      "learning_rate": 2.4557725058959895e-05,
+      "loss": 0.5719,
+      "step": 705
+    },
+    {
+      "epoch": 0.19578480310593455,
+      "grad_norm": 0.1873910129070282,
+      "learning_rate": 2.455626809387028e-05,
+      "loss": 0.5642,
+      "step": 706
+    },
+    {
+      "epoch": 0.19606211869107043,
+      "grad_norm": 0.19268092513084412,
+      "learning_rate": 2.4554808776283334e-05,
+      "loss": 0.5555,
+      "step": 707
+    },
+    {
+      "epoch": 0.19633943427620631,
+      "grad_norm": 0.20540937781333923,
+      "learning_rate": 2.4553347106483808e-05,
+      "loss": 0.6076,
+      "step": 708
+    },
+    {
+      "epoch": 0.1966167498613422,
+      "grad_norm": 0.19650278985500336,
+      "learning_rate": 2.4551883084756917e-05,
+      "loss": 0.5866,
+      "step": 709
+    },
+    {
+      "epoch": 0.19689406544647808,
+      "grad_norm": 0.221206396818161,
+      "learning_rate": 2.4550416711388327e-05,
+      "loss": 0.581,
+      "step": 710
+    },
+    {
+      "epoch": 0.19717138103161397,
+      "grad_norm": 0.19788506627082825,
+      "learning_rate": 2.4548947986664167e-05,
+      "loss": 0.5667,
+      "step": 711
+    },
+    {
+      "epoch": 0.19744869661674985,
+      "grad_norm": 0.22713138163089752,
+      "learning_rate": 2.454747691087102e-05,
+      "loss": 0.5732,
+      "step": 712
+    },
+    {
+      "epoch": 0.19772601220188574,
+      "grad_norm": 0.19035859405994415,
+      "learning_rate": 2.454600348429594e-05,
+      "loss": 0.6072,
+      "step": 713
+    },
+    {
+      "epoch": 0.19800332778702162,
+      "grad_norm": 0.19724468886852264,
+      "learning_rate": 2.4544527707226428e-05,
+      "loss": 0.5958,
+      "step": 714
+    },
+    {
+      "epoch": 0.1982806433721575,
+      "grad_norm": 0.20074382424354553,
+      "learning_rate": 2.4543049579950445e-05,
+      "loss": 0.6006,
+      "step": 715
+    },
+    {
+      "epoch": 0.1985579589572934,
+      "grad_norm": 0.19603832066059113,
+      "learning_rate": 2.4541569102756414e-05,
+      "loss": 0.5901,
+      "step": 716
+    },
+    {
+      "epoch": 0.19883527454242927,
+      "grad_norm": 0.1956927478313446,
+      "learning_rate": 2.4540086275933215e-05,
+      "loss": 0.5731,
+      "step": 717
+    },
+    {
+      "epoch": 0.19911259012756516,
+      "grad_norm": 0.19697944819927216,
+      "learning_rate": 2.4538601099770187e-05,
+      "loss": 0.5778,
+      "step": 718
+    },
+    {
+      "epoch": 0.19938990571270104,
+      "grad_norm": 0.19818982481956482,
+      "learning_rate": 2.453711357455713e-05,
+      "loss": 0.5588,
+      "step": 719
+    },
+    {
+      "epoch": 0.19966722129783693,
+      "grad_norm": 0.18997104465961456,
+      "learning_rate": 2.4535623700584297e-05,
+      "loss": 0.5789,
+      "step": 720
+    },
+    {
+      "epoch": 0.1999445368829728,
+      "grad_norm": 0.21167294681072235,
+      "learning_rate": 2.4534131478142402e-05,
+      "loss": 0.5804,
+      "step": 721
+    },
+    {
+      "epoch": 0.2002218524681087,
+      "grad_norm": 0.21374750137329102,
+      "learning_rate": 2.4532636907522617e-05,
+      "loss": 0.5727,
+      "step": 722
+    },
+    {
+      "epoch": 0.20049916805324458,
+      "grad_norm": 0.20970353484153748,
+      "learning_rate": 2.453113998901657e-05,
+      "loss": 0.5766,
+      "step": 723
+    },
+    {
+      "epoch": 0.20077648363838047,
+      "grad_norm": 0.19419021904468536,
+      "learning_rate": 2.4529640722916355e-05,
+      "loss": 0.5755,
+      "step": 724
+    },
+    {
+      "epoch": 0.20105379922351635,
+      "grad_norm": 0.20143002271652222,
+      "learning_rate": 2.4528139109514513e-05,
+      "loss": 0.5627,
+      "step": 725
+    },
+    {
+      "epoch": 0.20133111480865223,
+      "grad_norm": 0.20216137170791626,
+      "learning_rate": 2.4526635149104056e-05,
+      "loss": 0.5771,
+      "step": 726
+    },
+    {
+      "epoch": 0.20160843039378812,
+      "grad_norm": 0.20338623225688934,
+      "learning_rate": 2.452512884197844e-05,
+      "loss": 0.605,
+      "step": 727
+    },
+    {
+      "epoch": 0.201885745978924,
+      "grad_norm": 0.20078277587890625,
+      "learning_rate": 2.4523620188431585e-05,
+      "loss": 0.5954,
+      "step": 728
+    },
+    {
+      "epoch": 0.2021630615640599,
+      "grad_norm": 0.2046244889497757,
+      "learning_rate": 2.4522109188757875e-05,
+      "loss": 0.5945,
+      "step": 729
+    },
+    {
+      "epoch": 0.20244037714919577,
+      "grad_norm": 0.20462092757225037,
+      "learning_rate": 2.4520595843252138e-05,
+      "loss": 0.5762,
+      "step": 730
+    },
+    {
+      "epoch": 0.20271769273433166,
+      "grad_norm": 0.20749370753765106,
+      "learning_rate": 2.4519080152209675e-05,
+      "loss": 0.5435,
+      "step": 731
+    },
+    {
+      "epoch": 0.20299500831946754,
+      "grad_norm": 0.21734094619750977,
+      "learning_rate": 2.4517562115926233e-05,
+      "loss": 0.5961,
+      "step": 732
+    },
+    {
+      "epoch": 0.20327232390460345,
+      "grad_norm": 0.20754100382328033,
+      "learning_rate": 2.4516041734698024e-05,
+      "loss": 0.5548,
+      "step": 733
+    },
+    {
+      "epoch": 0.20354963948973934,
+      "grad_norm": 0.20619900524616241,
+      "learning_rate": 2.451451900882172e-05,
+      "loss": 0.5897,
+      "step": 734
+    },
+    {
+      "epoch": 0.20382695507487522,
+      "grad_norm": 0.28535279631614685,
+      "learning_rate": 2.451299393859443e-05,
+      "loss": 0.5815,
+      "step": 735
+    },
+    {
+      "epoch": 0.2041042706600111,
+      "grad_norm": 0.19404634833335876,
+      "learning_rate": 2.4511466524313748e-05,
+      "loss": 0.5743,
+      "step": 736
+    },
+    {
+      "epoch": 0.204381586245147,
+      "grad_norm": 0.18895412981510162,
+      "learning_rate": 2.4509936766277706e-05,
+      "loss": 0.5876,
+      "step": 737
+    },
+    {
+      "epoch": 0.20465890183028287,
+      "grad_norm": 0.22120583057403564,
+      "learning_rate": 2.4508404664784808e-05,
+      "loss": 0.5873,
+      "step": 738
+    },
+    {
+      "epoch": 0.20493621741541876,
+      "grad_norm": 0.21361954510211945,
+      "learning_rate": 2.4506870220134e-05,
+      "loss": 0.6002,
+      "step": 739
+    },
+    {
+      "epoch": 0.20521353300055464,
+      "grad_norm": 0.19417433440685272,
+      "learning_rate": 2.4505333432624694e-05,
+      "loss": 0.5673,
+      "step": 740
+    },
+    {
+      "epoch": 0.20549084858569053,
+      "grad_norm": 0.19543422758579254,
+      "learning_rate": 2.4503794302556765e-05,
+      "loss": 0.5628,
+      "step": 741
+    },
+    {
+      "epoch": 0.2057681641708264,
+      "grad_norm": 0.1954490691423416,
+      "learning_rate": 2.450225283023053e-05,
+      "loss": 0.6246,
+      "step": 742
+    },
+    {
+      "epoch": 0.2060454797559623,
+      "grad_norm": 0.1868993192911148,
+      "learning_rate": 2.4500709015946776e-05,
+      "loss": 0.5858,
+      "step": 743
+    },
+    {
+      "epoch": 0.20632279534109818,
+      "grad_norm": 0.2035941481590271,
+      "learning_rate": 2.449916286000674e-05,
+      "loss": 0.5974,
+      "step": 744
+    },
+    {
+      "epoch": 0.20660011092623407,
+      "grad_norm": 0.1918855458498001,
+      "learning_rate": 2.4497614362712118e-05,
+      "loss": 0.5843,
+      "step": 745
+    },
+    {
+      "epoch": 0.20687742651136995,
+      "grad_norm": 0.19244706630706787,
+      "learning_rate": 2.4496063524365063e-05,
+      "loss": 0.5535,
+      "step": 746
+    },
+    {
+      "epoch": 0.20715474209650583,
+      "grad_norm": 0.20424753427505493,
+      "learning_rate": 2.4494510345268185e-05,
+      "loss": 0.5835,
+      "step": 747
+    },
+    {
+      "epoch": 0.20743205768164172,
+      "grad_norm": 0.19604821503162384,
+      "learning_rate": 2.4492954825724544e-05,
+      "loss": 0.5701,
+      "step": 748
+    },
+    {
+      "epoch": 0.2077093732667776,
+      "grad_norm": 0.19546863436698914,
+      "learning_rate": 2.4491396966037678e-05,
+      "loss": 0.5898,
+      "step": 749
+    },
+    {
+      "epoch": 0.2079866888519135,
+      "grad_norm": 0.19861635565757751,
+      "learning_rate": 2.4489836766511555e-05,
+      "loss": 0.587,
+      "step": 750
+    },
+    {
+      "epoch": 0.20826400443704937,
+      "grad_norm": 0.19369752705097198,
+      "learning_rate": 2.4488274227450613e-05,
+      "loss": 0.6027,
+      "step": 751
+    },
+    {
+      "epoch": 0.20854132002218526,
+      "grad_norm": 0.4158318340778351,
+      "learning_rate": 2.448670934915975e-05,
+      "loss": 0.602,
+      "step": 752
+    },
+    {
+      "epoch": 0.20881863560732114,
+      "grad_norm": 0.20547251403331757,
+      "learning_rate": 2.4485142131944306e-05,
+      "loss": 0.5949,
+      "step": 753
+    },
+    {
+      "epoch": 0.20909595119245702,
+      "grad_norm": 0.21317002177238464,
+      "learning_rate": 2.4483572576110093e-05,
+      "loss": 0.5862,
+      "step": 754
+    },
+    {
+      "epoch": 0.2093732667775929,
+      "grad_norm": 0.19712896645069122,
+      "learning_rate": 2.448200068196337e-05,
+      "loss": 0.5983,
+      "step": 755
+    },
+    {
+      "epoch": 0.2096505823627288,
+      "grad_norm": 0.2054811269044876,
+      "learning_rate": 2.448042644981086e-05,
+      "loss": 0.5983,
+      "step": 756
+    },
+    {
+      "epoch": 0.20992789794786468,
+      "grad_norm": 0.19795221090316772,
+      "learning_rate": 2.447884987995973e-05,
+      "loss": 0.6208,
+      "step": 757
+    },
+    {
+      "epoch": 0.21020521353300056,
+      "grad_norm": 0.21395504474639893,
+      "learning_rate": 2.447727097271762e-05,
+      "loss": 0.5983,
+      "step": 758
+    },
+    {
+      "epoch": 0.21048252911813645,
+      "grad_norm": 0.19311439990997314,
+      "learning_rate": 2.447568972839261e-05,
+      "loss": 0.592,
+      "step": 759
+    },
+    {
+      "epoch": 0.21075984470327233,
+      "grad_norm": 0.21382609009742737,
+      "learning_rate": 2.4474106147293242e-05,
+      "loss": 0.5752,
+      "step": 760
+    },
+    {
+      "epoch": 0.21103716028840822,
+      "grad_norm": 0.19354097545146942,
+      "learning_rate": 2.447252022972852e-05,
+      "loss": 0.5911,
+      "step": 761
+    },
+    {
+      "epoch": 0.2113144758735441,
+      "grad_norm": 0.19984754920005798,
+      "learning_rate": 2.4470931976007894e-05,
+      "loss": 0.5995,
+      "step": 762
+    },
+    {
+      "epoch": 0.21159179145867998,
+      "grad_norm": 0.20404407382011414,
+      "learning_rate": 2.4469341386441274e-05,
+      "loss": 0.5551,
+      "step": 763
+    },
+    {
+      "epoch": 0.21186910704381587,
+      "grad_norm": 0.2025006264448166,
+      "learning_rate": 2.446774846133903e-05,
+      "loss": 0.6105,
+      "step": 764
+    },
+    {
+      "epoch": 0.21214642262895175,
+      "grad_norm": 0.20010975003242493,
+      "learning_rate": 2.446615320101198e-05,
+      "loss": 0.5788,
+      "step": 765
+    },
+    {
+      "epoch": 0.21242373821408764,
+      "grad_norm": 0.20225434005260468,
+      "learning_rate": 2.4464555605771404e-05,
+      "loss": 0.5636,
+      "step": 766
+    },
+    {
+      "epoch": 0.21270105379922352,
+      "grad_norm": 0.19845524430274963,
+      "learning_rate": 2.4462955675929032e-05,
+      "loss": 0.5758,
+      "step": 767
+    },
+    {
+      "epoch": 0.2129783693843594,
+      "grad_norm": 0.19598202407360077,
+      "learning_rate": 2.446135341179706e-05,
+      "loss": 0.5456,
+      "step": 768
+    },
+    {
+      "epoch": 0.2132556849694953,
+      "grad_norm": 0.2050497829914093,
+      "learning_rate": 2.445974881368812e-05,
+      "loss": 0.5912,
+      "step": 769
+    },
+    {
+      "epoch": 0.21353300055463117,
+      "grad_norm": 0.19924525916576385,
+      "learning_rate": 2.4458141881915324e-05,
+      "loss": 0.5479,
+      "step": 770
+    },
+    {
+      "epoch": 0.21381031613976706,
+      "grad_norm": 0.20329277217388153,
+      "learning_rate": 2.445653261679222e-05,
+      "loss": 0.6006,
+      "step": 771
+    },
+    {
+      "epoch": 0.21408763172490294,
+      "grad_norm": 0.19327110052108765,
+      "learning_rate": 2.4454921018632827e-05,
+      "loss": 0.5739,
+      "step": 772
+    },
+    {
+      "epoch": 0.21436494731003883,
+      "grad_norm": 0.19316452741622925,
+      "learning_rate": 2.4453307087751594e-05,
+      "loss": 0.5953,
+      "step": 773
+    },
+    {
+      "epoch": 0.2146422628951747,
+      "grad_norm": 0.19617030024528503,
+      "learning_rate": 2.4451690824463457e-05,
+      "loss": 0.5686,
+      "step": 774
+    },
+    {
+      "epoch": 0.2149195784803106,
+      "grad_norm": 0.20528316497802734,
+      "learning_rate": 2.4450072229083786e-05,
+      "loss": 0.5691,
+      "step": 775
+    },
+    {
+      "epoch": 0.21519689406544648,
+      "grad_norm": 0.20544420182704926,
+      "learning_rate": 2.4448451301928408e-05,
+      "loss": 0.5776,
+      "step": 776
+    },
+    {
+      "epoch": 0.21547420965058237,
+      "grad_norm": 0.21979959309101105,
+      "learning_rate": 2.4446828043313614e-05,
+      "loss": 0.5947,
+      "step": 777
+    },
+    {
+      "epoch": 0.21575152523571825,
+      "grad_norm": 0.2081802487373352,
+      "learning_rate": 2.4445202453556145e-05,
+      "loss": 0.5752,
+      "step": 778
+    },
+    {
+      "epoch": 0.21602884082085413,
+      "grad_norm": 0.2012367993593216,
+      "learning_rate": 2.4443574532973195e-05,
+      "loss": 0.5671,
+      "step": 779
+    },
+    {
+      "epoch": 0.21630615640599002,
+      "grad_norm": 0.1998508721590042,
+      "learning_rate": 2.4441944281882415e-05,
+      "loss": 0.6154,
+      "step": 780
+    },
+    {
+      "epoch": 0.2165834719911259,
+      "grad_norm": 0.20325055718421936,
+      "learning_rate": 2.444031170060191e-05,
+      "loss": 0.5743,
+      "step": 781
+    },
+    {
+      "epoch": 0.2168607875762618,
+      "grad_norm": 0.20255804061889648,
+      "learning_rate": 2.443867678945024e-05,
+      "loss": 0.5748,
+      "step": 782
+    },
+    {
+      "epoch": 0.21713810316139767,
+      "grad_norm": 0.1919908970594406,
+      "learning_rate": 2.4437039548746415e-05,
+      "loss": 0.5709,
+      "step": 783
+    },
+    {
+      "epoch": 0.21741541874653356,
+      "grad_norm": 0.20014292001724243,
+      "learning_rate": 2.443539997880991e-05,
+      "loss": 0.5636,
+      "step": 784
+    },
+    {
+      "epoch": 0.21769273433166944,
+      "grad_norm": 0.19818776845932007,
+      "learning_rate": 2.4433758079960647e-05,
+      "loss": 0.5649,
+      "step": 785
+    },
+    {
+      "epoch": 0.21797004991680533,
+      "grad_norm": 0.18718703091144562,
+      "learning_rate": 2.4432113852519005e-05,
+      "loss": 0.5917,
+      "step": 786
+    },
+    {
+      "epoch": 0.2182473655019412,
+      "grad_norm": 0.32280299067497253,
+      "learning_rate": 2.4430467296805816e-05,
+      "loss": 0.5864,
+      "step": 787
+    },
+    {
+      "epoch": 0.2185246810870771,
+      "grad_norm": 0.20851466059684753,
+      "learning_rate": 2.442881841314236e-05,
+      "loss": 0.5837,
+      "step": 788
+    },
+    {
+      "epoch": 0.21880199667221298,
+      "grad_norm": 0.1917923539876938,
+      "learning_rate": 2.442716720185039e-05,
+      "loss": 0.6032,
+      "step": 789
+    },
+    {
+      "epoch": 0.21907931225734886,
+      "grad_norm": 0.25185203552246094,
+      "learning_rate": 2.442551366325209e-05,
+      "loss": 0.5873,
+      "step": 790
+    },
+    {
+      "epoch": 0.21935662784248475,
+      "grad_norm": 0.1962638646364212,
+      "learning_rate": 2.4423857797670118e-05,
+      "loss": 0.577,
+      "step": 791
+    },
+    {
+      "epoch": 0.21963394342762063,
+      "grad_norm": 0.19475746154785156,
+      "learning_rate": 2.4422199605427572e-05,
+      "loss": 0.5677,
+      "step": 792
+    },
+    {
+      "epoch": 0.21991125901275652,
+      "grad_norm": 0.517663836479187,
+      "learning_rate": 2.4420539086848007e-05,
+      "loss": 0.5718,
+      "step": 793
+    },
+    {
+      "epoch": 0.2201885745978924,
+      "grad_norm": 0.1952415555715561,
+      "learning_rate": 2.441887624225544e-05,
+      "loss": 0.5801,
+      "step": 794
+    },
+    {
+      "epoch": 0.22046589018302828,
+      "grad_norm": 0.20376408100128174,
+      "learning_rate": 2.441721107197433e-05,
+      "loss": 0.6038,
+      "step": 795
+    },
+    {
+      "epoch": 0.22074320576816417,
+      "grad_norm": 0.21492497622966766,
+      "learning_rate": 2.4415543576329604e-05,
+      "loss": 0.5626,
+      "step": 796
+    },
+    {
+      "epoch": 0.22102052135330005,
+      "grad_norm": 0.21810825169086456,
+      "learning_rate": 2.4413873755646627e-05,
+      "loss": 0.566,
+      "step": 797
+    },
+    {
+      "epoch": 0.22129783693843594,
+      "grad_norm": 0.2002691626548767,
+      "learning_rate": 2.4412201610251232e-05,
+      "loss": 0.5706,
+      "step": 798
+    },
+    {
+      "epoch": 0.22157515252357182,
+      "grad_norm": 0.24929803609848022,
+      "learning_rate": 2.441052714046969e-05,
+      "loss": 0.5878,
+      "step": 799
+    },
+    {
+      "epoch": 0.2218524681087077,
+      "grad_norm": 0.20125854015350342,
+      "learning_rate": 2.440885034662874e-05,
+      "loss": 0.5869,
+      "step": 800
+    },
+    {
+      "epoch": 0.2221297836938436,
+      "grad_norm": 0.19206437468528748,
+      "learning_rate": 2.4407171229055574e-05,
+      "loss": 0.5911,
+      "step": 801
+    },
+    {
+      "epoch": 0.22240709927897948,
+      "grad_norm": 0.22323836386203766,
+      "learning_rate": 2.4405489788077823e-05,
+      "loss": 0.5725,
+      "step": 802
+    },
+    {
+      "epoch": 0.22268441486411536,
+      "grad_norm": 0.2044333517551422,
+      "learning_rate": 2.4403806024023584e-05,
+      "loss": 0.6243,
+      "step": 803
+    },
+    {
+      "epoch": 0.22296173044925124,
+      "grad_norm": 0.20750725269317627,
+      "learning_rate": 2.44021199372214e-05,
+      "loss": 0.6002,
+      "step": 804
+    },
+    {
+      "epoch": 0.22323904603438713,
+      "grad_norm": 0.2218470573425293,
+      "learning_rate": 2.4400431528000284e-05,
+      "loss": 0.5886,
+      "step": 805
+    },
+    {
+      "epoch": 0.223516361619523,
+      "grad_norm": 0.20135878026485443,
+      "learning_rate": 2.4398740796689676e-05,
+      "loss": 0.5711,
+      "step": 806
+    },
+    {
+      "epoch": 0.2237936772046589,
+      "grad_norm": 0.19776999950408936,
+      "learning_rate": 2.439704774361949e-05,
+      "loss": 0.5706,
+      "step": 807
+    },
+    {
+      "epoch": 0.22407099278979478,
+      "grad_norm": 0.22162839770317078,
+      "learning_rate": 2.4395352369120078e-05,
+      "loss": 0.593,
+      "step": 808
+    },
+    {
+      "epoch": 0.22434830837493067,
+      "grad_norm": 0.1974382847547531,
+      "learning_rate": 2.4393654673522264e-05,
+      "loss": 0.5657,
+      "step": 809
+    },
+    {
+      "epoch": 0.22462562396006655,
+      "grad_norm": 0.2362552434206009,
+      "learning_rate": 2.4391954657157302e-05,
+      "loss": 0.5844,
+      "step": 810
+    },
+    {
+      "epoch": 0.22490293954520243,
+      "grad_norm": 0.2056739628314972,
+      "learning_rate": 2.4390252320356915e-05,
+      "loss": 0.5959,
+      "step": 811
+    },
+    {
+      "epoch": 0.22518025513033832,
+      "grad_norm": 0.19037006795406342,
+      "learning_rate": 2.4388547663453275e-05,
+      "loss": 0.5638,
+      "step": 812
+    },
+    {
+      "epoch": 0.2254575707154742,
+      "grad_norm": 0.21129223704338074,
+      "learning_rate": 2.4386840686779004e-05,
+      "loss": 0.5954,
+      "step": 813
+    },
+    {
+      "epoch": 0.2257348863006101,
+      "grad_norm": 0.2718643546104431,
+      "learning_rate": 2.4385131390667184e-05,
+      "loss": 0.5881,
+      "step": 814
+    },
+    {
+      "epoch": 0.22601220188574597,
+      "grad_norm": 0.2181466519832611,
+      "learning_rate": 2.4383419775451334e-05,
+      "loss": 0.5858,
+      "step": 815
+    },
+    {
+      "epoch": 0.22628951747088186,
+      "grad_norm": 0.22737659513950348,
+      "learning_rate": 2.438170584146544e-05,
+      "loss": 0.5732,
+      "step": 816
+    },
+    {
+      "epoch": 0.22656683305601774,
+      "grad_norm": 0.20686711370944977,
+      "learning_rate": 2.437998958904394e-05,
+      "loss": 0.5524,
+      "step": 817
+    },
+    {
+      "epoch": 0.22684414864115363,
+      "grad_norm": 0.2140977680683136,
+      "learning_rate": 2.4378271018521714e-05,
+      "loss": 0.5628,
+      "step": 818
+    },
+    {
+      "epoch": 0.2271214642262895,
+      "grad_norm": 0.19793452322483063,
+      "learning_rate": 2.4376550130234104e-05,
+      "loss": 0.5796,
+      "step": 819
+    },
+    {
+      "epoch": 0.2273987798114254,
+      "grad_norm": 0.20289914309978485,
+      "learning_rate": 2.4374826924516903e-05,
+      "loss": 0.5848,
+      "step": 820
+    },
+    {
+      "epoch": 0.22767609539656128,
+      "grad_norm": 0.19081373512744904,
+      "learning_rate": 2.437310140170635e-05,
+      "loss": 0.6011,
+      "step": 821
+    },
+    {
+      "epoch": 0.22795341098169716,
+      "grad_norm": 0.20546457171440125,
+      "learning_rate": 2.437137356213914e-05,
+      "loss": 0.5476,
+      "step": 822
+    },
+    {
+      "epoch": 0.22823072656683305,
+      "grad_norm": 0.22531366348266602,
+      "learning_rate": 2.4369643406152422e-05,
+      "loss": 0.577,
+      "step": 823
+    },
+    {
+      "epoch": 0.22850804215196893,
+      "grad_norm": 0.1964918076992035,
+      "learning_rate": 2.4367910934083795e-05,
+      "loss": 0.5733,
+      "step": 824
+    },
+    {
+      "epoch": 0.22878535773710482,
+      "grad_norm": 0.1976742297410965,
+      "learning_rate": 2.4366176146271313e-05,
+      "loss": 0.576,
+      "step": 825
+    },
+    {
+      "epoch": 0.2290626733222407,
+      "grad_norm": 0.20134706795215607,
+      "learning_rate": 2.4364439043053475e-05,
+      "loss": 0.5715,
+      "step": 826
+    },
+    {
+      "epoch": 0.22933998890737659,
+      "grad_norm": 0.28538307547569275,
+      "learning_rate": 2.4362699624769236e-05,
+      "loss": 0.6009,
+      "step": 827
+    },
+    {
+      "epoch": 0.22961730449251247,
+      "grad_norm": 0.21845568716526031,
+      "learning_rate": 2.4360957891758006e-05,
+      "loss": 0.5936,
+      "step": 828
+    },
+    {
+      "epoch": 0.22989462007764835,
+      "grad_norm": 0.1977756768465042,
+      "learning_rate": 2.435921384435964e-05,
+      "loss": 0.5651,
+      "step": 829
+    },
+    {
+      "epoch": 0.23017193566278424,
+      "grad_norm": 0.20483648777008057,
+      "learning_rate": 2.4357467482914447e-05,
+      "loss": 0.5861,
+      "step": 830
+    },
+    {
+      "epoch": 0.23044925124792012,
+      "grad_norm": 0.191145658493042,
+      "learning_rate": 2.4355718807763196e-05,
+      "loss": 0.5951,
+      "step": 831
+    },
+    {
+      "epoch": 0.230726566833056,
+      "grad_norm": 0.20611602067947388,
+      "learning_rate": 2.4353967819247093e-05,
+      "loss": 0.5762,
+      "step": 832
+    },
+    {
+      "epoch": 0.2310038824181919,
+      "grad_norm": 0.34163740277290344,
+      "learning_rate": 2.43522145177078e-05,
+      "loss": 0.5573,
+      "step": 833
+    },
+    {
+      "epoch": 0.23128119800332778,
+      "grad_norm": 0.1989511400461197,
+      "learning_rate": 2.4350458903487438e-05,
+      "loss": 0.5799,
+      "step": 834
+    },
+    {
+      "epoch": 0.23155851358846366,
+      "grad_norm": 0.1951713114976883,
+      "learning_rate": 2.434870097692857e-05,
+      "loss": 0.5763,
+      "step": 835
+    },
+    {
+      "epoch": 0.23183582917359954,
+      "grad_norm": 0.19492702186107635,
+      "learning_rate": 2.4346940738374217e-05,
+      "loss": 0.5751,
+      "step": 836
+    },
+    {
+      "epoch": 0.23211314475873543,
+      "grad_norm": 0.20524460077285767,
+      "learning_rate": 2.434517818816785e-05,
+      "loss": 0.5959,
+      "step": 837
+    },
+    {
+      "epoch": 0.2323904603438713,
+      "grad_norm": 0.19833968579769135,
+      "learning_rate": 2.4343413326653384e-05,
+      "loss": 0.5881,
+      "step": 838
+    },
+    {
+      "epoch": 0.2326677759290072,
+      "grad_norm": 0.19568949937820435,
+      "learning_rate": 2.4341646154175192e-05,
+      "loss": 0.5449,
+      "step": 839
+    },
+    {
+      "epoch": 0.23294509151414308,
+      "grad_norm": 0.19156986474990845,
+      "learning_rate": 2.43398766710781e-05,
+      "loss": 0.5703,
+      "step": 840
+    },
+    {
+      "epoch": 0.23322240709927897,
+      "grad_norm": 0.20440654456615448,
+      "learning_rate": 2.4338104877707372e-05,
+      "loss": 0.5783,
+      "step": 841
+    },
+    {
+      "epoch": 0.23349972268441485,
+      "grad_norm": 0.20396758615970612,
+      "learning_rate": 2.4336330774408744e-05,
+      "loss": 0.5911,
+      "step": 842
+    },
+    {
+      "epoch": 0.23377703826955074,
+      "grad_norm": 0.19637946784496307,
+      "learning_rate": 2.4334554361528376e-05,
+      "loss": 0.5708,
+      "step": 843
+    },
+    {
+      "epoch": 0.23405435385468662,
+      "grad_norm": 0.21237438917160034,
+      "learning_rate": 2.433277563941291e-05,
+      "loss": 0.6104,
+      "step": 844
+    },
+    {
+      "epoch": 0.2343316694398225,
+      "grad_norm": 0.18440033495426178,
+      "learning_rate": 2.433099460840941e-05,
+      "loss": 0.5745,
+      "step": 845
+    },
+    {
+      "epoch": 0.23460898502495842,
+      "grad_norm": 0.19301645457744598,
+      "learning_rate": 2.4329211268865406e-05,
+      "loss": 0.5621,
+      "step": 846
+    },
+    {
+      "epoch": 0.2348863006100943,
+      "grad_norm": 0.2056163102388382,
+      "learning_rate": 2.4327425621128873e-05,
+      "loss": 0.5973,
+      "step": 847
+    },
+    {
+      "epoch": 0.23516361619523019,
+      "grad_norm": 0.20398737490177155,
+      "learning_rate": 2.432563766554824e-05,
+      "loss": 0.5795,
+      "step": 848
+    },
+    {
+      "epoch": 0.23544093178036607,
+      "grad_norm": 0.19015717506408691,
+      "learning_rate": 2.432384740247239e-05,
+      "loss": 0.5563,
+      "step": 849
+    },
+    {
+      "epoch": 0.23571824736550195,
+      "grad_norm": 0.19576147198677063,
+      "learning_rate": 2.4322054832250636e-05,
+      "loss": 0.5757,
+      "step": 850
+    },
+    {
+      "epoch": 0.23599556295063784,
+      "grad_norm": 0.1978127360343933,
+      "learning_rate": 2.4320259955232773e-05,
+      "loss": 0.5832,
+      "step": 851
+    },
+    {
+      "epoch": 0.23627287853577372,
+      "grad_norm": 0.19191214442253113,
+      "learning_rate": 2.4318462771769012e-05,
+      "loss": 0.5812,
+      "step": 852
+    },
+    {
+      "epoch": 0.2365501941209096,
+      "grad_norm": 0.18717870116233826,
+      "learning_rate": 2.4316663282210046e-05,
+      "loss": 0.5753,
+      "step": 853
+    },
+    {
+      "epoch": 0.2368275097060455,
+      "grad_norm": 0.18739578127861023,
+      "learning_rate": 2.4314861486906996e-05,
+      "loss": 0.5665,
+      "step": 854
+    },
+    {
+      "epoch": 0.23710482529118138,
+      "grad_norm": 0.1928299367427826,
+      "learning_rate": 2.431305738621144e-05,
+      "loss": 0.5836,
+      "step": 855
+    },
+    {
+      "epoch": 0.23738214087631726,
+      "grad_norm": 0.20048747956752777,
+      "learning_rate": 2.4311250980475408e-05,
+      "loss": 0.5673,
+      "step": 856
+    },
+    {
+      "epoch": 0.23765945646145314,
+      "grad_norm": 0.2141515463590622,
+      "learning_rate": 2.4309442270051376e-05,
+      "loss": 0.5426,
+      "step": 857
+    },
+    {
+      "epoch": 0.23793677204658903,
+      "grad_norm": 0.19637706875801086,
+      "learning_rate": 2.4307631255292273e-05,
+      "loss": 0.5542,
+      "step": 858
+    },
+    {
+      "epoch": 0.2382140876317249,
+      "grad_norm": 0.21245527267456055,
+      "learning_rate": 2.4305817936551472e-05,
+      "loss": 0.5867,
+      "step": 859
+    },
+    {
+      "epoch": 0.2384914032168608,
+      "grad_norm": 0.19474704563617706,
+      "learning_rate": 2.4304002314182804e-05,
+      "loss": 0.5716,
+      "step": 860
+    },
+    {
+      "epoch": 0.23876871880199668,
+      "grad_norm": 0.19590826332569122,
+      "learning_rate": 2.4302184388540544e-05,
+      "loss": 0.5746,
+      "step": 861
+    },
+    {
+      "epoch": 0.23904603438713257,
+      "grad_norm": 0.21298062801361084,
+      "learning_rate": 2.4300364159979418e-05,
+      "loss": 0.5788,
+      "step": 862
+    },
+    {
+      "epoch": 0.23932334997226845,
+      "grad_norm": 0.21597878634929657,
+      "learning_rate": 2.4298541628854597e-05,
+      "loss": 0.5868,
+      "step": 863
+    },
+    {
+      "epoch": 0.23960066555740434,
+      "grad_norm": 0.2077784389257431,
+      "learning_rate": 2.4296716795521707e-05,
+      "loss": 0.5878,
+      "step": 864
+    },
+    {
+      "epoch": 0.23987798114254022,
+      "grad_norm": 0.1982557773590088,
+      "learning_rate": 2.4294889660336823e-05,
+      "loss": 0.5734,
+      "step": 865
+    },
+    {
+      "epoch": 0.2401552967276761,
+      "grad_norm": 0.19443267583847046,
+      "learning_rate": 2.4293060223656465e-05,
+      "loss": 0.5845,
+      "step": 866
+    },
+    {
+      "epoch": 0.240432612312812,
+      "grad_norm": 0.2007235586643219,
+      "learning_rate": 2.4291228485837613e-05,
+      "loss": 0.5829,
+      "step": 867
+    },
+    {
+      "epoch": 0.24070992789794787,
+      "grad_norm": 0.18996436893939972,
+      "learning_rate": 2.4289394447237674e-05,
+      "loss": 0.5918,
+      "step": 868
+    },
+    {
+      "epoch": 0.24098724348308376,
+      "grad_norm": 0.19309580326080322,
+      "learning_rate": 2.4287558108214527e-05,
+      "loss": 0.5962,
+      "step": 869
+    },
+    {
+      "epoch": 0.24126455906821964,
+      "grad_norm": 0.19195586442947388,
+      "learning_rate": 2.428571946912649e-05,
+      "loss": 0.5738,
+      "step": 870
+    },
+    {
+      "epoch": 0.24154187465335553,
+      "grad_norm": 0.19671426713466644,
+      "learning_rate": 2.4283878530332322e-05,
+      "loss": 0.5551,
+      "step": 871
+    },
+    {
+      "epoch": 0.2418191902384914,
+      "grad_norm": 0.19544430077075958,
+      "learning_rate": 2.4282035292191247e-05,
+      "loss": 0.5876,
+      "step": 872
+    },
+    {
+      "epoch": 0.2420965058236273,
+      "grad_norm": 0.19258378446102142,
+      "learning_rate": 2.4280189755062928e-05,
+      "loss": 0.559,
+      "step": 873
+    },
+    {
+      "epoch": 0.24237382140876318,
+      "grad_norm": 0.21361422538757324,
+      "learning_rate": 2.427834191930748e-05,
+      "loss": 0.5753,
+      "step": 874
+    },
+    {
+      "epoch": 0.24265113699389906,
+      "grad_norm": 0.20279040932655334,
+      "learning_rate": 2.4276491785285457e-05,
+      "loss": 0.5724,
+      "step": 875
+    },
+    {
+      "epoch": 0.24292845257903495,
+      "grad_norm": 0.20467157661914825,
+      "learning_rate": 2.427463935335788e-05,
+      "loss": 0.5989,
+      "step": 876
+    },
+    {
+      "epoch": 0.24320576816417083,
+      "grad_norm": 0.20144453644752502,
+      "learning_rate": 2.4272784623886195e-05,
+      "loss": 0.5943,
+      "step": 877
+    },
+    {
+      "epoch": 0.24348308374930672,
+      "grad_norm": 0.19686299562454224,
+      "learning_rate": 2.4270927597232325e-05,
+      "loss": 0.5692,
+      "step": 878
+    },
+    {
+      "epoch": 0.2437603993344426,
+      "grad_norm": 0.22722774744033813,
+      "learning_rate": 2.426906827375861e-05,
+      "loss": 0.566,
+      "step": 879
+    },
+    {
+      "epoch": 0.24403771491957849,
+      "grad_norm": 0.19703295826911926,
+      "learning_rate": 2.4267206653827856e-05,
+      "loss": 0.5627,
+      "step": 880
+    },
+    {
+      "epoch": 0.24431503050471437,
+      "grad_norm": 0.2020971179008484,
+      "learning_rate": 2.4265342737803327e-05,
+      "loss": 0.5836,
+      "step": 881
+    },
+    {
+      "epoch": 0.24459234608985025,
+      "grad_norm": 0.1921062171459198,
+      "learning_rate": 2.4263476526048707e-05,
+      "loss": 0.5651,
+      "step": 882
+    },
+    {
+      "epoch": 0.24486966167498614,
+      "grad_norm": 0.20176348090171814,
+      "learning_rate": 2.4261608018928147e-05,
+      "loss": 0.5883,
+      "step": 883
+    },
+    {
+      "epoch": 0.24514697726012202,
+      "grad_norm": 0.19450893998146057,
+      "learning_rate": 2.425973721680625e-05,
+      "loss": 0.5591,
+      "step": 884
+    },
+    {
+      "epoch": 0.2454242928452579,
+      "grad_norm": 0.8580565452575684,
+      "learning_rate": 2.425786412004805e-05,
+      "loss": 0.5592,
+      "step": 885
+    },
+    {
+      "epoch": 0.2457016084303938,
+      "grad_norm": 0.24260735511779785,
+      "learning_rate": 2.4255988729019042e-05,
+      "loss": 0.5902,
+      "step": 886
+    },
+    {
+      "epoch": 0.24597892401552968,
+      "grad_norm": 0.19789321720600128,
+      "learning_rate": 2.4254111044085163e-05,
+      "loss": 0.5745,
+      "step": 887
+    },
+    {
+      "epoch": 0.24625623960066556,
+      "grad_norm": 0.21979939937591553,
+      "learning_rate": 2.4252231065612805e-05,
+      "loss": 0.5551,
+      "step": 888
+    },
+    {
+      "epoch": 0.24653355518580145,
+      "grad_norm": 0.20352937281131744,
+      "learning_rate": 2.425034879396879e-05,
+      "loss": 0.5591,
+      "step": 889
+    },
+    {
+      "epoch": 0.24681087077093733,
+      "grad_norm": 0.2068743109703064,
+      "learning_rate": 2.424846422952041e-05,
+      "loss": 0.5848,
+      "step": 890
+    },
+    {
+      "epoch": 0.24708818635607321,
+      "grad_norm": 0.19195155799388885,
+      "learning_rate": 2.4246577372635387e-05,
+      "loss": 0.5621,
+      "step": 891
+    },
+    {
+      "epoch": 0.2473655019412091,
+      "grad_norm": 0.2014138102531433,
+      "learning_rate": 2.42446882236819e-05,
+      "loss": 0.5689,
+      "step": 892
+    },
+    {
+      "epoch": 0.24764281752634498,
+      "grad_norm": 0.2104417085647583,
+      "learning_rate": 2.4242796783028573e-05,
+      "loss": 0.5898,
+      "step": 893
+    },
+    {
+      "epoch": 0.24792013311148087,
+      "grad_norm": 0.22085507214069366,
+      "learning_rate": 2.4240903051044474e-05,
+      "loss": 0.5762,
+      "step": 894
+    },
+    {
+      "epoch": 0.24819744869661675,
+      "grad_norm": 0.20655465126037598,
+      "learning_rate": 2.4239007028099117e-05,
+      "loss": 0.5654,
+      "step": 895
+    },
+    {
+      "epoch": 0.24847476428175264,
+      "grad_norm": 0.2050492912530899,
+      "learning_rate": 2.4237108714562474e-05,
+      "loss": 0.5859,
+      "step": 896
+    },
+    {
+      "epoch": 0.24875207986688852,
+      "grad_norm": 0.21355165541172028,
+      "learning_rate": 2.4235208110804947e-05,
+      "loss": 0.5695,
+      "step": 897
+    },
+    {
+      "epoch": 0.2490293954520244,
+      "grad_norm": 0.20924112200737,
+      "learning_rate": 2.42333052171974e-05,
+      "loss": 0.5593,
+      "step": 898
+    },
+    {
+      "epoch": 0.2493067110371603,
+      "grad_norm": 0.22572918236255646,
+      "learning_rate": 2.423140003411114e-05,
+      "loss": 0.6217,
+      "step": 899
+    },
+    {
+      "epoch": 0.24958402662229617,
+      "grad_norm": 0.2063211351633072,
+      "learning_rate": 2.4229492561917914e-05,
+      "loss": 0.5765,
+      "step": 900
+    },
+    {
+      "epoch": 0.24986134220743206,
+      "grad_norm": 0.21796129643917084,
+      "learning_rate": 2.4227582800989923e-05,
+      "loss": 0.5932,
+      "step": 901
+    },
+    {
+      "epoch": 0.25013865779256794,
+      "grad_norm": 0.20169825851917267,
+      "learning_rate": 2.4225670751699808e-05,
+      "loss": 0.5858,
+      "step": 902
+    },
+    {
+      "epoch": 0.2504159733777038,
+      "grad_norm": 0.2097984254360199,
+      "learning_rate": 2.4223756414420668e-05,
+      "loss": 0.5888,
+      "step": 903
+    },
+    {
+      "epoch": 0.2506932889628397,
+      "grad_norm": 0.2516496479511261,
+      "learning_rate": 2.4221839789526033e-05,
+      "loss": 0.5812,
+      "step": 904
+    },
+    {
+      "epoch": 0.2509706045479756,
+      "grad_norm": 0.2110574096441269,
+      "learning_rate": 2.421992087738989e-05,
+      "loss": 0.5859,
+      "step": 905
+    },
+    {
+      "epoch": 0.2512479201331115,
+      "grad_norm": 0.1935090869665146,
+      "learning_rate": 2.4217999678386673e-05,
+      "loss": 0.5768,
+      "step": 906
+    },
+    {
+      "epoch": 0.25152523571824736,
+      "grad_norm": 0.20760700106620789,
+      "learning_rate": 2.4216076192891257e-05,
+      "loss": 0.5856,
+      "step": 907
+    },
+    {
+      "epoch": 0.25180255130338325,
+      "grad_norm": 0.20197226107120514,
+      "learning_rate": 2.4214150421278964e-05,
+      "loss": 0.6041,
+      "step": 908
+    },
+    {
+      "epoch": 0.25207986688851913,
+      "grad_norm": 0.20894859731197357,
+      "learning_rate": 2.4212222363925563e-05,
+      "loss": 0.5821,
+      "step": 909
+    },
+    {
+      "epoch": 0.252357182473655,
+      "grad_norm": 0.20673821866512299,
+      "learning_rate": 2.4210292021207268e-05,
+      "loss": 0.5813,
+      "step": 910
+    },
+    {
+      "epoch": 0.2526344980587909,
+      "grad_norm": 0.23159608244895935,
+      "learning_rate": 2.420835939350074e-05,
+      "loss": 0.5803,
+      "step": 911
+    },
+    {
+      "epoch": 0.2529118136439268,
+      "grad_norm": 0.19740775227546692,
+      "learning_rate": 2.420642448118309e-05,
+      "loss": 0.5721,
+      "step": 912
+    },
+    {
+      "epoch": 0.25318912922906267,
+      "grad_norm": 0.20442472398281097,
+      "learning_rate": 2.4204487284631866e-05,
+      "loss": 0.5681,
+      "step": 913
+    },
+    {
+      "epoch": 0.25346644481419855,
+      "grad_norm": 0.19745062291622162,
+      "learning_rate": 2.420254780422507e-05,
+      "loss": 0.573,
+      "step": 914
+    },
+    {
+      "epoch": 0.25374376039933444,
+      "grad_norm": 0.19819265604019165,
+      "learning_rate": 2.4200606040341147e-05,
+      "loss": 0.6015,
+      "step": 915
+    },
+    {
+      "epoch": 0.2540210759844703,
+      "grad_norm": 0.20158332586288452,
+      "learning_rate": 2.4198661993358976e-05,
+      "loss": 0.5557,
+      "step": 916
+    },
+    {
+      "epoch": 0.2542983915696062,
+      "grad_norm": 0.19300688803195953,
+      "learning_rate": 2.4196715663657903e-05,
+      "loss": 0.5691,
+      "step": 917
+    },
+    {
+      "epoch": 0.2545757071547421,
+      "grad_norm": 0.19178220629692078,
+      "learning_rate": 2.4194767051617707e-05,
+      "loss": 0.5877,
+      "step": 918
+    },
+    {
+      "epoch": 0.254853022739878,
+      "grad_norm": 0.19551022350788116,
+      "learning_rate": 2.4192816157618615e-05,
+      "loss": 0.5719,
+      "step": 919
+    },
+    {
+      "epoch": 0.25513033832501386,
+      "grad_norm": 0.2008076012134552,
+      "learning_rate": 2.419086298204129e-05,
+      "loss": 0.579,
+      "step": 920
+    },
+    {
+      "epoch": 0.25540765391014975,
+      "grad_norm": 0.19526442885398865,
+      "learning_rate": 2.4188907525266856e-05,
+      "loss": 0.546,
+      "step": 921
+    },
+    {
+      "epoch": 0.25568496949528563,
+      "grad_norm": 0.1844739466905594,
+      "learning_rate": 2.418694978767687e-05,
+      "loss": 0.5759,
+      "step": 922
+    },
+    {
+      "epoch": 0.2559622850804215,
+      "grad_norm": 0.20165039598941803,
+      "learning_rate": 2.4184989769653343e-05,
+      "loss": 0.5618,
+      "step": 923
+    },
+    {
+      "epoch": 0.2562396006655574,
+      "grad_norm": 0.19225285947322845,
+      "learning_rate": 2.418302747157872e-05,
+      "loss": 0.5627,
+      "step": 924
+    },
+    {
+      "epoch": 0.2565169162506933,
+      "grad_norm": 0.19688788056373596,
+      "learning_rate": 2.418106289383591e-05,
+      "loss": 0.5714,
+      "step": 925
+    },
+    {
+      "epoch": 0.25679423183582917,
+      "grad_norm": 0.2039179801940918,
+      "learning_rate": 2.417909603680824e-05,
+      "loss": 0.5975,
+      "step": 926
+    },
+    {
+      "epoch": 0.25707154742096505,
+      "grad_norm": 0.21120522916316986,
+      "learning_rate": 2.41771269008795e-05,
+      "loss": 0.6026,
+      "step": 927
+    },
+    {
+      "epoch": 0.25734886300610094,
+      "grad_norm": 0.21704575419425964,
+      "learning_rate": 2.4175155486433927e-05,
+      "loss": 0.5798,
+      "step": 928
+    },
+    {
+      "epoch": 0.2576261785912368,
+      "grad_norm": 0.18682295083999634,
+      "learning_rate": 2.4173181793856187e-05,
+      "loss": 0.5872,
+      "step": 929
+    },
+    {
+      "epoch": 0.2579034941763727,
+      "grad_norm": 0.200609490275383,
+      "learning_rate": 2.4171205823531402e-05,
+      "loss": 0.5568,
+      "step": 930
+    },
+    {
+      "epoch": 0.2581808097615086,
+      "grad_norm": 0.1996905356645584,
+      "learning_rate": 2.416922757584514e-05,
+      "loss": 0.5965,
+      "step": 931
+    },
+    {
+      "epoch": 0.2584581253466445,
+      "grad_norm": 0.1988278329372406,
+      "learning_rate": 2.4167247051183412e-05,
+      "loss": 0.5819,
+      "step": 932
+    },
+    {
+      "epoch": 0.25873544093178036,
+      "grad_norm": 0.20553693175315857,
+      "learning_rate": 2.4165264249932662e-05,
+      "loss": 0.591,
+      "step": 933
+    },
+    {
+      "epoch": 0.25901275651691624,
+      "grad_norm": 0.19097572565078735,
+      "learning_rate": 2.416327917247979e-05,
+      "loss": 0.5517,
+      "step": 934
+    },
+    {
+      "epoch": 0.2592900721020521,
+      "grad_norm": 0.191832035779953,
+      "learning_rate": 2.4161291819212144e-05,
+      "loss": 0.5811,
+      "step": 935
+    },
+    {
+      "epoch": 0.259567387687188,
+      "grad_norm": 0.19829009473323822,
+      "learning_rate": 2.4159302190517496e-05,
+      "loss": 0.5966,
+      "step": 936
+    },
+    {
+      "epoch": 0.2598447032723239,
+      "grad_norm": 0.1965586543083191,
+      "learning_rate": 2.415731028678409e-05,
+      "loss": 0.5458,
+      "step": 937
+    },
+    {
+      "epoch": 0.2601220188574598,
+      "grad_norm": 0.20748619735240936,
+      "learning_rate": 2.4155316108400593e-05,
+      "loss": 0.5843,
+      "step": 938
+    },
+    {
+      "epoch": 0.26039933444259566,
+      "grad_norm": 0.20656223595142365,
+      "learning_rate": 2.415331965575612e-05,
+      "loss": 0.5749,
+      "step": 939
+    },
+    {
+      "epoch": 0.26067665002773155,
+      "grad_norm": 0.186055988073349,
+      "learning_rate": 2.4151320929240227e-05,
+      "loss": 0.5749,
+      "step": 940
+    },
+    {
+      "epoch": 0.26095396561286743,
+      "grad_norm": 0.21665525436401367,
+      "learning_rate": 2.4149319929242934e-05,
+      "loss": 0.5825,
+      "step": 941
+    },
+    {
+      "epoch": 0.2612312811980033,
+      "grad_norm": 0.2007439136505127,
+      "learning_rate": 2.4147316656154674e-05,
+      "loss": 0.5904,
+      "step": 942
+    },
+    {
+      "epoch": 0.2615085967831392,
+      "grad_norm": 0.20290826261043549,
+      "learning_rate": 2.4145311110366347e-05,
+      "loss": 0.6019,
+      "step": 943
+    },
+    {
+      "epoch": 0.2617859123682751,
+      "grad_norm": 0.20062971115112305,
+      "learning_rate": 2.4143303292269286e-05,
+      "loss": 0.5571,
+      "step": 944
+    },
+    {
+      "epoch": 0.26206322795341097,
+      "grad_norm": 0.2056947946548462,
+      "learning_rate": 2.414129320225527e-05,
+      "loss": 0.5683,
+      "step": 945
+    },
+    {
+      "epoch": 0.26234054353854686,
+      "grad_norm": 0.18966248631477356,
+      "learning_rate": 2.4139280840716517e-05,
+      "loss": 0.5962,
+      "step": 946
+    },
+    {
+      "epoch": 0.26261785912368274,
+      "grad_norm": 0.20114421844482422,
+      "learning_rate": 2.41372662080457e-05,
+      "loss": 0.5528,
+      "step": 947
+    },
+    {
+      "epoch": 0.2628951747088186,
+      "grad_norm": 0.1974896341562271,
+      "learning_rate": 2.4135249304635914e-05,
+      "loss": 0.5802,
+      "step": 948
+    },
+    {
+      "epoch": 0.2631724902939545,
+      "grad_norm": 0.1997508406639099,
+      "learning_rate": 2.4133230130880726e-05,
+      "loss": 0.6122,
+      "step": 949
+    },
+    {
+      "epoch": 0.2634498058790904,
+      "grad_norm": 0.18641312420368195,
+      "learning_rate": 2.413120868717412e-05,
+      "loss": 0.5705,
+      "step": 950
+    },
+    {
+      "epoch": 0.2637271214642263,
+      "grad_norm": 0.19226068258285522,
+      "learning_rate": 2.4129184973910533e-05,
+      "loss": 0.5737,
+      "step": 951
+    },
+    {
+      "epoch": 0.26400443704936216,
+      "grad_norm": 0.18802867829799652,
+      "learning_rate": 2.4127158991484855e-05,
+      "loss": 0.562,
+      "step": 952
+    },
+    {
+      "epoch": 0.26428175263449805,
+      "grad_norm": 0.19342157244682312,
+      "learning_rate": 2.41251307402924e-05,
+      "loss": 0.5483,
+      "step": 953
+    },
+    {
+      "epoch": 0.26455906821963393,
+      "grad_norm": 0.1865961253643036,
+      "learning_rate": 2.4123100220728935e-05,
+      "loss": 0.5638,
+      "step": 954
+    },
+    {
+      "epoch": 0.2648363838047698,
+      "grad_norm": 0.22456099092960358,
+      "learning_rate": 2.4121067433190666e-05,
+      "loss": 0.5918,
+      "step": 955
+    },
+    {
+      "epoch": 0.2651136993899057,
+      "grad_norm": 0.24317651987075806,
+      "learning_rate": 2.4119032378074245e-05,
+      "loss": 0.5477,
+      "step": 956
+    },
+    {
+      "epoch": 0.2653910149750416,
+      "grad_norm": 0.19641970098018646,
+      "learning_rate": 2.411699505577677e-05,
+      "loss": 0.5554,
+      "step": 957
+    },
+    {
+      "epoch": 0.26566833056017747,
+      "grad_norm": 0.20111410319805145,
+      "learning_rate": 2.4114955466695773e-05,
+      "loss": 0.562,
+      "step": 958
+    },
+    {
+      "epoch": 0.26594564614531335,
+      "grad_norm": 0.19789332151412964,
+      "learning_rate": 2.411291361122923e-05,
+      "loss": 0.5919,
+      "step": 959
+    },
+    {
+      "epoch": 0.26622296173044924,
+      "grad_norm": 0.20293276011943817,
+      "learning_rate": 2.4110869489775567e-05,
+      "loss": 0.5839,
+      "step": 960
+    },
+    {
+      "epoch": 0.2665002773155851,
+      "grad_norm": 0.18560869991779327,
+      "learning_rate": 2.410882310273364e-05,
+      "loss": 0.5476,
+      "step": 961
+    },
+    {
+      "epoch": 0.266777592900721,
+      "grad_norm": 0.1932571828365326,
+      "learning_rate": 2.410677445050276e-05,
+      "loss": 0.5736,
+      "step": 962
+    },
+    {
+      "epoch": 0.2670549084858569,
+      "grad_norm": 0.21673519909381866,
+      "learning_rate": 2.4104723533482664e-05,
+      "loss": 0.5558,
+      "step": 963
+    },
+    {
+      "epoch": 0.2673322240709928,
+      "grad_norm": 0.21258150041103363,
+      "learning_rate": 2.4102670352073548e-05,
+      "loss": 0.591,
+      "step": 964
+    },
+    {
+      "epoch": 0.26760953965612866,
+      "grad_norm": 0.20553314685821533,
+      "learning_rate": 2.4100614906676036e-05,
+      "loss": 0.5901,
+      "step": 965
+    },
+    {
+      "epoch": 0.26788685524126454,
+      "grad_norm": 0.23216207325458527,
+      "learning_rate": 2.4098557197691204e-05,
+      "loss": 0.5677,
+      "step": 966
+    },
+    {
+      "epoch": 0.2681641708264004,
+      "grad_norm": 0.20041371881961823,
+      "learning_rate": 2.4096497225520564e-05,
+      "loss": 0.566,
+      "step": 967
+    },
+    {
+      "epoch": 0.2684414864115363,
+      "grad_norm": 0.20803777873516083,
+      "learning_rate": 2.4094434990566076e-05,
+      "loss": 0.5714,
+      "step": 968
+    },
+    {
+      "epoch": 0.2687188019966722,
+      "grad_norm": 0.19973017275333405,
+      "learning_rate": 2.409237049323013e-05,
+      "loss": 0.5805,
+      "step": 969
+    },
+    {
+      "epoch": 0.2689961175818081,
+      "grad_norm": 0.20199733972549438,
+      "learning_rate": 2.4090303733915567e-05,
+      "loss": 0.5475,
+      "step": 970
+    },
+    {
+      "epoch": 0.26927343316694397,
+      "grad_norm": 0.2112300544977188,
+      "learning_rate": 2.4088234713025664e-05,
+      "loss": 0.5821,
+      "step": 971
+    },
+    {
+      "epoch": 0.26955074875207985,
+      "grad_norm": 0.18687258660793304,
+      "learning_rate": 2.408616343096415e-05,
+      "loss": 0.5369,
+      "step": 972
+    },
+    {
+      "epoch": 0.26982806433721573,
+      "grad_norm": 0.20027992129325867,
+      "learning_rate": 2.4084089888135176e-05,
+      "loss": 0.551,
+      "step": 973
+    },
+    {
+      "epoch": 0.2701053799223516,
+      "grad_norm": 0.1915608048439026,
+      "learning_rate": 2.408201408494335e-05,
+      "loss": 0.5653,
+      "step": 974
+    },
+    {
+      "epoch": 0.2703826955074875,
+      "grad_norm": 0.2044133096933365,
+      "learning_rate": 2.407993602179372e-05,
+      "loss": 0.597,
+      "step": 975
+    },
+    {
+      "epoch": 0.2706600110926234,
+      "grad_norm": 0.20056426525115967,
+      "learning_rate": 2.4077855699091764e-05,
+      "loss": 0.5864,
+      "step": 976
+    },
+    {
+      "epoch": 0.27093732667775927,
+      "grad_norm": 0.19527383148670197,
+      "learning_rate": 2.407577311724341e-05,
+      "loss": 0.565,
+      "step": 977
+    },
+    {
+      "epoch": 0.27121464226289516,
+      "grad_norm": 0.2120949625968933,
+      "learning_rate": 2.407368827665503e-05,
+      "loss": 0.5621,
+      "step": 978
+    },
+    {
+      "epoch": 0.27149195784803104,
+      "grad_norm": 0.18631702661514282,
+      "learning_rate": 2.407160117773343e-05,
+      "loss": 0.5657,
+      "step": 979
+    },
+    {
+      "epoch": 0.2717692734331669,
+      "grad_norm": 0.19784550368785858,
+      "learning_rate": 2.4069511820885854e-05,
+      "loss": 0.5547,
+      "step": 980
+    },
+    {
+      "epoch": 0.2720465890183028,
+      "grad_norm": 0.19714047014713287,
+      "learning_rate": 2.4067420206519993e-05,
+      "loss": 0.5793,
+      "step": 981
+    },
+    {
+      "epoch": 0.2723239046034387,
+      "grad_norm": 0.2170424610376358,
+      "learning_rate": 2.4065326335043976e-05,
+      "loss": 0.5745,
+      "step": 982
+    },
+    {
+      "epoch": 0.2726012201885746,
+      "grad_norm": 0.20346680283546448,
+      "learning_rate": 2.4063230206866377e-05,
+      "loss": 0.5828,
+      "step": 983
+    },
+    {
+      "epoch": 0.27287853577371046,
+      "grad_norm": 0.2216998040676117,
+      "learning_rate": 2.40611318223962e-05,
+      "loss": 0.5762,
+      "step": 984
+    },
+    {
+      "epoch": 0.27315585135884635,
+      "grad_norm": 0.20975996553897858,
+      "learning_rate": 2.4059031182042897e-05,
+      "loss": 0.5442,
+      "step": 985
+    },
+    {
+      "epoch": 0.27343316694398223,
+      "grad_norm": 0.19896754622459412,
+      "learning_rate": 2.405692828621636e-05,
+      "loss": 0.5977,
+      "step": 986
+    },
+    {
+      "epoch": 0.2737104825291181,
+      "grad_norm": 0.18813903629779816,
+      "learning_rate": 2.4054823135326922e-05,
+      "loss": 0.5641,
+      "step": 987
+    },
+    {
+      "epoch": 0.273987798114254,
+      "grad_norm": 0.19095094501972198,
+      "learning_rate": 2.4052715729785348e-05,
+      "loss": 0.5427,
+      "step": 988
+    },
+    {
+      "epoch": 0.2742651136993899,
+      "grad_norm": 0.19185671210289001,
+      "learning_rate": 2.405060607000285e-05,
+      "loss": 0.557,
+      "step": 989
+    },
+    {
+      "epoch": 0.27454242928452577,
+      "grad_norm": 0.19244584441184998,
+      "learning_rate": 2.4048494156391087e-05,
+      "loss": 0.5558,
+      "step": 990
+    },
+    {
+      "epoch": 0.27481974486966165,
+      "grad_norm": 0.20083992183208466,
+      "learning_rate": 2.404637998936214e-05,
+      "loss": 0.5635,
+      "step": 991
+    },
+    {
+      "epoch": 0.27509706045479754,
+      "grad_norm": 0.19767695665359497,
+      "learning_rate": 2.404426356932854e-05,
+      "loss": 0.5814,
+      "step": 992
+    },
+    {
+      "epoch": 0.2753743760399334,
+      "grad_norm": 0.19133426249027252,
+      "learning_rate": 2.4042144896703256e-05,
+      "loss": 0.5951,
+      "step": 993
+    },
+    {
+      "epoch": 0.2756516916250693,
+      "grad_norm": 0.19364149868488312,
+      "learning_rate": 2.40400239718997e-05,
+      "loss": 0.5695,
+      "step": 994
+    },
+    {
+      "epoch": 0.2759290072102052,
+      "grad_norm": 0.19669091701507568,
+      "learning_rate": 2.4037900795331722e-05,
+      "loss": 0.5801,
+      "step": 995
+    },
+    {
+      "epoch": 0.2762063227953411,
+      "grad_norm": 0.2011607140302658,
+      "learning_rate": 2.403577536741361e-05,
+      "loss": 0.5736,
+      "step": 996
+    },
+    {
+      "epoch": 0.27648363838047696,
+      "grad_norm": 0.19536298513412476,
+      "learning_rate": 2.4033647688560084e-05,
+      "loss": 0.5404,
+      "step": 997
+    },
+    {
+      "epoch": 0.27676095396561284,
+      "grad_norm": 0.1903197318315506,
+      "learning_rate": 2.403151775918632e-05,
+      "loss": 0.5939,
+      "step": 998
+    },
+    {
+      "epoch": 0.27703826955074873,
+      "grad_norm": 0.19172310829162598,
+      "learning_rate": 2.4029385579707916e-05,
+      "loss": 0.5688,
+      "step": 999
+    },
+    {
+      "epoch": 0.2773155851358846,
+      "grad_norm": 0.22239771485328674,
+      "learning_rate": 2.402725115054092e-05,
+      "loss": 0.5754,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2775929007210205,
+      "grad_norm": 0.18384471535682678,
+      "learning_rate": 2.402511447210182e-05,
+      "loss": 0.5668,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2778702163061564,
+      "grad_norm": 0.2017565667629242,
+      "learning_rate": 2.402297554480753e-05,
+      "loss": 0.5711,
+      "step": 1002
+    },
+    {
+      "epoch": 0.27814753189129227,
+      "grad_norm": 0.1890055239200592,
+      "learning_rate": 2.402083436907542e-05,
+      "loss": 0.5552,
+      "step": 1003
+    },
+    {
+      "epoch": 0.27842484747642815,
+      "grad_norm": 0.1961050033569336,
+      "learning_rate": 2.4018690945323284e-05,
+      "loss": 0.5744,
+      "step": 1004
+    },
+    {
+      "epoch": 0.27870216306156403,
+      "grad_norm": 0.2047930508852005,
+      "learning_rate": 2.401654527396936e-05,
+      "loss": 0.5794,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2789794786466999,
+      "grad_norm": 0.19990523159503937,
+      "learning_rate": 2.4014397355432335e-05,
+      "loss": 0.6046,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2792567942318358,
+      "grad_norm": 0.19778995215892792,
+      "learning_rate": 2.401224719013131e-05,
+      "loss": 0.5621,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2795341098169717,
+      "grad_norm": 0.1909160017967224,
+      "learning_rate": 2.4010094778485846e-05,
+      "loss": 0.5943,
+      "step": 1008
+    },
+    {
+      "epoch": 0.27981142540210757,
+      "grad_norm": 0.20408110320568085,
+      "learning_rate": 2.4007940120915946e-05,
+      "loss": 0.5991,
+      "step": 1009
+    },
+    {
+      "epoch": 0.28008874098724346,
+      "grad_norm": 0.20102624595165253,
+      "learning_rate": 2.4005783217842024e-05,
+      "loss": 0.6022,
+      "step": 1010
+    },
+    {
+      "epoch": 0.28036605657237934,
+      "grad_norm": 0.1910308301448822,
+      "learning_rate": 2.4003624069684957e-05,
+      "loss": 0.5874,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2806433721575152,
+      "grad_norm": 0.189175084233284,
+      "learning_rate": 2.4001462676866054e-05,
+      "loss": 0.5698,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2809206877426511,
+      "grad_norm": 0.19003014266490936,
+      "learning_rate": 2.3999299039807055e-05,
+      "loss": 0.5819,
+      "step": 1013
+    },
+    {
+      "epoch": 0.281198003327787,
+      "grad_norm": 0.2033187299966812,
+      "learning_rate": 2.3997133158930145e-05,
+      "loss": 0.5979,
+      "step": 1014
+    },
+    {
+      "epoch": 0.28147531891292293,
+      "grad_norm": 0.1880473643541336,
+      "learning_rate": 2.3994965034657946e-05,
+      "loss": 0.5472,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2817526344980588,
+      "grad_norm": 0.19091346859931946,
+      "learning_rate": 2.3992794667413514e-05,
+      "loss": 0.5698,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2820299500831947,
+      "grad_norm": 0.19986368715763092,
+      "learning_rate": 2.399062205762035e-05,
+      "loss": 0.5956,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2823072656683306,
+      "grad_norm": 0.19207067787647247,
+      "learning_rate": 2.398844720570238e-05,
+      "loss": 0.549,
+      "step": 1018
+    },
+    {
+      "epoch": 0.28258458125346647,
+      "grad_norm": 0.19408905506134033,
+      "learning_rate": 2.398627011208398e-05,
+      "loss": 0.5938,
+      "step": 1019
+    },
+    {
+      "epoch": 0.28286189683860236,
+      "grad_norm": 0.20354047417640686,
+      "learning_rate": 2.398409077718996e-05,
+      "loss": 0.5488,
+      "step": 1020
+    },
+    {
+      "epoch": 0.28313921242373824,
+      "grad_norm": 0.19870953261852264,
+      "learning_rate": 2.3981909201445563e-05,
+      "loss": 0.565,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2834165280088741,
+      "grad_norm": 0.19812439382076263,
+      "learning_rate": 2.3979725385276475e-05,
+      "loss": 0.5455,
+      "step": 1022
+    },
+    {
+      "epoch": 0.28369384359401,
+      "grad_norm": 0.1886581927537918,
+      "learning_rate": 2.3977539329108813e-05,
+      "loss": 0.5458,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2839711591791459,
+      "grad_norm": 0.1987125426530838,
+      "learning_rate": 2.3975351033369138e-05,
+      "loss": 0.5669,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2842484747642818,
+      "grad_norm": 0.22528746724128723,
+      "learning_rate": 2.397316049848444e-05,
+      "loss": 0.579,
+      "step": 1025
+    },
+    {
+      "epoch": 0.28452579034941766,
+      "grad_norm": 0.20103305578231812,
+      "learning_rate": 2.3970967724882154e-05,
+      "loss": 0.5843,
+      "step": 1026
+    },
+    {
+      "epoch": 0.28480310593455355,
+      "grad_norm": 0.21639437973499298,
+      "learning_rate": 2.396877271299015e-05,
+      "loss": 0.5709,
+      "step": 1027
+    },
+    {
+      "epoch": 0.28508042151968943,
+      "grad_norm": 0.24263976514339447,
+      "learning_rate": 2.3966575463236725e-05,
+      "loss": 0.5926,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2853577371048253,
+      "grad_norm": 0.19815106689929962,
+      "learning_rate": 2.396437597605063e-05,
+      "loss": 0.5722,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2856350526899612,
+      "grad_norm": 0.19260184466838837,
+      "learning_rate": 2.396217425186104e-05,
+      "loss": 0.5621,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2859123682750971,
+      "grad_norm": 0.20668724179267883,
+      "learning_rate": 2.3959970291097566e-05,
+      "loss": 0.5924,
+      "step": 1031
+    },
+    {
+      "epoch": 0.28618968386023297,
+      "grad_norm": 0.2019844353199005,
+      "learning_rate": 2.3957764094190265e-05,
+      "loss": 0.5821,
+      "step": 1032
+    },
+    {
+      "epoch": 0.28646699944536885,
+      "grad_norm": 0.1919315904378891,
+      "learning_rate": 2.3955555661569617e-05,
+      "loss": 0.5985,
+      "step": 1033
+    },
+    {
+      "epoch": 0.28674431503050474,
+      "grad_norm": 0.19516989588737488,
+      "learning_rate": 2.3953344993666555e-05,
+      "loss": 0.5993,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2870216306156406,
+      "grad_norm": 0.18881164491176605,
+      "learning_rate": 2.3951132090912432e-05,
+      "loss": 0.5911,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2872989462007765,
+      "grad_norm": 0.20282314717769623,
+      "learning_rate": 2.3948916953739045e-05,
+      "loss": 0.5939,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2875762617859124,
+      "grad_norm": 0.19400392472743988,
+      "learning_rate": 2.394669958257863e-05,
+      "loss": 0.5993,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2878535773710483,
+      "grad_norm": 0.21767151355743408,
+      "learning_rate": 2.3944479977863847e-05,
+      "loss": 0.5796,
+      "step": 1038
+    },
+    {
+      "epoch": 0.28813089295618416,
+      "grad_norm": 0.21109527349472046,
+      "learning_rate": 2.3942258140027805e-05,
+      "loss": 0.5678,
+      "step": 1039
+    },
+    {
+      "epoch": 0.28840820854132004,
+      "grad_norm": 0.19805195927619934,
+      "learning_rate": 2.3940034069504048e-05,
+      "loss": 0.5645,
+      "step": 1040
+    },
+    {
+      "epoch": 0.28868552412645593,
+      "grad_norm": 0.18621553480625153,
+      "learning_rate": 2.3937807766726545e-05,
+      "loss": 0.5639,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2889628397115918,
+      "grad_norm": 0.19867920875549316,
+      "learning_rate": 2.3935579232129705e-05,
+      "loss": 0.578,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2892401552967277,
+      "grad_norm": 0.1965349018573761,
+      "learning_rate": 2.393334846614838e-05,
+      "loss": 0.5895,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2895174708818636,
+      "grad_norm": 0.21346546709537506,
+      "learning_rate": 2.3931115469217848e-05,
+      "loss": 0.5592,
+      "step": 1044
+    },
+    {
+      "epoch": 0.28979478646699947,
+      "grad_norm": 0.19675497710704803,
+      "learning_rate": 2.392888024177382e-05,
+      "loss": 0.5709,
+      "step": 1045
+    },
+    {
+      "epoch": 0.29007210205213535,
+      "grad_norm": 0.18906576931476593,
+      "learning_rate": 2.392664278425246e-05,
+      "loss": 0.5721,
+      "step": 1046
+    },
+    {
+      "epoch": 0.29034941763727123,
+      "grad_norm": 0.1987171620130539,
+      "learning_rate": 2.3924403097090348e-05,
+      "loss": 0.5796,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2906267332224071,
+      "grad_norm": 0.20017191767692566,
+      "learning_rate": 2.392216118072451e-05,
+      "loss": 0.5803,
+      "step": 1048
+    },
+    {
+      "epoch": 0.290904048807543,
+      "grad_norm": 0.1904933750629425,
+      "learning_rate": 2.39199170355924e-05,
+      "loss": 0.5847,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2911813643926789,
+      "grad_norm": 0.20761634409427643,
+      "learning_rate": 2.3917670662131914e-05,
+      "loss": 0.5678,
+      "step": 1050
+    },
+    {
+      "epoch": 0.29145867997781477,
+      "grad_norm": 0.1986733376979828,
+      "learning_rate": 2.391542206078137e-05,
+      "loss": 0.5886,
+      "step": 1051
+    },
+    {
+      "epoch": 0.29173599556295066,
+      "grad_norm": 0.2128080427646637,
+      "learning_rate": 2.3913171231979543e-05,
+      "loss": 0.5649,
+      "step": 1052
+    },
+    {
+      "epoch": 0.29201331114808654,
+      "grad_norm": 0.1908857524394989,
+      "learning_rate": 2.391091817616562e-05,
+      "loss": 0.5951,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2922906267332224,
+      "grad_norm": 0.19354763627052307,
+      "learning_rate": 2.3908662893779228e-05,
+      "loss": 0.5878,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2925679423183583,
+      "grad_norm": 0.19191789627075195,
+      "learning_rate": 2.3906405385260443e-05,
+      "loss": 0.5842,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2928452579034942,
+      "grad_norm": 0.19070293009281158,
+      "learning_rate": 2.3904145651049764e-05,
+      "loss": 0.5707,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2931225734886301,
+      "grad_norm": 0.20022979378700256,
+      "learning_rate": 2.3901883691588116e-05,
+      "loss": 0.5838,
+      "step": 1057
+    },
+    {
+      "epoch": 0.29339988907376596,
+      "grad_norm": 0.18544460833072662,
+      "learning_rate": 2.3899619507316878e-05,
+      "loss": 0.5732,
+      "step": 1058
+    },
+    {
+      "epoch": 0.29367720465890185,
+      "grad_norm": 0.19487732648849487,
+      "learning_rate": 2.3897353098677845e-05,
+      "loss": 0.5681,
+      "step": 1059
+    },
+    {
+      "epoch": 0.29395452024403773,
+      "grad_norm": 0.19073253870010376,
+      "learning_rate": 2.3895084466113253e-05,
+      "loss": 0.5638,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2942318358291736,
+      "grad_norm": 0.19357110559940338,
+      "learning_rate": 2.3892813610065778e-05,
+      "loss": 0.5617,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2945091514143095,
+      "grad_norm": 0.19819645583629608,
+      "learning_rate": 2.3890540530978518e-05,
+      "loss": 0.5601,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2947864669994454,
+      "grad_norm": 0.20150421559810638,
+      "learning_rate": 2.3888265229295014e-05,
+      "loss": 0.5675,
+      "step": 1063
+    },
+    {
+      "epoch": 0.29506378258458127,
+      "grad_norm": 0.19517284631729126,
+      "learning_rate": 2.388598770545924e-05,
+      "loss": 0.5516,
+      "step": 1064
+    },
+    {
+      "epoch": 0.29534109816971715,
+      "grad_norm": 0.2073058784008026,
+      "learning_rate": 2.3883707959915594e-05,
+      "loss": 0.5729,
+      "step": 1065
+    },
+    {
+      "epoch": 0.29561841375485304,
+      "grad_norm": 0.190653994679451,
+      "learning_rate": 2.3881425993108922e-05,
+      "loss": 0.5932,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2958957293399889,
+      "grad_norm": 0.19685585796833038,
+      "learning_rate": 2.3879141805484492e-05,
+      "loss": 0.5579,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2961730449251248,
+      "grad_norm": 0.19154155254364014,
+      "learning_rate": 2.3876855397488014e-05,
+      "loss": 0.5622,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2964503605102607,
+      "grad_norm": 0.23048245906829834,
+      "learning_rate": 2.387456676956562e-05,
+      "loss": 0.5951,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2967276760953966,
+      "grad_norm": 0.2001733034849167,
+      "learning_rate": 2.387227592216389e-05,
+      "loss": 0.5723,
+      "step": 1070
+    },
+    {
+      "epoch": 0.29700499168053246,
+      "grad_norm": 0.20289377868175507,
+      "learning_rate": 2.3869982855729822e-05,
+      "loss": 0.6023,
+      "step": 1071
+    },
+    {
+      "epoch": 0.29728230726566834,
+      "grad_norm": 0.18517981469631195,
+      "learning_rate": 2.386768757071086e-05,
+      "loss": 0.6075,
+      "step": 1072
+    },
+    {
+      "epoch": 0.29755962285080423,
+      "grad_norm": 0.18674753606319427,
+      "learning_rate": 2.3865390067554865e-05,
+      "loss": 0.5725,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2978369384359401,
+      "grad_norm": 0.19436419010162354,
+      "learning_rate": 2.3863090346710153e-05,
+      "loss": 0.595,
+      "step": 1074
+    },
+    {
+      "epoch": 0.298114254021076,
+      "grad_norm": 0.1953345537185669,
+      "learning_rate": 2.3860788408625456e-05,
+      "loss": 0.5845,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2983915696062119,
+      "grad_norm": 0.23673701286315918,
+      "learning_rate": 2.385848425374994e-05,
+      "loss": 0.5716,
+      "step": 1076
+    },
+    {
+      "epoch": 0.29866888519134777,
+      "grad_norm": 0.20462685823440552,
+      "learning_rate": 2.385617788253321e-05,
+      "loss": 0.5486,
+      "step": 1077
+    },
+    {
+      "epoch": 0.29894620077648365,
+      "grad_norm": 0.19857102632522583,
+      "learning_rate": 2.3853869295425296e-05,
+      "loss": 0.5744,
+      "step": 1078
+    },
+    {
+      "epoch": 0.29922351636161953,
+      "grad_norm": 0.21950219571590424,
+      "learning_rate": 2.385155849287667e-05,
+      "loss": 0.5644,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2995008319467554,
+      "grad_norm": 0.25962188839912415,
+      "learning_rate": 2.384924547533823e-05,
+      "loss": 0.5826,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2997781475318913,
+      "grad_norm": 0.21206709742546082,
+      "learning_rate": 2.3846930243261302e-05,
+      "loss": 0.5726,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3000554631170272,
+      "grad_norm": 0.20144885778427124,
+      "learning_rate": 2.384461279709765e-05,
+      "loss": 0.5736,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3003327787021631,
+      "grad_norm": 0.19098758697509766,
+      "learning_rate": 2.3842293137299475e-05,
+      "loss": 0.57,
+      "step": 1083
+    },
+    {
+      "epoch": 0.30061009428729896,
+      "grad_norm": 0.20457082986831665,
+      "learning_rate": 2.38399712643194e-05,
+      "loss": 0.5527,
+      "step": 1084
+    },
+    {
+      "epoch": 0.30088740987243484,
+      "grad_norm": 0.22459912300109863,
+      "learning_rate": 2.3837647178610482e-05,
+      "loss": 0.5907,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3011647254575707,
+      "grad_norm": 0.2092091143131256,
+      "learning_rate": 2.3835320880626216e-05,
+      "loss": 0.5796,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3014420410427066,
+      "grad_norm": 0.18621404469013214,
+      "learning_rate": 2.3832992370820523e-05,
+      "loss": 0.6085,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3017193566278425,
+      "grad_norm": 0.1938057541847229,
+      "learning_rate": 2.3830661649647757e-05,
+      "loss": 0.5642,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3019966722129784,
+      "grad_norm": 0.2136821299791336,
+      "learning_rate": 2.3828328717562704e-05,
+      "loss": 0.5621,
+      "step": 1089
+    },
+    {
+      "epoch": 0.30227398779811426,
+      "grad_norm": 0.200357124209404,
+      "learning_rate": 2.3825993575020577e-05,
+      "loss": 0.5728,
+      "step": 1090
+    },
+    {
+      "epoch": 0.30255130338325015,
+      "grad_norm": 0.19953665137290955,
+      "learning_rate": 2.382365622247703e-05,
+      "loss": 0.5677,
+      "step": 1091
+    },
+    {
+      "epoch": 0.30282861896838603,
+      "grad_norm": 0.19972378015518188,
+      "learning_rate": 2.382131666038814e-05,
+      "loss": 0.5623,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3031059345535219,
+      "grad_norm": 0.21744661033153534,
+      "learning_rate": 2.381897488921041e-05,
+      "loss": 0.5508,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3033832501386578,
+      "grad_norm": 0.18918602168560028,
+      "learning_rate": 2.3816630909400793e-05,
+      "loss": 0.5805,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3036605657237937,
+      "grad_norm": 0.1894434541463852,
+      "learning_rate": 2.3814284721416656e-05,
+      "loss": 0.563,
+      "step": 1095
+    },
+    {
+      "epoch": 0.30393788130892957,
+      "grad_norm": 0.19141651690006256,
+      "learning_rate": 2.3811936325715807e-05,
+      "loss": 0.5839,
+      "step": 1096
+    },
+    {
+      "epoch": 0.30421519689406545,
+      "grad_norm": 0.1895507425069809,
+      "learning_rate": 2.3809585722756472e-05,
+      "loss": 0.5864,
+      "step": 1097
+    },
+    {
+      "epoch": 0.30449251247920134,
+      "grad_norm": 0.21037045121192932,
+      "learning_rate": 2.3807232912997324e-05,
+      "loss": 0.5806,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3047698280643372,
+      "grad_norm": 0.20355623960494995,
+      "learning_rate": 2.3804877896897455e-05,
+      "loss": 0.5677,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3050471436494731,
+      "grad_norm": 0.20115728676319122,
+      "learning_rate": 2.380252067491639e-05,
+      "loss": 0.5635,
+      "step": 1100
+    },
+    {
+      "epoch": 0.305324459234609,
+      "grad_norm": 0.19460982084274292,
+      "learning_rate": 2.3800161247514086e-05,
+      "loss": 0.5774,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3056017748197449,
+      "grad_norm": 0.18929602205753326,
+      "learning_rate": 2.3797799615150934e-05,
+      "loss": 0.5708,
+      "step": 1102
+    },
+    {
+      "epoch": 0.30587909040488076,
+      "grad_norm": 0.20288972556591034,
+      "learning_rate": 2.3795435778287745e-05,
+      "loss": 0.5852,
+      "step": 1103
+    },
+    {
+      "epoch": 0.30615640599001664,
+      "grad_norm": 0.20964893698692322,
+      "learning_rate": 2.379306973738577e-05,
+      "loss": 0.5731,
+      "step": 1104
+    },
+    {
+      "epoch": 0.30643372157515253,
+      "grad_norm": 0.2250620722770691,
+      "learning_rate": 2.379070149290668e-05,
+      "loss": 0.5741,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3067110371602884,
+      "grad_norm": 0.19434000551700592,
+      "learning_rate": 2.3788331045312592e-05,
+      "loss": 0.5949,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3069883527454243,
+      "grad_norm": 0.1980692446231842,
+      "learning_rate": 2.3785958395066037e-05,
+      "loss": 0.5591,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3072656683305602,
+      "grad_norm": 0.20279406011104584,
+      "learning_rate": 2.3783583542629984e-05,
+      "loss": 0.5925,
+      "step": 1108
+    },
+    {
+      "epoch": 0.30754298391569607,
+      "grad_norm": 0.19802772998809814,
+      "learning_rate": 2.378120648846783e-05,
+      "loss": 0.5756,
+      "step": 1109
+    },
+    {
+      "epoch": 0.30782029950083195,
+      "grad_norm": 0.19455523788928986,
+      "learning_rate": 2.37788272330434e-05,
+      "loss": 0.5846,
+      "step": 1110
+    },
+    {
+      "epoch": 0.30809761508596784,
+      "grad_norm": 0.19493591785430908,
+      "learning_rate": 2.3776445776820948e-05,
+      "loss": 0.5788,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3083749306711037,
+      "grad_norm": 0.19869014620780945,
+      "learning_rate": 2.3774062120265163e-05,
+      "loss": 0.5836,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3086522462562396,
+      "grad_norm": 0.19301962852478027,
+      "learning_rate": 2.3771676263841157e-05,
+      "loss": 0.5689,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3089295618413755,
+      "grad_norm": 0.2125353217124939,
+      "learning_rate": 2.3769288208014473e-05,
+      "loss": 0.579,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3092068774265114,
+      "grad_norm": 0.20093408226966858,
+      "learning_rate": 2.376689795325109e-05,
+      "loss": 0.5489,
+      "step": 1115
+    },
+    {
+      "epoch": 0.30948419301164726,
+      "grad_norm": 0.20457975566387177,
+      "learning_rate": 2.37645055000174e-05,
+      "loss": 0.5734,
+      "step": 1116
+    },
+    {
+      "epoch": 0.30976150859678314,
+      "grad_norm": 0.2070866972208023,
+      "learning_rate": 2.376211084878024e-05,
+      "loss": 0.5711,
+      "step": 1117
+    },
+    {
+      "epoch": 0.310038824181919,
+      "grad_norm": 0.19698989391326904,
+      "learning_rate": 2.375971400000687e-05,
+      "loss": 0.5781,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3103161397670549,
+      "grad_norm": 0.19639791548252106,
+      "learning_rate": 2.3757314954164982e-05,
+      "loss": 0.5496,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3105934553521908,
+      "grad_norm": 0.19890196621418,
+      "learning_rate": 2.3754913711722687e-05,
+      "loss": 0.5658,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3108707709373267,
+      "grad_norm": 0.2007942795753479,
+      "learning_rate": 2.3752510273148533e-05,
+      "loss": 0.5679,
+      "step": 1121
+    },
+    {
+      "epoch": 0.31114808652246256,
+      "grad_norm": 0.21742717921733856,
+      "learning_rate": 2.3750104638911493e-05,
+      "loss": 0.5603,
+      "step": 1122
+    },
+    {
+      "epoch": 0.31142540210759845,
+      "grad_norm": 0.19708824157714844,
+      "learning_rate": 2.3747696809480974e-05,
+      "loss": 0.561,
+      "step": 1123
+    },
+    {
+      "epoch": 0.31170271769273433,
+      "grad_norm": 0.20064733922481537,
+      "learning_rate": 2.374528678532681e-05,
+      "loss": 0.5535,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3119800332778702,
+      "grad_norm": 0.19142742455005646,
+      "learning_rate": 2.3742874566919248e-05,
+      "loss": 0.5621,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3122573488630061,
+      "grad_norm": 0.20165249705314636,
+      "learning_rate": 2.3740460154728987e-05,
+      "loss": 0.5862,
+      "step": 1126
+    },
+    {
+      "epoch": 0.312534664448142,
+      "grad_norm": 0.19451411068439484,
+      "learning_rate": 2.373804354922714e-05,
+      "loss": 0.5611,
+      "step": 1127
+    },
+    {
+      "epoch": 0.31281198003327787,
+      "grad_norm": 0.20111672580242157,
+      "learning_rate": 2.373562475088525e-05,
+      "loss": 0.5629,
+      "step": 1128
+    },
+    {
+      "epoch": 0.31308929561841375,
+      "grad_norm": 0.21810267865657806,
+      "learning_rate": 2.3733203760175292e-05,
+      "loss": 0.5746,
+      "step": 1129
+    },
+    {
+      "epoch": 0.31336661120354964,
+      "grad_norm": 0.20461341738700867,
+      "learning_rate": 2.3730780577569654e-05,
+      "loss": 0.5429,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3136439267886855,
+      "grad_norm": 0.18515266478061676,
+      "learning_rate": 2.3728355203541182e-05,
+      "loss": 0.5547,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3139212423738214,
+      "grad_norm": 0.1941545456647873,
+      "learning_rate": 2.3725927638563112e-05,
+      "loss": 0.5566,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3141985579589573,
+      "grad_norm": 0.20392583310604095,
+      "learning_rate": 2.3723497883109137e-05,
+      "loss": 0.5506,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3144758735440932,
+      "grad_norm": 0.19422155618667603,
+      "learning_rate": 2.3721065937653363e-05,
+      "loss": 0.5826,
+      "step": 1134
+    },
+    {
+      "epoch": 0.31475318912922906,
+      "grad_norm": 0.19887231290340424,
+      "learning_rate": 2.3718631802670334e-05,
+      "loss": 0.5651,
+      "step": 1135
+    },
+    {
+      "epoch": 0.31503050471436495,
+      "grad_norm": 0.20396895706653595,
+      "learning_rate": 2.3716195478635e-05,
+      "loss": 0.5396,
+      "step": 1136
+    },
+    {
+      "epoch": 0.31530782029950083,
+      "grad_norm": 0.2106340378522873,
+      "learning_rate": 2.3713756966022766e-05,
+      "loss": 0.552,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3155851358846367,
+      "grad_norm": 0.20762008428573608,
+      "learning_rate": 2.371131626530944e-05,
+      "loss": 0.5942,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3158624514697726,
+      "grad_norm": 0.19259460270404816,
+      "learning_rate": 2.3708873376971277e-05,
+      "loss": 0.5364,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3161397670549085,
+      "grad_norm": 0.19473238289356232,
+      "learning_rate": 2.3706428301484946e-05,
+      "loss": 0.5862,
+      "step": 1140
+    },
+    {
+      "epoch": 0.31641708264004437,
+      "grad_norm": 0.20146487653255463,
+      "learning_rate": 2.370398103932754e-05,
+      "loss": 0.5924,
+      "step": 1141
+    },
+    {
+      "epoch": 0.31669439822518025,
+      "grad_norm": 0.19463180005550385,
+      "learning_rate": 2.370153159097659e-05,
+      "loss": 0.554,
+      "step": 1142
+    },
+    {
+      "epoch": 0.31697171381031614,
+      "grad_norm": 0.22867369651794434,
+      "learning_rate": 2.3699079956910052e-05,
+      "loss": 0.58,
+      "step": 1143
+    },
+    {
+      "epoch": 0.317249029395452,
+      "grad_norm": 0.1938031017780304,
+      "learning_rate": 2.3696626137606297e-05,
+      "loss": 0.5998,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3175263449805879,
+      "grad_norm": 0.1907264143228531,
+      "learning_rate": 2.369417013354413e-05,
+      "loss": 0.5601,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3178036605657238,
+      "grad_norm": 0.21385140717029572,
+      "learning_rate": 2.369171194520279e-05,
+      "loss": 0.5919,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3180809761508597,
+      "grad_norm": 0.23154176771640778,
+      "learning_rate": 2.3689251573061932e-05,
+      "loss": 0.5901,
+      "step": 1147
+    },
+    {
+      "epoch": 0.31835829173599556,
+      "grad_norm": 0.19392211735248566,
+      "learning_rate": 2.3686789017601634e-05,
+      "loss": 0.5514,
+      "step": 1148
+    },
+    {
+      "epoch": 0.31863560732113144,
+      "grad_norm": 0.19515374302864075,
+      "learning_rate": 2.3684324279302418e-05,
+      "loss": 0.558,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3189129229062673,
+      "grad_norm": 0.18750956654548645,
+      "learning_rate": 2.3681857358645205e-05,
+      "loss": 0.5544,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3191902384914032,
+      "grad_norm": 0.20247051119804382,
+      "learning_rate": 2.3679388256111368e-05,
+      "loss": 0.5724,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3194675540765391,
+      "grad_norm": 0.19525887072086334,
+      "learning_rate": 2.3676916972182686e-05,
+      "loss": 0.583,
+      "step": 1152
+    },
+    {
+      "epoch": 0.319744869661675,
+      "grad_norm": 0.20457209646701813,
+      "learning_rate": 2.3674443507341377e-05,
+      "loss": 0.5592,
+      "step": 1153
+    },
+    {
+      "epoch": 0.32002218524681086,
+      "grad_norm": 0.19369947910308838,
+      "learning_rate": 2.367196786207008e-05,
+      "loss": 0.5889,
+      "step": 1154
+    },
+    {
+      "epoch": 0.32029950083194675,
+      "grad_norm": 0.18851204216480255,
+      "learning_rate": 2.3669490036851856e-05,
+      "loss": 0.5804,
+      "step": 1155
+    },
+    {
+      "epoch": 0.32057681641708263,
+      "grad_norm": 0.1838391125202179,
+      "learning_rate": 2.3667010032170196e-05,
+      "loss": 0.5251,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3208541320022185,
+      "grad_norm": 0.1949407011270523,
+      "learning_rate": 2.3664527848509015e-05,
+      "loss": 0.5719,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3211314475873544,
+      "grad_norm": 0.2245536744594574,
+      "learning_rate": 2.3662043486352653e-05,
+      "loss": 0.5699,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3214087631724903,
+      "grad_norm": 0.18889588117599487,
+      "learning_rate": 2.3659556946185875e-05,
+      "loss": 0.5449,
+      "step": 1159
+    },
+    {
+      "epoch": 0.32168607875762617,
+      "grad_norm": 0.18529140949249268,
+      "learning_rate": 2.3657068228493863e-05,
+      "loss": 0.5469,
+      "step": 1160
+    },
+    {
+      "epoch": 0.32196339434276205,
+      "grad_norm": 0.19042176008224487,
+      "learning_rate": 2.3654577333762246e-05,
+      "loss": 0.5662,
+      "step": 1161
+    },
+    {
+      "epoch": 0.32224070992789794,
+      "grad_norm": 0.22052356600761414,
+      "learning_rate": 2.3652084262477055e-05,
+      "loss": 0.5603,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3225180255130338,
+      "grad_norm": 0.19664537906646729,
+      "learning_rate": 2.364958901512475e-05,
+      "loss": 0.5837,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3227953410981697,
+      "grad_norm": 0.1836601048707962,
+      "learning_rate": 2.3647091592192234e-05,
+      "loss": 0.5626,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3230726566833056,
+      "grad_norm": 0.19754944741725922,
+      "learning_rate": 2.3644591994166805e-05,
+      "loss": 0.5867,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3233499722684415,
+      "grad_norm": 0.18801425397396088,
+      "learning_rate": 2.364209022153621e-05,
+      "loss": 0.5447,
+      "step": 1166
+    },
+    {
+      "epoch": 0.32362728785357736,
+      "grad_norm": 0.19709810614585876,
+      "learning_rate": 2.363958627478861e-05,
+      "loss": 0.5953,
+      "step": 1167
+    },
+    {
+      "epoch": 0.32390460343871325,
+      "grad_norm": 0.19867432117462158,
+      "learning_rate": 2.3637080154412588e-05,
+      "loss": 0.6041,
+      "step": 1168
+    },
+    {
+      "epoch": 0.32418191902384913,
+      "grad_norm": 0.19217659533023834,
+      "learning_rate": 2.363457186089716e-05,
+      "loss": 0.5693,
+      "step": 1169
+    },
+    {
+      "epoch": 0.324459234608985,
+      "grad_norm": 0.20366674661636353,
+      "learning_rate": 2.3632061394731753e-05,
+      "loss": 0.5957,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3247365501941209,
+      "grad_norm": 0.19761385023593903,
+      "learning_rate": 2.362954875640623e-05,
+      "loss": 0.5722,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3250138657792568,
+      "grad_norm": 0.20664581656455994,
+      "learning_rate": 2.362703394641087e-05,
+      "loss": 0.5763,
+      "step": 1172
+    },
+    {
+      "epoch": 0.32529118136439267,
+      "grad_norm": 0.18691927194595337,
+      "learning_rate": 2.3624516965236386e-05,
+      "loss": 0.5707,
+      "step": 1173
+    },
+    {
+      "epoch": 0.32556849694952855,
+      "grad_norm": 0.18426918983459473,
+      "learning_rate": 2.36219978133739e-05,
+      "loss": 0.5337,
+      "step": 1174
+    },
+    {
+      "epoch": 0.32584581253466444,
+      "grad_norm": 0.19540777802467346,
+      "learning_rate": 2.3619476491314977e-05,
+      "loss": 0.5549,
+      "step": 1175
+    },
+    {
+      "epoch": 0.3261231281198003,
+      "grad_norm": 0.2000686079263687,
+      "learning_rate": 2.3616952999551576e-05,
+      "loss": 0.5765,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3264004437049362,
+      "grad_norm": 0.19267399609088898,
+      "learning_rate": 2.3614427338576114e-05,
+      "loss": 0.5585,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3266777592900721,
+      "grad_norm": 0.21964752674102783,
+      "learning_rate": 2.3611899508881403e-05,
+      "loss": 0.5742,
+      "step": 1178
+    },
+    {
+      "epoch": 0.326955074875208,
+      "grad_norm": 0.21694490313529968,
+      "learning_rate": 2.3609369510960696e-05,
+      "loss": 0.5586,
+      "step": 1179
+    },
+    {
+      "epoch": 0.32723239046034386,
+      "grad_norm": 0.2035355418920517,
+      "learning_rate": 2.360683734530766e-05,
+      "loss": 0.5712,
+      "step": 1180
+    },
+    {
+      "epoch": 0.32750970604547974,
+      "grad_norm": 0.1984809935092926,
+      "learning_rate": 2.3604303012416383e-05,
+      "loss": 0.5846,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3277870216306156,
+      "grad_norm": 0.19246408343315125,
+      "learning_rate": 2.360176651278139e-05,
+      "loss": 0.5931,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3280643372157515,
+      "grad_norm": 0.18938469886779785,
+      "learning_rate": 2.3599227846897615e-05,
+      "loss": 0.5722,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3283416528008874,
+      "grad_norm": 0.1882466822862625,
+      "learning_rate": 2.359668701526042e-05,
+      "loss": 0.5693,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3286189683860233,
+      "grad_norm": 0.18115603923797607,
+      "learning_rate": 2.3594144018365584e-05,
+      "loss": 0.5694,
+      "step": 1185
+    },
+    {
+      "epoch": 0.32889628397115916,
+      "grad_norm": 0.21563617885112762,
+      "learning_rate": 2.3591598856709317e-05,
+      "loss": 0.5541,
+      "step": 1186
+    },
+    {
+      "epoch": 0.32917359955629505,
+      "grad_norm": 0.19328701496124268,
+      "learning_rate": 2.3589051530788246e-05,
+      "loss": 0.5469,
+      "step": 1187
+    },
+    {
+      "epoch": 0.32945091514143093,
+      "grad_norm": 0.19828177988529205,
+      "learning_rate": 2.358650204109942e-05,
+      "loss": 0.5785,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3297282307265668,
+      "grad_norm": 0.1821170300245285,
+      "learning_rate": 2.358395038814032e-05,
+      "loss": 0.5727,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3300055463117027,
+      "grad_norm": 0.19181084632873535,
+      "learning_rate": 2.3581396572408833e-05,
+      "loss": 0.5699,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3302828618968386,
+      "grad_norm": 0.18721237778663635,
+      "learning_rate": 2.3578840594403275e-05,
+      "loss": 0.5786,
+      "step": 1191
+    },
+    {
+      "epoch": 0.33056017748197447,
+      "grad_norm": 0.2080426812171936,
+      "learning_rate": 2.3576282454622394e-05,
+      "loss": 0.6082,
+      "step": 1192
+    },
+    {
+      "epoch": 0.33083749306711036,
+      "grad_norm": 0.19874081015586853,
+      "learning_rate": 2.3573722153565343e-05,
+      "loss": 0.5802,
+      "step": 1193
+    },
+    {
+      "epoch": 0.33111480865224624,
+      "grad_norm": 0.20292651653289795,
+      "learning_rate": 2.357115969173171e-05,
+      "loss": 0.5562,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3313921242373821,
+      "grad_norm": 0.1992533951997757,
+      "learning_rate": 2.356859506962149e-05,
+      "loss": 0.5491,
+      "step": 1195
+    },
+    {
+      "epoch": 0.331669439822518,
+      "grad_norm": 0.19271281361579895,
+      "learning_rate": 2.356602828773512e-05,
+      "loss": 0.5713,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3319467554076539,
+      "grad_norm": 0.19548431038856506,
+      "learning_rate": 2.356345934657344e-05,
+      "loss": 0.5455,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3322240709927898,
+      "grad_norm": 0.19355355203151703,
+      "learning_rate": 2.3560888246637726e-05,
+      "loss": 0.576,
+      "step": 1198
+    },
+    {
+      "epoch": 0.33250138657792566,
+      "grad_norm": 0.19047684967517853,
+      "learning_rate": 2.3558314988429657e-05,
+      "loss": 0.5569,
+      "step": 1199
+    },
+    {
+      "epoch": 0.33277870216306155,
+      "grad_norm": 0.19171833992004395,
+      "learning_rate": 2.3555739572451353e-05,
+      "loss": 0.5974,
+      "step": 1200
+    },
+    {
+      "epoch": 0.33305601774819743,
+      "grad_norm": 0.18777206540107727,
+      "learning_rate": 2.3553161999205337e-05,
+      "loss": 0.5614,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.1937248855829239,
+      "learning_rate": 2.355058226919457e-05,
+      "loss": 0.5725,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3336106489184692,
+      "grad_norm": 0.19912956655025482,
+      "learning_rate": 2.3548000382922422e-05,
+      "loss": 0.5595,
+      "step": 1203
+    },
+    {
+      "epoch": 0.3338879645036051,
+      "grad_norm": 0.19780032336711884,
+      "learning_rate": 2.354541634089269e-05,
+      "loss": 0.5766,
+      "step": 1204
+    },
+    {
+      "epoch": 0.33416528008874097,
+      "grad_norm": 0.19286835193634033,
+      "learning_rate": 2.3542830143609584e-05,
+      "loss": 0.5529,
+      "step": 1205
+    },
+    {
+      "epoch": 0.33444259567387685,
+      "grad_norm": 0.19875440001487732,
+      "learning_rate": 2.3540241791577745e-05,
+      "loss": 0.5616,
+      "step": 1206
+    },
+    {
+      "epoch": 0.33471991125901274,
+      "grad_norm": 0.20458170771598816,
+      "learning_rate": 2.3537651285302224e-05,
+      "loss": 0.5754,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3349972268441486,
+      "grad_norm": 0.19896847009658813,
+      "learning_rate": 2.3535058625288503e-05,
+      "loss": 0.586,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3352745424292845,
+      "grad_norm": 0.2005191296339035,
+      "learning_rate": 2.3532463812042478e-05,
+      "loss": 0.57,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3355518580144204,
+      "grad_norm": 0.19629941880702972,
+      "learning_rate": 2.3529866846070457e-05,
+      "loss": 0.5632,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3358291735995563,
+      "grad_norm": 0.2009962797164917,
+      "learning_rate": 2.3527267727879187e-05,
+      "loss": 0.5926,
+      "step": 1211
+    },
+    {
+      "epoch": 0.33610648918469216,
+      "grad_norm": 0.18961815536022186,
+      "learning_rate": 2.3524666457975826e-05,
+      "loss": 0.566,
+      "step": 1212
+    },
+    {
+      "epoch": 0.33638380476982804,
+      "grad_norm": 0.19457431137561798,
+      "learning_rate": 2.3522063036867938e-05,
+      "loss": 0.5302,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3366611203549639,
+      "grad_norm": 0.20213200151920319,
+      "learning_rate": 2.351945746506353e-05,
+      "loss": 0.5726,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3369384359400998,
+      "grad_norm": 0.1976713389158249,
+      "learning_rate": 2.351684974307102e-05,
+      "loss": 0.5425,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3372157515252357,
+      "grad_norm": 0.19058452546596527,
+      "learning_rate": 2.3514239871399235e-05,
+      "loss": 0.5695,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3374930671103716,
+      "grad_norm": 0.18397174775600433,
+      "learning_rate": 2.3511627850557432e-05,
+      "loss": 0.535,
+      "step": 1217
+    },
+    {
+      "epoch": 0.33777038269550747,
+      "grad_norm": 0.2004639059305191,
+      "learning_rate": 2.3509013681055293e-05,
+      "loss": 0.5698,
+      "step": 1218
+    },
+    {
+      "epoch": 0.33804769828064335,
+      "grad_norm": 0.1970200389623642,
+      "learning_rate": 2.3506397363402905e-05,
+      "loss": 0.5831,
+      "step": 1219
+    },
+    {
+      "epoch": 0.33832501386577923,
+      "grad_norm": 0.24607634544372559,
+      "learning_rate": 2.3503778898110782e-05,
+      "loss": 0.6011,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3386023294509151,
+      "grad_norm": 0.1974974125623703,
+      "learning_rate": 2.3501158285689857e-05,
+      "loss": 0.5402,
+      "step": 1221
+    },
+    {
+      "epoch": 0.338879645036051,
+      "grad_norm": 0.20804527401924133,
+      "learning_rate": 2.3498535526651478e-05,
+      "loss": 0.5804,
+      "step": 1222
+    },
+    {
+      "epoch": 0.3391569606211869,
+      "grad_norm": 0.21283218264579773,
+      "learning_rate": 2.349591062150742e-05,
+      "loss": 0.5498,
+      "step": 1223
+    },
+    {
+      "epoch": 0.33943427620632277,
+      "grad_norm": 0.20082467794418335,
+      "learning_rate": 2.3493283570769863e-05,
+      "loss": 0.5589,
+      "step": 1224
+    },
+    {
+      "epoch": 0.33971159179145866,
+      "grad_norm": 0.1998477727174759,
+      "learning_rate": 2.3490654374951426e-05,
+      "loss": 0.5496,
+      "step": 1225
+    },
+    {
+      "epoch": 0.33998890737659454,
+      "grad_norm": 0.18160581588745117,
+      "learning_rate": 2.3488023034565127e-05,
+      "loss": 0.569,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3402662229617304,
+      "grad_norm": 0.18750114738941193,
+      "learning_rate": 2.3485389550124413e-05,
+      "loss": 0.5768,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3405435385468663,
+      "grad_norm": 0.19347476959228516,
+      "learning_rate": 2.3482753922143143e-05,
+      "loss": 0.5354,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3408208541320022,
+      "grad_norm": 0.19742092490196228,
+      "learning_rate": 2.34801161511356e-05,
+      "loss": 0.5656,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3410981697171381,
+      "grad_norm": 0.19319242238998413,
+      "learning_rate": 2.3477476237616487e-05,
+      "loss": 0.5654,
+      "step": 1230
+    },
+    {
+      "epoch": 0.34137548530227396,
+      "grad_norm": 0.18535441160202026,
+      "learning_rate": 2.3474834182100914e-05,
+      "loss": 0.5419,
+      "step": 1231
+    },
+    {
+      "epoch": 0.34165280088740985,
+      "grad_norm": 0.2077123373746872,
+      "learning_rate": 2.347218998510442e-05,
+      "loss": 0.5719,
+      "step": 1232
+    },
+    {
+      "epoch": 0.34193011647254573,
+      "grad_norm": 0.1924944818019867,
+      "learning_rate": 2.3469543647142954e-05,
+      "loss": 0.5493,
+      "step": 1233
+    },
+    {
+      "epoch": 0.3422074320576816,
+      "grad_norm": 0.1943088173866272,
+      "learning_rate": 2.3466895168732894e-05,
+      "loss": 0.5418,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3424847476428175,
+      "grad_norm": 0.19199106097221375,
+      "learning_rate": 2.3464244550391023e-05,
+      "loss": 0.5463,
+      "step": 1235
+    },
+    {
+      "epoch": 0.3427620632279534,
+      "grad_norm": 0.19151833653450012,
+      "learning_rate": 2.3461591792634548e-05,
+      "loss": 0.5595,
+      "step": 1236
+    },
+    {
+      "epoch": 0.34303937881308927,
+      "grad_norm": 0.20270255208015442,
+      "learning_rate": 2.3458936895981093e-05,
+      "loss": 0.5963,
+      "step": 1237
+    },
+    {
+      "epoch": 0.34331669439822515,
+      "grad_norm": 0.1867821365594864,
+      "learning_rate": 2.3456279860948696e-05,
+      "loss": 0.5716,
+      "step": 1238
+    },
+    {
+      "epoch": 0.34359400998336104,
+      "grad_norm": 0.20384635031223297,
+      "learning_rate": 2.3453620688055817e-05,
+      "loss": 0.5726,
+      "step": 1239
+    },
+    {
+      "epoch": 0.343871325568497,
+      "grad_norm": 0.19507844746112823,
+      "learning_rate": 2.3450959377821334e-05,
+      "loss": 0.5461,
+      "step": 1240
+    },
+    {
+      "epoch": 0.34414864115363286,
+      "grad_norm": 0.18854713439941406,
+      "learning_rate": 2.3448295930764536e-05,
+      "loss": 0.5617,
+      "step": 1241
+    },
+    {
+      "epoch": 0.34442595673876875,
+      "grad_norm": 0.1924966722726822,
+      "learning_rate": 2.344563034740513e-05,
+      "loss": 0.5716,
+      "step": 1242
+    },
+    {
+      "epoch": 0.34470327232390463,
+      "grad_norm": 0.1878724843263626,
+      "learning_rate": 2.3442962628263245e-05,
+      "loss": 0.5588,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3449805879090405,
+      "grad_norm": 0.20984706282615662,
+      "learning_rate": 2.3440292773859422e-05,
+      "loss": 0.5623,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3452579034941764,
+      "grad_norm": 0.19068847596645355,
+      "learning_rate": 2.343762078471462e-05,
+      "loss": 0.5573,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3455352190793123,
+      "grad_norm": 0.1986820548772812,
+      "learning_rate": 2.343494666135022e-05,
+      "loss": 0.5799,
+      "step": 1246
+    },
+    {
+      "epoch": 0.34581253466444817,
+      "grad_norm": 0.21772533655166626,
+      "learning_rate": 2.343227040428801e-05,
+      "loss": 0.5674,
+      "step": 1247
+    },
+    {
+      "epoch": 0.34608985024958405,
+      "grad_norm": 0.20012469589710236,
+      "learning_rate": 2.3429592014050198e-05,
+      "loss": 0.5713,
+      "step": 1248
+    },
+    {
+      "epoch": 0.34636716583471994,
+      "grad_norm": 0.21233108639717102,
+      "learning_rate": 2.3426911491159408e-05,
+      "loss": 0.5232,
+      "step": 1249
+    },
+    {
+      "epoch": 0.3466444814198558,
+      "grad_norm": 0.19460223615169525,
+      "learning_rate": 2.3424228836138686e-05,
+      "loss": 0.5547,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3469217970049917,
+      "grad_norm": 0.18516409397125244,
+      "learning_rate": 2.3421544049511484e-05,
+      "loss": 0.5445,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3471991125901276,
+      "grad_norm": 0.20658938586711884,
+      "learning_rate": 2.341885713180168e-05,
+      "loss": 0.5551,
+      "step": 1252
+    },
+    {
+      "epoch": 0.3474764281752635,
+      "grad_norm": 0.19541315734386444,
+      "learning_rate": 2.3416168083533556e-05,
+      "loss": 0.5608,
+      "step": 1253
+    },
+    {
+      "epoch": 0.34775374376039936,
+      "grad_norm": 0.22885319590568542,
+      "learning_rate": 2.3413476905231825e-05,
+      "loss": 0.5836,
+      "step": 1254
+    },
+    {
+      "epoch": 0.34803105934553524,
+      "grad_norm": 0.2285439521074295,
+      "learning_rate": 2.3410783597421597e-05,
+      "loss": 0.5444,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3483083749306711,
+      "grad_norm": 0.200783833861351,
+      "learning_rate": 2.3408088160628422e-05,
+      "loss": 0.5601,
+      "step": 1256
+    },
+    {
+      "epoch": 0.348585690515807,
+      "grad_norm": 0.19225940108299255,
+      "learning_rate": 2.3405390595378236e-05,
+      "loss": 0.5602,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3488630061009429,
+      "grad_norm": 0.19936451315879822,
+      "learning_rate": 2.340269090219741e-05,
+      "loss": 0.5593,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3491403216860788,
+      "grad_norm": 0.20055221021175385,
+      "learning_rate": 2.3399989081612732e-05,
+      "loss": 0.568,
+      "step": 1259
+    },
+    {
+      "epoch": 0.34941763727121466,
+      "grad_norm": 0.1941661238670349,
+      "learning_rate": 2.3397285134151394e-05,
+      "loss": 0.5688,
+      "step": 1260
+    },
+    {
+      "epoch": 0.34969495285635055,
+      "grad_norm": 0.19209027290344238,
+      "learning_rate": 2.3394579060341008e-05,
+      "loss": 0.5635,
+      "step": 1261
+    },
+    {
+      "epoch": 0.34997226844148643,
+      "grad_norm": 0.20127955079078674,
+      "learning_rate": 2.33918708607096e-05,
+      "loss": 0.5833,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3502495840266223,
+      "grad_norm": 0.20718038082122803,
+      "learning_rate": 2.3389160535785612e-05,
+      "loss": 0.5883,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3505268996117582,
+      "grad_norm": 0.19489611685276031,
+      "learning_rate": 2.3386448086097902e-05,
+      "loss": 0.5618,
+      "step": 1264
+    },
+    {
+      "epoch": 0.3508042151968941,
+      "grad_norm": 0.19366618990898132,
+      "learning_rate": 2.338373351217574e-05,
+      "loss": 0.5813,
+      "step": 1265
+    },
+    {
+      "epoch": 0.35108153078202997,
+      "grad_norm": 0.19742882251739502,
+      "learning_rate": 2.3381016814548806e-05,
+      "loss": 0.5866,
+      "step": 1266
+    },
+    {
+      "epoch": 0.35135884636716586,
+      "grad_norm": 0.2014235556125641,
+      "learning_rate": 2.337829799374721e-05,
+      "loss": 0.5503,
+      "step": 1267
+    },
+    {
+      "epoch": 0.35163616195230174,
+      "grad_norm": 0.19284577667713165,
+      "learning_rate": 2.337557705030146e-05,
+      "loss": 0.5683,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3519134775374376,
+      "grad_norm": 0.19096529483795166,
+      "learning_rate": 2.3372853984742482e-05,
+      "loss": 0.5535,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3521907931225735,
+      "grad_norm": 0.19602730870246887,
+      "learning_rate": 2.337012879760162e-05,
+      "loss": 0.5717,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3524681087077094,
+      "grad_norm": 0.19658134877681732,
+      "learning_rate": 2.3367401489410635e-05,
+      "loss": 0.5945,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3527454242928453,
+      "grad_norm": 0.20233921706676483,
+      "learning_rate": 2.3364672060701688e-05,
+      "loss": 0.5887,
+      "step": 1272
+    },
+    {
+      "epoch": 0.35302273987798116,
+      "grad_norm": 0.20138201117515564,
+      "learning_rate": 2.3361940512007368e-05,
+      "loss": 0.5424,
+      "step": 1273
+    },
+    {
+      "epoch": 0.35330005546311705,
+      "grad_norm": 0.1879206895828247,
+      "learning_rate": 2.3359206843860675e-05,
+      "loss": 0.546,
+      "step": 1274
+    },
+    {
+      "epoch": 0.35357737104825293,
+      "grad_norm": 0.19074855744838715,
+      "learning_rate": 2.335647105679502e-05,
+      "loss": 0.5486,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3538546866333888,
+      "grad_norm": 0.2010781466960907,
+      "learning_rate": 2.335373315134422e-05,
+      "loss": 0.6002,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3541320022185247,
+      "grad_norm": 0.19462116062641144,
+      "learning_rate": 2.3350993128042523e-05,
+      "loss": 0.5937,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3544093178036606,
+      "grad_norm": 0.21550050377845764,
+      "learning_rate": 2.3348250987424573e-05,
+      "loss": 0.5783,
+      "step": 1278
+    },
+    {
+      "epoch": 0.35468663338879647,
+      "grad_norm": 0.20210538804531097,
+      "learning_rate": 2.3345506730025434e-05,
+      "loss": 0.5784,
+      "step": 1279
+    },
+    {
+      "epoch": 0.35496394897393235,
+      "grad_norm": 0.19225560128688812,
+      "learning_rate": 2.3342760356380588e-05,
+      "loss": 0.566,
+      "step": 1280
+    },
+    {
+      "epoch": 0.35524126455906824,
+      "grad_norm": 0.2126346081495285,
+      "learning_rate": 2.3340011867025924e-05,
+      "loss": 0.5369,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3555185801442041,
+      "grad_norm": 0.19172579050064087,
+      "learning_rate": 2.333726126249774e-05,
+      "loss": 0.5736,
+      "step": 1282
+    },
+    {
+      "epoch": 0.35579589572934,
+      "grad_norm": 0.20066869258880615,
+      "learning_rate": 2.333450854333276e-05,
+      "loss": 0.5538,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3560732113144759,
+      "grad_norm": 0.2043704241514206,
+      "learning_rate": 2.3331753710068106e-05,
+      "loss": 0.5682,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3563505268996118,
+      "grad_norm": 0.1963237076997757,
+      "learning_rate": 2.3328996763241323e-05,
+      "loss": 0.5342,
+      "step": 1285
+    },
+    {
+      "epoch": 0.35662784248474766,
+      "grad_norm": 0.18448245525360107,
+      "learning_rate": 2.332623770339036e-05,
+      "loss": 0.5717,
+      "step": 1286
+    },
+    {
+      "epoch": 0.35690515806988354,
+      "grad_norm": 0.1940702348947525,
+      "learning_rate": 2.3323476531053587e-05,
+      "loss": 0.5491,
+      "step": 1287
+    },
+    {
+      "epoch": 0.35718247365501943,
+      "grad_norm": 0.1945616453886032,
+      "learning_rate": 2.3320713246769782e-05,
+      "loss": 0.5464,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3574597892401553,
+      "grad_norm": 0.19436432421207428,
+      "learning_rate": 2.331794785107813e-05,
+      "loss": 0.5475,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3577371048252912,
+      "grad_norm": 0.19309785962104797,
+      "learning_rate": 2.3315180344518236e-05,
+      "loss": 0.5413,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3580144204104271,
+      "grad_norm": 0.19348269701004028,
+      "learning_rate": 2.331241072763012e-05,
+      "loss": 0.5713,
+      "step": 1291
+    },
+    {
+      "epoch": 0.35829173599556297,
+      "grad_norm": 0.18745654821395874,
+      "learning_rate": 2.330963900095419e-05,
+      "loss": 0.5661,
+      "step": 1292
+    },
+    {
+      "epoch": 0.35856905158069885,
+      "grad_norm": 0.20293056964874268,
+      "learning_rate": 2.3306865165031305e-05,
+      "loss": 0.5855,
+      "step": 1293
+    },
+    {
+      "epoch": 0.35884636716583473,
+      "grad_norm": 0.202471524477005,
+      "learning_rate": 2.3304089220402702e-05,
+      "loss": 0.5905,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3591236827509706,
+      "grad_norm": 0.18649210035800934,
+      "learning_rate": 2.330131116761004e-05,
+      "loss": 0.5768,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3594009983361065,
+      "grad_norm": 0.18877407908439636,
+      "learning_rate": 2.3298531007195398e-05,
+      "loss": 0.576,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3596783139212424,
+      "grad_norm": 0.18760234117507935,
+      "learning_rate": 2.329574873970125e-05,
+      "loss": 0.54,
+      "step": 1297
+    },
+    {
+      "epoch": 0.35995562950637827,
+      "grad_norm": 0.2042498141527176,
+      "learning_rate": 2.32929643656705e-05,
+      "loss": 0.5595,
+      "step": 1298
+    },
+    {
+      "epoch": 0.36023294509151416,
+      "grad_norm": 0.19519071280956268,
+      "learning_rate": 2.3290177885646448e-05,
+      "loss": 0.5446,
+      "step": 1299
+    },
+    {
+      "epoch": 0.36051026067665004,
+      "grad_norm": 0.19533094763755798,
+      "learning_rate": 2.3287389300172806e-05,
+      "loss": 0.619,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3607875762617859,
+      "grad_norm": 0.18442007899284363,
+      "learning_rate": 2.3284598609793705e-05,
+      "loss": 0.5651,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3610648918469218,
+      "grad_norm": 0.1934802085161209,
+      "learning_rate": 2.3281805815053688e-05,
+      "loss": 0.5894,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3613422074320577,
+      "grad_norm": 0.20552469789981842,
+      "learning_rate": 2.327901091649769e-05,
+      "loss": 0.5943,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3616195230171936,
+      "grad_norm": 0.19476066529750824,
+      "learning_rate": 2.3276213914671084e-05,
+      "loss": 0.5729,
+      "step": 1304
+    },
+    {
+      "epoch": 0.36189683860232946,
+      "grad_norm": 0.19621802866458893,
+      "learning_rate": 2.3273414810119632e-05,
+      "loss": 0.5616,
+      "step": 1305
+    },
+    {
+      "epoch": 0.36217415418746535,
+      "grad_norm": 0.192935511469841,
+      "learning_rate": 2.3270613603389513e-05,
+      "loss": 0.5507,
+      "step": 1306
+    },
+    {
+      "epoch": 0.36245146977260123,
+      "grad_norm": 0.18750174343585968,
+      "learning_rate": 2.3267810295027317e-05,
+      "loss": 0.5866,
+      "step": 1307
+    },
+    {
+      "epoch": 0.3627287853577371,
+      "grad_norm": 0.18611189723014832,
+      "learning_rate": 2.3265004885580047e-05,
+      "loss": 0.5638,
+      "step": 1308
+    },
+    {
+      "epoch": 0.363006100942873,
+      "grad_norm": 0.205692857503891,
+      "learning_rate": 2.3262197375595108e-05,
+      "loss": 0.5628,
+      "step": 1309
+    },
+    {
+      "epoch": 0.3632834165280089,
+      "grad_norm": 0.19303978979587555,
+      "learning_rate": 2.3259387765620322e-05,
+      "loss": 0.5785,
+      "step": 1310
+    },
+    {
+      "epoch": 0.36356073211314477,
+      "grad_norm": 0.19016319513320923,
+      "learning_rate": 2.325657605620392e-05,
+      "loss": 0.5637,
+      "step": 1311
+    },
+    {
+      "epoch": 0.36383804769828065,
+      "grad_norm": 0.18235303461551666,
+      "learning_rate": 2.325376224789454e-05,
+      "loss": 0.5518,
+      "step": 1312
+    },
+    {
+      "epoch": 0.36411536328341654,
+      "grad_norm": 0.1889420747756958,
+      "learning_rate": 2.325094634124123e-05,
+      "loss": 0.5539,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3643926788685524,
+      "grad_norm": 0.1905851662158966,
+      "learning_rate": 2.3248128336793444e-05,
+      "loss": 0.5892,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3646699944536883,
+      "grad_norm": 0.18385791778564453,
+      "learning_rate": 2.324530823510106e-05,
+      "loss": 0.5746,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3649473100388242,
+      "grad_norm": 0.18778762221336365,
+      "learning_rate": 2.3242486036714343e-05,
+      "loss": 0.5578,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3652246256239601,
+      "grad_norm": 0.18907684087753296,
+      "learning_rate": 2.3239661742183984e-05,
+      "loss": 0.5912,
+      "step": 1317
+    },
+    {
+      "epoch": 0.36550194120909596,
+      "grad_norm": 0.19764885306358337,
+      "learning_rate": 2.3236835352061076e-05,
+      "loss": 0.5719,
+      "step": 1318
+    },
+    {
+      "epoch": 0.36577925679423184,
+      "grad_norm": 0.1823858916759491,
+      "learning_rate": 2.3234006866897125e-05,
+      "loss": 0.5619,
+      "step": 1319
+    },
+    {
+      "epoch": 0.36605657237936773,
+      "grad_norm": 0.31975796818733215,
+      "learning_rate": 2.3231176287244044e-05,
+      "loss": 0.5935,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3663338879645036,
+      "grad_norm": 0.17738473415374756,
+      "learning_rate": 2.322834361365415e-05,
+      "loss": 0.56,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3666112035496395,
+      "grad_norm": 0.19397957623004913,
+      "learning_rate": 2.3225508846680173e-05,
+      "loss": 0.5767,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3668885191347754,
+      "grad_norm": 0.1985812485218048,
+      "learning_rate": 2.3222671986875255e-05,
+      "loss": 0.5322,
+      "step": 1323
+    },
+    {
+      "epoch": 0.36716583471991127,
+      "grad_norm": 0.19608476758003235,
+      "learning_rate": 2.3219833034792943e-05,
+      "loss": 0.5758,
+      "step": 1324
+    },
+    {
+      "epoch": 0.36744315030504715,
+      "grad_norm": 0.196056067943573,
+      "learning_rate": 2.3216991990987186e-05,
+      "loss": 0.5658,
+      "step": 1325
+    },
+    {
+      "epoch": 0.36772046589018303,
+      "grad_norm": 0.18968339264392853,
+      "learning_rate": 2.3214148856012354e-05,
+      "loss": 0.5472,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3679977814753189,
+      "grad_norm": 0.19283819198608398,
+      "learning_rate": 2.3211303630423208e-05,
+      "loss": 0.5554,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3682750970604548,
+      "grad_norm": 0.19521844387054443,
+      "learning_rate": 2.320845631477494e-05,
+      "loss": 0.5252,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3685524126455907,
+      "grad_norm": 0.19548355042934418,
+      "learning_rate": 2.3205606909623122e-05,
+      "loss": 0.569,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3688297282307266,
+      "grad_norm": 0.18432483077049255,
+      "learning_rate": 2.3202755415523763e-05,
+      "loss": 0.5612,
+      "step": 1330
+    },
+    {
+      "epoch": 0.36910704381586246,
+      "grad_norm": 0.19396983087062836,
+      "learning_rate": 2.3199901833033255e-05,
+      "loss": 0.586,
+      "step": 1331
+    },
+    {
+      "epoch": 0.36938435940099834,
+      "grad_norm": 0.18808341026306152,
+      "learning_rate": 2.3197046162708413e-05,
+      "loss": 0.566,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3696616749861342,
+      "grad_norm": 0.19177477061748505,
+      "learning_rate": 2.3194188405106453e-05,
+      "loss": 0.5673,
+      "step": 1333
+    },
+    {
+      "epoch": 0.3699389905712701,
+      "grad_norm": 0.19362227618694305,
+      "learning_rate": 2.3191328560784992e-05,
+      "loss": 0.5367,
+      "step": 1334
+    },
+    {
+      "epoch": 0.370216306156406,
+      "grad_norm": 0.1922491490840912,
+      "learning_rate": 2.3188466630302072e-05,
+      "loss": 0.5466,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3704936217415419,
+      "grad_norm": 0.19545325636863708,
+      "learning_rate": 2.3185602614216125e-05,
+      "loss": 0.5861,
+      "step": 1336
+    },
+    {
+      "epoch": 0.37077093732667776,
+      "grad_norm": 0.1878175288438797,
+      "learning_rate": 2.3182736513086002e-05,
+      "loss": 0.5429,
+      "step": 1337
+    },
+    {
+      "epoch": 0.37104825291181365,
+      "grad_norm": 0.20909211039543152,
+      "learning_rate": 2.3179868327470948e-05,
+      "loss": 0.579,
+      "step": 1338
+    },
+    {
+      "epoch": 0.37132556849694953,
+      "grad_norm": 0.2049614042043686,
+      "learning_rate": 2.3176998057930626e-05,
+      "loss": 0.5877,
+      "step": 1339
+    },
+    {
+      "epoch": 0.3716028840820854,
+      "grad_norm": 0.20033860206604004,
+      "learning_rate": 2.3174125705025103e-05,
+      "loss": 0.5576,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3718801996672213,
+      "grad_norm": 0.19559535384178162,
+      "learning_rate": 2.3171251269314846e-05,
+      "loss": 0.581,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3721575152523572,
+      "grad_norm": 0.19223865866661072,
+      "learning_rate": 2.3168374751360737e-05,
+      "loss": 0.5742,
+      "step": 1342
+    },
+    {
+      "epoch": 0.37243483083749307,
+      "grad_norm": 0.18235072493553162,
+      "learning_rate": 2.316549615172406e-05,
+      "loss": 0.5839,
+      "step": 1343
+    },
+    {
+      "epoch": 0.37271214642262895,
+      "grad_norm": 0.1987343281507492,
+      "learning_rate": 2.3162615470966512e-05,
+      "loss": 0.5531,
+      "step": 1344
+    },
+    {
+      "epoch": 0.37298946200776484,
+      "grad_norm": 0.20500993728637695,
+      "learning_rate": 2.3159732709650182e-05,
+      "loss": 0.5849,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3732667775929007,
+      "grad_norm": 0.19410103559494019,
+      "learning_rate": 2.3156847868337574e-05,
+      "loss": 0.602,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3735440931780366,
+      "grad_norm": 0.19419234991073608,
+      "learning_rate": 2.31539609475916e-05,
+      "loss": 0.5639,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3738214087631725,
+      "grad_norm": 0.19289056956768036,
+      "learning_rate": 2.3151071947975578e-05,
+      "loss": 0.5833,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3740987243483084,
+      "grad_norm": 0.1862688809633255,
+      "learning_rate": 2.314818087005322e-05,
+      "loss": 0.5587,
+      "step": 1349
+    },
+    {
+      "epoch": 0.37437603993344426,
+      "grad_norm": 0.1773182898759842,
+      "learning_rate": 2.314528771438866e-05,
+      "loss": 0.5297,
+      "step": 1350
+    },
+    {
+      "epoch": 0.37465335551858014,
+      "grad_norm": 0.18808980286121368,
+      "learning_rate": 2.314239248154642e-05,
+      "loss": 0.5478,
+      "step": 1351
+    },
+    {
+      "epoch": 0.37493067110371603,
+      "grad_norm": 0.1850731521844864,
+      "learning_rate": 2.3139495172091447e-05,
+      "loss": 0.5631,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3752079866888519,
+      "grad_norm": 0.233089417219162,
+      "learning_rate": 2.313659578658907e-05,
+      "loss": 0.5688,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3754853022739878,
+      "grad_norm": 0.19679111242294312,
+      "learning_rate": 2.313369432560505e-05,
+      "loss": 0.5713,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3757626178591237,
+      "grad_norm": 0.22906848788261414,
+      "learning_rate": 2.3130790789705535e-05,
+      "loss": 0.5727,
+      "step": 1355
+    },
+    {
+      "epoch": 0.37603993344425957,
+      "grad_norm": 0.20593827962875366,
+      "learning_rate": 2.3127885179457077e-05,
+      "loss": 0.5382,
+      "step": 1356
+    },
+    {
+      "epoch": 0.37631724902939545,
+      "grad_norm": 0.18781960010528564,
+      "learning_rate": 2.3124977495426637e-05,
+      "loss": 0.5746,
+      "step": 1357
+    },
+    {
+      "epoch": 0.37659456461453134,
+      "grad_norm": 0.2294550985097885,
+      "learning_rate": 2.3122067738181587e-05,
+      "loss": 0.5783,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3768718801996672,
+      "grad_norm": 0.2007582187652588,
+      "learning_rate": 2.311915590828969e-05,
+      "loss": 0.5586,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3771491957848031,
+      "grad_norm": 0.20126573741436005,
+      "learning_rate": 2.3116242006319132e-05,
+      "loss": 0.5621,
+      "step": 1360
+    },
+    {
+      "epoch": 0.377426511369939,
+      "grad_norm": 0.19607853889465332,
+      "learning_rate": 2.3113326032838487e-05,
+      "loss": 0.5305,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3777038269550749,
+      "grad_norm": 0.193894624710083,
+      "learning_rate": 2.3110407988416736e-05,
+      "loss": 0.5578,
+      "step": 1362
+    },
+    {
+      "epoch": 0.37798114254021076,
+      "grad_norm": 0.1813306212425232,
+      "learning_rate": 2.310748787362327e-05,
+      "loss": 0.5787,
+      "step": 1363
+    },
+    {
+      "epoch": 0.37825845812534664,
+      "grad_norm": 0.19636552035808563,
+      "learning_rate": 2.3104565689027875e-05,
+      "loss": 0.5615,
+      "step": 1364
+    },
+    {
+      "epoch": 0.3785357737104825,
+      "grad_norm": 0.1957857310771942,
+      "learning_rate": 2.3101641435200756e-05,
+      "loss": 0.5821,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3788130892956184,
+      "grad_norm": 0.18413352966308594,
+      "learning_rate": 2.3098715112712507e-05,
+      "loss": 0.5388,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3790904048807543,
+      "grad_norm": 0.20153377950191498,
+      "learning_rate": 2.3095786722134133e-05,
+      "loss": 0.5748,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3793677204658902,
+      "grad_norm": 0.22044618427753448,
+      "learning_rate": 2.309285626403704e-05,
+      "loss": 0.5668,
+      "step": 1368
+    },
+    {
+      "epoch": 0.37964503605102606,
+      "grad_norm": 0.19130001962184906,
+      "learning_rate": 2.3089923738993034e-05,
+      "loss": 0.5752,
+      "step": 1369
+    },
+    {
+      "epoch": 0.37992235163616195,
+      "grad_norm": 0.18104785680770874,
+      "learning_rate": 2.3086989147574333e-05,
+      "loss": 0.5577,
+      "step": 1370
+    },
+    {
+      "epoch": 0.38019966722129783,
+      "grad_norm": 0.18427520990371704,
+      "learning_rate": 2.3084052490353553e-05,
+      "loss": 0.5537,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3804769828064337,
+      "grad_norm": 0.1886986941099167,
+      "learning_rate": 2.3081113767903713e-05,
+      "loss": 0.5646,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3807542983915696,
+      "grad_norm": 0.19212138652801514,
+      "learning_rate": 2.3078172980798236e-05,
+      "loss": 0.5415,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3810316139767055,
+      "grad_norm": 0.18980136513710022,
+      "learning_rate": 2.3075230129610946e-05,
+      "loss": 0.5725,
+      "step": 1374
+    },
+    {
+      "epoch": 0.38130892956184137,
+      "grad_norm": 0.1848769187927246,
+      "learning_rate": 2.3072285214916072e-05,
+      "loss": 0.536,
+      "step": 1375
+    },
+    {
+      "epoch": 0.38158624514697725,
+      "grad_norm": 0.18508492410182953,
+      "learning_rate": 2.3069338237288247e-05,
+      "loss": 0.5753,
+      "step": 1376
+    },
+    {
+      "epoch": 0.38186356073211314,
+      "grad_norm": 0.1909710019826889,
+      "learning_rate": 2.30663891973025e-05,
+      "loss": 0.563,
+      "step": 1377
+    },
+    {
+      "epoch": 0.382140876317249,
+      "grad_norm": 0.20639832317829132,
+      "learning_rate": 2.3063438095534272e-05,
+      "loss": 0.5713,
+      "step": 1378
+    },
+    {
+      "epoch": 0.3824181919023849,
+      "grad_norm": 0.1900801658630371,
+      "learning_rate": 2.3060484932559395e-05,
+      "loss": 0.5579,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3826955074875208,
+      "grad_norm": 0.19208571314811707,
+      "learning_rate": 2.305752970895412e-05,
+      "loss": 0.5571,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3829728230726567,
+      "grad_norm": 0.1912533938884735,
+      "learning_rate": 2.3054572425295075e-05,
+      "loss": 0.5452,
+      "step": 1381
+    },
+    {
+      "epoch": 0.38325013865779256,
+      "grad_norm": 0.19920608401298523,
+      "learning_rate": 2.3051613082159313e-05,
+      "loss": 0.5799,
+      "step": 1382
+    },
+    {
+      "epoch": 0.38352745424292845,
+      "grad_norm": 0.19880720973014832,
+      "learning_rate": 2.3048651680124283e-05,
+      "loss": 0.5504,
+      "step": 1383
+    },
+    {
+      "epoch": 0.38380476982806433,
+      "grad_norm": 0.1889009028673172,
+      "learning_rate": 2.3045688219767824e-05,
+      "loss": 0.5751,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3840820854132002,
+      "grad_norm": 0.20132310688495636,
+      "learning_rate": 2.3042722701668194e-05,
+      "loss": 0.5723,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3843594009983361,
+      "grad_norm": 0.20646221935749054,
+      "learning_rate": 2.3039755126404037e-05,
+      "loss": 0.581,
+      "step": 1386
+    },
+    {
+      "epoch": 0.384636716583472,
+      "grad_norm": 0.19051332771778107,
+      "learning_rate": 2.3036785494554415e-05,
+      "loss": 0.5609,
+      "step": 1387
+    },
+    {
+      "epoch": 0.38491403216860787,
+      "grad_norm": 0.1971728354692459,
+      "learning_rate": 2.303381380669877e-05,
+      "loss": 0.563,
+      "step": 1388
+    },
+    {
+      "epoch": 0.38519134775374375,
+      "grad_norm": 0.19891361892223358,
+      "learning_rate": 2.303084006341697e-05,
+      "loss": 0.571,
+      "step": 1389
+    },
+    {
+      "epoch": 0.38546866333887964,
+      "grad_norm": 0.18801699578762054,
+      "learning_rate": 2.302786426528926e-05,
+      "loss": 0.5772,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3857459789240155,
+      "grad_norm": 0.1934468299150467,
+      "learning_rate": 2.3024886412896302e-05,
+      "loss": 0.5426,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3860232945091514,
+      "grad_norm": 0.20585696399211884,
+      "learning_rate": 2.3021906506819152e-05,
+      "loss": 0.5521,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3863006100942873,
+      "grad_norm": 0.19114629924297333,
+      "learning_rate": 2.3018924547639272e-05,
+      "loss": 0.597,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3865779256794232,
+      "grad_norm": 0.18134154379367828,
+      "learning_rate": 2.301594053593852e-05,
+      "loss": 0.5473,
+      "step": 1394
+    },
+    {
+      "epoch": 0.38685524126455906,
+      "grad_norm": 0.18706971406936646,
+      "learning_rate": 2.301295447229915e-05,
+      "loss": 0.5843,
+      "step": 1395
+    },
+    {
+      "epoch": 0.38713255684969494,
+      "grad_norm": 0.183674156665802,
+      "learning_rate": 2.300996635730383e-05,
+      "loss": 0.5268,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3874098724348308,
+      "grad_norm": 0.18946042656898499,
+      "learning_rate": 2.3006976191535616e-05,
+      "loss": 0.5529,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3876871880199667,
+      "grad_norm": 0.20170167088508606,
+      "learning_rate": 2.3003983975577975e-05,
+      "loss": 0.5781,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3879645036051026,
+      "grad_norm": 0.1854201704263687,
+      "learning_rate": 2.300098971001476e-05,
+      "loss": 0.536,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3882418191902385,
+      "grad_norm": 0.19159002602100372,
+      "learning_rate": 2.299799339543023e-05,
+      "loss": 0.584,
+      "step": 1400
+    },
+    {
+      "epoch": 0.38851913477537436,
+      "grad_norm": 0.18878300487995148,
+      "learning_rate": 2.299499503240905e-05,
+      "loss": 0.5613,
+      "step": 1401
+    },
+    {
+      "epoch": 0.38879645036051025,
+      "grad_norm": 0.2083284854888916,
+      "learning_rate": 2.2991994621536283e-05,
+      "loss": 0.5607,
+      "step": 1402
+    },
+    {
+      "epoch": 0.38907376594564613,
+      "grad_norm": 0.19140368700027466,
+      "learning_rate": 2.2988992163397386e-05,
+      "loss": 0.5723,
+      "step": 1403
+    },
+    {
+      "epoch": 0.389351081530782,
+      "grad_norm": 0.19407185912132263,
+      "learning_rate": 2.2985987658578217e-05,
+      "loss": 0.5592,
+      "step": 1404
+    },
+    {
+      "epoch": 0.3896283971159179,
+      "grad_norm": 0.202662855386734,
+      "learning_rate": 2.298298110766503e-05,
+      "loss": 0.5682,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3899057127010538,
+      "grad_norm": 0.1902836710214615,
+      "learning_rate": 2.2979972511244493e-05,
+      "loss": 0.5626,
+      "step": 1406
+    },
+    {
+      "epoch": 0.39018302828618967,
+      "grad_norm": 0.19113852083683014,
+      "learning_rate": 2.2976961869903657e-05,
+      "loss": 0.5713,
+      "step": 1407
+    },
+    {
+      "epoch": 0.39046034387132555,
+      "grad_norm": 0.19765017926692963,
+      "learning_rate": 2.2973949184229975e-05,
+      "loss": 0.5811,
+      "step": 1408
+    },
+    {
+      "epoch": 0.39073765945646144,
+      "grad_norm": 0.1867290586233139,
+      "learning_rate": 2.2970934454811306e-05,
+      "loss": 0.5094,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3910149750415973,
+      "grad_norm": 0.19188763201236725,
+      "learning_rate": 2.2967917682235905e-05,
+      "loss": 0.5736,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3912922906267332,
+      "grad_norm": 0.19784973561763763,
+      "learning_rate": 2.296489886709242e-05,
+      "loss": 0.5739,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3915696062118691,
+      "grad_norm": 0.19761349260807037,
+      "learning_rate": 2.2961878009969904e-05,
+      "loss": 0.5702,
+      "step": 1412
+    },
+    {
+      "epoch": 0.391846921797005,
+      "grad_norm": 0.19187502562999725,
+      "learning_rate": 2.2958855111457804e-05,
+      "loss": 0.5575,
+      "step": 1413
+    },
+    {
+      "epoch": 0.39212423738214086,
+      "grad_norm": 0.19723129272460938,
+      "learning_rate": 2.2955830172145975e-05,
+      "loss": 0.5757,
+      "step": 1414
+    },
+    {
+      "epoch": 0.39240155296727675,
+      "grad_norm": 0.1946565806865692,
+      "learning_rate": 2.2952803192624653e-05,
+      "loss": 0.5696,
+      "step": 1415
+    },
+    {
+      "epoch": 0.39267886855241263,
+      "grad_norm": 0.1714453399181366,
+      "learning_rate": 2.2949774173484488e-05,
+      "loss": 0.5783,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3929561841375485,
+      "grad_norm": 0.1907140165567398,
+      "learning_rate": 2.2946743115316518e-05,
+      "loss": 0.5642,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3932334997226844,
+      "grad_norm": 0.18922662734985352,
+      "learning_rate": 2.294371001871219e-05,
+      "loss": 0.5403,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3935108153078203,
+      "grad_norm": 0.18016541004180908,
+      "learning_rate": 2.294067488426333e-05,
+      "loss": 0.5428,
+      "step": 1419
+    },
+    {
+      "epoch": 0.39378813089295617,
+      "grad_norm": 0.18581193685531616,
+      "learning_rate": 2.293763771256218e-05,
+      "loss": 0.5671,
+      "step": 1420
+    },
+    {
+      "epoch": 0.39406544647809205,
+      "grad_norm": 0.19479569792747498,
+      "learning_rate": 2.293459850420138e-05,
+      "loss": 0.5471,
+      "step": 1421
+    },
+    {
+      "epoch": 0.39434276206322794,
+      "grad_norm": 0.19193242490291595,
+      "learning_rate": 2.2931557259773944e-05,
+      "loss": 0.5422,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3946200776483638,
+      "grad_norm": 0.1958051174879074,
+      "learning_rate": 2.2928513979873312e-05,
+      "loss": 0.5605,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3948973932334997,
+      "grad_norm": 0.1831037700176239,
+      "learning_rate": 2.2925468665093304e-05,
+      "loss": 0.5632,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3951747088186356,
+      "grad_norm": 0.19089363515377045,
+      "learning_rate": 2.2922421316028142e-05,
+      "loss": 0.5534,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3954520244037715,
+      "grad_norm": 0.18920765817165375,
+      "learning_rate": 2.2919371933272445e-05,
+      "loss": 0.5708,
+      "step": 1426
+    },
+    {
+      "epoch": 0.39572933998890736,
+      "grad_norm": 0.1790907382965088,
+      "learning_rate": 2.2916320517421224e-05,
+      "loss": 0.54,
+      "step": 1427
+    },
+    {
+      "epoch": 0.39600665557404324,
+      "grad_norm": 0.19448032975196838,
+      "learning_rate": 2.29132670690699e-05,
+      "loss": 0.5713,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3962839711591791,
+      "grad_norm": 0.18597151339054108,
+      "learning_rate": 2.2910211588814272e-05,
+      "loss": 0.5619,
+      "step": 1429
+    },
+    {
+      "epoch": 0.396561286744315,
+      "grad_norm": 0.19088095426559448,
+      "learning_rate": 2.2907154077250554e-05,
+      "loss": 0.5567,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3968386023294509,
+      "grad_norm": 0.2470846325159073,
+      "learning_rate": 2.290409453497534e-05,
+      "loss": 0.5567,
+      "step": 1431
+    },
+    {
+      "epoch": 0.3971159179145868,
+      "grad_norm": 0.189782977104187,
+      "learning_rate": 2.2901032962585633e-05,
+      "loss": 0.5689,
+      "step": 1432
+    },
+    {
+      "epoch": 0.39739323349972266,
+      "grad_norm": 0.18323121964931488,
+      "learning_rate": 2.289796936067882e-05,
+      "loss": 0.5695,
+      "step": 1433
+    },
+    {
+      "epoch": 0.39767054908485855,
+      "grad_norm": 0.19183968007564545,
+      "learning_rate": 2.28949037298527e-05,
+      "loss": 0.5527,
+      "step": 1434
+    },
+    {
+      "epoch": 0.39794786466999443,
+      "grad_norm": 0.2490936517715454,
+      "learning_rate": 2.2891836070705454e-05,
+      "loss": 0.5705,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3982251802551303,
+      "grad_norm": 0.18432928621768951,
+      "learning_rate": 2.2888766383835664e-05,
+      "loss": 0.5542,
+      "step": 1436
+    },
+    {
+      "epoch": 0.3985024958402662,
+      "grad_norm": 0.18222655355930328,
+      "learning_rate": 2.2885694669842305e-05,
+      "loss": 0.5693,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3987798114254021,
+      "grad_norm": 0.18472431600093842,
+      "learning_rate": 2.2882620929324758e-05,
+      "loss": 0.5664,
+      "step": 1438
+    },
+    {
+      "epoch": 0.39905712701053797,
+      "grad_norm": 0.19013690948486328,
+      "learning_rate": 2.2879545162882782e-05,
+      "loss": 0.534,
+      "step": 1439
+    },
+    {
+      "epoch": 0.39933444259567386,
+      "grad_norm": 0.1890943944454193,
+      "learning_rate": 2.2876467371116546e-05,
+      "loss": 0.5537,
+      "step": 1440
+    },
+    {
+      "epoch": 0.39961175818080974,
+      "grad_norm": 0.19494980573654175,
+      "learning_rate": 2.28733875546266e-05,
+      "loss": 0.5502,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3998890737659456,
+      "grad_norm": 0.215170755982399,
+      "learning_rate": 2.2870305714013908e-05,
+      "loss": 0.57,
+      "step": 1442
+    },
+    {
+      "epoch": 0.4001663893510815,
+      "grad_norm": 0.20500238239765167,
+      "learning_rate": 2.2867221849879816e-05,
+      "loss": 0.5648,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4004437049362174,
+      "grad_norm": 0.1898437738418579,
+      "learning_rate": 2.2864135962826067e-05,
+      "loss": 0.5685,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4007210205213533,
+      "grad_norm": 0.1926756501197815,
+      "learning_rate": 2.2861048053454797e-05,
+      "loss": 0.5514,
+      "step": 1445
+    },
+    {
+      "epoch": 0.40099833610648916,
+      "grad_norm": 0.19258753955364227,
+      "learning_rate": 2.2857958122368545e-05,
+      "loss": 0.5397,
+      "step": 1446
+    },
+    {
+      "epoch": 0.40127565169162505,
+      "grad_norm": 0.18636788427829742,
+      "learning_rate": 2.285486617017023e-05,
+      "loss": 0.5421,
+      "step": 1447
+    },
+    {
+      "epoch": 0.40155296727676093,
+      "grad_norm": 0.20043087005615234,
+      "learning_rate": 2.2851772197463184e-05,
+      "loss": 0.5241,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4018302828618968,
+      "grad_norm": 0.2062230408191681,
+      "learning_rate": 2.284867620485111e-05,
+      "loss": 0.5702,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4021075984470327,
+      "grad_norm": 0.1930972784757614,
+      "learning_rate": 2.284557819293813e-05,
+      "loss": 0.5533,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4023849140321686,
+      "grad_norm": 0.20658205449581146,
+      "learning_rate": 2.284247816232874e-05,
+      "loss": 0.5613,
+      "step": 1451
+    },
+    {
+      "epoch": 0.40266222961730447,
+      "grad_norm": 0.2103574126958847,
+      "learning_rate": 2.2839376113627848e-05,
+      "loss": 0.5668,
+      "step": 1452
+    },
+    {
+      "epoch": 0.40293954520244035,
+      "grad_norm": 0.19861139357089996,
+      "learning_rate": 2.2836272047440733e-05,
+      "loss": 0.5549,
+      "step": 1453
+    },
+    {
+      "epoch": 0.40321686078757624,
+      "grad_norm": 0.1885387897491455,
+      "learning_rate": 2.2833165964373093e-05,
+      "loss": 0.565,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4034941763727121,
+      "grad_norm": 0.18594177067279816,
+      "learning_rate": 2.2830057865030997e-05,
+      "loss": 0.5129,
+      "step": 1455
+    },
+    {
+      "epoch": 0.403771491957848,
+      "grad_norm": 0.19150151312351227,
+      "learning_rate": 2.282694775002092e-05,
+      "loss": 0.5591,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4040488075429839,
+      "grad_norm": 0.1973596066236496,
+      "learning_rate": 2.2823835619949735e-05,
+      "loss": 0.5795,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4043261231281198,
+      "grad_norm": 0.21641992032527924,
+      "learning_rate": 2.2820721475424693e-05,
+      "loss": 0.5286,
+      "step": 1458
+    },
+    {
+      "epoch": 0.40460343871325566,
+      "grad_norm": 0.20730352401733398,
+      "learning_rate": 2.281760531705345e-05,
+      "loss": 0.578,
+      "step": 1459
+    },
+    {
+      "epoch": 0.40488075429839154,
+      "grad_norm": 0.19733300805091858,
+      "learning_rate": 2.281448714544405e-05,
+      "loss": 0.6048,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4051580698835274,
+      "grad_norm": 0.2007417231798172,
+      "learning_rate": 2.281136696120493e-05,
+      "loss": 0.588,
+      "step": 1461
+    },
+    {
+      "epoch": 0.4054353854686633,
+      "grad_norm": 0.1785038709640503,
+      "learning_rate": 2.280824476494492e-05,
+      "loss": 0.5505,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4057127010537992,
+      "grad_norm": 0.18670029938220978,
+      "learning_rate": 2.2805120557273246e-05,
+      "loss": 0.5533,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4059900166389351,
+      "grad_norm": 0.18775032460689545,
+      "learning_rate": 2.2801994338799525e-05,
+      "loss": 0.5457,
+      "step": 1464
+    },
+    {
+      "epoch": 0.406267332224071,
+      "grad_norm": 0.18645282089710236,
+      "learning_rate": 2.2798866110133758e-05,
+      "loss": 0.5595,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4065446478092069,
+      "grad_norm": 0.2067873477935791,
+      "learning_rate": 2.279573587188635e-05,
+      "loss": 0.5438,
+      "step": 1466
+    },
+    {
+      "epoch": 0.4068219633943428,
+      "grad_norm": 0.18860718607902527,
+      "learning_rate": 2.2792603624668097e-05,
+      "loss": 0.541,
+      "step": 1467
+    },
+    {
+      "epoch": 0.4070992789794787,
+      "grad_norm": 0.18750540912151337,
+      "learning_rate": 2.2789469369090173e-05,
+      "loss": 0.5567,
+      "step": 1468
+    },
+    {
+      "epoch": 0.40737659456461456,
+      "grad_norm": 0.18958985805511475,
+      "learning_rate": 2.2786333105764162e-05,
+      "loss": 0.5413,
+      "step": 1469
+    },
+    {
+      "epoch": 0.40765391014975044,
+      "grad_norm": 0.27570971846580505,
+      "learning_rate": 2.2783194835302035e-05,
+      "loss": 0.5548,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4079312257348863,
+      "grad_norm": 0.1889680027961731,
+      "learning_rate": 2.2780054558316146e-05,
+      "loss": 0.5507,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4082085413200222,
+      "grad_norm": 0.18410347402095795,
+      "learning_rate": 2.277691227541925e-05,
+      "loss": 0.5689,
+      "step": 1472
+    },
+    {
+      "epoch": 0.4084858569051581,
+      "grad_norm": 0.18935447931289673,
+      "learning_rate": 2.277376798722448e-05,
+      "loss": 0.5631,
+      "step": 1473
+    },
+    {
+      "epoch": 0.408763172490294,
+      "grad_norm": 0.19542036950588226,
+      "learning_rate": 2.2770621694345385e-05,
+      "loss": 0.5614,
+      "step": 1474
+    },
+    {
+      "epoch": 0.40904048807542986,
+      "grad_norm": 0.2004593461751938,
+      "learning_rate": 2.2767473397395876e-05,
+      "loss": 0.5707,
+      "step": 1475
+    },
+    {
+      "epoch": 0.40931780366056575,
+      "grad_norm": 0.19551704823970795,
+      "learning_rate": 2.276432309699028e-05,
+      "loss": 0.5656,
+      "step": 1476
+    },
+    {
+      "epoch": 0.40959511924570163,
+      "grad_norm": 0.18432258069515228,
+      "learning_rate": 2.27611707937433e-05,
+      "loss": 0.5696,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4098724348308375,
+      "grad_norm": 0.20155014097690582,
+      "learning_rate": 2.2758016488270033e-05,
+      "loss": 0.581,
+      "step": 1478
+    },
+    {
+      "epoch": 0.4101497504159734,
+      "grad_norm": 0.1944228559732437,
+      "learning_rate": 2.2754860181185967e-05,
+      "loss": 0.565,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4104270660011093,
+      "grad_norm": 0.20183973014354706,
+      "learning_rate": 2.2751701873106983e-05,
+      "loss": 0.5743,
+      "step": 1480
+    },
+    {
+      "epoch": 0.41070438158624517,
+      "grad_norm": 0.18106213212013245,
+      "learning_rate": 2.274854156464935e-05,
+      "loss": 0.5726,
+      "step": 1481
+    },
+    {
+      "epoch": 0.41098169717138106,
+      "grad_norm": 0.19121475517749786,
+      "learning_rate": 2.2745379256429728e-05,
+      "loss": 0.5856,
+      "step": 1482
+    },
+    {
+      "epoch": 0.41125901275651694,
+      "grad_norm": 0.1916564702987671,
+      "learning_rate": 2.2742214949065166e-05,
+      "loss": 0.5494,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4115363283416528,
+      "grad_norm": 0.2309359312057495,
+      "learning_rate": 2.2739048643173105e-05,
+      "loss": 0.575,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4118136439267887,
+      "grad_norm": 0.3025970757007599,
+      "learning_rate": 2.2735880339371373e-05,
+      "loss": 0.5717,
+      "step": 1485
+    },
+    {
+      "epoch": 0.4120909595119246,
+      "grad_norm": 0.19141024351119995,
+      "learning_rate": 2.27327100382782e-05,
+      "loss": 0.5743,
+      "step": 1486
+    },
+    {
+      "epoch": 0.4123682750970605,
+      "grad_norm": 0.19072705507278442,
+      "learning_rate": 2.272953774051218e-05,
+      "loss": 0.5498,
+      "step": 1487
+    },
+    {
+      "epoch": 0.41264559068219636,
+      "grad_norm": 0.23282243311405182,
+      "learning_rate": 2.2726363446692324e-05,
+      "loss": 0.5653,
+      "step": 1488
+    },
+    {
+      "epoch": 0.41292290626733225,
+      "grad_norm": 0.1904718577861786,
+      "learning_rate": 2.2723187157438015e-05,
+      "loss": 0.5557,
+      "step": 1489
+    },
+    {
+      "epoch": 0.41320022185246813,
+      "grad_norm": 0.18210548162460327,
+      "learning_rate": 2.2720008873369036e-05,
+      "loss": 0.5648,
+      "step": 1490
+    },
+    {
+      "epoch": 0.413477537437604,
+      "grad_norm": 0.1940220296382904,
+      "learning_rate": 2.271682859510555e-05,
+      "loss": 0.5824,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4137548530227399,
+      "grad_norm": 0.18632103502750397,
+      "learning_rate": 2.2713646323268113e-05,
+      "loss": 0.5694,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4140321686078758,
+      "grad_norm": 0.19417639076709747,
+      "learning_rate": 2.2710462058477676e-05,
+      "loss": 0.5385,
+      "step": 1493
+    },
+    {
+      "epoch": 0.41430948419301167,
+      "grad_norm": 0.1876698136329651,
+      "learning_rate": 2.270727580135557e-05,
+      "loss": 0.5441,
+      "step": 1494
+    },
+    {
+      "epoch": 0.41458679977814755,
+      "grad_norm": 0.19295859336853027,
+      "learning_rate": 2.270408755252352e-05,
+      "loss": 0.5587,
+      "step": 1495
+    },
+    {
+      "epoch": 0.41486411536328344,
+      "grad_norm": 0.1876155138015747,
+      "learning_rate": 2.2700897312603635e-05,
+      "loss": 0.5597,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4151414309484193,
+      "grad_norm": 0.18888817727565765,
+      "learning_rate": 2.2697705082218417e-05,
+      "loss": 0.58,
+      "step": 1497
+    },
+    {
+      "epoch": 0.4154187465335552,
+      "grad_norm": 0.1931556612253189,
+      "learning_rate": 2.2694510861990755e-05,
+      "loss": 0.5195,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4156960621186911,
+      "grad_norm": 0.20737329125404358,
+      "learning_rate": 2.2691314652543922e-05,
+      "loss": 0.5742,
+      "step": 1499
+    },
+    {
+      "epoch": 0.415973377703827,
+      "grad_norm": 0.18405389785766602,
+      "learning_rate": 2.268811645450159e-05,
+      "loss": 0.5572,
+      "step": 1500
+    },
+    {
+      "epoch": 0.41625069328896286,
+      "grad_norm": 0.2047012746334076,
+      "learning_rate": 2.2684916268487805e-05,
+      "loss": 0.5682,
+      "step": 1501
+    },
+    {
+      "epoch": 0.41652800887409874,
+      "grad_norm": 0.19159288704395294,
+      "learning_rate": 2.2681714095127016e-05,
+      "loss": 0.5642,
+      "step": 1502
+    },
+    {
+      "epoch": 0.4168053244592346,
+      "grad_norm": 0.19872340559959412,
+      "learning_rate": 2.2678509935044046e-05,
+      "loss": 0.5801,
+      "step": 1503
+    },
+    {
+      "epoch": 0.4170826400443705,
+      "grad_norm": 0.22284448146820068,
+      "learning_rate": 2.267530378886411e-05,
+      "loss": 0.5468,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4173599556295064,
+      "grad_norm": 0.18655410408973694,
+      "learning_rate": 2.2672095657212822e-05,
+      "loss": 0.5557,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4176372712146423,
+      "grad_norm": 0.1884729266166687,
+      "learning_rate": 2.266888554071616e-05,
+      "loss": 0.5641,
+      "step": 1506
+    },
+    {
+      "epoch": 0.41791458679977816,
+      "grad_norm": 0.1828029453754425,
+      "learning_rate": 2.2665673440000512e-05,
+      "loss": 0.5295,
+      "step": 1507
+    },
+    {
+      "epoch": 0.41819190238491405,
+      "grad_norm": 0.19014927744865417,
+      "learning_rate": 2.2662459355692645e-05,
+      "loss": 0.5385,
+      "step": 1508
+    },
+    {
+      "epoch": 0.41846921797004993,
+      "grad_norm": 0.20038922131061554,
+      "learning_rate": 2.26592432884197e-05,
+      "loss": 0.5763,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4187465335551858,
+      "grad_norm": 0.19587905704975128,
+      "learning_rate": 2.2656025238809233e-05,
+      "loss": 0.5642,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4190238491403217,
+      "grad_norm": 0.18529456853866577,
+      "learning_rate": 2.265280520748916e-05,
+      "loss": 0.5467,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4193011647254576,
+      "grad_norm": 0.19396060705184937,
+      "learning_rate": 2.26495831950878e-05,
+      "loss": 0.5756,
+      "step": 1512
+    },
+    {
+      "epoch": 0.41957848031059347,
+      "grad_norm": 0.19712162017822266,
+      "learning_rate": 2.2646359202233848e-05,
+      "loss": 0.5856,
+      "step": 1513
+    },
+    {
+      "epoch": 0.41985579589572936,
+      "grad_norm": 0.1929578334093094,
+      "learning_rate": 2.264313322955639e-05,
+      "loss": 0.5677,
+      "step": 1514
+    },
+    {
+      "epoch": 0.42013311148086524,
+      "grad_norm": 0.19393184781074524,
+      "learning_rate": 2.263990527768491e-05,
+      "loss": 0.5409,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4204104270660011,
+      "grad_norm": 0.19869175553321838,
+      "learning_rate": 2.2636675347249252e-05,
+      "loss": 0.5344,
+      "step": 1516
+    },
+    {
+      "epoch": 0.420687742651137,
+      "grad_norm": 0.19600443542003632,
+      "learning_rate": 2.263344343887967e-05,
+      "loss": 0.5627,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4209650582362729,
+      "grad_norm": 0.19098101556301117,
+      "learning_rate": 2.263020955320679e-05,
+      "loss": 0.5704,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4212423738214088,
+      "grad_norm": 0.20919351279735565,
+      "learning_rate": 2.2626973690861635e-05,
+      "loss": 0.55,
+      "step": 1519
+    },
+    {
+      "epoch": 0.42151968940654466,
+      "grad_norm": 0.21177563071250916,
+      "learning_rate": 2.2623735852475602e-05,
+      "loss": 0.5723,
+      "step": 1520
+    },
+    {
+      "epoch": 0.42179700499168055,
+      "grad_norm": 0.19463296234607697,
+      "learning_rate": 2.262049603868048e-05,
+      "loss": 0.5601,
+      "step": 1521
+    },
+    {
+      "epoch": 0.42207432057681643,
+      "grad_norm": 0.20798444747924805,
+      "learning_rate": 2.2617254250108445e-05,
+      "loss": 0.5606,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4223516361619523,
+      "grad_norm": 0.20048515498638153,
+      "learning_rate": 2.2614010487392053e-05,
+      "loss": 0.5628,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4226289517470882,
+      "grad_norm": 0.1931045949459076,
+      "learning_rate": 2.2610764751164253e-05,
+      "loss": 0.5662,
+      "step": 1524
+    },
+    {
+      "epoch": 0.4229062673322241,
+      "grad_norm": 0.19520334899425507,
+      "learning_rate": 2.2607517042058367e-05,
+      "loss": 0.5552,
+      "step": 1525
+    },
+    {
+      "epoch": 0.42318358291735997,
+      "grad_norm": 0.1984749138355255,
+      "learning_rate": 2.2604267360708113e-05,
+      "loss": 0.5672,
+      "step": 1526
+    },
+    {
+      "epoch": 0.42346089850249585,
+      "grad_norm": 0.19392019510269165,
+      "learning_rate": 2.2601015707747585e-05,
+      "loss": 0.5689,
+      "step": 1527
+    },
+    {
+      "epoch": 0.42373821408763174,
+      "grad_norm": 0.1880037933588028,
+      "learning_rate": 2.2597762083811276e-05,
+      "loss": 0.5606,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4240155296727676,
+      "grad_norm": 0.19556698203086853,
+      "learning_rate": 2.259450648953405e-05,
+      "loss": 0.5626,
+      "step": 1529
+    },
+    {
+      "epoch": 0.4242928452579035,
+      "grad_norm": 0.18473385274410248,
+      "learning_rate": 2.2591248925551156e-05,
+      "loss": 0.541,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4245701608430394,
+      "grad_norm": 0.2530158460140228,
+      "learning_rate": 2.2587989392498237e-05,
+      "loss": 0.5429,
+      "step": 1531
+    },
+    {
+      "epoch": 0.4248474764281753,
+      "grad_norm": 0.22607286274433136,
+      "learning_rate": 2.258472789101131e-05,
+      "loss": 0.5578,
+      "step": 1532
+    },
+    {
+      "epoch": 0.42512479201331116,
+      "grad_norm": 0.19067604839801788,
+      "learning_rate": 2.258146442172678e-05,
+      "loss": 0.5474,
+      "step": 1533
+    },
+    {
+      "epoch": 0.42540210759844704,
+      "grad_norm": 0.18085655570030212,
+      "learning_rate": 2.257819898528144e-05,
+      "loss": 0.5506,
+      "step": 1534
+    },
+    {
+      "epoch": 0.42567942318358293,
+      "grad_norm": 0.18807153403759003,
+      "learning_rate": 2.257493158231246e-05,
+      "loss": 0.5461,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4259567387687188,
+      "grad_norm": 0.18786919116973877,
+      "learning_rate": 2.25716622134574e-05,
+      "loss": 0.5781,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4262340543538547,
+      "grad_norm": 0.19066756963729858,
+      "learning_rate": 2.2568390879354195e-05,
+      "loss": 0.5378,
+      "step": 1537
+    },
+    {
+      "epoch": 0.4265113699389906,
+      "grad_norm": 0.1895960569381714,
+      "learning_rate": 2.2565117580641175e-05,
+      "loss": 0.5661,
+      "step": 1538
+    },
+    {
+      "epoch": 0.42678868552412647,
+      "grad_norm": 0.21506737172603607,
+      "learning_rate": 2.2561842317957045e-05,
+      "loss": 0.661,
+      "step": 1539
+    },
+    {
+      "epoch": 0.42706600110926235,
+      "grad_norm": 0.1904166042804718,
+      "learning_rate": 2.2558565091940895e-05,
+      "loss": 0.5643,
+      "step": 1540
+    },
+    {
+      "epoch": 0.42734331669439823,
+      "grad_norm": 0.18517723679542542,
+      "learning_rate": 2.2555285903232197e-05,
+      "loss": 0.5509,
+      "step": 1541
+    },
+    {
+      "epoch": 0.4276206322795341,
+      "grad_norm": 0.2003047913312912,
+      "learning_rate": 2.2552004752470814e-05,
+      "loss": 0.5487,
+      "step": 1542
+    },
+    {
+      "epoch": 0.42789794786467,
+      "grad_norm": 0.19767479598522186,
+      "learning_rate": 2.2548721640296976e-05,
+      "loss": 0.5534,
+      "step": 1543
+    },
+    {
+      "epoch": 0.4281752634498059,
+      "grad_norm": 0.1890031397342682,
+      "learning_rate": 2.2545436567351312e-05,
+      "loss": 0.5762,
+      "step": 1544
+    },
+    {
+      "epoch": 0.42845257903494177,
+      "grad_norm": 0.22215117514133453,
+      "learning_rate": 2.2542149534274827e-05,
+      "loss": 0.5433,
+      "step": 1545
+    },
+    {
+      "epoch": 0.42872989462007766,
+      "grad_norm": 0.17825603485107422,
+      "learning_rate": 2.2538860541708902e-05,
+      "loss": 0.5724,
+      "step": 1546
+    },
+    {
+      "epoch": 0.42900721020521354,
+      "grad_norm": 0.1967187076807022,
+      "learning_rate": 2.2535569590295313e-05,
+      "loss": 0.5632,
+      "step": 1547
+    },
+    {
+      "epoch": 0.4292845257903494,
+      "grad_norm": 0.18680186569690704,
+      "learning_rate": 2.253227668067621e-05,
+      "loss": 0.5704,
+      "step": 1548
+    },
+    {
+      "epoch": 0.4295618413754853,
+      "grad_norm": 0.17610576748847961,
+      "learning_rate": 2.2528981813494127e-05,
+      "loss": 0.5295,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4298391569606212,
+      "grad_norm": 0.19055548310279846,
+      "learning_rate": 2.2525684989391975e-05,
+      "loss": 0.5651,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4301164725457571,
+      "grad_norm": 0.19191017746925354,
+      "learning_rate": 2.2522386209013062e-05,
+      "loss": 0.5366,
+      "step": 1551
+    },
+    {
+      "epoch": 0.43039378813089296,
+      "grad_norm": 0.20109041035175323,
+      "learning_rate": 2.2519085473001055e-05,
+      "loss": 0.5508,
+      "step": 1552
+    },
+    {
+      "epoch": 0.43067110371602885,
+      "grad_norm": 0.18693628907203674,
+      "learning_rate": 2.2515782782000027e-05,
+      "loss": 0.5603,
+      "step": 1553
+    },
+    {
+      "epoch": 0.43094841930116473,
+      "grad_norm": 0.18805643916130066,
+      "learning_rate": 2.2512478136654412e-05,
+      "loss": 0.5197,
+      "step": 1554
+    },
+    {
+      "epoch": 0.4312257348863006,
+      "grad_norm": 0.19016428291797638,
+      "learning_rate": 2.2509171537609042e-05,
+      "loss": 0.5719,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4315030504714365,
+      "grad_norm": 0.18083550035953522,
+      "learning_rate": 2.2505862985509112e-05,
+      "loss": 0.5502,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4317803660565724,
+      "grad_norm": 0.18382135033607483,
+      "learning_rate": 2.2502552481000218e-05,
+      "loss": 0.5437,
+      "step": 1557
+    },
+    {
+      "epoch": 0.43205768164170827,
+      "grad_norm": 0.20305176079273224,
+      "learning_rate": 2.2499240024728316e-05,
+      "loss": 0.5559,
+      "step": 1558
+    },
+    {
+      "epoch": 0.43233499722684415,
+      "grad_norm": 0.19593636691570282,
+      "learning_rate": 2.2495925617339765e-05,
+      "loss": 0.5695,
+      "step": 1559
+    },
+    {
+      "epoch": 0.43261231281198004,
+      "grad_norm": 0.18708528578281403,
+      "learning_rate": 2.2492609259481283e-05,
+      "loss": 0.5828,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4328896283971159,
+      "grad_norm": 0.20516738295555115,
+      "learning_rate": 2.248929095179999e-05,
+      "loss": 0.5691,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4331669439822518,
+      "grad_norm": 0.19292466342449188,
+      "learning_rate": 2.248597069494337e-05,
+      "loss": 0.5567,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4334442595673877,
+      "grad_norm": 0.18612238764762878,
+      "learning_rate": 2.2482648489559296e-05,
+      "loss": 0.5497,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4337215751525236,
+      "grad_norm": 0.18284805119037628,
+      "learning_rate": 2.2479324336296016e-05,
+      "loss": 0.5305,
+      "step": 1564
+    },
+    {
+      "epoch": 0.43399889073765946,
+      "grad_norm": 0.19658204913139343,
+      "learning_rate": 2.247599823580216e-05,
+      "loss": 0.5631,
+      "step": 1565
+    },
+    {
+      "epoch": 0.43427620632279534,
+      "grad_norm": 0.19752554595470428,
+      "learning_rate": 2.2472670188726737e-05,
+      "loss": 0.5647,
+      "step": 1566
+    },
+    {
+      "epoch": 0.43455352190793123,
+      "grad_norm": 0.18531207740306854,
+      "learning_rate": 2.246934019571914e-05,
+      "loss": 0.5667,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4348308374930671,
+      "grad_norm": 0.18967878818511963,
+      "learning_rate": 2.2466008257429142e-05,
+      "loss": 0.5572,
+      "step": 1568
+    },
+    {
+      "epoch": 0.435108153078203,
+      "grad_norm": 0.20551453530788422,
+      "learning_rate": 2.2462674374506886e-05,
+      "loss": 0.5554,
+      "step": 1569
+    },
+    {
+      "epoch": 0.4353854686633389,
+      "grad_norm": 0.1888595074415207,
+      "learning_rate": 2.2459338547602905e-05,
+      "loss": 0.5986,
+      "step": 1570
+    },
+    {
+      "epoch": 0.43566278424847477,
+      "grad_norm": 0.1824692189693451,
+      "learning_rate": 2.2456000777368102e-05,
+      "loss": 0.5223,
+      "step": 1571
+    },
+    {
+      "epoch": 0.43594009983361065,
+      "grad_norm": 0.189620703458786,
+      "learning_rate": 2.245266106445377e-05,
+      "loss": 0.5465,
+      "step": 1572
+    },
+    {
+      "epoch": 0.43621741541874653,
+      "grad_norm": 0.20219853520393372,
+      "learning_rate": 2.2449319409511574e-05,
+      "loss": 0.5655,
+      "step": 1573
+    },
+    {
+      "epoch": 0.4364947310038824,
+      "grad_norm": 0.18667539954185486,
+      "learning_rate": 2.244597581319356e-05,
+      "loss": 0.5307,
+      "step": 1574
+    },
+    {
+      "epoch": 0.4367720465890183,
+      "grad_norm": 0.19490988552570343,
+      "learning_rate": 2.2442630276152148e-05,
+      "loss": 0.5666,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4370493621741542,
+      "grad_norm": 0.18786631524562836,
+      "learning_rate": 2.2439282799040146e-05,
+      "loss": 0.535,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4373266777592901,
+      "grad_norm": 0.1902604103088379,
+      "learning_rate": 2.2435933382510735e-05,
+      "loss": 0.5362,
+      "step": 1577
+    },
+    {
+      "epoch": 0.43760399334442596,
+      "grad_norm": 0.17974701523780823,
+      "learning_rate": 2.2432582027217473e-05,
+      "loss": 0.5538,
+      "step": 1578
+    },
+    {
+      "epoch": 0.43788130892956184,
+      "grad_norm": 0.1953812539577484,
+      "learning_rate": 2.2429228733814294e-05,
+      "loss": 0.5584,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4381586245146977,
+      "grad_norm": 0.17894363403320312,
+      "learning_rate": 2.2425873502955524e-05,
+      "loss": 0.5667,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4384359400998336,
+      "grad_norm": 0.2096855342388153,
+      "learning_rate": 2.2422516335295852e-05,
+      "loss": 0.5634,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4387132556849695,
+      "grad_norm": 0.1870642900466919,
+      "learning_rate": 2.241915723149035e-05,
+      "loss": 0.5691,
+      "step": 1582
+    },
+    {
+      "epoch": 0.4389905712701054,
+      "grad_norm": 0.2048119157552719,
+      "learning_rate": 2.241579619219447e-05,
+      "loss": 0.5645,
+      "step": 1583
+    },
+    {
+      "epoch": 0.43926788685524126,
+      "grad_norm": 0.19508272409439087,
+      "learning_rate": 2.2412433218064037e-05,
+      "loss": 0.5543,
+      "step": 1584
+    },
+    {
+      "epoch": 0.43954520244037715,
+      "grad_norm": 0.1895490139722824,
+      "learning_rate": 2.240906830975526e-05,
+      "loss": 0.5522,
+      "step": 1585
+    },
+    {
+      "epoch": 0.43982251802551303,
+      "grad_norm": 0.18243685364723206,
+      "learning_rate": 2.240570146792472e-05,
+      "loss": 0.5695,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4400998336106489,
+      "grad_norm": 0.18209905922412872,
+      "learning_rate": 2.2402332693229377e-05,
+      "loss": 0.5447,
+      "step": 1587
+    },
+    {
+      "epoch": 0.4403771491957848,
+      "grad_norm": 0.1850498467683792,
+      "learning_rate": 2.2398961986326567e-05,
+      "loss": 0.5289,
+      "step": 1588
+    },
+    {
+      "epoch": 0.4406544647809207,
+      "grad_norm": 0.19210520386695862,
+      "learning_rate": 2.2395589347874005e-05,
+      "loss": 0.5792,
+      "step": 1589
+    },
+    {
+      "epoch": 0.44093178036605657,
+      "grad_norm": 0.19306769967079163,
+      "learning_rate": 2.239221477852978e-05,
+      "loss": 0.5771,
+      "step": 1590
+    },
+    {
+      "epoch": 0.44120909595119245,
+      "grad_norm": 0.18809685111045837,
+      "learning_rate": 2.2388838278952367e-05,
+      "loss": 0.5648,
+      "step": 1591
+    },
+    {
+      "epoch": 0.44148641153632834,
+      "grad_norm": 0.19211184978485107,
+      "learning_rate": 2.2385459849800606e-05,
+      "loss": 0.5867,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4417637271214642,
+      "grad_norm": 0.2126152366399765,
+      "learning_rate": 2.2382079491733715e-05,
+      "loss": 0.5705,
+      "step": 1593
+    },
+    {
+      "epoch": 0.4420410427066001,
+      "grad_norm": 0.18234777450561523,
+      "learning_rate": 2.23786972054113e-05,
+      "loss": 0.5373,
+      "step": 1594
+    },
+    {
+      "epoch": 0.442318358291736,
+      "grad_norm": 0.18619130551815033,
+      "learning_rate": 2.2375312991493324e-05,
+      "loss": 0.5525,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4425956738768719,
+      "grad_norm": 0.19054090976715088,
+      "learning_rate": 2.237192685064014e-05,
+      "loss": 0.5606,
+      "step": 1596
+    },
+    {
+      "epoch": 0.44287298946200776,
+      "grad_norm": 0.19798876345157623,
+      "learning_rate": 2.236853878351248e-05,
+      "loss": 0.5389,
+      "step": 1597
+    },
+    {
+      "epoch": 0.44315030504714364,
+      "grad_norm": 0.20198500156402588,
+      "learning_rate": 2.2365148790771442e-05,
+      "loss": 0.577,
+      "step": 1598
+    },
+    {
+      "epoch": 0.44342762063227953,
+      "grad_norm": 0.22898751497268677,
+      "learning_rate": 2.2361756873078502e-05,
+      "loss": 0.5516,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4437049362174154,
+      "grad_norm": 0.1942586600780487,
+      "learning_rate": 2.2358363031095513e-05,
+      "loss": 0.5583,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4439822518025513,
+      "grad_norm": 0.18665020167827606,
+      "learning_rate": 2.23549672654847e-05,
+      "loss": 0.562,
+      "step": 1601
+    },
+    {
+      "epoch": 0.4442595673876872,
+      "grad_norm": 0.20378893613815308,
+      "learning_rate": 2.2351569576908675e-05,
+      "loss": 0.5783,
+      "step": 1602
+    },
+    {
+      "epoch": 0.44453688297282307,
+      "grad_norm": 0.21167699992656708,
+      "learning_rate": 2.2348169966030416e-05,
+      "loss": 0.5453,
+      "step": 1603
+    },
+    {
+      "epoch": 0.44481419855795895,
+      "grad_norm": 0.19410103559494019,
+      "learning_rate": 2.234476843351327e-05,
+      "loss": 0.5527,
+      "step": 1604
+    },
+    {
+      "epoch": 0.44509151414309484,
+      "grad_norm": 0.17932045459747314,
+      "learning_rate": 2.2341364980020973e-05,
+      "loss": 0.568,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4453688297282307,
+      "grad_norm": 0.2007569819688797,
+      "learning_rate": 2.2337959606217624e-05,
+      "loss": 0.5699,
+      "step": 1606
+    },
+    {
+      "epoch": 0.4456461453133666,
+      "grad_norm": 0.18869513273239136,
+      "learning_rate": 2.2334552312767705e-05,
+      "loss": 0.5506,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4459234608985025,
+      "grad_norm": 0.1867865025997162,
+      "learning_rate": 2.2331143100336072e-05,
+      "loss": 0.5758,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4462007764836384,
+      "grad_norm": 0.19275221228599548,
+      "learning_rate": 2.2327731969587947e-05,
+      "loss": 0.5408,
+      "step": 1609
+    },
+    {
+      "epoch": 0.44647809206877426,
+      "grad_norm": 0.19611743092536926,
+      "learning_rate": 2.2324318921188932e-05,
+      "loss": 0.5974,
+      "step": 1610
+    },
+    {
+      "epoch": 0.44675540765391014,
+      "grad_norm": 0.20041632652282715,
+      "learning_rate": 2.2320903955805e-05,
+      "loss": 0.5598,
+      "step": 1611
+    },
+    {
+      "epoch": 0.447032723239046,
+      "grad_norm": 0.18512395024299622,
+      "learning_rate": 2.2317487074102514e-05,
+      "loss": 0.5661,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4473100388241819,
+      "grad_norm": 0.1952909678220749,
+      "learning_rate": 2.2314068276748188e-05,
+      "loss": 0.5635,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4475873544093178,
+      "grad_norm": 0.1981881707906723,
+      "learning_rate": 2.231064756440912e-05,
+      "loss": 0.5601,
+      "step": 1614
+    },
+    {
+      "epoch": 0.4478646699944537,
+      "grad_norm": 0.1989242285490036,
+      "learning_rate": 2.230722493775279e-05,
+      "loss": 0.5635,
+      "step": 1615
+    },
+    {
+      "epoch": 0.44814198557958956,
+      "grad_norm": 0.18769480288028717,
+      "learning_rate": 2.2303800397447034e-05,
+      "loss": 0.5589,
+      "step": 1616
+    },
+    {
+      "epoch": 0.44841930116472545,
+      "grad_norm": 0.19833675026893616,
+      "learning_rate": 2.230037394416007e-05,
+      "loss": 0.5622,
+      "step": 1617
+    },
+    {
+      "epoch": 0.44869661674986133,
+      "grad_norm": 0.19801415503025055,
+      "learning_rate": 2.2296945578560498e-05,
+      "loss": 0.5862,
+      "step": 1618
+    },
+    {
+      "epoch": 0.4489739323349972,
+      "grad_norm": 0.1936810463666916,
+      "learning_rate": 2.2293515301317274e-05,
+      "loss": 0.5452,
+      "step": 1619
+    },
+    {
+      "epoch": 0.4492512479201331,
+      "grad_norm": 0.20558279752731323,
+      "learning_rate": 2.2290083113099748e-05,
+      "loss": 0.5573,
+      "step": 1620
+    },
+    {
+      "epoch": 0.449528563505269,
+      "grad_norm": 0.2063780575990677,
+      "learning_rate": 2.2286649014577615e-05,
+      "loss": 0.6017,
+      "step": 1621
+    },
+    {
+      "epoch": 0.44980587909040487,
+      "grad_norm": 0.18592418730258942,
+      "learning_rate": 2.2283213006420973e-05,
+      "loss": 0.5421,
+      "step": 1622
+    },
+    {
+      "epoch": 0.45008319467554075,
+      "grad_norm": 0.19346946477890015,
+      "learning_rate": 2.227977508930027e-05,
+      "loss": 0.53,
+      "step": 1623
+    },
+    {
+      "epoch": 0.45036051026067664,
+      "grad_norm": 0.2029414027929306,
+      "learning_rate": 2.2276335263886336e-05,
+      "loss": 0.5808,
+      "step": 1624
+    },
+    {
+      "epoch": 0.4506378258458125,
+      "grad_norm": 0.21214988827705383,
+      "learning_rate": 2.2272893530850373e-05,
+      "loss": 0.564,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4509151414309484,
+      "grad_norm": 0.19748808443546295,
+      "learning_rate": 2.2269449890863956e-05,
+      "loss": 0.5562,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4511924570160843,
+      "grad_norm": 0.19283835589885712,
+      "learning_rate": 2.2266004344599028e-05,
+      "loss": 0.5511,
+      "step": 1627
+    },
+    {
+      "epoch": 0.4514697726012202,
+      "grad_norm": 0.1917610466480255,
+      "learning_rate": 2.2262556892727904e-05,
+      "loss": 0.5744,
+      "step": 1628
+    },
+    {
+      "epoch": 0.45174708818635606,
+      "grad_norm": 0.20705603063106537,
+      "learning_rate": 2.225910753592328e-05,
+      "loss": 0.5786,
+      "step": 1629
+    },
+    {
+      "epoch": 0.45202440377149194,
+      "grad_norm": 0.19348299503326416,
+      "learning_rate": 2.225565627485821e-05,
+      "loss": 0.5899,
+      "step": 1630
+    },
+    {
+      "epoch": 0.45230171935662783,
+      "grad_norm": 0.19649489223957062,
+      "learning_rate": 2.2252203110206134e-05,
+      "loss": 0.5317,
+      "step": 1631
+    },
+    {
+      "epoch": 0.4525790349417637,
+      "grad_norm": 0.18169252574443817,
+      "learning_rate": 2.224874804264085e-05,
+      "loss": 0.5243,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4528563505268996,
+      "grad_norm": 0.188733771443367,
+      "learning_rate": 2.224529107283653e-05,
+      "loss": 0.5376,
+      "step": 1633
+    },
+    {
+      "epoch": 0.4531336661120355,
+      "grad_norm": 0.18059036135673523,
+      "learning_rate": 2.2241832201467727e-05,
+      "loss": 0.5603,
+      "step": 1634
+    },
+    {
+      "epoch": 0.45341098169717137,
+      "grad_norm": 0.19139643013477325,
+      "learning_rate": 2.223837142920936e-05,
+      "loss": 0.5603,
+      "step": 1635
+    },
+    {
+      "epoch": 0.45368829728230725,
+      "grad_norm": 0.19182981550693512,
+      "learning_rate": 2.2234908756736712e-05,
+      "loss": 0.5805,
+      "step": 1636
+    },
+    {
+      "epoch": 0.45396561286744314,
+      "grad_norm": 0.187539204955101,
+      "learning_rate": 2.223144418472544e-05,
+      "loss": 0.5546,
+      "step": 1637
+    },
+    {
+      "epoch": 0.454242928452579,
+      "grad_norm": 0.19611942768096924,
+      "learning_rate": 2.2227977713851587e-05,
+      "loss": 0.5349,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4545202440377149,
+      "grad_norm": 0.19672465324401855,
+      "learning_rate": 2.2224509344791536e-05,
+      "loss": 0.5342,
+      "step": 1639
+    },
+    {
+      "epoch": 0.4547975596228508,
+      "grad_norm": 0.17679478228092194,
+      "learning_rate": 2.222103907822207e-05,
+      "loss": 0.5473,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4550748752079867,
+      "grad_norm": 0.1936340481042862,
+      "learning_rate": 2.2217566914820322e-05,
+      "loss": 0.5543,
+      "step": 1641
+    },
+    {
+      "epoch": 0.45535219079312256,
+      "grad_norm": 0.18610352277755737,
+      "learning_rate": 2.2214092855263813e-05,
+      "loss": 0.5412,
+      "step": 1642
+    },
+    {
+      "epoch": 0.45562950637825844,
+      "grad_norm": 0.18969598412513733,
+      "learning_rate": 2.2210616900230412e-05,
+      "loss": 0.5707,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4559068219633943,
+      "grad_norm": 0.18808507919311523,
+      "learning_rate": 2.220713905039838e-05,
+      "loss": 0.5288,
+      "step": 1644
+    },
+    {
+      "epoch": 0.4561841375485302,
+      "grad_norm": 0.19103705883026123,
+      "learning_rate": 2.220365930644633e-05,
+      "loss": 0.5925,
+      "step": 1645
+    },
+    {
+      "epoch": 0.4564614531336661,
+      "grad_norm": 0.1837342530488968,
+      "learning_rate": 2.2200177669053258e-05,
+      "loss": 0.5893,
+      "step": 1646
+    },
+    {
+      "epoch": 0.456738768718802,
+      "grad_norm": 0.1928233504295349,
+      "learning_rate": 2.2196694138898517e-05,
+      "loss": 0.5445,
+      "step": 1647
+    },
+    {
+      "epoch": 0.45701608430393786,
+      "grad_norm": 0.195438414812088,
+      "learning_rate": 2.2193208716661846e-05,
+      "loss": 0.5561,
+      "step": 1648
+    },
+    {
+      "epoch": 0.45729339988907375,
+      "grad_norm": 0.18883578479290009,
+      "learning_rate": 2.2189721403023334e-05,
+      "loss": 0.5463,
+      "step": 1649
+    },
+    {
+      "epoch": 0.45757071547420963,
+      "grad_norm": 0.19664356112480164,
+      "learning_rate": 2.2186232198663455e-05,
+      "loss": 0.576,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4578480310593455,
+      "grad_norm": 0.5387895703315735,
+      "learning_rate": 2.218274110426304e-05,
+      "loss": 0.562,
+      "step": 1651
+    },
+    {
+      "epoch": 0.4581253466444814,
+      "grad_norm": 0.18329556286334991,
+      "learning_rate": 2.21792481205033e-05,
+      "loss": 0.5737,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4584026622296173,
+      "grad_norm": 0.20161500573158264,
+      "learning_rate": 2.21757532480658e-05,
+      "loss": 0.5711,
+      "step": 1653
+    },
+    {
+      "epoch": 0.45867997781475317,
+      "grad_norm": 0.20053423941135406,
+      "learning_rate": 2.2172256487632488e-05,
+      "loss": 0.5842,
+      "step": 1654
+    },
+    {
+      "epoch": 0.45895729339988905,
+      "grad_norm": 0.1900150030851364,
+      "learning_rate": 2.2168757839885672e-05,
+      "loss": 0.5702,
+      "step": 1655
+    },
+    {
+      "epoch": 0.45923460898502494,
+      "grad_norm": 0.19815103709697723,
+      "learning_rate": 2.2165257305508035e-05,
+      "loss": 0.5661,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4595119245701608,
+      "grad_norm": 0.18877148628234863,
+      "learning_rate": 2.2161754885182623e-05,
+      "loss": 0.5278,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4597892401552967,
+      "grad_norm": 0.1913890838623047,
+      "learning_rate": 2.215825057959285e-05,
+      "loss": 0.5342,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4600665557404326,
+      "grad_norm": 0.18911883234977722,
+      "learning_rate": 2.2154744389422493e-05,
+      "loss": 0.5473,
+      "step": 1659
+    },
+    {
+      "epoch": 0.4603438713255685,
+      "grad_norm": 0.2087319940328598,
+      "learning_rate": 2.2151236315355714e-05,
+      "loss": 0.5839,
+      "step": 1660
+    },
+    {
+      "epoch": 0.46062118691070436,
+      "grad_norm": 0.19037386775016785,
+      "learning_rate": 2.214772635807702e-05,
+      "loss": 0.5518,
+      "step": 1661
+    },
+    {
+      "epoch": 0.46089850249584025,
+      "grad_norm": 0.1889956146478653,
+      "learning_rate": 2.2144214518271307e-05,
+      "loss": 0.5527,
+      "step": 1662
+    },
+    {
+      "epoch": 0.46117581808097613,
+      "grad_norm": 0.24889707565307617,
+      "learning_rate": 2.214070079662382e-05,
+      "loss": 0.5766,
+      "step": 1663
+    },
+    {
+      "epoch": 0.461453133666112,
+      "grad_norm": 0.19265350699424744,
+      "learning_rate": 2.213718519382018e-05,
+      "loss": 0.5721,
+      "step": 1664
+    },
+    {
+      "epoch": 0.4617304492512479,
+      "grad_norm": 0.19039899110794067,
+      "learning_rate": 2.213366771054638e-05,
+      "loss": 0.5678,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4620077648363838,
+      "grad_norm": 0.203168585896492,
+      "learning_rate": 2.2130148347488773e-05,
+      "loss": 0.5486,
+      "step": 1666
+    },
+    {
+      "epoch": 0.46228508042151967,
+      "grad_norm": 0.19051045179367065,
+      "learning_rate": 2.2126627105334073e-05,
+      "loss": 0.544,
+      "step": 1667
+    },
+    {
+      "epoch": 0.46256239600665555,
+      "grad_norm": 0.18888403475284576,
+      "learning_rate": 2.212310398476937e-05,
+      "loss": 0.5436,
+      "step": 1668
+    },
+    {
+      "epoch": 0.46283971159179144,
+      "grad_norm": 0.1800607144832611,
+      "learning_rate": 2.2119578986482127e-05,
+      "loss": 0.5659,
+      "step": 1669
+    },
+    {
+      "epoch": 0.4631170271769273,
+      "grad_norm": 0.19722715020179749,
+      "learning_rate": 2.211605211116015e-05,
+      "loss": 0.5582,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4633943427620632,
+      "grad_norm": 0.19088581204414368,
+      "learning_rate": 2.2112523359491637e-05,
+      "loss": 0.5471,
+      "step": 1671
+    },
+    {
+      "epoch": 0.4636716583471991,
+      "grad_norm": 0.19284148514270782,
+      "learning_rate": 2.210899273216514e-05,
+      "loss": 0.5408,
+      "step": 1672
+    },
+    {
+      "epoch": 0.463948973932335,
+      "grad_norm": 0.1880495250225067,
+      "learning_rate": 2.2105460229869574e-05,
+      "loss": 0.561,
+      "step": 1673
+    },
+    {
+      "epoch": 0.46422628951747086,
+      "grad_norm": 0.186056986451149,
+      "learning_rate": 2.2101925853294226e-05,
+      "loss": 0.5811,
+      "step": 1674
+    },
+    {
+      "epoch": 0.46450360510260674,
+      "grad_norm": 0.1980651617050171,
+      "learning_rate": 2.2098389603128744e-05,
+      "loss": 0.5456,
+      "step": 1675
+    },
+    {
+      "epoch": 0.4647809206877426,
+      "grad_norm": 0.2039021998643875,
+      "learning_rate": 2.2094851480063143e-05,
+      "loss": 0.6018,
+      "step": 1676
+    },
+    {
+      "epoch": 0.4650582362728785,
+      "grad_norm": 0.19567370414733887,
+      "learning_rate": 2.2091311484787815e-05,
+      "loss": 0.5499,
+      "step": 1677
+    },
+    {
+      "epoch": 0.4653355518580144,
+      "grad_norm": 0.19234254956245422,
+      "learning_rate": 2.208776961799349e-05,
+      "loss": 0.5522,
+      "step": 1678
+    },
+    {
+      "epoch": 0.4656128674431503,
+      "grad_norm": 0.1954323649406433,
+      "learning_rate": 2.20842258803713e-05,
+      "loss": 0.5702,
+      "step": 1679
+    },
+    {
+      "epoch": 0.46589018302828616,
+      "grad_norm": 0.1974237710237503,
+      "learning_rate": 2.20806802726127e-05,
+      "loss": 0.5604,
+      "step": 1680
+    },
+    {
+      "epoch": 0.46616749861342205,
+      "grad_norm": 0.19580432772636414,
+      "learning_rate": 2.2077132795409552e-05,
+      "loss": 0.5184,
+      "step": 1681
+    },
+    {
+      "epoch": 0.46644481419855793,
+      "grad_norm": 0.19890138506889343,
+      "learning_rate": 2.207358344945405e-05,
+      "loss": 0.5767,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4667221297836938,
+      "grad_norm": 0.18458129465579987,
+      "learning_rate": 2.2070032235438776e-05,
+      "loss": 0.5316,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4669994453688297,
+      "grad_norm": 0.1938208043575287,
+      "learning_rate": 2.206647915405665e-05,
+      "loss": 0.5809,
+      "step": 1684
+    },
+    {
+      "epoch": 0.4672767609539656,
+      "grad_norm": 0.20221275091171265,
+      "learning_rate": 2.206292420600099e-05,
+      "loss": 0.5731,
+      "step": 1685
+    },
+    {
+      "epoch": 0.46755407653910147,
+      "grad_norm": 0.194106787443161,
+      "learning_rate": 2.205936739196545e-05,
+      "loss": 0.5624,
+      "step": 1686
+    },
+    {
+      "epoch": 0.46783139212423736,
+      "grad_norm": 0.1918519288301468,
+      "learning_rate": 2.205580871264406e-05,
+      "loss": 0.535,
+      "step": 1687
+    },
+    {
+      "epoch": 0.46810870770937324,
+      "grad_norm": 0.19345992803573608,
+      "learning_rate": 2.2052248168731216e-05,
+      "loss": 0.5686,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4683860232945091,
+      "grad_norm": 0.20070746541023254,
+      "learning_rate": 2.2048685760921674e-05,
+      "loss": 0.567,
+      "step": 1689
+    },
+    {
+      "epoch": 0.468663338879645,
+      "grad_norm": 0.1972619593143463,
+      "learning_rate": 2.204512148991055e-05,
+      "loss": 0.555,
+      "step": 1690
+    },
+    {
+      "epoch": 0.46894065446478095,
+      "grad_norm": 0.19085273146629333,
+      "learning_rate": 2.2041555356393327e-05,
+      "loss": 0.5985,
+      "step": 1691
+    },
+    {
+      "epoch": 0.46921797004991683,
+      "grad_norm": 0.20294739305973053,
+      "learning_rate": 2.2037987361065855e-05,
+      "loss": 0.5762,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4694952856350527,
+      "grad_norm": 0.1896994560956955,
+      "learning_rate": 2.203441750462435e-05,
+      "loss": 0.5857,
+      "step": 1693
+    },
+    {
+      "epoch": 0.4697726012201886,
+      "grad_norm": 0.19331398606300354,
+      "learning_rate": 2.2030845787765377e-05,
+      "loss": 0.5654,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4700499168053245,
+      "grad_norm": 0.20668423175811768,
+      "learning_rate": 2.2027272211185875e-05,
+      "loss": 0.5812,
+      "step": 1695
+    },
+    {
+      "epoch": 0.47032723239046037,
+      "grad_norm": 0.1910814493894577,
+      "learning_rate": 2.2023696775583146e-05,
+      "loss": 0.5479,
+      "step": 1696
+    },
+    {
+      "epoch": 0.47060454797559625,
+      "grad_norm": 0.1902536004781723,
+      "learning_rate": 2.2020119481654848e-05,
+      "loss": 0.5647,
+      "step": 1697
+    },
+    {
+      "epoch": 0.47088186356073214,
+      "grad_norm": 0.1955711394548416,
+      "learning_rate": 2.201654033009901e-05,
+      "loss": 0.581,
+      "step": 1698
+    },
+    {
+      "epoch": 0.471159179145868,
+      "grad_norm": 0.1895892173051834,
+      "learning_rate": 2.2012959321614018e-05,
+      "loss": 0.5658,
+      "step": 1699
+    },
+    {
+      "epoch": 0.4714364947310039,
+      "grad_norm": 0.1803213357925415,
+      "learning_rate": 2.2009376456898622e-05,
+      "loss": 0.558,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4717138103161398,
+      "grad_norm": 0.19255508482456207,
+      "learning_rate": 2.200579173665193e-05,
+      "loss": 0.5649,
+      "step": 1701
+    },
+    {
+      "epoch": 0.4719911259012757,
+      "grad_norm": 0.20155562460422516,
+      "learning_rate": 2.2002205161573426e-05,
+      "loss": 0.5592,
+      "step": 1702
+    },
+    {
+      "epoch": 0.47226844148641156,
+      "grad_norm": 0.20530745387077332,
+      "learning_rate": 2.1998616732362935e-05,
+      "loss": 0.5677,
+      "step": 1703
+    },
+    {
+      "epoch": 0.47254575707154745,
+      "grad_norm": 0.2158524990081787,
+      "learning_rate": 2.1995026449720657e-05,
+      "loss": 0.5476,
+      "step": 1704
+    },
+    {
+      "epoch": 0.47282307265668333,
+      "grad_norm": 0.2383381724357605,
+      "learning_rate": 2.1991434314347155e-05,
+      "loss": 0.5413,
+      "step": 1705
+    },
+    {
+      "epoch": 0.4731003882418192,
+      "grad_norm": 0.19787725806236267,
+      "learning_rate": 2.1987840326943343e-05,
+      "loss": 0.5637,
+      "step": 1706
+    },
+    {
+      "epoch": 0.4733777038269551,
+      "grad_norm": 0.18500499427318573,
+      "learning_rate": 2.1984244488210508e-05,
+      "loss": 0.5333,
+      "step": 1707
+    },
+    {
+      "epoch": 0.473655019412091,
+      "grad_norm": 0.19429421424865723,
+      "learning_rate": 2.1980646798850295e-05,
+      "loss": 0.5611,
+      "step": 1708
+    },
+    {
+      "epoch": 0.47393233499722687,
+      "grad_norm": 0.18553559482097626,
+      "learning_rate": 2.197704725956471e-05,
+      "loss": 0.5516,
+      "step": 1709
+    },
+    {
+      "epoch": 0.47420965058236275,
+      "grad_norm": 0.1934727132320404,
+      "learning_rate": 2.197344587105611e-05,
+      "loss": 0.5464,
+      "step": 1710
+    },
+    {
+      "epoch": 0.47448696616749864,
+      "grad_norm": 0.20638912916183472,
+      "learning_rate": 2.1969842634027233e-05,
+      "loss": 0.5664,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4747642817526345,
+      "grad_norm": 0.19581542909145355,
+      "learning_rate": 2.196623754918115e-05,
+      "loss": 0.5597,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4750415973377704,
+      "grad_norm": 0.19786013662815094,
+      "learning_rate": 2.1962630617221325e-05,
+      "loss": 0.5729,
+      "step": 1713
+    },
+    {
+      "epoch": 0.4753189129229063,
+      "grad_norm": 0.19448676705360413,
+      "learning_rate": 2.1959021838851556e-05,
+      "loss": 0.5573,
+      "step": 1714
+    },
+    {
+      "epoch": 0.4755962285080422,
+      "grad_norm": 0.2131812423467636,
+      "learning_rate": 2.1955411214776015e-05,
+      "loss": 0.5862,
+      "step": 1715
+    },
+    {
+      "epoch": 0.47587354409317806,
+      "grad_norm": 0.2251943200826645,
+      "learning_rate": 2.195179874569923e-05,
+      "loss": 0.5847,
+      "step": 1716
+    },
+    {
+      "epoch": 0.47615085967831394,
+      "grad_norm": 0.20042872428894043,
+      "learning_rate": 2.1948184432326084e-05,
+      "loss": 0.5742,
+      "step": 1717
+    },
+    {
+      "epoch": 0.4764281752634498,
+      "grad_norm": 0.193080872297287,
+      "learning_rate": 2.1944568275361838e-05,
+      "loss": 0.5441,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4767054908485857,
+      "grad_norm": 0.20040108263492584,
+      "learning_rate": 2.194095027551209e-05,
+      "loss": 0.5635,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4769828064337216,
+      "grad_norm": 0.19599542021751404,
+      "learning_rate": 2.193733043348281e-05,
+      "loss": 0.5852,
+      "step": 1720
+    },
+    {
+      "epoch": 0.4772601220188575,
+      "grad_norm": 0.1938834935426712,
+      "learning_rate": 2.1933708749980324e-05,
+      "loss": 0.5644,
+      "step": 1721
+    },
+    {
+      "epoch": 0.47753743760399336,
+      "grad_norm": 0.19517837464809418,
+      "learning_rate": 2.1930085225711317e-05,
+      "loss": 0.5568,
+      "step": 1722
+    },
+    {
+      "epoch": 0.47781475318912925,
+      "grad_norm": 0.1954992413520813,
+      "learning_rate": 2.1926459861382843e-05,
+      "loss": 0.5459,
+      "step": 1723
+    },
+    {
+      "epoch": 0.47809206877426513,
+      "grad_norm": 0.17570015788078308,
+      "learning_rate": 2.1922832657702297e-05,
+      "loss": 0.539,
+      "step": 1724
+    },
+    {
+      "epoch": 0.478369384359401,
+      "grad_norm": 0.24834416806697845,
+      "learning_rate": 2.1919203615377442e-05,
+      "loss": 0.5401,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4786466999445369,
+      "grad_norm": 0.19633722305297852,
+      "learning_rate": 2.1915572735116413e-05,
+      "loss": 0.5469,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4789240155296728,
+      "grad_norm": 0.19104620814323425,
+      "learning_rate": 2.1911940017627676e-05,
+      "loss": 0.5472,
+      "step": 1727
+    },
+    {
+      "epoch": 0.47920133111480867,
+      "grad_norm": 0.18845802545547485,
+      "learning_rate": 2.1908305463620084e-05,
+      "loss": 0.5742,
+      "step": 1728
+    },
+    {
+      "epoch": 0.47947864669994456,
+      "grad_norm": 0.20180946588516235,
+      "learning_rate": 2.190466907380282e-05,
+      "loss": 0.5402,
+      "step": 1729
+    },
+    {
+      "epoch": 0.47975596228508044,
+      "grad_norm": 0.19500130414962769,
+      "learning_rate": 2.190103084888545e-05,
+      "loss": 0.5476,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4800332778702163,
+      "grad_norm": 0.1933142989873886,
+      "learning_rate": 2.1897390789577887e-05,
+      "loss": 0.5426,
+      "step": 1731
+    },
+    {
+      "epoch": 0.4803105934553522,
+      "grad_norm": 0.1977783739566803,
+      "learning_rate": 2.1893748896590404e-05,
+      "loss": 0.5614,
+      "step": 1732
+    },
+    {
+      "epoch": 0.4805879090404881,
+      "grad_norm": 0.2301134616136551,
+      "learning_rate": 2.1890105170633624e-05,
+      "loss": 0.564,
+      "step": 1733
+    },
+    {
+      "epoch": 0.480865224625624,
+      "grad_norm": 0.219647616147995,
+      "learning_rate": 2.1886459612418542e-05,
+      "loss": 0.5289,
+      "step": 1734
+    },
+    {
+      "epoch": 0.48114254021075986,
+      "grad_norm": 0.19821031391620636,
+      "learning_rate": 2.18828122226565e-05,
+      "loss": 0.5903,
+      "step": 1735
+    },
+    {
+      "epoch": 0.48141985579589575,
+      "grad_norm": 0.24226263165473938,
+      "learning_rate": 2.18791630020592e-05,
+      "loss": 0.5795,
+      "step": 1736
+    },
+    {
+      "epoch": 0.48169717138103163,
+      "grad_norm": 0.200203076004982,
+      "learning_rate": 2.18755119513387e-05,
+      "loss": 0.568,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4819744869661675,
+      "grad_norm": 0.19729411602020264,
+      "learning_rate": 2.1871859071207425e-05,
+      "loss": 0.5633,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4822518025513034,
+      "grad_norm": 0.19362856447696686,
+      "learning_rate": 2.1868204362378136e-05,
+      "loss": 0.5709,
+      "step": 1739
+    },
+    {
+      "epoch": 0.4825291181364393,
+      "grad_norm": 0.21311257779598236,
+      "learning_rate": 2.1864547825563968e-05,
+      "loss": 0.5481,
+      "step": 1740
+    },
+    {
+      "epoch": 0.48280643372157517,
+      "grad_norm": 0.1957651972770691,
+      "learning_rate": 2.1860889461478416e-05,
+      "loss": 0.5481,
+      "step": 1741
+    },
+    {
+      "epoch": 0.48308374930671105,
+      "grad_norm": 0.20088225603103638,
+      "learning_rate": 2.1857229270835316e-05,
+      "loss": 0.5774,
+      "step": 1742
+    },
+    {
+      "epoch": 0.48336106489184694,
+      "grad_norm": 0.19715815782546997,
+      "learning_rate": 2.1853567254348873e-05,
+      "loss": 0.5737,
+      "step": 1743
+    },
+    {
+      "epoch": 0.4836383804769828,
+      "grad_norm": 0.2071049064397812,
+      "learning_rate": 2.184990341273364e-05,
+      "loss": 0.5745,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4839156960621187,
+      "grad_norm": 0.19320468604564667,
+      "learning_rate": 2.1846237746704526e-05,
+      "loss": 0.5662,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4841930116472546,
+      "grad_norm": 0.18520797789096832,
+      "learning_rate": 2.1842570256976807e-05,
+      "loss": 0.5808,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4844703272323905,
+      "grad_norm": 0.20254836976528168,
+      "learning_rate": 2.18389009442661e-05,
+      "loss": 0.5558,
+      "step": 1747
+    },
+    {
+      "epoch": 0.48474764281752636,
+      "grad_norm": 0.2153664231300354,
+      "learning_rate": 2.1835229809288393e-05,
+      "loss": 0.5661,
+      "step": 1748
+    },
+    {
+      "epoch": 0.48502495840266224,
+      "grad_norm": 0.1884908676147461,
+      "learning_rate": 2.183155685276002e-05,
+      "loss": 0.5577,
+      "step": 1749
+    },
+    {
+      "epoch": 0.4853022739877981,
+      "grad_norm": 0.19069699943065643,
+      "learning_rate": 2.1827882075397664e-05,
+      "loss": 0.5417,
+      "step": 1750
+    },
+    {
+      "epoch": 0.485579589572934,
+      "grad_norm": 0.2055320143699646,
+      "learning_rate": 2.182420547791838e-05,
+      "loss": 0.5887,
+      "step": 1751
+    },
+    {
+      "epoch": 0.4858569051580699,
+      "grad_norm": 0.19550320506095886,
+      "learning_rate": 2.182052706103957e-05,
+      "loss": 0.5348,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4861342207432058,
+      "grad_norm": 0.19508466124534607,
+      "learning_rate": 2.1816846825478988e-05,
+      "loss": 0.5506,
+      "step": 1753
+    },
+    {
+      "epoch": 0.48641153632834166,
+      "grad_norm": 0.18409934639930725,
+      "learning_rate": 2.181316477195474e-05,
+      "loss": 0.5629,
+      "step": 1754
+    },
+    {
+      "epoch": 0.48668885191347755,
+      "grad_norm": 0.18185651302337646,
+      "learning_rate": 2.1809480901185302e-05,
+      "loss": 0.5471,
+      "step": 1755
+    },
+    {
+      "epoch": 0.48696616749861343,
+      "grad_norm": 0.18371707201004028,
+      "learning_rate": 2.180579521388949e-05,
+      "loss": 0.5747,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4872434830837493,
+      "grad_norm": 0.1873805671930313,
+      "learning_rate": 2.1802107710786476e-05,
+      "loss": 0.5606,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4875207986688852,
+      "grad_norm": 0.19243435561656952,
+      "learning_rate": 2.1798418392595794e-05,
+      "loss": 0.5638,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4877981142540211,
+      "grad_norm": 0.184648796916008,
+      "learning_rate": 2.179472726003733e-05,
+      "loss": 0.5584,
+      "step": 1759
+    },
+    {
+      "epoch": 0.48807542983915697,
+      "grad_norm": 0.17338646948337555,
+      "learning_rate": 2.1791034313831316e-05,
+      "loss": 0.556,
+      "step": 1760
+    },
+    {
+      "epoch": 0.48835274542429286,
+      "grad_norm": 0.18127377331256866,
+      "learning_rate": 2.1787339554698344e-05,
+      "loss": 0.5631,
+      "step": 1761
+    },
+    {
+      "epoch": 0.48863006100942874,
+      "grad_norm": 0.1888059824705124,
+      "learning_rate": 2.1783642983359364e-05,
+      "loss": 0.5611,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4889073765945646,
+      "grad_norm": 0.17712418735027313,
+      "learning_rate": 2.1779944600535672e-05,
+      "loss": 0.5462,
+      "step": 1763
+    },
+    {
+      "epoch": 0.4891846921797005,
+      "grad_norm": 0.19322241842746735,
+      "learning_rate": 2.177624440694892e-05,
+      "loss": 0.5511,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4894620077648364,
+      "grad_norm": 0.18574179708957672,
+      "learning_rate": 2.1772542403321118e-05,
+      "loss": 0.5531,
+      "step": 1765
+    },
+    {
+      "epoch": 0.4897393233499723,
+      "grad_norm": 0.18718282878398895,
+      "learning_rate": 2.1768838590374617e-05,
+      "loss": 0.5683,
+      "step": 1766
+    },
+    {
+      "epoch": 0.49001663893510816,
+      "grad_norm": 0.18598803877830505,
+      "learning_rate": 2.1765132968832135e-05,
+      "loss": 0.5488,
+      "step": 1767
+    },
+    {
+      "epoch": 0.49029395452024405,
+      "grad_norm": 0.18899311125278473,
+      "learning_rate": 2.1761425539416737e-05,
+      "loss": 0.5449,
+      "step": 1768
+    },
+    {
+      "epoch": 0.49057127010537993,
+      "grad_norm": 0.1895790696144104,
+      "learning_rate": 2.175771630285184e-05,
+      "loss": 0.58,
+      "step": 1769
+    },
+    {
+      "epoch": 0.4908485856905158,
+      "grad_norm": 0.17997822165489197,
+      "learning_rate": 2.1754005259861217e-05,
+      "loss": 0.5734,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4911259012756517,
+      "grad_norm": 0.19107869267463684,
+      "learning_rate": 2.175029241116898e-05,
+      "loss": 0.5707,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4914032168607876,
+      "grad_norm": 0.22478626668453217,
+      "learning_rate": 2.1746577757499613e-05,
+      "loss": 0.5667,
+      "step": 1772
+    },
+    {
+      "epoch": 0.49168053244592347,
+      "grad_norm": 0.2036799043416977,
+      "learning_rate": 2.1742861299577947e-05,
+      "loss": 0.5505,
+      "step": 1773
+    },
+    {
+      "epoch": 0.49195784803105935,
+      "grad_norm": 0.1856662929058075,
+      "learning_rate": 2.1739143038129152e-05,
+      "loss": 0.538,
+      "step": 1774
+    },
+    {
+      "epoch": 0.49223516361619524,
+      "grad_norm": 0.18573297560214996,
+      "learning_rate": 2.1735422973878766e-05,
+      "loss": 0.5507,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4925124792013311,
+      "grad_norm": 0.19560420513153076,
+      "learning_rate": 2.1731701107552673e-05,
+      "loss": 0.5395,
+      "step": 1776
+    },
+    {
+      "epoch": 0.492789794786467,
+      "grad_norm": 0.18675924837589264,
+      "learning_rate": 2.1727977439877094e-05,
+      "loss": 0.5523,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4930671103716029,
+      "grad_norm": 0.19126398861408234,
+      "learning_rate": 2.1724251971578636e-05,
+      "loss": 0.5736,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4933444259567388,
+      "grad_norm": 0.17955103516578674,
+      "learning_rate": 2.1720524703384222e-05,
+      "loss": 0.5398,
+      "step": 1779
+    },
+    {
+      "epoch": 0.49362174154187466,
+      "grad_norm": 0.1961311250925064,
+      "learning_rate": 2.1716795636021148e-05,
+      "loss": 0.5565,
+      "step": 1780
+    },
+    {
+      "epoch": 0.49389905712701054,
+      "grad_norm": 0.1914500594139099,
+      "learning_rate": 2.171306477021705e-05,
+      "loss": 0.5296,
+      "step": 1781
+    },
+    {
+      "epoch": 0.49417637271214643,
+      "grad_norm": 0.2172551453113556,
+      "learning_rate": 2.170933210669992e-05,
+      "loss": 0.5711,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4944536882972823,
+      "grad_norm": 0.18967878818511963,
+      "learning_rate": 2.1705597646198098e-05,
+      "loss": 0.5719,
+      "step": 1783
+    },
+    {
+      "epoch": 0.4947310038824182,
+      "grad_norm": 0.19041703641414642,
+      "learning_rate": 2.1701861389440277e-05,
+      "loss": 0.5431,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4950083194675541,
+      "grad_norm": 0.19202065467834473,
+      "learning_rate": 2.1698123337155503e-05,
+      "loss": 0.5392,
+      "step": 1785
+    },
+    {
+      "epoch": 0.49528563505268997,
+      "grad_norm": 0.1918521374464035,
+      "learning_rate": 2.1694383490073162e-05,
+      "loss": 0.5268,
+      "step": 1786
+    },
+    {
+      "epoch": 0.49556295063782585,
+      "grad_norm": 0.19112561643123627,
+      "learning_rate": 2.1690641848923004e-05,
+      "loss": 0.5741,
+      "step": 1787
+    },
+    {
+      "epoch": 0.49584026622296173,
+      "grad_norm": 0.28441137075424194,
+      "learning_rate": 2.168689841443512e-05,
+      "loss": 0.5628,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4961175818080976,
+      "grad_norm": 0.1971031278371811,
+      "learning_rate": 2.1683153187339955e-05,
+      "loss": 0.5336,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4963948973932335,
+      "grad_norm": 0.18874448537826538,
+      "learning_rate": 2.16794061683683e-05,
+      "loss": 0.5717,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4966722129783694,
+      "grad_norm": 0.19406016170978546,
+      "learning_rate": 2.1675657358251293e-05,
+      "loss": 0.5641,
+      "step": 1791
+    },
+    {
+      "epoch": 0.49694952856350527,
+      "grad_norm": 0.19491691887378693,
+      "learning_rate": 2.1671906757720433e-05,
+      "loss": 0.5598,
+      "step": 1792
+    },
+    {
+      "epoch": 0.49722684414864116,
+      "grad_norm": 0.1898011416196823,
+      "learning_rate": 2.166815436750756e-05,
+      "loss": 0.5748,
+      "step": 1793
+    },
+    {
+      "epoch": 0.49750415973377704,
+      "grad_norm": 0.18792784214019775,
+      "learning_rate": 2.1664400188344863e-05,
+      "loss": 0.5383,
+      "step": 1794
+    },
+    {
+      "epoch": 0.4977814753189129,
+      "grad_norm": 0.1921299546957016,
+      "learning_rate": 2.1660644220964886e-05,
+      "loss": 0.5649,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4980587909040488,
+      "grad_norm": 0.1881396770477295,
+      "learning_rate": 2.1656886466100514e-05,
+      "loss": 0.5525,
+      "step": 1796
+    },
+    {
+      "epoch": 0.4983361064891847,
+      "grad_norm": 0.19252420961856842,
+      "learning_rate": 2.1653126924484985e-05,
+      "loss": 0.5308,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4986134220743206,
+      "grad_norm": 0.2611597180366516,
+      "learning_rate": 2.1649365596851884e-05,
+      "loss": 0.5664,
+      "step": 1798
+    },
+    {
+      "epoch": 0.49889073765945646,
+      "grad_norm": 0.18851755559444427,
+      "learning_rate": 2.164560248393515e-05,
+      "loss": 0.5314,
+      "step": 1799
+    },
+    {
+      "epoch": 0.49916805324459235,
+      "grad_norm": 0.19385330379009247,
+      "learning_rate": 2.164183758646906e-05,
+      "loss": 0.5615,
+      "step": 1800
+    },
+    {
+      "epoch": 0.49944536882972823,
+      "grad_norm": 0.20486874878406525,
+      "learning_rate": 2.163807090518825e-05,
+      "loss": 0.5833,
+      "step": 1801
+    },
+    {
+      "epoch": 0.4997226844148641,
+      "grad_norm": 0.21984978020191193,
+      "learning_rate": 2.16343024408277e-05,
+      "loss": 0.5494,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1872473657131195,
+      "learning_rate": 2.1630532194122733e-05,
+      "loss": 0.5388,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5002773155851359,
+      "grad_norm": 0.19348090887069702,
+      "learning_rate": 2.1626760165809022e-05,
+      "loss": 0.5615,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5005546311702718,
+      "grad_norm": 0.18968060612678528,
+      "learning_rate": 2.16229863566226e-05,
+      "loss": 0.5401,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5008319467554077,
+      "grad_norm": 0.1913541853427887,
+      "learning_rate": 2.161921076729983e-05,
+      "loss": 0.5797,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5011092623405435,
+      "grad_norm": 0.18872712552547455,
+      "learning_rate": 2.1615433398577428e-05,
+      "loss": 0.5385,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5013865779256794,
+      "grad_norm": 0.19591103494167328,
+      "learning_rate": 2.1611654251192465e-05,
+      "loss": 0.5568,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5016638935108153,
+      "grad_norm": 0.19541212916374207,
+      "learning_rate": 2.1607873325882343e-05,
+      "loss": 0.5679,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5019412090959512,
+      "grad_norm": 0.1972798854112625,
+      "learning_rate": 2.160409062338483e-05,
+      "loss": 0.5513,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5022185246810871,
+      "grad_norm": 0.19507858157157898,
+      "learning_rate": 2.1600306144438027e-05,
+      "loss": 0.5509,
+      "step": 1811
+    },
+    {
+      "epoch": 0.502495840266223,
+      "grad_norm": 0.18637265264987946,
+      "learning_rate": 2.1596519889780387e-05,
+      "loss": 0.5632,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5027731558513588,
+      "grad_norm": 0.17938151955604553,
+      "learning_rate": 2.159273186015071e-05,
+      "loss": 0.5583,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5030504714364947,
+      "grad_norm": 0.19754791259765625,
+      "learning_rate": 2.158894205628814e-05,
+      "loss": 0.5705,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5033277870216306,
+      "grad_norm": 0.19835114479064941,
+      "learning_rate": 2.1585150478932165e-05,
+      "loss": 0.6031,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5036051026067665,
+      "grad_norm": 0.1887637823820114,
+      "learning_rate": 2.1581357128822627e-05,
+      "loss": 0.5551,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5038824181919024,
+      "grad_norm": 0.18568859994411469,
+      "learning_rate": 2.157756200669971e-05,
+      "loss": 0.5384,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5041597337770383,
+      "grad_norm": 0.1866898089647293,
+      "learning_rate": 2.1573765113303936e-05,
+      "loss": 0.5474,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5044370493621742,
+      "grad_norm": 0.18115058541297913,
+      "learning_rate": 2.156996644937618e-05,
+      "loss": 0.5459,
+      "step": 1819
+    },
+    {
+      "epoch": 0.50471436494731,
+      "grad_norm": 0.188395157456398,
+      "learning_rate": 2.1566166015657672e-05,
+      "loss": 0.5379,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5049916805324459,
+      "grad_norm": 0.2012917846441269,
+      "learning_rate": 2.156236381288997e-05,
+      "loss": 0.5771,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5052689961175818,
+      "grad_norm": 0.18014481663703918,
+      "learning_rate": 2.1558559841814986e-05,
+      "loss": 0.5508,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5055463117027177,
+      "grad_norm": 0.1956920027732849,
+      "learning_rate": 2.1554754103174972e-05,
+      "loss": 0.5625,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5058236272878536,
+      "grad_norm": 0.1985912024974823,
+      "learning_rate": 2.1550946597712536e-05,
+      "loss": 0.532,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5061009428729895,
+      "grad_norm": 0.1955011934041977,
+      "learning_rate": 2.1547137326170613e-05,
+      "loss": 0.5641,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5063782584581253,
+      "grad_norm": 0.1937127411365509,
+      "learning_rate": 2.1543326289292497e-05,
+      "loss": 0.5369,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5066555740432612,
+      "grad_norm": 0.22294695675373077,
+      "learning_rate": 2.153951348782183e-05,
+      "loss": 0.5754,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5069328896283971,
+      "grad_norm": 0.1839090883731842,
+      "learning_rate": 2.1535698922502582e-05,
+      "loss": 0.5344,
+      "step": 1828
+    },
+    {
+      "epoch": 0.507210205213533,
+      "grad_norm": 0.18940389156341553,
+      "learning_rate": 2.1531882594079074e-05,
+      "loss": 0.5399,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5074875207986689,
+      "grad_norm": 0.19242316484451294,
+      "learning_rate": 2.152806450329598e-05,
+      "loss": 0.5473,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5077648363838048,
+      "grad_norm": 0.19500425457954407,
+      "learning_rate": 2.1524244650898308e-05,
+      "loss": 0.5812,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5080421519689406,
+      "grad_norm": 0.19228143990039825,
+      "learning_rate": 2.1520423037631408e-05,
+      "loss": 0.5518,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5083194675540765,
+      "grad_norm": 0.1868668794631958,
+      "learning_rate": 2.1516599664240985e-05,
+      "loss": 0.5534,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5085967831392124,
+      "grad_norm": 0.1996205449104309,
+      "learning_rate": 2.151277453147308e-05,
+      "loss": 0.5283,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5088740987243483,
+      "grad_norm": 0.18492724001407623,
+      "learning_rate": 2.150894764007407e-05,
+      "loss": 0.563,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5091514143094842,
+      "grad_norm": 0.1847442388534546,
+      "learning_rate": 2.150511899079069e-05,
+      "loss": 0.5478,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5094287298946201,
+      "grad_norm": 0.19469550251960754,
+      "learning_rate": 2.1501288584370006e-05,
+      "loss": 0.5388,
+      "step": 1837
+    },
+    {
+      "epoch": 0.509706045479756,
+      "grad_norm": 0.18692530691623688,
+      "learning_rate": 2.1497456421559436e-05,
+      "loss": 0.523,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5099833610648918,
+      "grad_norm": 0.18751968443393707,
+      "learning_rate": 2.1493622503106736e-05,
+      "loss": 0.561,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5102606766500277,
+      "grad_norm": 0.18232478201389313,
+      "learning_rate": 2.1489786829760005e-05,
+      "loss": 0.5579,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5105379922351636,
+      "grad_norm": 0.20106928050518036,
+      "learning_rate": 2.1485949402267684e-05,
+      "loss": 0.5445,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5108153078202995,
+      "grad_norm": 0.19342289865016937,
+      "learning_rate": 2.1482110221378555e-05,
+      "loss": 0.5627,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5110926234054354,
+      "grad_norm": 0.1977401226758957,
+      "learning_rate": 2.1478269287841747e-05,
+      "loss": 0.5949,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5113699389905713,
+      "grad_norm": 0.18564966320991516,
+      "learning_rate": 2.1474426602406722e-05,
+      "loss": 0.5598,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5116472545757071,
+      "grad_norm": 0.19453705847263336,
+      "learning_rate": 2.1470582165823296e-05,
+      "loss": 0.5876,
+      "step": 1845
+    },
+    {
+      "epoch": 0.511924570160843,
+      "grad_norm": 0.1874392032623291,
+      "learning_rate": 2.146673597884162e-05,
+      "loss": 0.5465,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5122018857459789,
+      "grad_norm": 0.19269202649593353,
+      "learning_rate": 2.1462888042212183e-05,
+      "loss": 0.5333,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5124792013311148,
+      "grad_norm": 0.21227677166461945,
+      "learning_rate": 2.1459038356685824e-05,
+      "loss": 0.5772,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5127565169162507,
+      "grad_norm": 0.18281084299087524,
+      "learning_rate": 2.1455186923013716e-05,
+      "loss": 0.5732,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5130338325013866,
+      "grad_norm": 0.18322256207466125,
+      "learning_rate": 2.1451333741947373e-05,
+      "loss": 0.5367,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5133111480865225,
+      "grad_norm": 0.19677676260471344,
+      "learning_rate": 2.1447478814238658e-05,
+      "loss": 0.5797,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5135884636716583,
+      "grad_norm": 0.2173527330160141,
+      "learning_rate": 2.1443622140639768e-05,
+      "loss": 0.5688,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5138657792567942,
+      "grad_norm": 0.240644633769989,
+      "learning_rate": 2.143976372190324e-05,
+      "loss": 0.5754,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5141430948419301,
+      "grad_norm": 0.19872795045375824,
+      "learning_rate": 2.1435903558781954e-05,
+      "loss": 0.5752,
+      "step": 1854
+    },
+    {
+      "epoch": 0.514420410427066,
+      "grad_norm": 0.18520157039165497,
+      "learning_rate": 2.143204165202914e-05,
+      "loss": 0.5564,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5146977260122019,
+      "grad_norm": 0.1843482255935669,
+      "learning_rate": 2.1428178002398342e-05,
+      "loss": 0.536,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5149750415973378,
+      "grad_norm": 0.18434906005859375,
+      "learning_rate": 2.1424312610643467e-05,
+      "loss": 0.5722,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5152523571824736,
+      "grad_norm": 0.18243710696697235,
+      "learning_rate": 2.1420445477518756e-05,
+      "loss": 0.5134,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5155296727676095,
+      "grad_norm": 0.1954345405101776,
+      "learning_rate": 2.14165766037788e-05,
+      "loss": 0.5465,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5158069883527454,
+      "grad_norm": 0.1804470270872116,
+      "learning_rate": 2.1412705990178496e-05,
+      "loss": 0.5529,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5160843039378813,
+      "grad_norm": 0.19080010056495667,
+      "learning_rate": 2.140883363747312e-05,
+      "loss": 0.54,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5163616195230172,
+      "grad_norm": 0.18515829741954803,
+      "learning_rate": 2.1404959546418268e-05,
+      "loss": 0.5409,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5166389351081531,
+      "grad_norm": 0.19277918338775635,
+      "learning_rate": 2.1401083717769876e-05,
+      "loss": 0.5703,
+      "step": 1863
+    },
+    {
+      "epoch": 0.516916250693289,
+      "grad_norm": 0.2017827183008194,
+      "learning_rate": 2.139720615228422e-05,
+      "loss": 0.5545,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5171935662784248,
+      "grad_norm": 0.19409088790416718,
+      "learning_rate": 2.1393326850717915e-05,
+      "loss": 0.5613,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5174708818635607,
+      "grad_norm": 0.18852104246616364,
+      "learning_rate": 2.138944581382792e-05,
+      "loss": 0.5784,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5177481974486966,
+      "grad_norm": 0.18418540060520172,
+      "learning_rate": 2.1385563042371525e-05,
+      "loss": 0.5291,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5180255130338325,
+      "grad_norm": 0.20222961902618408,
+      "learning_rate": 2.138167853710636e-05,
+      "loss": 0.5432,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5183028286189684,
+      "grad_norm": 0.1890316754579544,
+      "learning_rate": 2.1377792298790396e-05,
+      "loss": 0.5859,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5185801442041043,
+      "grad_norm": 0.19611713290214539,
+      "learning_rate": 2.1373904328181946e-05,
+      "loss": 0.5468,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5188574597892401,
+      "grad_norm": 0.18812295794487,
+      "learning_rate": 2.1370014626039648e-05,
+      "loss": 0.5356,
+      "step": 1871
+    },
+    {
+      "epoch": 0.519134775374376,
+      "grad_norm": 0.18963585793972015,
+      "learning_rate": 2.136612319312249e-05,
+      "loss": 0.5646,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5194120909595119,
+      "grad_norm": 0.1885354369878769,
+      "learning_rate": 2.1362230030189795e-05,
+      "loss": 0.5578,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5196894065446478,
+      "grad_norm": 0.1933208405971527,
+      "learning_rate": 2.1358335138001224e-05,
+      "loss": 0.5379,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5199667221297837,
+      "grad_norm": 0.1998060792684555,
+      "learning_rate": 2.1354438517316767e-05,
+      "loss": 0.5741,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5202440377149196,
+      "grad_norm": 0.18762660026550293,
+      "learning_rate": 2.135054016889676e-05,
+      "loss": 0.5618,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5205213533000554,
+      "grad_norm": 0.2009068727493286,
+      "learning_rate": 2.1346640093501872e-05,
+      "loss": 0.5499,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5207986688851913,
+      "grad_norm": 0.1818576157093048,
+      "learning_rate": 2.1342738291893122e-05,
+      "loss": 0.5276,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5210759844703272,
+      "grad_norm": 0.2910788357257843,
+      "learning_rate": 2.1338834764831845e-05,
+      "loss": 0.5333,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5213533000554631,
+      "grad_norm": 0.20181319117546082,
+      "learning_rate": 2.1334929513079722e-05,
+      "loss": 0.567,
+      "step": 1880
+    },
+    {
+      "epoch": 0.521630615640599,
+      "grad_norm": 0.19496895372867584,
+      "learning_rate": 2.133102253739878e-05,
+      "loss": 0.5369,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5219079312257349,
+      "grad_norm": 0.18957465887069702,
+      "learning_rate": 2.1327113838551362e-05,
+      "loss": 0.5359,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5221852468108708,
+      "grad_norm": 0.19057804346084595,
+      "learning_rate": 2.132320341730017e-05,
+      "loss": 0.5595,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5224625623960066,
+      "grad_norm": 0.18731848895549774,
+      "learning_rate": 2.131929127440822e-05,
+      "loss": 0.5258,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5227398779811425,
+      "grad_norm": 0.19227994978427887,
+      "learning_rate": 2.131537741063888e-05,
+      "loss": 0.5927,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5230171935662784,
+      "grad_norm": 0.18374542891979218,
+      "learning_rate": 2.1311461826755847e-05,
+      "loss": 0.543,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5232945091514143,
+      "grad_norm": 0.1936596930027008,
+      "learning_rate": 2.1307544523523156e-05,
+      "loss": 0.5677,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5235718247365502,
+      "grad_norm": 0.18285861611366272,
+      "learning_rate": 2.1303625501705183e-05,
+      "loss": 0.5738,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5238491403216861,
+      "grad_norm": 0.23504310846328735,
+      "learning_rate": 2.1299704762066618e-05,
+      "loss": 0.5785,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5241264559068219,
+      "grad_norm": 0.1995551437139511,
+      "learning_rate": 2.129578230537252e-05,
+      "loss": 0.5637,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5244037714919578,
+      "grad_norm": 0.18941858410835266,
+      "learning_rate": 2.1291858132388248e-05,
+      "loss": 0.5523,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5246810870770937,
+      "grad_norm": 0.18262585997581482,
+      "learning_rate": 2.1287932243879523e-05,
+      "loss": 0.5587,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5249584026622296,
+      "grad_norm": 0.1891373097896576,
+      "learning_rate": 2.1284004640612376e-05,
+      "loss": 0.5494,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5252357182473655,
+      "grad_norm": 0.1890500783920288,
+      "learning_rate": 2.1280075323353206e-05,
+      "loss": 0.5411,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5255130338325014,
+      "grad_norm": 0.182157963514328,
+      "learning_rate": 2.127614429286871e-05,
+      "loss": 0.5265,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5257903494176372,
+      "grad_norm": 0.18624311685562134,
+      "learning_rate": 2.1272211549925946e-05,
+      "loss": 0.5309,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5260676650027731,
+      "grad_norm": 0.19566656649112701,
+      "learning_rate": 2.1268277095292292e-05,
+      "loss": 0.5694,
+      "step": 1897
+    },
+    {
+      "epoch": 0.526344980587909,
+      "grad_norm": 0.19160917401313782,
+      "learning_rate": 2.1264340929735467e-05,
+      "loss": 0.5324,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5266222961730449,
+      "grad_norm": 0.18548499047756195,
+      "learning_rate": 2.126040305402352e-05,
+      "loss": 0.5326,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5268996117581808,
+      "grad_norm": 0.18145646154880524,
+      "learning_rate": 2.1256463468924837e-05,
+      "loss": 0.5491,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5271769273433167,
+      "grad_norm": 0.1829940527677536,
+      "learning_rate": 2.125252217520813e-05,
+      "loss": 0.5235,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5274542429284526,
+      "grad_norm": 0.19107188284397125,
+      "learning_rate": 2.1248579173642453e-05,
+      "loss": 0.539,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5277315585135884,
+      "grad_norm": 0.18394434452056885,
+      "learning_rate": 2.1244634464997188e-05,
+      "loss": 0.5672,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5280088740987243,
+      "grad_norm": 0.19489826261997223,
+      "learning_rate": 2.1240688050042058e-05,
+      "loss": 0.5514,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5282861896838602,
+      "grad_norm": 0.18582908809185028,
+      "learning_rate": 2.1236739929547105e-05,
+      "loss": 0.5484,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5285635052689961,
+      "grad_norm": 0.19354532659053802,
+      "learning_rate": 2.123279010428272e-05,
+      "loss": 0.5542,
+      "step": 1906
+    },
+    {
+      "epoch": 0.528840820854132,
+      "grad_norm": 0.19419358670711517,
+      "learning_rate": 2.1228838575019612e-05,
+      "loss": 0.5349,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5291181364392679,
+      "grad_norm": 0.18452230095863342,
+      "learning_rate": 2.1224885342528834e-05,
+      "loss": 0.5676,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5293954520244037,
+      "grad_norm": 0.19760510325431824,
+      "learning_rate": 2.1220930407581762e-05,
+      "loss": 0.5439,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5296727676095396,
+      "grad_norm": 0.18620994687080383,
+      "learning_rate": 2.121697377095011e-05,
+      "loss": 0.5612,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5299500831946755,
+      "grad_norm": 0.1856573075056076,
+      "learning_rate": 2.121301543340593e-05,
+      "loss": 0.5267,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5302273987798114,
+      "grad_norm": 0.19919690489768982,
+      "learning_rate": 2.1209055395721586e-05,
+      "loss": 0.5499,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5305047143649473,
+      "grad_norm": 0.18541640043258667,
+      "learning_rate": 2.1205093658669793e-05,
+      "loss": 0.5693,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5307820299500832,
+      "grad_norm": 0.18908998370170593,
+      "learning_rate": 2.120113022302359e-05,
+      "loss": 0.5421,
+      "step": 1914
+    },
+    {
+      "epoch": 0.531059345535219,
+      "grad_norm": 0.19005346298217773,
+      "learning_rate": 2.119716508955635e-05,
+      "loss": 0.5475,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5313366611203549,
+      "grad_norm": 0.19358742237091064,
+      "learning_rate": 2.1193198259041774e-05,
+      "loss": 0.5671,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5316139767054908,
+      "grad_norm": 0.19891729950904846,
+      "learning_rate": 2.1189229732253894e-05,
+      "loss": 0.5623,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5318912922906267,
+      "grad_norm": 0.1928594410419464,
+      "learning_rate": 2.1185259509967082e-05,
+      "loss": 0.5467,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5321686078757626,
+      "grad_norm": 0.21051675081253052,
+      "learning_rate": 2.118128759295602e-05,
+      "loss": 0.5504,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5324459234608985,
+      "grad_norm": 0.1916693150997162,
+      "learning_rate": 2.1177313981995745e-05,
+      "loss": 0.5376,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5327232390460344,
+      "grad_norm": 0.18833374977111816,
+      "learning_rate": 2.1173338677861616e-05,
+      "loss": 0.526,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5330005546311702,
+      "grad_norm": 0.1901186853647232,
+      "learning_rate": 2.116936168132931e-05,
+      "loss": 0.5658,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5332778702163061,
+      "grad_norm": 0.19015184044837952,
+      "learning_rate": 2.1165382993174848e-05,
+      "loss": 0.5703,
+      "step": 1923
+    },
+    {
+      "epoch": 0.533555185801442,
+      "grad_norm": 0.17683614790439606,
+      "learning_rate": 2.116140261417458e-05,
+      "loss": 0.5384,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5338325013865779,
+      "grad_norm": 0.18865369260311127,
+      "learning_rate": 2.1157420545105187e-05,
+      "loss": 0.5473,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5341098169717138,
+      "grad_norm": 0.1918121576309204,
+      "learning_rate": 2.1153436786743668e-05,
+      "loss": 0.5587,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5343871325568497,
+      "grad_norm": 0.20620866119861603,
+      "learning_rate": 2.1149451339867363e-05,
+      "loss": 0.5358,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5346644481419855,
+      "grad_norm": 0.19819065928459167,
+      "learning_rate": 2.114546420525394e-05,
+      "loss": 0.5681,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5349417637271214,
+      "grad_norm": 0.19372405111789703,
+      "learning_rate": 2.114147538368139e-05,
+      "loss": 0.5615,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5352190793122573,
+      "grad_norm": 0.19177958369255066,
+      "learning_rate": 2.1137484875928048e-05,
+      "loss": 0.5336,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5354963948973932,
+      "grad_norm": 0.1978660523891449,
+      "learning_rate": 2.1133492682772556e-05,
+      "loss": 0.5465,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5357737104825291,
+      "grad_norm": 0.19225792586803436,
+      "learning_rate": 2.1129498804993902e-05,
+      "loss": 0.5548,
+      "step": 1932
+    },
+    {
+      "epoch": 0.536051026067665,
+      "grad_norm": 0.19737133383750916,
+      "learning_rate": 2.1125503243371398e-05,
+      "loss": 0.5636,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5363283416528009,
+      "grad_norm": 0.19044062495231628,
+      "learning_rate": 2.112150599868468e-05,
+      "loss": 0.5595,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5366056572379367,
+      "grad_norm": 0.18862438201904297,
+      "learning_rate": 2.1117507071713724e-05,
+      "loss": 0.5594,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5368829728230726,
+      "grad_norm": 0.20726455748081207,
+      "learning_rate": 2.111350646323882e-05,
+      "loss": 0.5678,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5371602884082085,
+      "grad_norm": 0.1967361867427826,
+      "learning_rate": 2.1109504174040594e-05,
+      "loss": 0.5518,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5374376039933444,
+      "grad_norm": 0.19529032707214355,
+      "learning_rate": 2.1105500204899997e-05,
+      "loss": 0.5775,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5377149195784803,
+      "grad_norm": 0.19610485434532166,
+      "learning_rate": 2.110149455659831e-05,
+      "loss": 0.563,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5379922351636162,
+      "grad_norm": 0.18098145723342896,
+      "learning_rate": 2.109748722991715e-05,
+      "loss": 0.5401,
+      "step": 1940
+    },
+    {
+      "epoch": 0.538269550748752,
+      "grad_norm": 0.21725578606128693,
+      "learning_rate": 2.109347822563844e-05,
+      "loss": 0.5671,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5385468663338879,
+      "grad_norm": 0.20272882282733917,
+      "learning_rate": 2.108946754454445e-05,
+      "loss": 0.5699,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5388241819190238,
+      "grad_norm": 0.18334084749221802,
+      "learning_rate": 2.108545518741777e-05,
+      "loss": 0.5508,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5391014975041597,
+      "grad_norm": 0.20137275755405426,
+      "learning_rate": 2.1081441155041314e-05,
+      "loss": 0.5315,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5393788130892956,
+      "grad_norm": 0.21601121127605438,
+      "learning_rate": 2.1077425448198327e-05,
+      "loss": 0.5415,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5396561286744315,
+      "grad_norm": 0.19099989533424377,
+      "learning_rate": 2.107340806767238e-05,
+      "loss": 0.5492,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5399334442595674,
+      "grad_norm": 0.18626771867275238,
+      "learning_rate": 2.106938901424737e-05,
+      "loss": 0.5622,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5402107598447032,
+      "grad_norm": 0.1979578286409378,
+      "learning_rate": 2.1065368288707523e-05,
+      "loss": 0.5659,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5404880754298391,
+      "grad_norm": 0.18457596004009247,
+      "learning_rate": 2.1061345891837393e-05,
+      "loss": 0.554,
+      "step": 1949
+    },
+    {
+      "epoch": 0.540765391014975,
+      "grad_norm": 0.1944621354341507,
+      "learning_rate": 2.1057321824421843e-05,
+      "loss": 0.5574,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5410427066001109,
+      "grad_norm": 0.20393145084381104,
+      "learning_rate": 2.1053296087246087e-05,
+      "loss": 0.5557,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5413200221852468,
+      "grad_norm": 0.19211652874946594,
+      "learning_rate": 2.1049268681095647e-05,
+      "loss": 0.5626,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5415973377703827,
+      "grad_norm": 0.18954886496067047,
+      "learning_rate": 2.1045239606756378e-05,
+      "loss": 0.5481,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5418746533555185,
+      "grad_norm": 0.19707705080509186,
+      "learning_rate": 2.1041208865014464e-05,
+      "loss": 0.5435,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5421519689406544,
+      "grad_norm": 0.20506185293197632,
+      "learning_rate": 2.10371764566564e-05,
+      "loss": 0.5322,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5424292845257903,
+      "grad_norm": 0.19055700302124023,
+      "learning_rate": 2.103314238246903e-05,
+      "loss": 0.5513,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5427066001109262,
+      "grad_norm": 0.2124052196741104,
+      "learning_rate": 2.102910664323949e-05,
+      "loss": 0.574,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5429839156960621,
+      "grad_norm": 0.19025616347789764,
+      "learning_rate": 2.1025069239755273e-05,
+      "loss": 0.5342,
+      "step": 1958
+    },
+    {
+      "epoch": 0.543261231281198,
+      "grad_norm": 0.18279728293418884,
+      "learning_rate": 2.102103017280418e-05,
+      "loss": 0.5542,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5435385468663338,
+      "grad_norm": 0.18862898647785187,
+      "learning_rate": 2.101698944317434e-05,
+      "loss": 0.5743,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5438158624514697,
+      "grad_norm": 0.18205633759498596,
+      "learning_rate": 2.101294705165421e-05,
+      "loss": 0.5597,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5440931780366056,
+      "grad_norm": 0.1904565840959549,
+      "learning_rate": 2.100890299903256e-05,
+      "loss": 0.5789,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5443704936217415,
+      "grad_norm": 0.19704897701740265,
+      "learning_rate": 2.1004857286098495e-05,
+      "loss": 0.5626,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5446478092068774,
+      "grad_norm": 0.1878540813922882,
+      "learning_rate": 2.1000809913641445e-05,
+      "loss": 0.5713,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5449251247920133,
+      "grad_norm": 0.2131820023059845,
+      "learning_rate": 2.0996760882451148e-05,
+      "loss": 0.5632,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5452024403771492,
+      "grad_norm": 0.19665499031543732,
+      "learning_rate": 2.0992710193317693e-05,
+      "loss": 0.5585,
+      "step": 1966
+    },
+    {
+      "epoch": 0.545479755962285,
+      "grad_norm": 0.18704085052013397,
+      "learning_rate": 2.0988657847031467e-05,
+      "loss": 0.5534,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5457570715474209,
+      "grad_norm": 0.1849927306175232,
+      "learning_rate": 2.0984603844383195e-05,
+      "loss": 0.5654,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5460343871325568,
+      "grad_norm": 0.18812572956085205,
+      "learning_rate": 2.0980548186163918e-05,
+      "loss": 0.5412,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5463117027176927,
+      "grad_norm": 0.18602755665779114,
+      "learning_rate": 2.0976490873165e-05,
+      "loss": 0.5503,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5465890183028286,
+      "grad_norm": 0.19630911946296692,
+      "learning_rate": 2.097243190617813e-05,
+      "loss": 0.5687,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5468663338879645,
+      "grad_norm": 0.18516132235527039,
+      "learning_rate": 2.0968371285995323e-05,
+      "loss": 0.5561,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5471436494731003,
+      "grad_norm": 0.18923065066337585,
+      "learning_rate": 2.0964309013408914e-05,
+      "loss": 0.5477,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5474209650582362,
+      "grad_norm": 0.18436457216739655,
+      "learning_rate": 2.096024508921156e-05,
+      "loss": 0.56,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5476982806433721,
+      "grad_norm": 0.18705391883850098,
+      "learning_rate": 2.095617951419624e-05,
+      "loss": 0.536,
+      "step": 1975
+    },
+    {
+      "epoch": 0.547975596228508,
+      "grad_norm": 0.18911142647266388,
+      "learning_rate": 2.095211228915625e-05,
+      "loss": 0.5431,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5482529118136439,
+      "grad_norm": 0.19025933742523193,
+      "learning_rate": 2.0948043414885222e-05,
+      "loss": 0.5352,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5485302273987798,
+      "grad_norm": 0.18421220779418945,
+      "learning_rate": 2.0943972892177094e-05,
+      "loss": 0.536,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5488075429839157,
+      "grad_norm": 0.19545422494411469,
+      "learning_rate": 2.0939900721826132e-05,
+      "loss": 0.5563,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5490848585690515,
+      "grad_norm": 0.19028547406196594,
+      "learning_rate": 2.0935826904626937e-05,
+      "loss": 0.5081,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5493621741541874,
+      "grad_norm": 0.20487383008003235,
+      "learning_rate": 2.0931751441374406e-05,
+      "loss": 0.5426,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5496394897393233,
+      "grad_norm": 0.20360392332077026,
+      "learning_rate": 2.0927674332863774e-05,
+      "loss": 0.5686,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5499168053244592,
+      "grad_norm": 0.19089289009571075,
+      "learning_rate": 2.092359557989059e-05,
+      "loss": 0.5574,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5501941209095951,
+      "grad_norm": 0.1797301024198532,
+      "learning_rate": 2.0919515183250736e-05,
+      "loss": 0.5666,
+      "step": 1984
+    },
+    {
+      "epoch": 0.550471436494731,
+      "grad_norm": 0.19856330752372742,
+      "learning_rate": 2.0915433143740393e-05,
+      "loss": 0.5373,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5507487520798668,
+      "grad_norm": 0.19353127479553223,
+      "learning_rate": 2.0911349462156082e-05,
+      "loss": 0.5454,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5510260676650027,
+      "grad_norm": 0.19544550776481628,
+      "learning_rate": 2.090726413929464e-05,
+      "loss": 0.5705,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5513033832501386,
+      "grad_norm": 0.199398934841156,
+      "learning_rate": 2.0903177175953216e-05,
+      "loss": 0.5431,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5515806988352745,
+      "grad_norm": 0.19894284009933472,
+      "learning_rate": 2.0899088572929286e-05,
+      "loss": 0.5658,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5518580144204104,
+      "grad_norm": 0.1806151121854782,
+      "learning_rate": 2.0894998331020645e-05,
+      "loss": 0.5748,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5521353300055463,
+      "grad_norm": 0.19128967821598053,
+      "learning_rate": 2.089090645102541e-05,
+      "loss": 0.5576,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5524126455906821,
+      "grad_norm": 0.20147933065891266,
+      "learning_rate": 2.0886812933742013e-05,
+      "loss": 0.5738,
+      "step": 1992
+    },
+    {
+      "epoch": 0.552689961175818,
+      "grad_norm": 0.17531876266002655,
+      "learning_rate": 2.0882717779969207e-05,
+      "loss": 0.544,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5529672767609539,
+      "grad_norm": 0.19415581226348877,
+      "learning_rate": 2.087862099050607e-05,
+      "loss": 0.5315,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5532445923460898,
+      "grad_norm": 0.22229987382888794,
+      "learning_rate": 2.087452256615199e-05,
+      "loss": 0.5324,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5535219079312257,
+      "grad_norm": 0.18078093230724335,
+      "learning_rate": 2.0870422507706676e-05,
+      "loss": 0.5378,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5537992235163616,
+      "grad_norm": 0.18564875423908234,
+      "learning_rate": 2.0866320815970157e-05,
+      "loss": 0.5238,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5540765391014975,
+      "grad_norm": 0.19045381247997284,
+      "learning_rate": 2.086221749174279e-05,
+      "loss": 0.5527,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5543538546866333,
+      "grad_norm": 0.18606480956077576,
+      "learning_rate": 2.0858112535825242e-05,
+      "loss": 0.5132,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5546311702717692,
+      "grad_norm": 0.19520308077335358,
+      "learning_rate": 2.0854005949018487e-05,
+      "loss": 0.554,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5549084858569051,
+      "grad_norm": 0.19881142675876617,
+      "learning_rate": 2.0849897732123838e-05,
+      "loss": 0.5327,
+      "step": 2001
+    },
+    {
+      "epoch": 0.555185801442041,
+      "grad_norm": 0.18831515312194824,
+      "learning_rate": 2.0845787885942917e-05,
+      "loss": 0.5541,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5554631170271769,
+      "grad_norm": 0.1954246610403061,
+      "learning_rate": 2.0841676411277662e-05,
+      "loss": 0.5744,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5557404326123128,
+      "grad_norm": 0.20168912410736084,
+      "learning_rate": 2.0837563308930325e-05,
+      "loss": 0.5704,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5560177481974486,
+      "grad_norm": 0.19541358947753906,
+      "learning_rate": 2.0833448579703492e-05,
+      "loss": 0.5555,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5562950637825845,
+      "grad_norm": 0.20079733431339264,
+      "learning_rate": 2.082933222440005e-05,
+      "loss": 0.5394,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5565723793677204,
+      "grad_norm": 0.1951141506433487,
+      "learning_rate": 2.082521424382321e-05,
+      "loss": 0.5314,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5568496949528563,
+      "grad_norm": 0.1864672750234604,
+      "learning_rate": 2.0821094638776497e-05,
+      "loss": 0.5365,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5571270105379922,
+      "grad_norm": 0.18978318572044373,
+      "learning_rate": 2.0816973410063754e-05,
+      "loss": 0.5199,
+      "step": 2009
+    },
+    {
+      "epoch": 0.5574043261231281,
+      "grad_norm": 0.1931847780942917,
+      "learning_rate": 2.0812850558489153e-05,
+      "loss": 0.5701,
+      "step": 2010
+    },
+    {
+      "epoch": 0.557681641708264,
+      "grad_norm": 0.18549029529094696,
+      "learning_rate": 2.0808726084857157e-05,
+      "loss": 0.4933,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5579589572933998,
+      "grad_norm": 0.18897327780723572,
+      "learning_rate": 2.0804599989972567e-05,
+      "loss": 0.58,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5582362728785357,
+      "grad_norm": 0.20237743854522705,
+      "learning_rate": 2.0800472274640494e-05,
+      "loss": 0.5577,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5585135884636716,
+      "grad_norm": 0.18905304372310638,
+      "learning_rate": 2.0796342939666362e-05,
+      "loss": 0.5763,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5587909040488075,
+      "grad_norm": 0.2023850679397583,
+      "learning_rate": 2.079221198585592e-05,
+      "loss": 0.5502,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5590682196339434,
+      "grad_norm": 0.19735179841518402,
+      "learning_rate": 2.0788079414015215e-05,
+      "loss": 0.5293,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5593455352190793,
+      "grad_norm": 0.18779948353767395,
+      "learning_rate": 2.078394522495063e-05,
+      "loss": 0.5524,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5596228508042151,
+      "grad_norm": 0.20868045091629028,
+      "learning_rate": 2.0779809419468854e-05,
+      "loss": 0.5724,
+      "step": 2018
+    },
+    {
+      "epoch": 0.559900166389351,
+      "grad_norm": 0.21078519523143768,
+      "learning_rate": 2.077567199837689e-05,
+      "loss": 0.5398,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5601774819744869,
+      "grad_norm": 0.18992477655410767,
+      "learning_rate": 2.0771532962482057e-05,
+      "loss": 0.5665,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5604547975596228,
+      "grad_norm": 0.1859736144542694,
+      "learning_rate": 2.0767392312591992e-05,
+      "loss": 0.5429,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5607321131447587,
+      "grad_norm": 0.18972904980182648,
+      "learning_rate": 2.0763250049514654e-05,
+      "loss": 0.5595,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5610094287298946,
+      "grad_norm": 0.189973384141922,
+      "learning_rate": 2.0759106174058293e-05,
+      "loss": 0.5669,
+      "step": 2023
+    },
+    {
+      "epoch": 0.5612867443150305,
+      "grad_norm": 0.19088168442249298,
+      "learning_rate": 2.07549606870315e-05,
+      "loss": 0.5654,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5615640599001663,
+      "grad_norm": 0.1857694834470749,
+      "learning_rate": 2.075081358924317e-05,
+      "loss": 0.5318,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5618413754853022,
+      "grad_norm": 0.19243502616882324,
+      "learning_rate": 2.0746664881502496e-05,
+      "loss": 0.549,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5621186910704381,
+      "grad_norm": 0.18214242160320282,
+      "learning_rate": 2.0742514564619022e-05,
+      "loss": 0.5318,
+      "step": 2027
+    },
+    {
+      "epoch": 0.562396006655574,
+      "grad_norm": 0.18616682291030884,
+      "learning_rate": 2.0738362639402574e-05,
+      "loss": 0.5397,
+      "step": 2028
+    },
+    {
+      "epoch": 0.56267332224071,
+      "grad_norm": 0.18527580797672272,
+      "learning_rate": 2.07342091066633e-05,
+      "loss": 0.5513,
+      "step": 2029
+    },
+    {
+      "epoch": 0.5629506378258459,
+      "grad_norm": 0.20299869775772095,
+      "learning_rate": 2.073005396721167e-05,
+      "loss": 0.5437,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5632279534109818,
+      "grad_norm": 0.18229471147060394,
+      "learning_rate": 2.072589722185846e-05,
+      "loss": 0.5411,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5635052689961176,
+      "grad_norm": 0.19586963951587677,
+      "learning_rate": 2.0721738871414763e-05,
+      "loss": 0.5674,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5637825845812535,
+      "grad_norm": 0.21511641144752502,
+      "learning_rate": 2.0717578916691977e-05,
+      "loss": 0.5762,
+      "step": 2033
+    },
+    {
+      "epoch": 0.5640599001663894,
+      "grad_norm": 0.18836332857608795,
+      "learning_rate": 2.071341735850183e-05,
+      "loss": 0.5599,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5643372157515253,
+      "grad_norm": 0.1830950826406479,
+      "learning_rate": 2.070925419765634e-05,
+      "loss": 0.539,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5646145313366612,
+      "grad_norm": 0.18749170005321503,
+      "learning_rate": 2.070508943496786e-05,
+      "loss": 0.5516,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5648918469217971,
+      "grad_norm": 0.18844857811927795,
+      "learning_rate": 2.070092307124904e-05,
+      "loss": 0.5301,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5651691625069329,
+      "grad_norm": 0.1957472264766693,
+      "learning_rate": 2.0696755107312845e-05,
+      "loss": 0.5612,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5654464780920688,
+      "grad_norm": 0.1901545226573944,
+      "learning_rate": 2.0692585543972566e-05,
+      "loss": 0.5529,
+      "step": 2039
+    },
+    {
+      "epoch": 0.5657237936772047,
+      "grad_norm": 0.1877562254667282,
+      "learning_rate": 2.0688414382041788e-05,
+      "loss": 0.5324,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5660011092623406,
+      "grad_norm": 0.18741729855537415,
+      "learning_rate": 2.068424162233441e-05,
+      "loss": 0.5522,
+      "step": 2041
+    },
+    {
+      "epoch": 0.5662784248474765,
+      "grad_norm": 0.18096144497394562,
+      "learning_rate": 2.068006726566466e-05,
+      "loss": 0.5382,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5665557404326124,
+      "grad_norm": 0.18831691145896912,
+      "learning_rate": 2.0675891312847064e-05,
+      "loss": 0.5547,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5668330560177482,
+      "grad_norm": 0.19189482927322388,
+      "learning_rate": 2.0671713764696445e-05,
+      "loss": 0.5699,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5671103716028841,
+      "grad_norm": 0.183754563331604,
+      "learning_rate": 2.0667534622027974e-05,
+      "loss": 0.5256,
+      "step": 2045
+    },
+    {
+      "epoch": 0.56738768718802,
+      "grad_norm": 0.19887655973434448,
+      "learning_rate": 2.0663353885657098e-05,
+      "loss": 0.5792,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5676650027731559,
+      "grad_norm": 0.18769381940364838,
+      "learning_rate": 2.0659171556399596e-05,
+      "loss": 0.5679,
+      "step": 2047
+    },
+    {
+      "epoch": 0.5679423183582918,
+      "grad_norm": 0.20135840773582458,
+      "learning_rate": 2.0654987635071554e-05,
+      "loss": 0.5766,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5682196339434277,
+      "grad_norm": 0.20204971730709076,
+      "learning_rate": 2.065080212248936e-05,
+      "loss": 0.602,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5684969495285636,
+      "grad_norm": 0.18772201240062714,
+      "learning_rate": 2.0646615019469724e-05,
+      "loss": 0.5584,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5687742651136994,
+      "grad_norm": 0.18956024944782257,
+      "learning_rate": 2.064242632682965e-05,
+      "loss": 0.573,
+      "step": 2051
+    },
+    {
+      "epoch": 0.5690515806988353,
+      "grad_norm": 0.20082207024097443,
+      "learning_rate": 2.0638236045386472e-05,
+      "loss": 0.5478,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5693288962839712,
+      "grad_norm": 0.1947973519563675,
+      "learning_rate": 2.063404417595783e-05,
+      "loss": 0.5543,
+      "step": 2053
+    },
+    {
+      "epoch": 0.5696062118691071,
+      "grad_norm": 0.19126874208450317,
+      "learning_rate": 2.0629850719361654e-05,
+      "loss": 0.5704,
+      "step": 2054
+    },
+    {
+      "epoch": 0.569883527454243,
+      "grad_norm": 0.3870353698730469,
+      "learning_rate": 2.062565567641621e-05,
+      "loss": 0.5323,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5701608430393789,
+      "grad_norm": 0.20603255927562714,
+      "learning_rate": 2.0621459047940056e-05,
+      "loss": 0.574,
+      "step": 2056
+    },
+    {
+      "epoch": 0.5704381586245147,
+      "grad_norm": 0.18535418808460236,
+      "learning_rate": 2.0617260834752068e-05,
+      "loss": 0.5358,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5707154742096506,
+      "grad_norm": 0.24727782607078552,
+      "learning_rate": 2.061306103767143e-05,
+      "loss": 0.5744,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5709927897947865,
+      "grad_norm": 0.19518110156059265,
+      "learning_rate": 2.0608859657517633e-05,
+      "loss": 0.5655,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5712701053799224,
+      "grad_norm": 0.19751004874706268,
+      "learning_rate": 2.0604656695110476e-05,
+      "loss": 0.5555,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5715474209650583,
+      "grad_norm": 0.18788260221481323,
+      "learning_rate": 2.0600452151270068e-05,
+      "loss": 0.5463,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5718247365501942,
+      "grad_norm": 0.23272345960140228,
+      "learning_rate": 2.0596246026816826e-05,
+      "loss": 0.5601,
+      "step": 2062
+    },
+    {
+      "epoch": 0.57210205213533,
+      "grad_norm": 0.19826073944568634,
+      "learning_rate": 2.059203832257148e-05,
+      "loss": 0.5654,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5723793677204659,
+      "grad_norm": 0.19848833978176117,
+      "learning_rate": 2.058782903935506e-05,
+      "loss": 0.5666,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5726566833056018,
+      "grad_norm": 0.21063697338104248,
+      "learning_rate": 2.0583618177988917e-05,
+      "loss": 0.5421,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5729339988907377,
+      "grad_norm": 0.19619859755039215,
+      "learning_rate": 2.0579405739294695e-05,
+      "loss": 0.5343,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5732113144758736,
+      "grad_norm": 0.18998843431472778,
+      "learning_rate": 2.057519172409435e-05,
+      "loss": 0.5486,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5734886300610095,
+      "grad_norm": 0.19432705640792847,
+      "learning_rate": 2.0570976133210152e-05,
+      "loss": 0.528,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5737659456461454,
+      "grad_norm": 0.18903441727161407,
+      "learning_rate": 2.0566758967464677e-05,
+      "loss": 0.5529,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5740432612312812,
+      "grad_norm": 0.19154904782772064,
+      "learning_rate": 2.05625402276808e-05,
+      "loss": 0.5519,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5743205768164171,
+      "grad_norm": 0.19544386863708496,
+      "learning_rate": 2.0558319914681713e-05,
+      "loss": 0.5444,
+      "step": 2071
+    },
+    {
+      "epoch": 0.574597892401553,
+      "grad_norm": 0.1918288618326187,
+      "learning_rate": 2.055409802929091e-05,
+      "loss": 0.5646,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5748752079866889,
+      "grad_norm": 0.19810417294502258,
+      "learning_rate": 2.054987457233219e-05,
+      "loss": 0.5799,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5751525235718248,
+      "grad_norm": 0.20202040672302246,
+      "learning_rate": 2.0545649544629665e-05,
+      "loss": 0.5555,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5754298391569607,
+      "grad_norm": 0.19962945580482483,
+      "learning_rate": 2.0541422947007748e-05,
+      "loss": 0.5245,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5757071547420965,
+      "grad_norm": 0.19925406575202942,
+      "learning_rate": 2.053719478029116e-05,
+      "loss": 0.5755,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5759844703272324,
+      "grad_norm": 0.1866733878850937,
+      "learning_rate": 2.0532965045304932e-05,
+      "loss": 0.5339,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5762617859123683,
+      "grad_norm": 0.22320881485939026,
+      "learning_rate": 2.052873374287439e-05,
+      "loss": 0.5111,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5765391014975042,
+      "grad_norm": 0.19458182156085968,
+      "learning_rate": 2.0524500873825182e-05,
+      "loss": 0.5625,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5768164170826401,
+      "grad_norm": 0.18524032831192017,
+      "learning_rate": 2.0520266438983242e-05,
+      "loss": 0.5876,
+      "step": 2080
+    },
+    {
+      "epoch": 0.577093732667776,
+      "grad_norm": 0.17904391884803772,
+      "learning_rate": 2.0516030439174833e-05,
+      "loss": 0.5541,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5773710482529119,
+      "grad_norm": 0.1819324642419815,
+      "learning_rate": 2.05117928752265e-05,
+      "loss": 0.5326,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5776483638380477,
+      "grad_norm": 0.18749113380908966,
+      "learning_rate": 2.0507553747965114e-05,
+      "loss": 0.5502,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5779256794231836,
+      "grad_norm": 0.1892707794904709,
+      "learning_rate": 2.050331305821783e-05,
+      "loss": 0.5533,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5782029950083195,
+      "grad_norm": 0.1937221884727478,
+      "learning_rate": 2.0499070806812126e-05,
+      "loss": 0.5615,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5784803105934554,
+      "grad_norm": 0.20081382989883423,
+      "learning_rate": 2.0494826994575777e-05,
+      "loss": 0.5424,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5787576261785913,
+      "grad_norm": 0.18762163817882538,
+      "learning_rate": 2.0490581622336863e-05,
+      "loss": 0.5621,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5790349417637272,
+      "grad_norm": 0.18031221628189087,
+      "learning_rate": 2.048633469092377e-05,
+      "loss": 0.5045,
+      "step": 2088
+    },
+    {
+      "epoch": 0.579312257348863,
+      "grad_norm": 0.21820896863937378,
+      "learning_rate": 2.048208620116518e-05,
+      "loss": 0.5676,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5795895729339989,
+      "grad_norm": 0.19855897128582,
+      "learning_rate": 2.0477836153890095e-05,
+      "loss": 0.5461,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5798668885191348,
+      "grad_norm": 0.1899833232164383,
+      "learning_rate": 2.0473584549927806e-05,
+      "loss": 0.5519,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5801442041042707,
+      "grad_norm": 0.1892709583044052,
+      "learning_rate": 2.0469331390107914e-05,
+      "loss": 0.5255,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5804215196894066,
+      "grad_norm": 0.19004952907562256,
+      "learning_rate": 2.0465076675260326e-05,
+      "loss": 0.5468,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5806988352745425,
+      "grad_norm": 0.18664851784706116,
+      "learning_rate": 2.0460820406215247e-05,
+      "loss": 0.5333,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5809761508596784,
+      "grad_norm": 0.18904832005500793,
+      "learning_rate": 2.045656258380319e-05,
+      "loss": 0.5653,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5812534664448142,
+      "grad_norm": 0.19123364984989166,
+      "learning_rate": 2.0452303208854966e-05,
+      "loss": 0.5368,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5815307820299501,
+      "grad_norm": 0.1906225085258484,
+      "learning_rate": 2.0448042282201694e-05,
+      "loss": 0.5672,
+      "step": 2097
+    },
+    {
+      "epoch": 0.581808097615086,
+      "grad_norm": 0.19360537827014923,
+      "learning_rate": 2.0443779804674796e-05,
+      "loss": 0.548,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5820854132002219,
+      "grad_norm": 0.21694041788578033,
+      "learning_rate": 2.0439515777105987e-05,
+      "loss": 0.5724,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5823627287853578,
+      "grad_norm": 0.191674143075943,
+      "learning_rate": 2.04352502003273e-05,
+      "loss": 0.5295,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5826400443704937,
+      "grad_norm": 0.19594216346740723,
+      "learning_rate": 2.0430983075171055e-05,
+      "loss": 0.5412,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5829173599556295,
+      "grad_norm": 0.19115525484085083,
+      "learning_rate": 2.0426714402469887e-05,
+      "loss": 0.5368,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5831946755407654,
+      "grad_norm": 0.19171211123466492,
+      "learning_rate": 2.042244418305673e-05,
+      "loss": 0.57,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5834719911259013,
+      "grad_norm": 0.18373924493789673,
+      "learning_rate": 2.0418172417764802e-05,
+      "loss": 0.5648,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5837493067110372,
+      "grad_norm": 0.20192669332027435,
+      "learning_rate": 2.0413899107427652e-05,
+      "loss": 0.5699,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5840266222961731,
+      "grad_norm": 0.18735186755657196,
+      "learning_rate": 2.0409624252879112e-05,
+      "loss": 0.551,
+      "step": 2106
+    },
+    {
+      "epoch": 0.584303937881309,
+      "grad_norm": 0.19031678140163422,
+      "learning_rate": 2.0405347854953316e-05,
+      "loss": 0.5313,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5845812534664449,
+      "grad_norm": 0.19912661612033844,
+      "learning_rate": 2.0401069914484707e-05,
+      "loss": 0.5815,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5848585690515807,
+      "grad_norm": 0.1985718458890915,
+      "learning_rate": 2.0396790432308025e-05,
+      "loss": 0.5364,
+      "step": 2109
+    },
+    {
+      "epoch": 0.5851358846367166,
+      "grad_norm": 0.20383597910404205,
+      "learning_rate": 2.0392509409258303e-05,
+      "loss": 0.5747,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5854132002218525,
+      "grad_norm": 0.19357682764530182,
+      "learning_rate": 2.038822684617089e-05,
+      "loss": 0.5508,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5856905158069884,
+      "grad_norm": 0.18165504932403564,
+      "learning_rate": 2.0383942743881425e-05,
+      "loss": 0.5234,
+      "step": 2112
+    },
+    {
+      "epoch": 0.5859678313921243,
+      "grad_norm": 0.18874984979629517,
+      "learning_rate": 2.0379657103225852e-05,
+      "loss": 0.5691,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5862451469772602,
+      "grad_norm": 0.18956932425498962,
+      "learning_rate": 2.0375369925040406e-05,
+      "loss": 0.5664,
+      "step": 2114
+    },
+    {
+      "epoch": 0.586522462562396,
+      "grad_norm": 0.17788472771644592,
+      "learning_rate": 2.0371081210161634e-05,
+      "loss": 0.5473,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5867997781475319,
+      "grad_norm": 0.22669924795627594,
+      "learning_rate": 2.0366790959426378e-05,
+      "loss": 0.5603,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5870770937326678,
+      "grad_norm": 0.1913762092590332,
+      "learning_rate": 2.0362499173671784e-05,
+      "loss": 0.5698,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5873544093178037,
+      "grad_norm": 0.19173979759216309,
+      "learning_rate": 2.0358205853735287e-05,
+      "loss": 0.5735,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5876317249029396,
+      "grad_norm": 0.18526272475719452,
+      "learning_rate": 2.035391100045462e-05,
+      "loss": 0.5462,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5879090404880755,
+      "grad_norm": 0.20534314215183258,
+      "learning_rate": 2.034961461466784e-05,
+      "loss": 0.5643,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5881863560732113,
+      "grad_norm": 0.19565671682357788,
+      "learning_rate": 2.0345316697213273e-05,
+      "loss": 0.5599,
+      "step": 2121
+    },
+    {
+      "epoch": 0.5884636716583472,
+      "grad_norm": 0.19610898196697235,
+      "learning_rate": 2.034101724892956e-05,
+      "loss": 0.5597,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5887409872434831,
+      "grad_norm": 0.19708271324634552,
+      "learning_rate": 2.033671627065564e-05,
+      "loss": 0.5494,
+      "step": 2123
+    },
+    {
+      "epoch": 0.589018302828619,
+      "grad_norm": 0.285875141620636,
+      "learning_rate": 2.033241376323075e-05,
+      "loss": 0.5553,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5892956184137549,
+      "grad_norm": 0.1879926174879074,
+      "learning_rate": 2.0328109727494417e-05,
+      "loss": 0.5464,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5895729339988908,
+      "grad_norm": 0.20556902885437012,
+      "learning_rate": 2.032380416428647e-05,
+      "loss": 0.5533,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5898502495840267,
+      "grad_norm": 0.1847870945930481,
+      "learning_rate": 2.0319497074447043e-05,
+      "loss": 0.5629,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5901275651691625,
+      "grad_norm": 0.19035299122333527,
+      "learning_rate": 2.0315188458816567e-05,
+      "loss": 0.5491,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5904048807542984,
+      "grad_norm": 0.17980261147022247,
+      "learning_rate": 2.031087831823576e-05,
+      "loss": 0.5281,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5906821963394343,
+      "grad_norm": 0.19888748228549957,
+      "learning_rate": 2.030656665354565e-05,
+      "loss": 0.5443,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5909595119245702,
+      "grad_norm": 0.17473064363002777,
+      "learning_rate": 2.0302253465587555e-05,
+      "loss": 0.5263,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5912368275097061,
+      "grad_norm": 0.1915203481912613,
+      "learning_rate": 2.0297938755203088e-05,
+      "loss": 0.5609,
+      "step": 2132
+    },
+    {
+      "epoch": 0.591514143094842,
+      "grad_norm": 0.1980845183134079,
+      "learning_rate": 2.029362252323417e-05,
+      "loss": 0.5549,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5917914586799778,
+      "grad_norm": 0.18366824090480804,
+      "learning_rate": 2.028930477052301e-05,
+      "loss": 0.5333,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5920687742651137,
+      "grad_norm": 0.1946541965007782,
+      "learning_rate": 2.0284985497912118e-05,
+      "loss": 0.588,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5923460898502496,
+      "grad_norm": 0.1904669553041458,
+      "learning_rate": 2.028066470624429e-05,
+      "loss": 0.5602,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5926234054353855,
+      "grad_norm": 0.1851152926683426,
+      "learning_rate": 2.0276342396362636e-05,
+      "loss": 0.5461,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5929007210205214,
+      "grad_norm": 0.18607455492019653,
+      "learning_rate": 2.0272018569110552e-05,
+      "loss": 0.542,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5931780366056573,
+      "grad_norm": 0.19023166596889496,
+      "learning_rate": 2.0267693225331726e-05,
+      "loss": 0.5564,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5934553521907932,
+      "grad_norm": 0.18977974355220795,
+      "learning_rate": 2.0263366365870152e-05,
+      "loss": 0.5778,
+      "step": 2140
+    },
+    {
+      "epoch": 0.593732667775929,
+      "grad_norm": 0.2097538560628891,
+      "learning_rate": 2.0259037991570116e-05,
+      "loss": 0.5822,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5940099833610649,
+      "grad_norm": 0.18402041494846344,
+      "learning_rate": 2.0254708103276193e-05,
+      "loss": 0.5207,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5942872989462008,
+      "grad_norm": 0.196335569024086,
+      "learning_rate": 2.025037670183326e-05,
+      "loss": 0.5509,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5945646145313367,
+      "grad_norm": 0.18417152762413025,
+      "learning_rate": 2.0246043788086498e-05,
+      "loss": 0.5418,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5948419301164726,
+      "grad_norm": 0.1963946372270584,
+      "learning_rate": 2.024170936288136e-05,
+      "loss": 0.5717,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5951192457016085,
+      "grad_norm": 0.20180796086788177,
+      "learning_rate": 2.023737342706361e-05,
+      "loss": 0.5457,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5953965612867443,
+      "grad_norm": 0.18826082348823547,
+      "learning_rate": 2.0233035981479316e-05,
+      "loss": 0.545,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5956738768718802,
+      "grad_norm": 0.1920921951532364,
+      "learning_rate": 2.0228697026974808e-05,
+      "loss": 0.5584,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5959511924570161,
+      "grad_norm": 0.20350618660449982,
+      "learning_rate": 2.0224356564396747e-05,
+      "loss": 0.5343,
+      "step": 2149
+    },
+    {
+      "epoch": 0.596228508042152,
+      "grad_norm": 0.19070284068584442,
+      "learning_rate": 2.0220014594592068e-05,
+      "loss": 0.5573,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5965058236272879,
+      "grad_norm": 0.1937059462070465,
+      "learning_rate": 2.0215671118408004e-05,
+      "loss": 0.5548,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5967831392124238,
+      "grad_norm": 0.18935304880142212,
+      "learning_rate": 2.021132613669208e-05,
+      "loss": 0.5746,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5970604547975596,
+      "grad_norm": 0.20879191160202026,
+      "learning_rate": 2.0206979650292117e-05,
+      "loss": 0.5602,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5973377703826955,
+      "grad_norm": 0.19923923909664154,
+      "learning_rate": 2.020263166005624e-05,
+      "loss": 0.5441,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5976150859678314,
+      "grad_norm": 0.1885758638381958,
+      "learning_rate": 2.019828216683284e-05,
+      "loss": 0.5362,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5978924015529673,
+      "grad_norm": 0.18912041187286377,
+      "learning_rate": 2.019393117147063e-05,
+      "loss": 0.5488,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5981697171381032,
+      "grad_norm": 0.18454459309577942,
+      "learning_rate": 2.0189578674818603e-05,
+      "loss": 0.5386,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5984470327232391,
+      "grad_norm": 0.1803792268037796,
+      "learning_rate": 2.018522467772604e-05,
+      "loss": 0.5444,
+      "step": 2158
+    },
+    {
+      "epoch": 0.598724348308375,
+      "grad_norm": 0.18980631232261658,
+      "learning_rate": 2.0180869181042532e-05,
+      "loss": 0.5521,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5990016638935108,
+      "grad_norm": 0.18810968101024628,
+      "learning_rate": 2.0176512185617945e-05,
+      "loss": 0.5342,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5992789794786467,
+      "grad_norm": 0.24775730073451996,
+      "learning_rate": 2.0172153692302445e-05,
+      "loss": 0.533,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5995562950637826,
+      "grad_norm": 0.19899478554725647,
+      "learning_rate": 2.0167793701946488e-05,
+      "loss": 0.5394,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5998336106489185,
+      "grad_norm": 0.19542957842350006,
+      "learning_rate": 2.0163432215400822e-05,
+      "loss": 0.5379,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6001109262340544,
+      "grad_norm": 0.2002883106470108,
+      "learning_rate": 2.0159069233516504e-05,
+      "loss": 0.5672,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6003882418191903,
+      "grad_norm": 0.18904021382331848,
+      "learning_rate": 2.0154704757144845e-05,
+      "loss": 0.5483,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6006655574043261,
+      "grad_norm": 0.19162911176681519,
+      "learning_rate": 2.0150338787137486e-05,
+      "loss": 0.5552,
+      "step": 2166
+    },
+    {
+      "epoch": 0.600942872989462,
+      "grad_norm": 0.18318617343902588,
+      "learning_rate": 2.014597132434633e-05,
+      "loss": 0.5548,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6012201885745979,
+      "grad_norm": 0.1864987015724182,
+      "learning_rate": 2.01416023696236e-05,
+      "loss": 0.5401,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6014975041597338,
+      "grad_norm": 0.18252375721931458,
+      "learning_rate": 2.0137231923821785e-05,
+      "loss": 0.5547,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6017748197448697,
+      "grad_norm": 0.196000874042511,
+      "learning_rate": 2.013285998779367e-05,
+      "loss": 0.5518,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6020521353300056,
+      "grad_norm": 0.17955927550792694,
+      "learning_rate": 2.0128486562392354e-05,
+      "loss": 0.5312,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6023294509151415,
+      "grad_norm": 0.1867658495903015,
+      "learning_rate": 2.0124111648471192e-05,
+      "loss": 0.5374,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6026067665002773,
+      "grad_norm": 0.20001055300235748,
+      "learning_rate": 2.0119735246883852e-05,
+      "loss": 0.5501,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6028840820854132,
+      "grad_norm": 0.19283756613731384,
+      "learning_rate": 2.011535735848428e-05,
+      "loss": 0.5505,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6031613976705491,
+      "grad_norm": 0.18642939627170563,
+      "learning_rate": 2.011097798412673e-05,
+      "loss": 0.5246,
+      "step": 2175
+    },
+    {
+      "epoch": 0.603438713255685,
+      "grad_norm": 0.1911175698041916,
+      "learning_rate": 2.0106597124665716e-05,
+      "loss": 0.5134,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6037160288408209,
+      "grad_norm": 0.18297746777534485,
+      "learning_rate": 2.0102214780956073e-05,
+      "loss": 0.5598,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6039933444259568,
+      "grad_norm": 0.19483189284801483,
+      "learning_rate": 2.0097830953852914e-05,
+      "loss": 0.5502,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6042706600110926,
+      "grad_norm": 0.18785777688026428,
+      "learning_rate": 2.009344564421163e-05,
+      "loss": 0.5384,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6045479755962285,
+      "grad_norm": 0.19045081734657288,
+      "learning_rate": 2.0089058852887923e-05,
+      "loss": 0.5311,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6048252911813644,
+      "grad_norm": 0.19081301987171173,
+      "learning_rate": 2.0084670580737758e-05,
+      "loss": 0.5593,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6051026067665003,
+      "grad_norm": 0.19843098521232605,
+      "learning_rate": 2.0080280828617414e-05,
+      "loss": 0.557,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6053799223516362,
+      "grad_norm": 0.1962948590517044,
+      "learning_rate": 2.0075889597383446e-05,
+      "loss": 0.5591,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6056572379367721,
+      "grad_norm": 0.193936288356781,
+      "learning_rate": 2.0071496887892693e-05,
+      "loss": 0.5721,
+      "step": 2184
+    },
+    {
+      "epoch": 0.605934553521908,
+      "grad_norm": 0.19298399984836578,
+      "learning_rate": 2.00671027010023e-05,
+      "loss": 0.5734,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6062118691070438,
+      "grad_norm": 0.20237858593463898,
+      "learning_rate": 2.006270703756968e-05,
+      "loss": 0.548,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6064891846921797,
+      "grad_norm": 0.19070473313331604,
+      "learning_rate": 2.0058309898452552e-05,
+      "loss": 0.5187,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6067665002773156,
+      "grad_norm": 0.18649962544441223,
+      "learning_rate": 2.0053911284508902e-05,
+      "loss": 0.5465,
+      "step": 2188
+    },
+    {
+      "epoch": 0.6070438158624515,
+      "grad_norm": 0.21800090372562408,
+      "learning_rate": 2.0049511196597027e-05,
+      "loss": 0.5665,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6073211314475874,
+      "grad_norm": 0.18259546160697937,
+      "learning_rate": 2.00451096355755e-05,
+      "loss": 0.5456,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6075984470327233,
+      "grad_norm": 0.19721095263957977,
+      "learning_rate": 2.0040706602303173e-05,
+      "loss": 0.5752,
+      "step": 2191
+    },
+    {
+      "epoch": 0.6078757626178591,
+      "grad_norm": 0.19083106517791748,
+      "learning_rate": 2.0036302097639204e-05,
+      "loss": 0.552,
+      "step": 2192
+    },
+    {
+      "epoch": 0.608153078202995,
+      "grad_norm": 0.18962670862674713,
+      "learning_rate": 2.0031896122443023e-05,
+      "loss": 0.5729,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6084303937881309,
+      "grad_norm": 0.19744956493377686,
+      "learning_rate": 2.0027488677574358e-05,
+      "loss": 0.5494,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6087077093732668,
+      "grad_norm": 0.19325025379657745,
+      "learning_rate": 2.0023079763893208e-05,
+      "loss": 0.5389,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6089850249584027,
+      "grad_norm": 0.1866646558046341,
+      "learning_rate": 2.0018669382259885e-05,
+      "loss": 0.5703,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6092623405435386,
+      "grad_norm": 0.21793098747730255,
+      "learning_rate": 2.001425753353496e-05,
+      "loss": 0.5585,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6095396561286744,
+      "grad_norm": 0.18613992631435394,
+      "learning_rate": 2.0009844218579298e-05,
+      "loss": 0.544,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6098169717138103,
+      "grad_norm": 0.1990228146314621,
+      "learning_rate": 2.0005429438254063e-05,
+      "loss": 0.5681,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6100942872989462,
+      "grad_norm": 0.19253648817539215,
+      "learning_rate": 2.000101319342069e-05,
+      "loss": 0.5451,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6103716028840821,
+      "grad_norm": 0.18924476206302643,
+      "learning_rate": 1.9996595484940915e-05,
+      "loss": 0.5684,
+      "step": 2201
+    },
+    {
+      "epoch": 0.610648918469218,
+      "grad_norm": 0.18552450835704803,
+      "learning_rate": 1.9992176313676737e-05,
+      "loss": 0.5307,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6109262340543539,
+      "grad_norm": 0.19528694450855255,
+      "learning_rate": 1.9987755680490456e-05,
+      "loss": 0.5598,
+      "step": 2203
+    },
+    {
+      "epoch": 0.6112035496394898,
+      "grad_norm": 0.1909715086221695,
+      "learning_rate": 1.998333358624466e-05,
+      "loss": 0.5647,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6114808652246256,
+      "grad_norm": 0.19347867369651794,
+      "learning_rate": 1.9978910031802218e-05,
+      "loss": 0.5726,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6117581808097615,
+      "grad_norm": 0.22693443298339844,
+      "learning_rate": 1.9974485018026273e-05,
+      "loss": 0.5403,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6120354963948974,
+      "grad_norm": 0.1854747086763382,
+      "learning_rate": 1.997005854578027e-05,
+      "loss": 0.5733,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6123128119800333,
+      "grad_norm": 0.2657473683357239,
+      "learning_rate": 1.9965630615927932e-05,
+      "loss": 0.5516,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6125901275651692,
+      "grad_norm": 0.19607201218605042,
+      "learning_rate": 1.996120122933326e-05,
+      "loss": 0.5525,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6128674431503051,
+      "grad_norm": 0.17779147624969482,
+      "learning_rate": 1.9956770386860547e-05,
+      "loss": 0.5377,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6131447587354409,
+      "grad_norm": 0.2034800499677658,
+      "learning_rate": 1.9952338089374366e-05,
+      "loss": 0.5434,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6134220743205768,
+      "grad_norm": 0.18624994158744812,
+      "learning_rate": 1.9947904337739582e-05,
+      "loss": 0.5301,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6136993899057127,
+      "grad_norm": 0.18844860792160034,
+      "learning_rate": 1.9943469132821334e-05,
+      "loss": 0.5508,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6139767054908486,
+      "grad_norm": 0.20432956516742706,
+      "learning_rate": 1.9939032475485043e-05,
+      "loss": 0.5565,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6142540210759845,
+      "grad_norm": 0.18925762176513672,
+      "learning_rate": 1.9934594366596423e-05,
+      "loss": 0.5745,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6145313366611204,
+      "grad_norm": 0.20550455152988434,
+      "learning_rate": 1.993015480702147e-05,
+      "loss": 0.5689,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6148086522462562,
+      "grad_norm": 0.18953469395637512,
+      "learning_rate": 1.992571379762645e-05,
+      "loss": 0.5393,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6150859678313921,
+      "grad_norm": 0.19926683604717255,
+      "learning_rate": 1.9921271339277935e-05,
+      "loss": 0.5563,
+      "step": 2218
+    },
+    {
+      "epoch": 0.615363283416528,
+      "grad_norm": 0.18314692378044128,
+      "learning_rate": 1.9916827432842756e-05,
+      "loss": 0.5353,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6156405990016639,
+      "grad_norm": 0.20692428946495056,
+      "learning_rate": 1.991238207918804e-05,
+      "loss": 0.57,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6159179145867998,
+      "grad_norm": 0.202706441283226,
+      "learning_rate": 1.990793527918119e-05,
+      "loss": 0.5496,
+      "step": 2221
+    },
+    {
+      "epoch": 0.6161952301719357,
+      "grad_norm": 0.18222178518772125,
+      "learning_rate": 1.99034870336899e-05,
+      "loss": 0.5553,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6164725457570716,
+      "grad_norm": 0.20487068593502045,
+      "learning_rate": 1.9899037343582135e-05,
+      "loss": 0.5611,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6167498613422074,
+      "grad_norm": 0.18165314197540283,
+      "learning_rate": 1.989458620972615e-05,
+      "loss": 0.5561,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6170271769273433,
+      "grad_norm": 0.1925044059753418,
+      "learning_rate": 1.9890133632990488e-05,
+      "loss": 0.5866,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6173044925124792,
+      "grad_norm": 0.19239196181297302,
+      "learning_rate": 1.988567961424395e-05,
+      "loss": 0.5629,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6175818080976151,
+      "grad_norm": 0.19511006772518158,
+      "learning_rate": 1.9881224154355638e-05,
+      "loss": 0.5484,
+      "step": 2227
+    },
+    {
+      "epoch": 0.617859123682751,
+      "grad_norm": 0.18884329497814178,
+      "learning_rate": 1.9876767254194932e-05,
+      "loss": 0.5541,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6181364392678869,
+      "grad_norm": 0.1917329877614975,
+      "learning_rate": 1.9872308914631494e-05,
+      "loss": 0.5535,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6184137548530227,
+      "grad_norm": 0.19358281791210175,
+      "learning_rate": 1.986784913653526e-05,
+      "loss": 0.5445,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6186910704381586,
+      "grad_norm": 0.1858266144990921,
+      "learning_rate": 1.9863387920776454e-05,
+      "loss": 0.5446,
+      "step": 2231
+    },
+    {
+      "epoch": 0.6189683860232945,
+      "grad_norm": 0.1879933476448059,
+      "learning_rate": 1.985892526822557e-05,
+      "loss": 0.5511,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6192457016084304,
+      "grad_norm": 0.19498036801815033,
+      "learning_rate": 1.9854461179753396e-05,
+      "loss": 0.5535,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6195230171935663,
+      "grad_norm": 0.1921449601650238,
+      "learning_rate": 1.9849995656231e-05,
+      "loss": 0.563,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6198003327787022,
+      "grad_norm": 0.1764960139989853,
+      "learning_rate": 1.984552869852971e-05,
+      "loss": 0.5604,
+      "step": 2235
+    },
+    {
+      "epoch": 0.620077648363838,
+      "grad_norm": 0.18862247467041016,
+      "learning_rate": 1.984106030752116e-05,
+      "loss": 0.5507,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6203549639489739,
+      "grad_norm": 0.19972002506256104,
+      "learning_rate": 1.9836590484077244e-05,
+      "loss": 0.5443,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6206322795341098,
+      "grad_norm": 0.18754172325134277,
+      "learning_rate": 1.983211922907014e-05,
+      "loss": 0.5361,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6209095951192457,
+      "grad_norm": 0.20183882117271423,
+      "learning_rate": 1.9827646543372322e-05,
+      "loss": 0.5537,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6211869107043816,
+      "grad_norm": 0.18864542245864868,
+      "learning_rate": 1.9823172427856518e-05,
+      "loss": 0.5515,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6214642262895175,
+      "grad_norm": 0.20876039564609528,
+      "learning_rate": 1.981869688339575e-05,
+      "loss": 0.5353,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6217415418746534,
+      "grad_norm": 0.20126941800117493,
+      "learning_rate": 1.9814219910863313e-05,
+      "loss": 0.5549,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6220188574597892,
+      "grad_norm": 0.1895267367362976,
+      "learning_rate": 1.9809741511132786e-05,
+      "loss": 0.5427,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6222961730449251,
+      "grad_norm": 0.1944306641817093,
+      "learning_rate": 1.980526168507802e-05,
+      "loss": 0.5627,
+      "step": 2244
+    },
+    {
+      "epoch": 0.622573488630061,
+      "grad_norm": 0.19707219302654266,
+      "learning_rate": 1.980078043357315e-05,
+      "loss": 0.573,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6228508042151969,
+      "grad_norm": 0.1903533786535263,
+      "learning_rate": 1.9796297757492587e-05,
+      "loss": 0.5534,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6231281198003328,
+      "grad_norm": 0.1989421546459198,
+      "learning_rate": 1.9791813657711022e-05,
+      "loss": 0.5343,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6234054353854687,
+      "grad_norm": 0.1919817328453064,
+      "learning_rate": 1.9787328135103418e-05,
+      "loss": 0.5493,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6236827509706045,
+      "grad_norm": 0.18703347444534302,
+      "learning_rate": 1.9782841190545024e-05,
+      "loss": 0.572,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6239600665557404,
+      "grad_norm": 0.18850034475326538,
+      "learning_rate": 1.9778352824911356e-05,
+      "loss": 0.5358,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6242373821408763,
+      "grad_norm": 0.22748717665672302,
+      "learning_rate": 1.9773863039078217e-05,
+      "loss": 0.5569,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6245146977260122,
+      "grad_norm": 0.1906166821718216,
+      "learning_rate": 1.976937183392168e-05,
+      "loss": 0.5226,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6247920133111481,
+      "grad_norm": 0.1964375227689743,
+      "learning_rate": 1.9764879210318098e-05,
+      "loss": 0.5467,
+      "step": 2253
+    },
+    {
+      "epoch": 0.625069328896284,
+      "grad_norm": 0.27289459109306335,
+      "learning_rate": 1.9760385169144108e-05,
+      "loss": 0.5597,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6253466444814199,
+      "grad_norm": 0.1879124641418457,
+      "learning_rate": 1.9755889711276603e-05,
+      "loss": 0.5509,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6256239600665557,
+      "grad_norm": 0.1970473676919937,
+      "learning_rate": 1.9751392837592782e-05,
+      "loss": 0.53,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6259012756516916,
+      "grad_norm": 0.20250020921230316,
+      "learning_rate": 1.9746894548970092e-05,
+      "loss": 0.5458,
+      "step": 2257
+    },
+    {
+      "epoch": 0.6261785912368275,
+      "grad_norm": 0.20085811614990234,
+      "learning_rate": 1.9742394846286277e-05,
+      "loss": 0.5483,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6264559068219634,
+      "grad_norm": 0.21487122774124146,
+      "learning_rate": 1.9737893730419337e-05,
+      "loss": 0.5522,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6267332224070993,
+      "grad_norm": 0.1989215910434723,
+      "learning_rate": 1.9733391202247577e-05,
+      "loss": 0.5269,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6270105379922352,
+      "grad_norm": 0.1842491775751114,
+      "learning_rate": 1.9728887262649536e-05,
+      "loss": 0.543,
+      "step": 2261
+    },
+    {
+      "epoch": 0.627287853577371,
+      "grad_norm": 0.1977192461490631,
+      "learning_rate": 1.972438191250407e-05,
+      "loss": 0.5457,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6275651691625069,
+      "grad_norm": 0.18922410905361176,
+      "learning_rate": 1.9719875152690288e-05,
+      "loss": 0.5508,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6278424847476428,
+      "grad_norm": 0.19666342437267303,
+      "learning_rate": 1.9715366984087575e-05,
+      "loss": 0.5359,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6281198003327787,
+      "grad_norm": 0.19078543782234192,
+      "learning_rate": 1.9710857407575595e-05,
+      "loss": 0.5349,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6283971159179146,
+      "grad_norm": 0.18407784402370453,
+      "learning_rate": 1.970634642403429e-05,
+      "loss": 0.5554,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6286744315030505,
+      "grad_norm": 0.1921215057373047,
+      "learning_rate": 1.9701834034343864e-05,
+      "loss": 0.583,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6289517470881864,
+      "grad_norm": 0.2015913873910904,
+      "learning_rate": 1.969732023938481e-05,
+      "loss": 0.5656,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6292290626733222,
+      "grad_norm": 0.19959089159965515,
+      "learning_rate": 1.969280504003789e-05,
+      "loss": 0.5404,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6295063782584581,
+      "grad_norm": 0.18149082362651825,
+      "learning_rate": 1.968828843718414e-05,
+      "loss": 0.5716,
+      "step": 2270
+    },
+    {
+      "epoch": 0.629783693843594,
+      "grad_norm": 0.20897513628005981,
+      "learning_rate": 1.9683770431704857e-05,
+      "loss": 0.5566,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6300610094287299,
+      "grad_norm": 0.18247532844543457,
+      "learning_rate": 1.9679251024481636e-05,
+      "loss": 0.5463,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6303383250138658,
+      "grad_norm": 0.18694794178009033,
+      "learning_rate": 1.9674730216396334e-05,
+      "loss": 0.5611,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6306156405990017,
+      "grad_norm": 0.1819789707660675,
+      "learning_rate": 1.9670208008331073e-05,
+      "loss": 0.5336,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6308929561841375,
+      "grad_norm": 0.18136881291866302,
+      "learning_rate": 1.9665684401168258e-05,
+      "loss": 0.5496,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6311702717692734,
+      "grad_norm": 0.1916753500699997,
+      "learning_rate": 1.9661159395790563e-05,
+      "loss": 0.56,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6314475873544093,
+      "grad_norm": 0.18821988999843597,
+      "learning_rate": 1.965663299308094e-05,
+      "loss": 0.5507,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6317249029395452,
+      "grad_norm": 0.2045837789773941,
+      "learning_rate": 1.965210519392261e-05,
+      "loss": 0.5495,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6320022185246811,
+      "grad_norm": 0.17228901386260986,
+      "learning_rate": 1.964757599919907e-05,
+      "loss": 0.5505,
+      "step": 2279
+    },
+    {
+      "epoch": 0.632279534109817,
+      "grad_norm": 0.19306735694408417,
+      "learning_rate": 1.9643045409794074e-05,
+      "loss": 0.5257,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6325568496949528,
+      "grad_norm": 0.19630911946296692,
+      "learning_rate": 1.9638513426591668e-05,
+      "loss": 0.5324,
+      "step": 2281
+    },
+    {
+      "epoch": 0.6328341652800887,
+      "grad_norm": 0.19862103462219238,
+      "learning_rate": 1.9633980050476164e-05,
+      "loss": 0.5556,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6331114808652246,
+      "grad_norm": 0.20335890352725983,
+      "learning_rate": 1.9629445282332136e-05,
+      "loss": 0.5583,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6333887964503605,
+      "grad_norm": 0.19751910865306854,
+      "learning_rate": 1.9624909123044448e-05,
+      "loss": 0.5844,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6336661120354964,
+      "grad_norm": 0.18284855782985687,
+      "learning_rate": 1.9620371573498212e-05,
+      "loss": 0.5393,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6339434276206323,
+      "grad_norm": 0.18531352281570435,
+      "learning_rate": 1.961583263457884e-05,
+      "loss": 0.5591,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6342207432057682,
+      "grad_norm": 0.18705891072750092,
+      "learning_rate": 1.9611292307171987e-05,
+      "loss": 0.5309,
+      "step": 2287
+    },
+    {
+      "epoch": 0.634498058790904,
+      "grad_norm": 0.18091407418251038,
+      "learning_rate": 1.9606750592163593e-05,
+      "loss": 0.5041,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6347753743760399,
+      "grad_norm": 0.19142916798591614,
+      "learning_rate": 1.960220749043987e-05,
+      "loss": 0.5601,
+      "step": 2289
+    },
+    {
+      "epoch": 0.6350526899611758,
+      "grad_norm": 0.18897870182991028,
+      "learning_rate": 1.9597663002887294e-05,
+      "loss": 0.5541,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6353300055463117,
+      "grad_norm": 0.19178354740142822,
+      "learning_rate": 1.959311713039262e-05,
+      "loss": 0.5217,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6356073211314476,
+      "grad_norm": 0.18749533593654633,
+      "learning_rate": 1.9588569873842864e-05,
+      "loss": 0.5291,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6358846367165835,
+      "grad_norm": 0.20121093094348907,
+      "learning_rate": 1.9584021234125323e-05,
+      "loss": 0.5414,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6361619523017193,
+      "grad_norm": 0.18628259003162384,
+      "learning_rate": 1.957947121212754e-05,
+      "loss": 0.511,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6364392678868552,
+      "grad_norm": 0.18616369366645813,
+      "learning_rate": 1.9574919808737364e-05,
+      "loss": 0.5496,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6367165834719911,
+      "grad_norm": 0.20249204337596893,
+      "learning_rate": 1.9570367024842888e-05,
+      "loss": 0.5585,
+      "step": 2296
+    },
+    {
+      "epoch": 0.636993899057127,
+      "grad_norm": 0.23987984657287598,
+      "learning_rate": 1.9565812861332477e-05,
+      "loss": 0.5659,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6372712146422629,
+      "grad_norm": 0.1872101128101349,
+      "learning_rate": 1.956125731909477e-05,
+      "loss": 0.5419,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6375485302273988,
+      "grad_norm": 0.18463543057441711,
+      "learning_rate": 1.955670039901868e-05,
+      "loss": 0.5302,
+      "step": 2299
+    },
+    {
+      "epoch": 0.6378258458125347,
+      "grad_norm": 0.20480939745903015,
+      "learning_rate": 1.955214210199338e-05,
+      "loss": 0.5343,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6381031613976705,
+      "grad_norm": 0.18215136229991913,
+      "learning_rate": 1.9547582428908306e-05,
+      "loss": 0.546,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6383804769828064,
+      "grad_norm": 0.17943714559078217,
+      "learning_rate": 1.954302138065318e-05,
+      "loss": 0.5334,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6386577925679423,
+      "grad_norm": 0.19718489050865173,
+      "learning_rate": 1.9538458958117982e-05,
+      "loss": 0.5621,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6389351081530782,
+      "grad_norm": 0.19121624529361725,
+      "learning_rate": 1.953389516219296e-05,
+      "loss": 0.5523,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6392124237382141,
+      "grad_norm": 0.1928092986345291,
+      "learning_rate": 1.9529329993768634e-05,
+      "loss": 0.5455,
+      "step": 2305
+    },
+    {
+      "epoch": 0.63948973932335,
+      "grad_norm": 0.1785450428724289,
+      "learning_rate": 1.952476345373579e-05,
+      "loss": 0.5643,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6397670549084858,
+      "grad_norm": 0.17965517938137054,
+      "learning_rate": 1.9520195542985476e-05,
+      "loss": 0.5266,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6400443704936217,
+      "grad_norm": 0.21014821529388428,
+      "learning_rate": 1.9515626262409016e-05,
+      "loss": 0.5327,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6403216860787576,
+      "grad_norm": 0.18984338641166687,
+      "learning_rate": 1.951105561289799e-05,
+      "loss": 0.5678,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6405990016638935,
+      "grad_norm": 0.18631823360919952,
+      "learning_rate": 1.9506483595344267e-05,
+      "loss": 0.5469,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6408763172490294,
+      "grad_norm": 0.18727704882621765,
+      "learning_rate": 1.9501910210639958e-05,
+      "loss": 0.5657,
+      "step": 2311
+    },
+    {
+      "epoch": 0.6411536328341653,
+      "grad_norm": 0.17807155847549438,
+      "learning_rate": 1.9497335459677458e-05,
+      "loss": 0.5044,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6414309484193012,
+      "grad_norm": 0.18843533098697662,
+      "learning_rate": 1.9492759343349415e-05,
+      "loss": 0.553,
+      "step": 2313
+    },
+    {
+      "epoch": 0.641708264004437,
+      "grad_norm": 0.1941610723733902,
+      "learning_rate": 1.9488181862548753e-05,
+      "loss": 0.587,
+      "step": 2314
+    },
+    {
+      "epoch": 0.6419855795895729,
+      "grad_norm": 0.1894078403711319,
+      "learning_rate": 1.9483603018168666e-05,
+      "loss": 0.5285,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6422628951747088,
+      "grad_norm": 0.19420726597309113,
+      "learning_rate": 1.9479022811102604e-05,
+      "loss": 0.5302,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6425402107598447,
+      "grad_norm": 0.205157071352005,
+      "learning_rate": 1.9474441242244284e-05,
+      "loss": 0.5539,
+      "step": 2317
+    },
+    {
+      "epoch": 0.6428175263449806,
+      "grad_norm": 0.1943119317293167,
+      "learning_rate": 1.9469858312487693e-05,
+      "loss": 0.5465,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6430948419301165,
+      "grad_norm": 0.19968454539775848,
+      "learning_rate": 1.946527402272708e-05,
+      "loss": 0.5438,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6433721575152523,
+      "grad_norm": 0.18584848940372467,
+      "learning_rate": 1.9460688373856967e-05,
+      "loss": 0.546,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6436494731003882,
+      "grad_norm": 0.17821067571640015,
+      "learning_rate": 1.945610136677213e-05,
+      "loss": 0.5148,
+      "step": 2321
+    },
+    {
+      "epoch": 0.6439267886855241,
+      "grad_norm": 0.19228345155715942,
+      "learning_rate": 1.945151300236762e-05,
+      "loss": 0.5368,
+      "step": 2322
+    },
+    {
+      "epoch": 0.64420410427066,
+      "grad_norm": 0.18330131471157074,
+      "learning_rate": 1.9446923281538747e-05,
+      "loss": 0.5611,
+      "step": 2323
+    },
+    {
+      "epoch": 0.6444814198557959,
+      "grad_norm": 0.18893574178218842,
+      "learning_rate": 1.9442332205181086e-05,
+      "loss": 0.54,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6447587354409318,
+      "grad_norm": 0.19229231774806976,
+      "learning_rate": 1.943773977419047e-05,
+      "loss": 0.5236,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6450360510260676,
+      "grad_norm": 0.20103448629379272,
+      "learning_rate": 1.9433145989463027e-05,
+      "loss": 0.554,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6453133666112035,
+      "grad_norm": 0.1895090788602829,
+      "learning_rate": 1.9428550851895098e-05,
+      "loss": 0.5676,
+      "step": 2327
+    },
+    {
+      "epoch": 0.6455906821963394,
+      "grad_norm": 0.18887649476528168,
+      "learning_rate": 1.9423954362383334e-05,
+      "loss": 0.5601,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6458679977814753,
+      "grad_norm": 0.2108272910118103,
+      "learning_rate": 1.941935652182463e-05,
+      "loss": 0.5748,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6461453133666112,
+      "grad_norm": 0.20968154072761536,
+      "learning_rate": 1.941475733111614e-05,
+      "loss": 0.5306,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6464226289517471,
+      "grad_norm": 0.18780824542045593,
+      "learning_rate": 1.9410156791155297e-05,
+      "loss": 0.5326,
+      "step": 2331
+    },
+    {
+      "epoch": 0.646699944536883,
+      "grad_norm": 0.19030767679214478,
+      "learning_rate": 1.9405554902839778e-05,
+      "loss": 0.5533,
+      "step": 2332
+    },
+    {
+      "epoch": 0.6469772601220188,
+      "grad_norm": 0.1849377304315567,
+      "learning_rate": 1.9400951667067542e-05,
+      "loss": 0.571,
+      "step": 2333
+    },
+    {
+      "epoch": 0.6472545757071547,
+      "grad_norm": 0.1965888887643814,
+      "learning_rate": 1.9396347084736794e-05,
+      "loss": 0.5591,
+      "step": 2334
+    },
+    {
+      "epoch": 0.6475318912922906,
+      "grad_norm": 0.1921495646238327,
+      "learning_rate": 1.9391741156746013e-05,
+      "loss": 0.5723,
+      "step": 2335
+    },
+    {
+      "epoch": 0.6478092068774265,
+      "grad_norm": 0.20073464512825012,
+      "learning_rate": 1.9387133883993948e-05,
+      "loss": 0.535,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6480865224625624,
+      "grad_norm": 0.18834145367145538,
+      "learning_rate": 1.938252526737958e-05,
+      "loss": 0.5256,
+      "step": 2337
+    },
+    {
+      "epoch": 0.6483638380476983,
+      "grad_norm": 0.1929401457309723,
+      "learning_rate": 1.9377915307802192e-05,
+      "loss": 0.5204,
+      "step": 2338
+    },
+    {
+      "epoch": 0.6486411536328341,
+      "grad_norm": 0.19703806936740875,
+      "learning_rate": 1.9373304006161298e-05,
+      "loss": 0.5603,
+      "step": 2339
+    },
+    {
+      "epoch": 0.64891846921797,
+      "grad_norm": 0.205661803483963,
+      "learning_rate": 1.9368691363356682e-05,
+      "loss": 0.522,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6491957848031059,
+      "grad_norm": 0.19013790786266327,
+      "learning_rate": 1.9364077380288408e-05,
+      "loss": 0.549,
+      "step": 2341
+    },
+    {
+      "epoch": 0.6494731003882418,
+      "grad_norm": 0.18775691092014313,
+      "learning_rate": 1.935946205785677e-05,
+      "loss": 0.5389,
+      "step": 2342
+    },
+    {
+      "epoch": 0.6497504159733777,
+      "grad_norm": 0.18783038854599,
+      "learning_rate": 1.9354845396962353e-05,
+      "loss": 0.5378,
+      "step": 2343
+    },
+    {
+      "epoch": 0.6500277315585136,
+      "grad_norm": 0.19113753736019135,
+      "learning_rate": 1.9350227398505976e-05,
+      "loss": 0.5461,
+      "step": 2344
+    },
+    {
+      "epoch": 0.6503050471436495,
+      "grad_norm": 0.17977707087993622,
+      "learning_rate": 1.9345608063388742e-05,
+      "loss": 0.5512,
+      "step": 2345
+    },
+    {
+      "epoch": 0.6505823627287853,
+      "grad_norm": 0.20450885593891144,
+      "learning_rate": 1.9340987392512006e-05,
+      "loss": 0.5501,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6508596783139212,
+      "grad_norm": 0.2484101504087448,
+      "learning_rate": 1.9336365386777376e-05,
+      "loss": 0.5223,
+      "step": 2347
+    },
+    {
+      "epoch": 0.6511369938990571,
+      "grad_norm": 0.20386487245559692,
+      "learning_rate": 1.9331742047086743e-05,
+      "loss": 0.5262,
+      "step": 2348
+    },
+    {
+      "epoch": 0.651414309484193,
+      "grad_norm": 0.18846935033798218,
+      "learning_rate": 1.9327117374342223e-05,
+      "loss": 0.5347,
+      "step": 2349
+    },
+    {
+      "epoch": 0.6516916250693289,
+      "grad_norm": 0.18970006704330444,
+      "learning_rate": 1.932249136944623e-05,
+      "loss": 0.5683,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6519689406544648,
+      "grad_norm": 0.1962558776140213,
+      "learning_rate": 1.9317864033301407e-05,
+      "loss": 0.5791,
+      "step": 2351
+    },
+    {
+      "epoch": 0.6522462562396006,
+      "grad_norm": 0.20231659710407257,
+      "learning_rate": 1.9313235366810676e-05,
+      "loss": 0.5627,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6525235718247365,
+      "grad_norm": 0.1958416849374771,
+      "learning_rate": 1.9308605370877215e-05,
+      "loss": 0.5375,
+      "step": 2353
+    },
+    {
+      "epoch": 0.6528008874098724,
+      "grad_norm": 0.20812073349952698,
+      "learning_rate": 1.9303974046404455e-05,
+      "loss": 0.5299,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6530782029950083,
+      "grad_norm": 0.1921248584985733,
+      "learning_rate": 1.929934139429609e-05,
+      "loss": 0.5544,
+      "step": 2355
+    },
+    {
+      "epoch": 0.6533555185801442,
+      "grad_norm": 0.20390520989894867,
+      "learning_rate": 1.929470741545607e-05,
+      "loss": 0.581,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6536328341652801,
+      "grad_norm": 0.18480364978313446,
+      "learning_rate": 1.9290072110788616e-05,
+      "loss": 0.5599,
+      "step": 2357
+    },
+    {
+      "epoch": 0.653910149750416,
+      "grad_norm": 0.1953095942735672,
+      "learning_rate": 1.928543548119819e-05,
+      "loss": 0.5961,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6541874653355518,
+      "grad_norm": 0.19727711379528046,
+      "learning_rate": 1.9280797527589527e-05,
+      "loss": 0.5585,
+      "step": 2359
+    },
+    {
+      "epoch": 0.6544647809206877,
+      "grad_norm": 0.19928644597530365,
+      "learning_rate": 1.927615825086761e-05,
+      "loss": 0.5522,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6547420965058236,
+      "grad_norm": 0.1891396939754486,
+      "learning_rate": 1.9271517651937688e-05,
+      "loss": 0.5586,
+      "step": 2361
+    },
+    {
+      "epoch": 0.6550194120909595,
+      "grad_norm": 0.1949121057987213,
+      "learning_rate": 1.9266875731705266e-05,
+      "loss": 0.5307,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6552967276760954,
+      "grad_norm": 0.21069341897964478,
+      "learning_rate": 1.9262232491076104e-05,
+      "loss": 0.5546,
+      "step": 2363
+    },
+    {
+      "epoch": 0.6555740432612313,
+      "grad_norm": 0.18490912020206451,
+      "learning_rate": 1.925758793095622e-05,
+      "loss": 0.5613,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6558513588463671,
+      "grad_norm": 0.1878899782896042,
+      "learning_rate": 1.9252942052251892e-05,
+      "loss": 0.5497,
+      "step": 2365
+    },
+    {
+      "epoch": 0.656128674431503,
+      "grad_norm": 0.19967246055603027,
+      "learning_rate": 1.9248294855869653e-05,
+      "loss": 0.5327,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6564059900166389,
+      "grad_norm": 0.19277790188789368,
+      "learning_rate": 1.9243646342716296e-05,
+      "loss": 0.5545,
+      "step": 2367
+    },
+    {
+      "epoch": 0.6566833056017748,
+      "grad_norm": 0.19045887887477875,
+      "learning_rate": 1.9238996513698864e-05,
+      "loss": 0.5071,
+      "step": 2368
+    },
+    {
+      "epoch": 0.6569606211869107,
+      "grad_norm": 0.19380688667297363,
+      "learning_rate": 1.923434536972467e-05,
+      "loss": 0.5437,
+      "step": 2369
+    },
+    {
+      "epoch": 0.6572379367720466,
+      "grad_norm": 0.18202729523181915,
+      "learning_rate": 1.9229692911701275e-05,
+      "loss": 0.5356,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6575152523571824,
+      "grad_norm": 0.18908429145812988,
+      "learning_rate": 1.9225039140536488e-05,
+      "loss": 0.5526,
+      "step": 2371
+    },
+    {
+      "epoch": 0.6577925679423183,
+      "grad_norm": 0.21290616691112518,
+      "learning_rate": 1.9220384057138386e-05,
+      "loss": 0.5306,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6580698835274542,
+      "grad_norm": 0.18612539768218994,
+      "learning_rate": 1.9215727662415303e-05,
+      "loss": 0.5387,
+      "step": 2373
+    },
+    {
+      "epoch": 0.6583471991125901,
+      "grad_norm": 0.182894766330719,
+      "learning_rate": 1.9211069957275822e-05,
+      "loss": 0.5365,
+      "step": 2374
+    },
+    {
+      "epoch": 0.658624514697726,
+      "grad_norm": 0.1848146617412567,
+      "learning_rate": 1.920641094262879e-05,
+      "loss": 0.5423,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6589018302828619,
+      "grad_norm": 0.18521788716316223,
+      "learning_rate": 1.92017506193833e-05,
+      "loss": 0.5534,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6591791458679978,
+      "grad_norm": 0.1895231306552887,
+      "learning_rate": 1.9197088988448703e-05,
+      "loss": 0.5447,
+      "step": 2377
+    },
+    {
+      "epoch": 0.6594564614531336,
+      "grad_norm": 0.18669599294662476,
+      "learning_rate": 1.9192426050734608e-05,
+      "loss": 0.5446,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6597337770382695,
+      "grad_norm": 0.1867615282535553,
+      "learning_rate": 1.9187761807150878e-05,
+      "loss": 0.5269,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6600110926234054,
+      "grad_norm": 0.19190800189971924,
+      "learning_rate": 1.918309625860763e-05,
+      "loss": 0.5833,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6602884082085413,
+      "grad_norm": 0.18087397515773773,
+      "learning_rate": 1.917842940601524e-05,
+      "loss": 0.5379,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6605657237936772,
+      "grad_norm": 0.18190105259418488,
+      "learning_rate": 1.9173761250284324e-05,
+      "loss": 0.5489,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6608430393788131,
+      "grad_norm": 0.19485372304916382,
+      "learning_rate": 1.9169091792325777e-05,
+      "loss": 0.5687,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6611203549639489,
+      "grad_norm": 0.19502972066402435,
+      "learning_rate": 1.9164421033050724e-05,
+      "loss": 0.5445,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6613976705490848,
+      "grad_norm": 0.18996240198612213,
+      "learning_rate": 1.915974897337056e-05,
+      "loss": 0.5621,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6616749861342207,
+      "grad_norm": 0.19751591980457306,
+      "learning_rate": 1.915507561419692e-05,
+      "loss": 0.5468,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6619523017193566,
+      "grad_norm": 0.2202579826116562,
+      "learning_rate": 1.915040095644171e-05,
+      "loss": 0.5515,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6622296173044925,
+      "grad_norm": 0.1802307367324829,
+      "learning_rate": 1.914572500101707e-05,
+      "loss": 0.5449,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6625069328896284,
+      "grad_norm": 0.18632298707962036,
+      "learning_rate": 1.914104774883541e-05,
+      "loss": 0.5612,
+      "step": 2389
+    },
+    {
+      "epoch": 0.6627842484747642,
+      "grad_norm": 0.19929082691669464,
+      "learning_rate": 1.9136369200809378e-05,
+      "loss": 0.5789,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6630615640599001,
+      "grad_norm": 0.18650726974010468,
+      "learning_rate": 1.913168935785189e-05,
+      "loss": 0.5618,
+      "step": 2391
+    },
+    {
+      "epoch": 0.663338879645036,
+      "grad_norm": 0.1912173330783844,
+      "learning_rate": 1.912700822087611e-05,
+      "loss": 0.5487,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6636161952301719,
+      "grad_norm": 0.19155828654766083,
+      "learning_rate": 1.912232579079544e-05,
+      "loss": 0.5421,
+      "step": 2393
+    },
+    {
+      "epoch": 0.6638935108153078,
+      "grad_norm": 0.21673326194286346,
+      "learning_rate": 1.9117642068523556e-05,
+      "loss": 0.5302,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6641708264004437,
+      "grad_norm": 0.1941951960325241,
+      "learning_rate": 1.9112957054974373e-05,
+      "loss": 0.5558,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6644481419855796,
+      "grad_norm": 0.18575075268745422,
+      "learning_rate": 1.9108270751062064e-05,
+      "loss": 0.5737,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6647254575707154,
+      "grad_norm": 0.19413797557353973,
+      "learning_rate": 1.9103583157701046e-05,
+      "loss": 0.5559,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6650027731558513,
+      "grad_norm": 0.19030508399009705,
+      "learning_rate": 1.9098894275805994e-05,
+      "loss": 0.5519,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6652800887409872,
+      "grad_norm": 0.18635134398937225,
+      "learning_rate": 1.9094204106291842e-05,
+      "loss": 0.5203,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6655574043261231,
+      "grad_norm": 0.1943938434123993,
+      "learning_rate": 1.908951265007375e-05,
+      "loss": 0.5664,
+      "step": 2400
+    },
+    {
+      "epoch": 0.665834719911259,
+      "grad_norm": 0.2069421112537384,
+      "learning_rate": 1.9084819908067156e-05,
+      "loss": 0.5561,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6661120354963949,
+      "grad_norm": 0.18940883874893188,
+      "learning_rate": 1.9080125881187737e-05,
+      "loss": 0.5833,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6663893510815307,
+      "grad_norm": 0.1828288733959198,
+      "learning_rate": 1.907543057035142e-05,
+      "loss": 0.5478,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.18825417757034302,
+      "learning_rate": 1.907073397647439e-05,
+      "loss": 0.5362,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6669439822518025,
+      "grad_norm": 0.19599126279354095,
+      "learning_rate": 1.906603610047307e-05,
+      "loss": 0.5425,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6672212978369384,
+      "grad_norm": 0.1776115894317627,
+      "learning_rate": 1.9061336943264145e-05,
+      "loss": 0.5395,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6674986134220743,
+      "grad_norm": 0.20096100866794586,
+      "learning_rate": 1.905663650576454e-05,
+      "loss": 0.547,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6677759290072102,
+      "grad_norm": 0.1988787204027176,
+      "learning_rate": 1.9051934788891443e-05,
+      "loss": 0.5336,
+      "step": 2408
+    },
+    {
+      "epoch": 0.668053244592346,
+      "grad_norm": 0.19204209744930267,
+      "learning_rate": 1.9047231793562276e-05,
+      "loss": 0.5449,
+      "step": 2409
+    },
+    {
+      "epoch": 0.6683305601774819,
+      "grad_norm": 0.1779128760099411,
+      "learning_rate": 1.904252752069472e-05,
+      "loss": 0.5023,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6686078757626178,
+      "grad_norm": 0.2877557575702667,
+      "learning_rate": 1.9037821971206703e-05,
+      "loss": 0.5446,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6688851913477537,
+      "grad_norm": 0.19600874185562134,
+      "learning_rate": 1.90331151460164e-05,
+      "loss": 0.5275,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6691625069328896,
+      "grad_norm": 0.28246966004371643,
+      "learning_rate": 1.9028407046042246e-05,
+      "loss": 0.5321,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6694398225180255,
+      "grad_norm": 0.1964629739522934,
+      "learning_rate": 1.9023697672202905e-05,
+      "loss": 0.5727,
+      "step": 2414
+    },
+    {
+      "epoch": 0.6697171381031614,
+      "grad_norm": 0.17945913970470428,
+      "learning_rate": 1.901898702541731e-05,
+      "loss": 0.5237,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6699944536882972,
+      "grad_norm": 0.190501868724823,
+      "learning_rate": 1.901427510660463e-05,
+      "loss": 0.5491,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6702717692734331,
+      "grad_norm": 0.19791793823242188,
+      "learning_rate": 1.9009561916684282e-05,
+      "loss": 0.5586,
+      "step": 2417
+    },
+    {
+      "epoch": 0.670549084858569,
+      "grad_norm": 0.18914659321308136,
+      "learning_rate": 1.900484745657594e-05,
+      "loss": 0.5302,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6708264004437049,
+      "grad_norm": 0.1981426477432251,
+      "learning_rate": 1.9000131727199513e-05,
+      "loss": 0.5609,
+      "step": 2419
+    },
+    {
+      "epoch": 0.6711037160288408,
+      "grad_norm": 0.1939757615327835,
+      "learning_rate": 1.8995414729475165e-05,
+      "loss": 0.5749,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6713810316139767,
+      "grad_norm": 0.19931401312351227,
+      "learning_rate": 1.899069646432332e-05,
+      "loss": 0.547,
+      "step": 2421
+    },
+    {
+      "epoch": 0.6716583471991125,
+      "grad_norm": 0.19219137728214264,
+      "learning_rate": 1.898597693266462e-05,
+      "loss": 0.5426,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6719356627842484,
+      "grad_norm": 0.199588343501091,
+      "learning_rate": 1.898125613541998e-05,
+      "loss": 0.5411,
+      "step": 2423
+    },
+    {
+      "epoch": 0.6722129783693843,
+      "grad_norm": 0.19906532764434814,
+      "learning_rate": 1.897653407351055e-05,
+      "loss": 0.5707,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6724902939545202,
+      "grad_norm": 0.18722088634967804,
+      "learning_rate": 1.8971810747857726e-05,
+      "loss": 0.5623,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6727676095396561,
+      "grad_norm": 0.18101942539215088,
+      "learning_rate": 1.8967086159383162e-05,
+      "loss": 0.5519,
+      "step": 2426
+    },
+    {
+      "epoch": 0.673044925124792,
+      "grad_norm": 0.19272929430007935,
+      "learning_rate": 1.8962360309008746e-05,
+      "loss": 0.5413,
+      "step": 2427
+    },
+    {
+      "epoch": 0.6733222407099279,
+      "grad_norm": 0.19695578515529633,
+      "learning_rate": 1.8957633197656615e-05,
+      "loss": 0.5299,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6735995562950637,
+      "grad_norm": 0.19021473824977875,
+      "learning_rate": 1.8952904826249158e-05,
+      "loss": 0.5453,
+      "step": 2429
+    },
+    {
+      "epoch": 0.6738768718801996,
+      "grad_norm": 0.1905011683702469,
+      "learning_rate": 1.8948175195709e-05,
+      "loss": 0.5458,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6741541874653355,
+      "grad_norm": 0.1979636549949646,
+      "learning_rate": 1.8943444306959017e-05,
+      "loss": 0.5379,
+      "step": 2431
+    },
+    {
+      "epoch": 0.6744315030504714,
+      "grad_norm": 0.19846384227275848,
+      "learning_rate": 1.8938712160922343e-05,
+      "loss": 0.5164,
+      "step": 2432
+    },
+    {
+      "epoch": 0.6747088186356073,
+      "grad_norm": 0.18356280028820038,
+      "learning_rate": 1.893397875852233e-05,
+      "loss": 0.5429,
+      "step": 2433
+    },
+    {
+      "epoch": 0.6749861342207432,
+      "grad_norm": 0.1836164891719818,
+      "learning_rate": 1.8929244100682597e-05,
+      "loss": 0.5569,
+      "step": 2434
+    },
+    {
+      "epoch": 0.675263449805879,
+      "grad_norm": 0.20087088644504547,
+      "learning_rate": 1.8924508188327e-05,
+      "loss": 0.5676,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6755407653910149,
+      "grad_norm": 0.1871204376220703,
+      "learning_rate": 1.891977102237964e-05,
+      "loss": 0.5314,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6758180809761508,
+      "grad_norm": 0.19372668862342834,
+      "learning_rate": 1.891503260376487e-05,
+      "loss": 0.543,
+      "step": 2437
+    },
+    {
+      "epoch": 0.6760953965612867,
+      "grad_norm": 0.20033282041549683,
+      "learning_rate": 1.891029293340727e-05,
+      "loss": 0.5395,
+      "step": 2438
+    },
+    {
+      "epoch": 0.6763727121464226,
+      "grad_norm": 0.1941455453634262,
+      "learning_rate": 1.8905552012231684e-05,
+      "loss": 0.5775,
+      "step": 2439
+    },
+    {
+      "epoch": 0.6766500277315585,
+      "grad_norm": 0.18717962503433228,
+      "learning_rate": 1.890080984116319e-05,
+      "loss": 0.5305,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6769273433166944,
+      "grad_norm": 0.1877082884311676,
+      "learning_rate": 1.8896066421127106e-05,
+      "loss": 0.5438,
+      "step": 2441
+    },
+    {
+      "epoch": 0.6772046589018302,
+      "grad_norm": 0.19558964669704437,
+      "learning_rate": 1.8891321753049008e-05,
+      "loss": 0.5452,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6774819744869661,
+      "grad_norm": 0.19550803303718567,
+      "learning_rate": 1.8886575837854696e-05,
+      "loss": 0.5409,
+      "step": 2443
+    },
+    {
+      "epoch": 0.677759290072102,
+      "grad_norm": 0.18728572130203247,
+      "learning_rate": 1.888182867647023e-05,
+      "loss": 0.5572,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6780366056572379,
+      "grad_norm": 0.184623122215271,
+      "learning_rate": 1.8877080269821906e-05,
+      "loss": 0.565,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6783139212423738,
+      "grad_norm": 0.19215163588523865,
+      "learning_rate": 1.8872330618836265e-05,
+      "loss": 0.532,
+      "step": 2446
+    },
+    {
+      "epoch": 0.6785912368275097,
+      "grad_norm": 0.19430597126483917,
+      "learning_rate": 1.886757972444009e-05,
+      "loss": 0.5409,
+      "step": 2447
+    },
+    {
+      "epoch": 0.6788685524126455,
+      "grad_norm": 0.1996561586856842,
+      "learning_rate": 1.88628275875604e-05,
+      "loss": 0.5514,
+      "step": 2448
+    },
+    {
+      "epoch": 0.6791458679977814,
+      "grad_norm": 0.20458458364009857,
+      "learning_rate": 1.8858074209124473e-05,
+      "loss": 0.5566,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6794231835829173,
+      "grad_norm": 0.20288583636283875,
+      "learning_rate": 1.885331959005981e-05,
+      "loss": 0.5237,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6797004991680532,
+      "grad_norm": 0.18585625290870667,
+      "learning_rate": 1.8848563731294172e-05,
+      "loss": 0.53,
+      "step": 2451
+    },
+    {
+      "epoch": 0.6799778147531891,
+      "grad_norm": 0.22308149933815002,
+      "learning_rate": 1.8843806633755544e-05,
+      "loss": 0.5513,
+      "step": 2452
+    },
+    {
+      "epoch": 0.680255130338325,
+      "grad_norm": 0.18869346380233765,
+      "learning_rate": 1.8839048298372165e-05,
+      "loss": 0.5527,
+      "step": 2453
+    },
+    {
+      "epoch": 0.6805324459234608,
+      "grad_norm": 0.1881789267063141,
+      "learning_rate": 1.8834288726072513e-05,
+      "loss": 0.5368,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6808097615085967,
+      "grad_norm": 0.18957830965518951,
+      "learning_rate": 1.882952791778531e-05,
+      "loss": 0.5235,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6810870770937326,
+      "grad_norm": 0.1874406933784485,
+      "learning_rate": 1.882476587443951e-05,
+      "loss": 0.5164,
+      "step": 2456
+    },
+    {
+      "epoch": 0.6813643926788685,
+      "grad_norm": 0.19134515523910522,
+      "learning_rate": 1.8820002596964316e-05,
+      "loss": 0.5606,
+      "step": 2457
+    },
+    {
+      "epoch": 0.6816417082640044,
+      "grad_norm": 0.18497633934020996,
+      "learning_rate": 1.881523808628917e-05,
+      "loss": 0.5681,
+      "step": 2458
+    },
+    {
+      "epoch": 0.6819190238491403,
+      "grad_norm": 0.18682947754859924,
+      "learning_rate": 1.881047234334376e-05,
+      "loss": 0.55,
+      "step": 2459
+    },
+    {
+      "epoch": 0.6821963394342762,
+      "grad_norm": 0.1982649862766266,
+      "learning_rate": 1.8805705369057993e-05,
+      "loss": 0.5321,
+      "step": 2460
+    },
+    {
+      "epoch": 0.682473655019412,
+      "grad_norm": 0.19017384946346283,
+      "learning_rate": 1.880093716436205e-05,
+      "loss": 0.5295,
+      "step": 2461
+    },
+    {
+      "epoch": 0.6827509706045479,
+      "grad_norm": 0.20334112644195557,
+      "learning_rate": 1.8796167730186322e-05,
+      "loss": 0.5601,
+      "step": 2462
+    },
+    {
+      "epoch": 0.6830282861896838,
+      "grad_norm": 0.1974753588438034,
+      "learning_rate": 1.8791397067461457e-05,
+      "loss": 0.5572,
+      "step": 2463
+    },
+    {
+      "epoch": 0.6833056017748197,
+      "grad_norm": 0.17885488271713257,
+      "learning_rate": 1.878662517711834e-05,
+      "loss": 0.5245,
+      "step": 2464
+    },
+    {
+      "epoch": 0.6835829173599556,
+      "grad_norm": 0.18409696221351624,
+      "learning_rate": 1.8781852060088083e-05,
+      "loss": 0.5321,
+      "step": 2465
+    },
+    {
+      "epoch": 0.6838602329450915,
+      "grad_norm": 0.19201841950416565,
+      "learning_rate": 1.877707771730206e-05,
+      "loss": 0.5563,
+      "step": 2466
+    },
+    {
+      "epoch": 0.6841375485302273,
+      "grad_norm": 0.1807066947221756,
+      "learning_rate": 1.8772302149691866e-05,
+      "loss": 0.5253,
+      "step": 2467
+    },
+    {
+      "epoch": 0.6844148641153632,
+      "grad_norm": 0.18462277948856354,
+      "learning_rate": 1.8767525358189343e-05,
+      "loss": 0.5315,
+      "step": 2468
+    },
+    {
+      "epoch": 0.6846921797004991,
+      "grad_norm": 0.18250201642513275,
+      "learning_rate": 1.876274734372656e-05,
+      "loss": 0.5383,
+      "step": 2469
+    },
+    {
+      "epoch": 0.684969495285635,
+      "grad_norm": 0.1986282765865326,
+      "learning_rate": 1.8757968107235853e-05,
+      "loss": 0.5282,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6852468108707709,
+      "grad_norm": 0.26469552516937256,
+      "learning_rate": 1.8753187649649757e-05,
+      "loss": 0.5564,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6855241264559068,
+      "grad_norm": 0.19594305753707886,
+      "learning_rate": 1.874840597190108e-05,
+      "loss": 0.5546,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6858014420410427,
+      "grad_norm": 0.18754026293754578,
+      "learning_rate": 1.8743623074922843e-05,
+      "loss": 0.5309,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6860787576261785,
+      "grad_norm": 0.1846148520708084,
+      "learning_rate": 1.873883895964833e-05,
+      "loss": 0.5288,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6863560732113144,
+      "grad_norm": 0.18642264604568481,
+      "learning_rate": 1.873405362701104e-05,
+      "loss": 0.5316,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6866333887964503,
+      "grad_norm": 0.20203615725040436,
+      "learning_rate": 1.8729267077944717e-05,
+      "loss": 0.5235,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6869107043815862,
+      "grad_norm": 0.18540050089359283,
+      "learning_rate": 1.872447931338335e-05,
+      "loss": 0.5488,
+      "step": 2477
+    },
+    {
+      "epoch": 0.6871880199667221,
+      "grad_norm": 0.19175854325294495,
+      "learning_rate": 1.8719690334261148e-05,
+      "loss": 0.5529,
+      "step": 2478
+    },
+    {
+      "epoch": 0.687465335551858,
+      "grad_norm": 0.18168555200099945,
+      "learning_rate": 1.8714900141512574e-05,
+      "loss": 0.5119,
+      "step": 2479
+    },
+    {
+      "epoch": 0.687742651136994,
+      "grad_norm": 0.1855335384607315,
+      "learning_rate": 1.871010873607233e-05,
+      "loss": 0.5448,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6880199667221298,
+      "grad_norm": 0.1834007054567337,
+      "learning_rate": 1.870531611887533e-05,
+      "loss": 0.5583,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6882972823072657,
+      "grad_norm": 0.1926104575395584,
+      "learning_rate": 1.870052229085675e-05,
+      "loss": 0.5549,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6885745978924016,
+      "grad_norm": 0.19106236100196838,
+      "learning_rate": 1.8695727252951995e-05,
+      "loss": 0.5146,
+      "step": 2483
+    },
+    {
+      "epoch": 0.6888519134775375,
+      "grad_norm": 0.18813811242580414,
+      "learning_rate": 1.8690931006096695e-05,
+      "loss": 0.5773,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6891292290626734,
+      "grad_norm": 0.19836729764938354,
+      "learning_rate": 1.8686133551226735e-05,
+      "loss": 0.5793,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6894065446478093,
+      "grad_norm": 0.1816731095314026,
+      "learning_rate": 1.8681334889278217e-05,
+      "loss": 0.5205,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6896838602329451,
+      "grad_norm": 0.1861170530319214,
+      "learning_rate": 1.8676535021187495e-05,
+      "loss": 0.5589,
+      "step": 2487
+    },
+    {
+      "epoch": 0.689961175818081,
+      "grad_norm": 0.18680687248706818,
+      "learning_rate": 1.867173394789114e-05,
+      "loss": 0.5283,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6902384914032169,
+      "grad_norm": 0.19478975236415863,
+      "learning_rate": 1.866693167032598e-05,
+      "loss": 0.5559,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6905158069883528,
+      "grad_norm": 0.19333085417747498,
+      "learning_rate": 1.8662128189429058e-05,
+      "loss": 0.516,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6907931225734887,
+      "grad_norm": 0.19102855026721954,
+      "learning_rate": 1.8657323506137668e-05,
+      "loss": 0.5587,
+      "step": 2491
+    },
+    {
+      "epoch": 0.6910704381586246,
+      "grad_norm": 0.19927440583705902,
+      "learning_rate": 1.8652517621389324e-05,
+      "loss": 0.5475,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6913477537437605,
+      "grad_norm": 0.19673167169094086,
+      "learning_rate": 1.8647710536121784e-05,
+      "loss": 0.5504,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6916250693288963,
+      "grad_norm": 0.19291207194328308,
+      "learning_rate": 1.8642902251273038e-05,
+      "loss": 0.555,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6919023849140322,
+      "grad_norm": 0.22686271369457245,
+      "learning_rate": 1.863809276778131e-05,
+      "loss": 0.5169,
+      "step": 2495
+    },
+    {
+      "epoch": 0.6921797004991681,
+      "grad_norm": 0.19432714581489563,
+      "learning_rate": 1.8633282086585057e-05,
+      "loss": 0.5287,
+      "step": 2496
+    },
+    {
+      "epoch": 0.692457016084304,
+      "grad_norm": 0.18512091040611267,
+      "learning_rate": 1.8628470208622972e-05,
+      "loss": 0.5452,
+      "step": 2497
+    },
+    {
+      "epoch": 0.6927343316694399,
+      "grad_norm": 0.18711940944194794,
+      "learning_rate": 1.8623657134833976e-05,
+      "loss": 0.5489,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6930116472545758,
+      "grad_norm": 0.19019177556037903,
+      "learning_rate": 1.8618842866157234e-05,
+      "loss": 0.5188,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6932889628397116,
+      "grad_norm": 0.18888919055461884,
+      "learning_rate": 1.861402740353213e-05,
+      "loss": 0.5563,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6935662784248475,
+      "grad_norm": 0.19487237930297852,
+      "learning_rate": 1.8609210747898293e-05,
+      "loss": 0.5627,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6938435940099834,
+      "grad_norm": 0.18301233649253845,
+      "learning_rate": 1.8604392900195573e-05,
+      "loss": 0.5539,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6941209095951193,
+      "grad_norm": 0.18450453877449036,
+      "learning_rate": 1.8599573861364074e-05,
+      "loss": 0.5238,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6943982251802552,
+      "grad_norm": 0.19734638929367065,
+      "learning_rate": 1.8594753632344104e-05,
+      "loss": 0.5619,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6946755407653911,
+      "grad_norm": 0.1974724531173706,
+      "learning_rate": 1.858993221407622e-05,
+      "loss": 0.5805,
+      "step": 2505
+    },
+    {
+      "epoch": 0.694952856350527,
+      "grad_norm": 0.20074540376663208,
+      "learning_rate": 1.858510960750122e-05,
+      "loss": 0.5515,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6952301719356628,
+      "grad_norm": 0.18881772458553314,
+      "learning_rate": 1.8580285813560104e-05,
+      "loss": 0.5371,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6955074875207987,
+      "grad_norm": 0.19910936057567596,
+      "learning_rate": 1.8575460833194142e-05,
+      "loss": 0.5288,
+      "step": 2508
+    },
+    {
+      "epoch": 0.6957848031059346,
+      "grad_norm": 0.18827465176582336,
+      "learning_rate": 1.8570634667344795e-05,
+      "loss": 0.5591,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6960621186910705,
+      "grad_norm": 0.18730634450912476,
+      "learning_rate": 1.8565807316953796e-05,
+      "loss": 0.5126,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6963394342762064,
+      "grad_norm": 0.18552148342132568,
+      "learning_rate": 1.856097878296307e-05,
+      "loss": 0.5208,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6966167498613423,
+      "grad_norm": 0.18370574712753296,
+      "learning_rate": 1.8556149066314803e-05,
+      "loss": 0.5122,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6968940654464781,
+      "grad_norm": 0.19410766661167145,
+      "learning_rate": 1.8551318167951403e-05,
+      "loss": 0.5062,
+      "step": 2513
+    },
+    {
+      "epoch": 0.697171381031614,
+      "grad_norm": 0.1899997889995575,
+      "learning_rate": 1.85464860888155e-05,
+      "loss": 0.5355,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6974486966167499,
+      "grad_norm": 0.2276785969734192,
+      "learning_rate": 1.854165282984996e-05,
+      "loss": 0.5434,
+      "step": 2515
+    },
+    {
+      "epoch": 0.6977260122018858,
+      "grad_norm": 0.20629100501537323,
+      "learning_rate": 1.8536818391997884e-05,
+      "loss": 0.5434,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6980033277870217,
+      "grad_norm": 0.1940404623746872,
+      "learning_rate": 1.8531982776202598e-05,
+      "loss": 0.5566,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6982806433721576,
+      "grad_norm": 0.18856097757816315,
+      "learning_rate": 1.8527145983407658e-05,
+      "loss": 0.5414,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6985579589572934,
+      "grad_norm": 0.18844252824783325,
+      "learning_rate": 1.8522308014556843e-05,
+      "loss": 0.5535,
+      "step": 2519
+    },
+    {
+      "epoch": 0.6988352745424293,
+      "grad_norm": 0.19054186344146729,
+      "learning_rate": 1.8517468870594188e-05,
+      "loss": 0.5436,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6991125901275652,
+      "grad_norm": 0.1948595494031906,
+      "learning_rate": 1.8512628552463917e-05,
+      "loss": 0.5589,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6993899057127011,
+      "grad_norm": 0.2020605355501175,
+      "learning_rate": 1.850778706111052e-05,
+      "loss": 0.5586,
+      "step": 2522
+    },
+    {
+      "epoch": 0.699667221297837,
+      "grad_norm": 0.1909698247909546,
+      "learning_rate": 1.8502944397478693e-05,
+      "loss": 0.5231,
+      "step": 2523
+    },
+    {
+      "epoch": 0.6999445368829729,
+      "grad_norm": 0.19061705470085144,
+      "learning_rate": 1.849810056251337e-05,
+      "loss": 0.5436,
+      "step": 2524
+    },
+    {
+      "epoch": 0.7002218524681088,
+      "grad_norm": 0.18788163363933563,
+      "learning_rate": 1.8493255557159704e-05,
+      "loss": 0.5223,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7004991680532446,
+      "grad_norm": 0.18889035284519196,
+      "learning_rate": 1.8488409382363095e-05,
+      "loss": 0.5178,
+      "step": 2526
+    },
+    {
+      "epoch": 0.7007764836383805,
+      "grad_norm": 0.18843677639961243,
+      "learning_rate": 1.8483562039069157e-05,
+      "loss": 0.5535,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7010537992235164,
+      "grad_norm": 0.195985808968544,
+      "learning_rate": 1.847871352822373e-05,
+      "loss": 0.5592,
+      "step": 2528
+    },
+    {
+      "epoch": 0.7013311148086523,
+      "grad_norm": 0.1854049116373062,
+      "learning_rate": 1.8473863850772897e-05,
+      "loss": 0.5272,
+      "step": 2529
+    },
+    {
+      "epoch": 0.7016084303937882,
+      "grad_norm": 0.17917431890964508,
+      "learning_rate": 1.8469013007662946e-05,
+      "loss": 0.5239,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7018857459789241,
+      "grad_norm": 0.18059198558330536,
+      "learning_rate": 1.8464160999840417e-05,
+      "loss": 0.565,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7021630615640599,
+      "grad_norm": 0.18871940672397614,
+      "learning_rate": 1.8459307828252052e-05,
+      "loss": 0.5422,
+      "step": 2532
+    },
+    {
+      "epoch": 0.7024403771491958,
+      "grad_norm": 0.1852055937051773,
+      "learning_rate": 1.845445349384485e-05,
+      "loss": 0.5211,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7027176927343317,
+      "grad_norm": 0.1867264062166214,
+      "learning_rate": 1.8449597997566005e-05,
+      "loss": 0.5318,
+      "step": 2534
+    },
+    {
+      "epoch": 0.7029950083194676,
+      "grad_norm": 0.1850994974374771,
+      "learning_rate": 1.844474134036296e-05,
+      "loss": 0.5375,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7032723239046035,
+      "grad_norm": 0.188653826713562,
+      "learning_rate": 1.8439883523183377e-05,
+      "loss": 0.5223,
+      "step": 2536
+    },
+    {
+      "epoch": 0.7035496394897394,
+      "grad_norm": 0.19549886882305145,
+      "learning_rate": 1.8435024546975142e-05,
+      "loss": 0.5946,
+      "step": 2537
+    },
+    {
+      "epoch": 0.7038269550748752,
+      "grad_norm": 0.19603100419044495,
+      "learning_rate": 1.8430164412686375e-05,
+      "loss": 0.5579,
+      "step": 2538
+    },
+    {
+      "epoch": 0.7041042706600111,
+      "grad_norm": 0.27463892102241516,
+      "learning_rate": 1.8425303121265414e-05,
+      "loss": 0.561,
+      "step": 2539
+    },
+    {
+      "epoch": 0.704381586245147,
+      "grad_norm": 0.1948249191045761,
+      "learning_rate": 1.842044067366082e-05,
+      "loss": 0.5715,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7046589018302829,
+      "grad_norm": 0.191276416182518,
+      "learning_rate": 1.8415577070821398e-05,
+      "loss": 0.5448,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7049362174154188,
+      "grad_norm": 0.19777309894561768,
+      "learning_rate": 1.841071231369616e-05,
+      "loss": 0.5333,
+      "step": 2542
+    },
+    {
+      "epoch": 0.7052135330005547,
+      "grad_norm": 0.1844799816608429,
+      "learning_rate": 1.8405846403234346e-05,
+      "loss": 0.5455,
+      "step": 2543
+    },
+    {
+      "epoch": 0.7054908485856906,
+      "grad_norm": 0.17811760306358337,
+      "learning_rate": 1.840097934038543e-05,
+      "loss": 0.5307,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7057681641708264,
+      "grad_norm": 0.1874844878911972,
+      "learning_rate": 1.8396111126099094e-05,
+      "loss": 0.5755,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7060454797559623,
+      "grad_norm": 0.20174047350883484,
+      "learning_rate": 1.839124176132527e-05,
+      "loss": 0.5422,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7063227953410982,
+      "grad_norm": 0.1862955093383789,
+      "learning_rate": 1.838637124701409e-05,
+      "loss": 0.5469,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7066001109262341,
+      "grad_norm": 0.18790222704410553,
+      "learning_rate": 1.8381499584115924e-05,
+      "loss": 0.5472,
+      "step": 2548
+    },
+    {
+      "epoch": 0.70687742651137,
+      "grad_norm": 0.17207522690296173,
+      "learning_rate": 1.8376626773581358e-05,
+      "loss": 0.5074,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7071547420965059,
+      "grad_norm": 0.19320419430732727,
+      "learning_rate": 1.8371752816361215e-05,
+      "loss": 0.5416,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7074320576816417,
+      "grad_norm": 0.19113574922084808,
+      "learning_rate": 1.8366877713406526e-05,
+      "loss": 0.533,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7077093732667776,
+      "grad_norm": 0.19141121208667755,
+      "learning_rate": 1.8362001465668554e-05,
+      "loss": 0.5435,
+      "step": 2552
+    },
+    {
+      "epoch": 0.7079866888519135,
+      "grad_norm": 0.17392635345458984,
+      "learning_rate": 1.8357124074098788e-05,
+      "loss": 0.5126,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7082640044370494,
+      "grad_norm": 0.18224339187145233,
+      "learning_rate": 1.8352245539648933e-05,
+      "loss": 0.5369,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7085413200221853,
+      "grad_norm": 0.193458691239357,
+      "learning_rate": 1.834736586327092e-05,
+      "loss": 0.5581,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7088186356073212,
+      "grad_norm": 0.19396451115608215,
+      "learning_rate": 1.8342485045916902e-05,
+      "loss": 0.546,
+      "step": 2556
+    },
+    {
+      "epoch": 0.709095951192457,
+      "grad_norm": 0.2662739157676697,
+      "learning_rate": 1.8337603088539263e-05,
+      "loss": 0.5557,
+      "step": 2557
+    },
+    {
+      "epoch": 0.7093732667775929,
+      "grad_norm": 0.18841521441936493,
+      "learning_rate": 1.8332719992090592e-05,
+      "loss": 0.5359,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7096505823627288,
+      "grad_norm": 0.20101507008075714,
+      "learning_rate": 1.8327835757523716e-05,
+      "loss": 0.54,
+      "step": 2559
+    },
+    {
+      "epoch": 0.7099278979478647,
+      "grad_norm": 0.17808422446250916,
+      "learning_rate": 1.832295038579168e-05,
+      "loss": 0.5263,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7102052135330006,
+      "grad_norm": 0.19543784856796265,
+      "learning_rate": 1.8318063877847747e-05,
+      "loss": 0.5342,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7104825291181365,
+      "grad_norm": 0.1944831758737564,
+      "learning_rate": 1.8313176234645406e-05,
+      "loss": 0.4973,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7107598447032724,
+      "grad_norm": 0.19557087123394012,
+      "learning_rate": 1.8308287457138362e-05,
+      "loss": 0.533,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7110371602884082,
+      "grad_norm": 0.19200831651687622,
+      "learning_rate": 1.8303397546280547e-05,
+      "loss": 0.5417,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7113144758735441,
+      "grad_norm": 0.1837347149848938,
+      "learning_rate": 1.829850650302612e-05,
+      "loss": 0.5594,
+      "step": 2565
+    },
+    {
+      "epoch": 0.71159179145868,
+      "grad_norm": 0.1859401911497116,
+      "learning_rate": 1.8293614328329437e-05,
+      "loss": 0.5249,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7118691070438159,
+      "grad_norm": 0.18670018017292023,
+      "learning_rate": 1.8288721023145105e-05,
+      "loss": 0.5339,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7121464226289518,
+      "grad_norm": 0.19364280998706818,
+      "learning_rate": 1.8283826588427927e-05,
+      "loss": 0.5729,
+      "step": 2568
+    },
+    {
+      "epoch": 0.7124237382140877,
+      "grad_norm": 0.19787278771400452,
+      "learning_rate": 1.827893102513295e-05,
+      "loss": 0.5544,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7127010537992235,
+      "grad_norm": 0.19540858268737793,
+      "learning_rate": 1.827403433421541e-05,
+      "loss": 0.5233,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7129783693843594,
+      "grad_norm": 0.20480972528457642,
+      "learning_rate": 1.8269136516630798e-05,
+      "loss": 0.545,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7132556849694953,
+      "grad_norm": 0.20005930960178375,
+      "learning_rate": 1.82642375733348e-05,
+      "loss": 0.561,
+      "step": 2572
+    },
+    {
+      "epoch": 0.7135330005546312,
+      "grad_norm": 0.18213188648223877,
+      "learning_rate": 1.825933750528333e-05,
+      "loss": 0.5277,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7138103161397671,
+      "grad_norm": 0.19505798816680908,
+      "learning_rate": 1.8254436313432522e-05,
+      "loss": 0.5283,
+      "step": 2574
+    },
+    {
+      "epoch": 0.714087631724903,
+      "grad_norm": 0.1885930299758911,
+      "learning_rate": 1.824953399873873e-05,
+      "loss": 0.5189,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7143649473100389,
+      "grad_norm": 0.19202375411987305,
+      "learning_rate": 1.824463056215852e-05,
+      "loss": 0.5197,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7146422628951747,
+      "grad_norm": 0.19426311552524567,
+      "learning_rate": 1.823972600464869e-05,
+      "loss": 0.5243,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7149195784803106,
+      "grad_norm": 0.18350745737552643,
+      "learning_rate": 1.8234820327166244e-05,
+      "loss": 0.5283,
+      "step": 2578
+    },
+    {
+      "epoch": 0.7151968940654465,
+      "grad_norm": 0.1888923943042755,
+      "learning_rate": 1.822991353066841e-05,
+      "loss": 0.5298,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7154742096505824,
+      "grad_norm": 0.19120195508003235,
+      "learning_rate": 1.8225005616112636e-05,
+      "loss": 0.5583,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7157515252357183,
+      "grad_norm": 0.18390871584415436,
+      "learning_rate": 1.8220096584456587e-05,
+      "loss": 0.5174,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7160288408208542,
+      "grad_norm": 0.19268232583999634,
+      "learning_rate": 1.8215186436658142e-05,
+      "loss": 0.5706,
+      "step": 2582
+    },
+    {
+      "epoch": 0.71630615640599,
+      "grad_norm": 0.2088870108127594,
+      "learning_rate": 1.82102751736754e-05,
+      "loss": 0.5467,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7165834719911259,
+      "grad_norm": 0.18982863426208496,
+      "learning_rate": 1.8205362796466682e-05,
+      "loss": 0.547,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7168607875762618,
+      "grad_norm": 0.20448660850524902,
+      "learning_rate": 1.820044930599052e-05,
+      "loss": 0.5523,
+      "step": 2585
+    },
+    {
+      "epoch": 0.7171381031613977,
+      "grad_norm": 0.19304388761520386,
+      "learning_rate": 1.8195534703205674e-05,
+      "loss": 0.5766,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7174154187465336,
+      "grad_norm": 0.20076531171798706,
+      "learning_rate": 1.81906189890711e-05,
+      "loss": 0.5327,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7176927343316695,
+      "grad_norm": 0.24745801091194153,
+      "learning_rate": 1.8185702164546e-05,
+      "loss": 0.5386,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7179700499168054,
+      "grad_norm": 0.1958095282316208,
+      "learning_rate": 1.8180784230589758e-05,
+      "loss": 0.5592,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7182473655019412,
+      "grad_norm": 0.1962784081697464,
+      "learning_rate": 1.8175865188162007e-05,
+      "loss": 0.5547,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7185246810870771,
+      "grad_norm": 0.21214796602725983,
+      "learning_rate": 1.8170945038222577e-05,
+      "loss": 0.5789,
+      "step": 2591
+    },
+    {
+      "epoch": 0.718801996672213,
+      "grad_norm": 0.18657919764518738,
+      "learning_rate": 1.8166023781731523e-05,
+      "loss": 0.5909,
+      "step": 2592
+    },
+    {
+      "epoch": 0.7190793122573489,
+      "grad_norm": 0.1949455291032791,
+      "learning_rate": 1.816110141964911e-05,
+      "loss": 0.5804,
+      "step": 2593
+    },
+    {
+      "epoch": 0.7193566278424848,
+      "grad_norm": 0.20440496504306793,
+      "learning_rate": 1.8156177952935824e-05,
+      "loss": 0.5819,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7196339434276207,
+      "grad_norm": 0.19375431537628174,
+      "learning_rate": 1.815125338255236e-05,
+      "loss": 0.5707,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7199112590127565,
+      "grad_norm": 0.2573017477989197,
+      "learning_rate": 1.8146327709459635e-05,
+      "loss": 0.5622,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7201885745978924,
+      "grad_norm": 0.19676506519317627,
+      "learning_rate": 1.8141400934618775e-05,
+      "loss": 0.5668,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7204658901830283,
+      "grad_norm": 0.19473743438720703,
+      "learning_rate": 1.8136473058991126e-05,
+      "loss": 0.5654,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7207432057681642,
+      "grad_norm": 0.18709680438041687,
+      "learning_rate": 1.8131544083538253e-05,
+      "loss": 0.5283,
+      "step": 2599
+    },
+    {
+      "epoch": 0.7210205213533001,
+      "grad_norm": 0.19263465702533722,
+      "learning_rate": 1.812661400922192e-05,
+      "loss": 0.5379,
+      "step": 2600
+    },
+    {
+      "epoch": 0.721297836938436,
+      "grad_norm": 0.19263778626918793,
+      "learning_rate": 1.8121682837004118e-05,
+      "loss": 0.5678,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7215751525235718,
+      "grad_norm": 0.19861197471618652,
+      "learning_rate": 1.8116750567847058e-05,
+      "loss": 0.5456,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7218524681087077,
+      "grad_norm": 0.1754927933216095,
+      "learning_rate": 1.8111817202713143e-05,
+      "loss": 0.5164,
+      "step": 2603
+    },
+    {
+      "epoch": 0.7221297836938436,
+      "grad_norm": 0.19112901389598846,
+      "learning_rate": 1.8106882742565008e-05,
+      "loss": 0.5362,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7224070992789795,
+      "grad_norm": 0.1980351209640503,
+      "learning_rate": 1.8101947188365503e-05,
+      "loss": 0.5687,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7226844148641154,
+      "grad_norm": 0.19299866259098053,
+      "learning_rate": 1.8097010541077678e-05,
+      "loss": 0.5589,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7229617304492513,
+      "grad_norm": 0.19257931411266327,
+      "learning_rate": 1.809207280166481e-05,
+      "loss": 0.5701,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7232390460343872,
+      "grad_norm": 0.1912074089050293,
+      "learning_rate": 1.8087133971090374e-05,
+      "loss": 0.544,
+      "step": 2608
+    },
+    {
+      "epoch": 0.723516361619523,
+      "grad_norm": 0.19741860032081604,
+      "learning_rate": 1.808219405031808e-05,
+      "loss": 0.5527,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7237936772046589,
+      "grad_norm": 0.18676535785198212,
+      "learning_rate": 1.807725304031182e-05,
+      "loss": 0.5539,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7240709927897948,
+      "grad_norm": 0.17912089824676514,
+      "learning_rate": 1.807231094203573e-05,
+      "loss": 0.5342,
+      "step": 2611
+    },
+    {
+      "epoch": 0.7243483083749307,
+      "grad_norm": 0.18593581020832062,
+      "learning_rate": 1.806736775645414e-05,
+      "loss": 0.5366,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7246256239600666,
+      "grad_norm": 0.20467206835746765,
+      "learning_rate": 1.8062423484531592e-05,
+      "loss": 0.527,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7249029395452025,
+      "grad_norm": 0.18463392555713654,
+      "learning_rate": 1.8057478127232854e-05,
+      "loss": 0.545,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7251802551303383,
+      "grad_norm": 0.20011630654335022,
+      "learning_rate": 1.805253168552289e-05,
+      "loss": 0.5242,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7254575707154742,
+      "grad_norm": 0.18936677277088165,
+      "learning_rate": 1.804758416036688e-05,
+      "loss": 0.5285,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7257348863006101,
+      "grad_norm": 0.19428247213363647,
+      "learning_rate": 1.804263555273022e-05,
+      "loss": 0.5448,
+      "step": 2617
+    },
+    {
+      "epoch": 0.726012201885746,
+      "grad_norm": 0.18848338723182678,
+      "learning_rate": 1.8037685863578514e-05,
+      "loss": 0.5673,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7262895174708819,
+      "grad_norm": 0.19170448184013367,
+      "learning_rate": 1.803273509387758e-05,
+      "loss": 0.5234,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7265668330560178,
+      "grad_norm": 0.18249256908893585,
+      "learning_rate": 1.8027783244593443e-05,
+      "loss": 0.5377,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7268441486411537,
+      "grad_norm": 0.18726296722888947,
+      "learning_rate": 1.8022830316692336e-05,
+      "loss": 0.5381,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7271214642262895,
+      "grad_norm": 0.1881718784570694,
+      "learning_rate": 1.801787631114071e-05,
+      "loss": 0.5453,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7273987798114254,
+      "grad_norm": 0.19469492137432098,
+      "learning_rate": 1.8012921228905225e-05,
+      "loss": 0.5231,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7276760953965613,
+      "grad_norm": 0.18475371599197388,
+      "learning_rate": 1.8007965070952743e-05,
+      "loss": 0.5323,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7279534109816972,
+      "grad_norm": 0.18544836342334747,
+      "learning_rate": 1.8003007838250343e-05,
+      "loss": 0.5345,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7282307265668331,
+      "grad_norm": 0.19587865471839905,
+      "learning_rate": 1.799804953176532e-05,
+      "loss": 0.5168,
+      "step": 2626
+    },
+    {
+      "epoch": 0.728508042151969,
+      "grad_norm": 0.23171131312847137,
+      "learning_rate": 1.7993090152465163e-05,
+      "loss": 0.5235,
+      "step": 2627
+    },
+    {
+      "epoch": 0.7287853577371048,
+      "grad_norm": 0.19391484558582306,
+      "learning_rate": 1.7988129701317582e-05,
+      "loss": 0.5345,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7290626733222407,
+      "grad_norm": 0.19040954113006592,
+      "learning_rate": 1.7983168179290488e-05,
+      "loss": 0.5226,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7293399889073766,
+      "grad_norm": 0.1814422607421875,
+      "learning_rate": 1.797820558735201e-05,
+      "loss": 0.5291,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7296173044925125,
+      "grad_norm": 0.19977906346321106,
+      "learning_rate": 1.797324192647048e-05,
+      "loss": 0.5641,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7298946200776484,
+      "grad_norm": 0.19043037295341492,
+      "learning_rate": 1.796827719761444e-05,
+      "loss": 0.5686,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7301719356627843,
+      "grad_norm": 0.19778837263584137,
+      "learning_rate": 1.7963311401752638e-05,
+      "loss": 0.5648,
+      "step": 2633
+    },
+    {
+      "epoch": 0.7304492512479202,
+      "grad_norm": 0.45009469985961914,
+      "learning_rate": 1.7958344539854034e-05,
+      "loss": 0.5244,
+      "step": 2634
+    },
+    {
+      "epoch": 0.730726566833056,
+      "grad_norm": 0.1936669498682022,
+      "learning_rate": 1.7953376612887793e-05,
+      "loss": 0.5296,
+      "step": 2635
+    },
+    {
+      "epoch": 0.7310038824181919,
+      "grad_norm": 0.19970230758190155,
+      "learning_rate": 1.7948407621823287e-05,
+      "loss": 0.5832,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7312811980033278,
+      "grad_norm": 0.19142326712608337,
+      "learning_rate": 1.794343756763011e-05,
+      "loss": 0.5478,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7315585135884637,
+      "grad_norm": 0.18312163650989532,
+      "learning_rate": 1.7938466451278034e-05,
+      "loss": 0.5382,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7318358291735996,
+      "grad_norm": 0.19482283294200897,
+      "learning_rate": 1.793349427373707e-05,
+      "loss": 0.5392,
+      "step": 2639
+    },
+    {
+      "epoch": 0.7321131447587355,
+      "grad_norm": 0.20726899802684784,
+      "learning_rate": 1.7928521035977413e-05,
+      "loss": 0.5597,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7323904603438713,
+      "grad_norm": 0.20082899928092957,
+      "learning_rate": 1.7923546738969478e-05,
+      "loss": 0.5293,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7326677759290072,
+      "grad_norm": 0.19428935647010803,
+      "learning_rate": 1.791857138368388e-05,
+      "loss": 0.5428,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7329450915141431,
+      "grad_norm": 0.19222451746463776,
+      "learning_rate": 1.791359497109144e-05,
+      "loss": 0.5417,
+      "step": 2643
+    },
+    {
+      "epoch": 0.733222407099279,
+      "grad_norm": 0.1904270201921463,
+      "learning_rate": 1.7908617502163188e-05,
+      "loss": 0.5368,
+      "step": 2644
+    },
+    {
+      "epoch": 0.7334997226844149,
+      "grad_norm": 0.1973213404417038,
+      "learning_rate": 1.7903638977870372e-05,
+      "loss": 0.5347,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7337770382695508,
+      "grad_norm": 0.1838080883026123,
+      "learning_rate": 1.7898659399184415e-05,
+      "loss": 0.5239,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7340543538546866,
+      "grad_norm": 0.18665340542793274,
+      "learning_rate": 1.7893678767076982e-05,
+      "loss": 0.5469,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7343316694398225,
+      "grad_norm": 0.18644295632839203,
+      "learning_rate": 1.788869708251991e-05,
+      "loss": 0.5423,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7346089850249584,
+      "grad_norm": 0.18912896513938904,
+      "learning_rate": 1.788371434648528e-05,
+      "loss": 0.5289,
+      "step": 2649
+    },
+    {
+      "epoch": 0.7348863006100943,
+      "grad_norm": 0.1896572709083557,
+      "learning_rate": 1.7878730559945327e-05,
+      "loss": 0.544,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7351636161952302,
+      "grad_norm": 0.18456673622131348,
+      "learning_rate": 1.7873745723872545e-05,
+      "loss": 0.5437,
+      "step": 2651
+    },
+    {
+      "epoch": 0.7354409317803661,
+      "grad_norm": 0.18309368193149567,
+      "learning_rate": 1.7868759839239596e-05,
+      "loss": 0.5079,
+      "step": 2652
+    },
+    {
+      "epoch": 0.735718247365502,
+      "grad_norm": 0.18934939801692963,
+      "learning_rate": 1.7863772907019356e-05,
+      "loss": 0.5473,
+      "step": 2653
+    },
+    {
+      "epoch": 0.7359955629506378,
+      "grad_norm": 0.19800709187984467,
+      "learning_rate": 1.7858784928184916e-05,
+      "loss": 0.5386,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7362728785357737,
+      "grad_norm": 0.20616304874420166,
+      "learning_rate": 1.7853795903709556e-05,
+      "loss": 0.5388,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7365501941209096,
+      "grad_norm": 0.18855001032352448,
+      "learning_rate": 1.7848805834566768e-05,
+      "loss": 0.5499,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7368275097060455,
+      "grad_norm": 0.18595397472381592,
+      "learning_rate": 1.7843814721730244e-05,
+      "loss": 0.5599,
+      "step": 2657
+    },
+    {
+      "epoch": 0.7371048252911814,
+      "grad_norm": 0.1929646134376526,
+      "learning_rate": 1.7838822566173894e-05,
+      "loss": 0.5412,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7373821408763173,
+      "grad_norm": 0.19367007911205292,
+      "learning_rate": 1.7833829368871808e-05,
+      "loss": 0.5328,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7376594564614531,
+      "grad_norm": 0.1873459815979004,
+      "learning_rate": 1.7828835130798296e-05,
+      "loss": 0.5444,
+      "step": 2660
+    },
+    {
+      "epoch": 0.737936772046589,
+      "grad_norm": 0.20193496346473694,
+      "learning_rate": 1.7823839852927867e-05,
+      "loss": 0.5287,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7382140876317249,
+      "grad_norm": 0.18771541118621826,
+      "learning_rate": 1.7818843536235224e-05,
+      "loss": 0.5404,
+      "step": 2662
+    },
+    {
+      "epoch": 0.7384914032168608,
+      "grad_norm": 0.1829247921705246,
+      "learning_rate": 1.781384618169529e-05,
+      "loss": 0.5512,
+      "step": 2663
+    },
+    {
+      "epoch": 0.7387687188019967,
+      "grad_norm": 0.2635699212551117,
+      "learning_rate": 1.7808847790283183e-05,
+      "loss": 0.5678,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7390460343871326,
+      "grad_norm": 0.1860508918762207,
+      "learning_rate": 1.780384836297421e-05,
+      "loss": 0.5348,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7393233499722685,
+      "grad_norm": 0.18465931713581085,
+      "learning_rate": 1.7798847900743904e-05,
+      "loss": 0.5043,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7396006655574043,
+      "grad_norm": 0.18881580233573914,
+      "learning_rate": 1.779384640456798e-05,
+      "loss": 0.5251,
+      "step": 2667
+    },
+    {
+      "epoch": 0.7398779811425402,
+      "grad_norm": 0.18016085028648376,
+      "learning_rate": 1.7788843875422367e-05,
+      "loss": 0.5585,
+      "step": 2668
+    },
+    {
+      "epoch": 0.7401552967276761,
+      "grad_norm": 0.19220809638500214,
+      "learning_rate": 1.7783840314283183e-05,
+      "loss": 0.5263,
+      "step": 2669
+    },
+    {
+      "epoch": 0.740432612312812,
+      "grad_norm": 0.18954598903656006,
+      "learning_rate": 1.7778835722126764e-05,
+      "loss": 0.542,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7407099278979479,
+      "grad_norm": 0.18674500286579132,
+      "learning_rate": 1.7773830099929635e-05,
+      "loss": 0.5247,
+      "step": 2671
+    },
+    {
+      "epoch": 0.7409872434830838,
+      "grad_norm": 0.18231706321239471,
+      "learning_rate": 1.776882344866853e-05,
+      "loss": 0.5115,
+      "step": 2672
+    },
+    {
+      "epoch": 0.7412645590682196,
+      "grad_norm": 0.18846355378627777,
+      "learning_rate": 1.776381576932037e-05,
+      "loss": 0.5186,
+      "step": 2673
+    },
+    {
+      "epoch": 0.7415418746533555,
+      "grad_norm": 0.18721552193164825,
+      "learning_rate": 1.7758807062862292e-05,
+      "loss": 0.5313,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7418191902384914,
+      "grad_norm": 0.191980242729187,
+      "learning_rate": 1.775379733027163e-05,
+      "loss": 0.5505,
+      "step": 2675
+    },
+    {
+      "epoch": 0.7420965058236273,
+      "grad_norm": 0.18773001432418823,
+      "learning_rate": 1.7748786572525907e-05,
+      "loss": 0.5334,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7423738214087632,
+      "grad_norm": 0.19638672471046448,
+      "learning_rate": 1.7743774790602864e-05,
+      "loss": 0.5718,
+      "step": 2677
+    },
+    {
+      "epoch": 0.7426511369938991,
+      "grad_norm": 0.19333893060684204,
+      "learning_rate": 1.7738761985480425e-05,
+      "loss": 0.5479,
+      "step": 2678
+    },
+    {
+      "epoch": 0.742928452579035,
+      "grad_norm": 0.20004448294639587,
+      "learning_rate": 1.7733748158136725e-05,
+      "loss": 0.5331,
+      "step": 2679
+    },
+    {
+      "epoch": 0.7432057681641708,
+      "grad_norm": 0.19896887242794037,
+      "learning_rate": 1.7728733309550097e-05,
+      "loss": 0.5484,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7434830837493067,
+      "grad_norm": 0.1829969435930252,
+      "learning_rate": 1.7723717440699066e-05,
+      "loss": 0.5459,
+      "step": 2681
+    },
+    {
+      "epoch": 0.7437603993344426,
+      "grad_norm": 0.1870088130235672,
+      "learning_rate": 1.771870055256236e-05,
+      "loss": 0.5628,
+      "step": 2682
+    },
+    {
+      "epoch": 0.7440377149195785,
+      "grad_norm": 0.18835529685020447,
+      "learning_rate": 1.7713682646118914e-05,
+      "loss": 0.5439,
+      "step": 2683
+    },
+    {
+      "epoch": 0.7443150305047144,
+      "grad_norm": 0.20633459091186523,
+      "learning_rate": 1.7708663722347845e-05,
+      "loss": 0.5677,
+      "step": 2684
+    },
+    {
+      "epoch": 0.7445923460898503,
+      "grad_norm": 0.18712858855724335,
+      "learning_rate": 1.7703643782228488e-05,
+      "loss": 0.5666,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7448696616749861,
+      "grad_norm": 0.18475639820098877,
+      "learning_rate": 1.769862282674036e-05,
+      "loss": 0.5548,
+      "step": 2686
+    },
+    {
+      "epoch": 0.745146977260122,
+      "grad_norm": 0.19311586022377014,
+      "learning_rate": 1.769360085686318e-05,
+      "loss": 0.5428,
+      "step": 2687
+    },
+    {
+      "epoch": 0.7454242928452579,
+      "grad_norm": 0.19635801017284393,
+      "learning_rate": 1.7688577873576872e-05,
+      "loss": 0.5228,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7457016084303938,
+      "grad_norm": 0.18878091871738434,
+      "learning_rate": 1.7683553877861554e-05,
+      "loss": 0.4995,
+      "step": 2689
+    },
+    {
+      "epoch": 0.7459789240155297,
+      "grad_norm": 0.1772637963294983,
+      "learning_rate": 1.7678528870697537e-05,
+      "loss": 0.5241,
+      "step": 2690
+    },
+    {
+      "epoch": 0.7462562396006656,
+      "grad_norm": 0.18778811395168304,
+      "learning_rate": 1.7673502853065335e-05,
+      "loss": 0.5247,
+      "step": 2691
+    },
+    {
+      "epoch": 0.7465335551858014,
+      "grad_norm": 0.20334573090076447,
+      "learning_rate": 1.7668475825945656e-05,
+      "loss": 0.5369,
+      "step": 2692
+    },
+    {
+      "epoch": 0.7468108707709373,
+      "grad_norm": 0.18709522485733032,
+      "learning_rate": 1.766344779031941e-05,
+      "loss": 0.5701,
+      "step": 2693
+    },
+    {
+      "epoch": 0.7470881863560732,
+      "grad_norm": 0.18577025830745697,
+      "learning_rate": 1.7658418747167694e-05,
+      "loss": 0.5409,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7473655019412091,
+      "grad_norm": 0.19616863131523132,
+      "learning_rate": 1.765338869747181e-05,
+      "loss": 0.5533,
+      "step": 2695
+    },
+    {
+      "epoch": 0.747642817526345,
+      "grad_norm": 0.20094148814678192,
+      "learning_rate": 1.764835764221326e-05,
+      "loss": 0.5252,
+      "step": 2696
+    },
+    {
+      "epoch": 0.7479201331114809,
+      "grad_norm": 0.19073578715324402,
+      "learning_rate": 1.7643325582373728e-05,
+      "loss": 0.52,
+      "step": 2697
+    },
+    {
+      "epoch": 0.7481974486966168,
+      "grad_norm": 0.19425810873508453,
+      "learning_rate": 1.7638292518935103e-05,
+      "loss": 0.5612,
+      "step": 2698
+    },
+    {
+      "epoch": 0.7484747642817526,
+      "grad_norm": 0.1896180808544159,
+      "learning_rate": 1.7633258452879475e-05,
+      "loss": 0.5504,
+      "step": 2699
+    },
+    {
+      "epoch": 0.7487520798668885,
+      "grad_norm": 0.19124028086662292,
+      "learning_rate": 1.762822338518912e-05,
+      "loss": 0.5335,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7490293954520244,
+      "grad_norm": 0.19054248929023743,
+      "learning_rate": 1.762318731684651e-05,
+      "loss": 0.552,
+      "step": 2701
+    },
+    {
+      "epoch": 0.7493067110371603,
+      "grad_norm": 0.18848967552185059,
+      "learning_rate": 1.761815024883432e-05,
+      "loss": 0.557,
+      "step": 2702
+    },
+    {
+      "epoch": 0.7495840266222962,
+      "grad_norm": 0.1953321397304535,
+      "learning_rate": 1.7613112182135406e-05,
+      "loss": 0.5916,
+      "step": 2703
+    },
+    {
+      "epoch": 0.7498613422074321,
+      "grad_norm": 0.18123508989810944,
+      "learning_rate": 1.7608073117732848e-05,
+      "loss": 0.5643,
+      "step": 2704
+    },
+    {
+      "epoch": 0.7501386577925679,
+      "grad_norm": 0.19151508808135986,
+      "learning_rate": 1.760303305660988e-05,
+      "loss": 0.5512,
+      "step": 2705
+    },
+    {
+      "epoch": 0.7504159733777038,
+      "grad_norm": 0.18512828648090363,
+      "learning_rate": 1.7597991999749967e-05,
+      "loss": 0.5627,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7506932889628397,
+      "grad_norm": 0.20286305248737335,
+      "learning_rate": 1.7592949948136737e-05,
+      "loss": 0.568,
+      "step": 2707
+    },
+    {
+      "epoch": 0.7509706045479756,
+      "grad_norm": 0.20783564448356628,
+      "learning_rate": 1.758790690275405e-05,
+      "loss": 0.5563,
+      "step": 2708
+    },
+    {
+      "epoch": 0.7512479201331115,
+      "grad_norm": 0.18659134209156036,
+      "learning_rate": 1.7582862864585913e-05,
+      "loss": 0.5285,
+      "step": 2709
+    },
+    {
+      "epoch": 0.7515252357182474,
+      "grad_norm": 0.17614908516407013,
+      "learning_rate": 1.757781783461657e-05,
+      "loss": 0.4952,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7518025513033832,
+      "grad_norm": 0.18984104692935944,
+      "learning_rate": 1.757277181383043e-05,
+      "loss": 0.5564,
+      "step": 2711
+    },
+    {
+      "epoch": 0.7520798668885191,
+      "grad_norm": 0.19199056923389435,
+      "learning_rate": 1.756772480321211e-05,
+      "loss": 0.5815,
+      "step": 2712
+    },
+    {
+      "epoch": 0.752357182473655,
+      "grad_norm": 0.5123929381370544,
+      "learning_rate": 1.7562676803746414e-05,
+      "loss": 0.537,
+      "step": 2713
+    },
+    {
+      "epoch": 0.7526344980587909,
+      "grad_norm": 0.1854097694158554,
+      "learning_rate": 1.7557627816418337e-05,
+      "loss": 0.5275,
+      "step": 2714
+    },
+    {
+      "epoch": 0.7529118136439268,
+      "grad_norm": 0.1879512071609497,
+      "learning_rate": 1.755257784221308e-05,
+      "loss": 0.5292,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7531891292290627,
+      "grad_norm": 0.18993008136749268,
+      "learning_rate": 1.7547526882116014e-05,
+      "loss": 0.5282,
+      "step": 2716
+    },
+    {
+      "epoch": 0.7534664448141986,
+      "grad_norm": 0.18528947234153748,
+      "learning_rate": 1.7542474937112725e-05,
+      "loss": 0.5457,
+      "step": 2717
+    },
+    {
+      "epoch": 0.7537437603993344,
+      "grad_norm": 0.1859760731458664,
+      "learning_rate": 1.753742200818898e-05,
+      "loss": 0.5448,
+      "step": 2718
+    },
+    {
+      "epoch": 0.7540210759844703,
+      "grad_norm": 0.17713625729084015,
+      "learning_rate": 1.753236809633073e-05,
+      "loss": 0.5479,
+      "step": 2719
+    },
+    {
+      "epoch": 0.7542983915696062,
+      "grad_norm": 0.17718899250030518,
+      "learning_rate": 1.7527313202524144e-05,
+      "loss": 0.5378,
+      "step": 2720
+    },
+    {
+      "epoch": 0.7545757071547421,
+      "grad_norm": 0.19346462190151215,
+      "learning_rate": 1.752225732775555e-05,
+      "loss": 0.552,
+      "step": 2721
+    },
+    {
+      "epoch": 0.754853022739878,
+      "grad_norm": 0.17717614769935608,
+      "learning_rate": 1.7517200473011488e-05,
+      "loss": 0.5348,
+      "step": 2722
+    },
+    {
+      "epoch": 0.7551303383250139,
+      "grad_norm": 0.1940585970878601,
+      "learning_rate": 1.751214263927869e-05,
+      "loss": 0.5405,
+      "step": 2723
+    },
+    {
+      "epoch": 0.7554076539101497,
+      "grad_norm": 0.19863441586494446,
+      "learning_rate": 1.7507083827544065e-05,
+      "loss": 0.5357,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7556849694952856,
+      "grad_norm": 0.18913887441158295,
+      "learning_rate": 1.7502024038794727e-05,
+      "loss": 0.5658,
+      "step": 2725
+    },
+    {
+      "epoch": 0.7559622850804215,
+      "grad_norm": 0.18873843550682068,
+      "learning_rate": 1.7496963274017975e-05,
+      "loss": 0.5774,
+      "step": 2726
+    },
+    {
+      "epoch": 0.7562396006655574,
+      "grad_norm": 0.1888992041349411,
+      "learning_rate": 1.7491901534201295e-05,
+      "loss": 0.5319,
+      "step": 2727
+    },
+    {
+      "epoch": 0.7565169162506933,
+      "grad_norm": 0.18841078877449036,
+      "learning_rate": 1.7486838820332362e-05,
+      "loss": 0.5542,
+      "step": 2728
+    },
+    {
+      "epoch": 0.7567942318358292,
+      "grad_norm": 0.19633720815181732,
+      "learning_rate": 1.7481775133399057e-05,
+      "loss": 0.5615,
+      "step": 2729
+    },
+    {
+      "epoch": 0.757071547420965,
+      "grad_norm": 0.19098395109176636,
+      "learning_rate": 1.7476710474389434e-05,
+      "loss": 0.5642,
+      "step": 2730
+    },
+    {
+      "epoch": 0.7573488630061009,
+      "grad_norm": 0.1895277500152588,
+      "learning_rate": 1.747164484429174e-05,
+      "loss": 0.557,
+      "step": 2731
+    },
+    {
+      "epoch": 0.7576261785912368,
+      "grad_norm": 0.18626920878887177,
+      "learning_rate": 1.7466578244094417e-05,
+      "loss": 0.5314,
+      "step": 2732
+    },
+    {
+      "epoch": 0.7579034941763727,
+      "grad_norm": 0.1883586198091507,
+      "learning_rate": 1.746151067478609e-05,
+      "loss": 0.5457,
+      "step": 2733
+    },
+    {
+      "epoch": 0.7581808097615086,
+      "grad_norm": 0.18349739909172058,
+      "learning_rate": 1.745644213735558e-05,
+      "loss": 0.5467,
+      "step": 2734
+    },
+    {
+      "epoch": 0.7584581253466445,
+      "grad_norm": 0.17938470840454102,
+      "learning_rate": 1.7451372632791888e-05,
+      "loss": 0.5444,
+      "step": 2735
+    },
+    {
+      "epoch": 0.7587354409317804,
+      "grad_norm": 0.19910098612308502,
+      "learning_rate": 1.7446302162084215e-05,
+      "loss": 0.5777,
+      "step": 2736
+    },
+    {
+      "epoch": 0.7590127565169162,
+      "grad_norm": 0.20055991411209106,
+      "learning_rate": 1.7441230726221936e-05,
+      "loss": 0.5546,
+      "step": 2737
+    },
+    {
+      "epoch": 0.7592900721020521,
+      "grad_norm": 0.18521596491336823,
+      "learning_rate": 1.743615832619463e-05,
+      "loss": 0.5239,
+      "step": 2738
+    },
+    {
+      "epoch": 0.759567387687188,
+      "grad_norm": 0.19067248702049255,
+      "learning_rate": 1.7431084962992052e-05,
+      "loss": 0.5288,
+      "step": 2739
+    },
+    {
+      "epoch": 0.7598447032723239,
+      "grad_norm": 0.18803349137306213,
+      "learning_rate": 1.7426010637604152e-05,
+      "loss": 0.5397,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7601220188574598,
+      "grad_norm": 0.1846192330121994,
+      "learning_rate": 1.7420935351021062e-05,
+      "loss": 0.5379,
+      "step": 2741
+    },
+    {
+      "epoch": 0.7603993344425957,
+      "grad_norm": 0.1797967106103897,
+      "learning_rate": 1.7415859104233108e-05,
+      "loss": 0.5528,
+      "step": 2742
+    },
+    {
+      "epoch": 0.7606766500277315,
+      "grad_norm": 0.18609130382537842,
+      "learning_rate": 1.7410781898230797e-05,
+      "loss": 0.5473,
+      "step": 2743
+    },
+    {
+      "epoch": 0.7609539656128674,
+      "grad_norm": 0.23831035196781158,
+      "learning_rate": 1.7405703734004837e-05,
+      "loss": 0.5114,
+      "step": 2744
+    },
+    {
+      "epoch": 0.7612312811980033,
+      "grad_norm": 0.16559574007987976,
+      "learning_rate": 1.74006246125461e-05,
+      "loss": 0.501,
+      "step": 2745
+    },
+    {
+      "epoch": 0.7615085967831392,
+      "grad_norm": 0.1880342662334442,
+      "learning_rate": 1.7395544534845663e-05,
+      "loss": 0.5344,
+      "step": 2746
+    },
+    {
+      "epoch": 0.7617859123682751,
+      "grad_norm": 0.1892349123954773,
+      "learning_rate": 1.7390463501894778e-05,
+      "loss": 0.5745,
+      "step": 2747
+    },
+    {
+      "epoch": 0.762063227953411,
+      "grad_norm": 0.2029358148574829,
+      "learning_rate": 1.7385381514684896e-05,
+      "loss": 0.5206,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7623405435385469,
+      "grad_norm": 0.2009795904159546,
+      "learning_rate": 1.7380298574207645e-05,
+      "loss": 0.5765,
+      "step": 2749
+    },
+    {
+      "epoch": 0.7626178591236827,
+      "grad_norm": 0.18493135273456573,
+      "learning_rate": 1.737521468145484e-05,
+      "loss": 0.567,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7628951747088186,
+      "grad_norm": 0.2014453411102295,
+      "learning_rate": 1.7370129837418487e-05,
+      "loss": 0.564,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7631724902939545,
+      "grad_norm": 0.19111500680446625,
+      "learning_rate": 1.7365044043090766e-05,
+      "loss": 0.5589,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7634498058790904,
+      "grad_norm": 0.19902296364307404,
+      "learning_rate": 1.7359957299464062e-05,
+      "loss": 0.5543,
+      "step": 2753
+    },
+    {
+      "epoch": 0.7637271214642263,
+      "grad_norm": 0.19618061184883118,
+      "learning_rate": 1.7354869607530923e-05,
+      "loss": 0.5639,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7640044370493622,
+      "grad_norm": 0.17767266929149628,
+      "learning_rate": 1.7349780968284094e-05,
+      "loss": 0.5041,
+      "step": 2755
+    },
+    {
+      "epoch": 0.764281752634498,
+      "grad_norm": 0.19212745130062103,
+      "learning_rate": 1.7344691382716508e-05,
+      "loss": 0.5416,
+      "step": 2756
+    },
+    {
+      "epoch": 0.7645590682196339,
+      "grad_norm": 0.19896987080574036,
+      "learning_rate": 1.7339600851821274e-05,
+      "loss": 0.5505,
+      "step": 2757
+    },
+    {
+      "epoch": 0.7648363838047698,
+      "grad_norm": 0.1840539574623108,
+      "learning_rate": 1.7334509376591695e-05,
+      "loss": 0.5373,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7651136993899057,
+      "grad_norm": 0.20208869874477386,
+      "learning_rate": 1.7329416958021247e-05,
+      "loss": 0.5553,
+      "step": 2759
+    },
+    {
+      "epoch": 0.7653910149750416,
+      "grad_norm": 0.18945381045341492,
+      "learning_rate": 1.7324323597103597e-05,
+      "loss": 0.536,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7656683305601775,
+      "grad_norm": 0.19363372027873993,
+      "learning_rate": 1.7319229294832597e-05,
+      "loss": 0.5607,
+      "step": 2761
+    },
+    {
+      "epoch": 0.7659456461453134,
+      "grad_norm": 0.20634472370147705,
+      "learning_rate": 1.7314134052202272e-05,
+      "loss": 0.5451,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7662229617304492,
+      "grad_norm": 0.19053952395915985,
+      "learning_rate": 1.730903787020685e-05,
+      "loss": 0.5661,
+      "step": 2763
+    },
+    {
+      "epoch": 0.7665002773155851,
+      "grad_norm": 0.18681201338768005,
+      "learning_rate": 1.7303940749840726e-05,
+      "loss": 0.5312,
+      "step": 2764
+    },
+    {
+      "epoch": 0.766777592900721,
+      "grad_norm": 0.19541014730930328,
+      "learning_rate": 1.7298842692098488e-05,
+      "loss": 0.5366,
+      "step": 2765
+    },
+    {
+      "epoch": 0.7670549084858569,
+      "grad_norm": 0.19399814307689667,
+      "learning_rate": 1.729374369797489e-05,
+      "loss": 0.547,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7673322240709928,
+      "grad_norm": 0.19590160250663757,
+      "learning_rate": 1.7288643768464892e-05,
+      "loss": 0.5286,
+      "step": 2767
+    },
+    {
+      "epoch": 0.7676095396561287,
+      "grad_norm": 0.18889807164669037,
+      "learning_rate": 1.7283542904563625e-05,
+      "loss": 0.5357,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7678868552412645,
+      "grad_norm": 0.1905566155910492,
+      "learning_rate": 1.7278441107266395e-05,
+      "loss": 0.5663,
+      "step": 2769
+    },
+    {
+      "epoch": 0.7681641708264004,
+      "grad_norm": 0.19881530106067657,
+      "learning_rate": 1.7273338377568707e-05,
+      "loss": 0.5829,
+      "step": 2770
+    },
+    {
+      "epoch": 0.7684414864115363,
+      "grad_norm": 0.17877154052257538,
+      "learning_rate": 1.726823471646623e-05,
+      "loss": 0.5607,
+      "step": 2771
+    },
+    {
+      "epoch": 0.7687188019966722,
+      "grad_norm": 0.1801327019929886,
+      "learning_rate": 1.7263130124954832e-05,
+      "loss": 0.5608,
+      "step": 2772
+    },
+    {
+      "epoch": 0.7689961175818081,
+      "grad_norm": 0.1976090669631958,
+      "learning_rate": 1.7258024604030547e-05,
+      "loss": 0.4987,
+      "step": 2773
+    },
+    {
+      "epoch": 0.769273433166944,
+      "grad_norm": 0.18582318723201752,
+      "learning_rate": 1.72529181546896e-05,
+      "loss": 0.5422,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7695507487520798,
+      "grad_norm": 0.1935378760099411,
+      "learning_rate": 1.7247810777928396e-05,
+      "loss": 0.5464,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7698280643372157,
+      "grad_norm": 0.187955841422081,
+      "learning_rate": 1.7242702474743517e-05,
+      "loss": 0.514,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7701053799223516,
+      "grad_norm": 0.18268531560897827,
+      "learning_rate": 1.7237593246131735e-05,
+      "loss": 0.517,
+      "step": 2777
+    },
+    {
+      "epoch": 0.7703826955074875,
+      "grad_norm": 0.18831866979599,
+      "learning_rate": 1.7232483093089986e-05,
+      "loss": 0.579,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7706600110926234,
+      "grad_norm": 0.20355799794197083,
+      "learning_rate": 1.7227372016615402e-05,
+      "loss": 0.5656,
+      "step": 2779
+    },
+    {
+      "epoch": 0.7709373266777593,
+      "grad_norm": 0.1870361566543579,
+      "learning_rate": 1.7222260017705286e-05,
+      "loss": 0.5528,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7712146422628952,
+      "grad_norm": 0.2032066434621811,
+      "learning_rate": 1.7217147097357127e-05,
+      "loss": 0.5684,
+      "step": 2781
+    },
+    {
+      "epoch": 0.771491957848031,
+      "grad_norm": 0.18692703545093536,
+      "learning_rate": 1.7212033256568595e-05,
+      "loss": 0.5654,
+      "step": 2782
+    },
+    {
+      "epoch": 0.7717692734331669,
+      "grad_norm": 0.1919548362493515,
+      "learning_rate": 1.7206918496337525e-05,
+      "loss": 0.5801,
+      "step": 2783
+    },
+    {
+      "epoch": 0.7720465890183028,
+      "grad_norm": 0.19089831411838531,
+      "learning_rate": 1.7201802817661955e-05,
+      "loss": 0.5617,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7723239046034387,
+      "grad_norm": 0.19473430514335632,
+      "learning_rate": 1.7196686221540077e-05,
+      "loss": 0.5376,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7726012201885746,
+      "grad_norm": 0.1860806941986084,
+      "learning_rate": 1.7191568708970286e-05,
+      "loss": 0.5541,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7728785357737105,
+      "grad_norm": 0.18971897661685944,
+      "learning_rate": 1.7186450280951137e-05,
+      "loss": 0.5514,
+      "step": 2787
+    },
+    {
+      "epoch": 0.7731558513588463,
+      "grad_norm": 0.18644990026950836,
+      "learning_rate": 1.7181330938481375e-05,
+      "loss": 0.5504,
+      "step": 2788
+    },
+    {
+      "epoch": 0.7734331669439822,
+      "grad_norm": 0.18728910386562347,
+      "learning_rate": 1.717621068255992e-05,
+      "loss": 0.5528,
+      "step": 2789
+    },
+    {
+      "epoch": 0.7737104825291181,
+      "grad_norm": 0.19061507284641266,
+      "learning_rate": 1.7171089514185857e-05,
+      "loss": 0.5356,
+      "step": 2790
+    },
+    {
+      "epoch": 0.773987798114254,
+      "grad_norm": 0.18289197981357574,
+      "learning_rate": 1.7165967434358483e-05,
+      "loss": 0.5487,
+      "step": 2791
+    },
+    {
+      "epoch": 0.7742651136993899,
+      "grad_norm": 0.19676977396011353,
+      "learning_rate": 1.716084444407723e-05,
+      "loss": 0.5625,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7745424292845258,
+      "grad_norm": 0.18007270991802216,
+      "learning_rate": 1.7155720544341746e-05,
+      "loss": 0.543,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7748197448696617,
+      "grad_norm": 0.19169475138187408,
+      "learning_rate": 1.715059573615183e-05,
+      "loss": 0.5378,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7750970604547975,
+      "grad_norm": 0.1876417100429535,
+      "learning_rate": 1.714547002050747e-05,
+      "loss": 0.5304,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7753743760399334,
+      "grad_norm": 0.18966078758239746,
+      "learning_rate": 1.714034339840883e-05,
+      "loss": 0.5302,
+      "step": 2796
+    },
+    {
+      "epoch": 0.7756516916250693,
+      "grad_norm": 0.1796397715806961,
+      "learning_rate": 1.7135215870856253e-05,
+      "loss": 0.5388,
+      "step": 2797
+    },
+    {
+      "epoch": 0.7759290072102052,
+      "grad_norm": 0.19350911676883698,
+      "learning_rate": 1.7130087438850252e-05,
+      "loss": 0.553,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7762063227953411,
+      "grad_norm": 0.19393737614154816,
+      "learning_rate": 1.7124958103391516e-05,
+      "loss": 0.514,
+      "step": 2799
+    },
+    {
+      "epoch": 0.776483638380477,
+      "grad_norm": 0.2177378535270691,
+      "learning_rate": 1.711982786548092e-05,
+      "loss": 0.5187,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7767609539656128,
+      "grad_norm": 0.18803465366363525,
+      "learning_rate": 1.7114696726119505e-05,
+      "loss": 0.5661,
+      "step": 2801
+    },
+    {
+      "epoch": 0.7770382695507487,
+      "grad_norm": 0.19588807225227356,
+      "learning_rate": 1.7109564686308498e-05,
+      "loss": 0.5463,
+      "step": 2802
+    },
+    {
+      "epoch": 0.7773155851358846,
+      "grad_norm": 0.18581606447696686,
+      "learning_rate": 1.710443174704929e-05,
+      "loss": 0.548,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7775929007210205,
+      "grad_norm": 0.18562249839305878,
+      "learning_rate": 1.7099297909343455e-05,
+      "loss": 0.5313,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7778702163061564,
+      "grad_norm": 0.17982806265354156,
+      "learning_rate": 1.7094163174192744e-05,
+      "loss": 0.5499,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7781475318912923,
+      "grad_norm": 0.18412619829177856,
+      "learning_rate": 1.708902754259908e-05,
+      "loss": 0.5495,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7784248474764282,
+      "grad_norm": 0.19657637178897858,
+      "learning_rate": 1.7083891015564555e-05,
+      "loss": 0.5565,
+      "step": 2807
+    },
+    {
+      "epoch": 0.778702163061564,
+      "grad_norm": 0.17935718595981598,
+      "learning_rate": 1.7078753594091445e-05,
+      "loss": 0.5446,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7789794786466999,
+      "grad_norm": 0.18647761642932892,
+      "learning_rate": 1.7073615279182198e-05,
+      "loss": 0.5416,
+      "step": 2809
+    },
+    {
+      "epoch": 0.7792567942318358,
+      "grad_norm": 0.19295147061347961,
+      "learning_rate": 1.7068476071839434e-05,
+      "loss": 0.5498,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7795341098169717,
+      "grad_norm": 0.19305361807346344,
+      "learning_rate": 1.706333597306595e-05,
+      "loss": 0.5613,
+      "step": 2811
+    },
+    {
+      "epoch": 0.7798114254021076,
+      "grad_norm": 0.1887744814157486,
+      "learning_rate": 1.7058194983864715e-05,
+      "loss": 0.5493,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7800887409872435,
+      "grad_norm": 0.18676388263702393,
+      "learning_rate": 1.7053053105238866e-05,
+      "loss": 0.527,
+      "step": 2813
+    },
+    {
+      "epoch": 0.7803660565723793,
+      "grad_norm": 0.1888217329978943,
+      "learning_rate": 1.7047910338191732e-05,
+      "loss": 0.5543,
+      "step": 2814
+    },
+    {
+      "epoch": 0.7806433721575152,
+      "grad_norm": 0.19341908395290375,
+      "learning_rate": 1.7042766683726793e-05,
+      "loss": 0.5188,
+      "step": 2815
+    },
+    {
+      "epoch": 0.7809206877426511,
+      "grad_norm": 0.18204385042190552,
+      "learning_rate": 1.7037622142847717e-05,
+      "loss": 0.541,
+      "step": 2816
+    },
+    {
+      "epoch": 0.781198003327787,
+      "grad_norm": 0.20170709490776062,
+      "learning_rate": 1.7032476716558338e-05,
+      "loss": 0.5925,
+      "step": 2817
+    },
+    {
+      "epoch": 0.7814753189129229,
+      "grad_norm": 0.18066342175006866,
+      "learning_rate": 1.7027330405862668e-05,
+      "loss": 0.5537,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7817526344980588,
+      "grad_norm": 0.18021415174007416,
+      "learning_rate": 1.7022183211764886e-05,
+      "loss": 0.5098,
+      "step": 2819
+    },
+    {
+      "epoch": 0.7820299500831946,
+      "grad_norm": 0.19226256012916565,
+      "learning_rate": 1.7017035135269345e-05,
+      "loss": 0.5626,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7823072656683305,
+      "grad_norm": 0.1880798488855362,
+      "learning_rate": 1.7011886177380572e-05,
+      "loss": 0.5601,
+      "step": 2821
+    },
+    {
+      "epoch": 0.7825845812534664,
+      "grad_norm": 0.19588159024715424,
+      "learning_rate": 1.7006736339103267e-05,
+      "loss": 0.56,
+      "step": 2822
+    },
+    {
+      "epoch": 0.7828618968386023,
+      "grad_norm": 0.21878387033939362,
+      "learning_rate": 1.7001585621442295e-05,
+      "loss": 0.5683,
+      "step": 2823
+    },
+    {
+      "epoch": 0.7831392124237382,
+      "grad_norm": 0.18837064504623413,
+      "learning_rate": 1.6996434025402706e-05,
+      "loss": 0.5827,
+      "step": 2824
+    },
+    {
+      "epoch": 0.7834165280088741,
+      "grad_norm": 0.1791253536939621,
+      "learning_rate": 1.6991281551989704e-05,
+      "loss": 0.5125,
+      "step": 2825
+    },
+    {
+      "epoch": 0.78369384359401,
+      "grad_norm": 0.19195421040058136,
+      "learning_rate": 1.698612820220868e-05,
+      "loss": 0.5348,
+      "step": 2826
+    },
+    {
+      "epoch": 0.7839711591791458,
+      "grad_norm": 0.1831100434064865,
+      "learning_rate": 1.6980973977065185e-05,
+      "loss": 0.5354,
+      "step": 2827
+    },
+    {
+      "epoch": 0.7842484747642817,
+      "grad_norm": 0.19085463881492615,
+      "learning_rate": 1.6975818877564945e-05,
+      "loss": 0.5724,
+      "step": 2828
+    },
+    {
+      "epoch": 0.7845257903494176,
+      "grad_norm": 0.18546035885810852,
+      "learning_rate": 1.6970662904713857e-05,
+      "loss": 0.5433,
+      "step": 2829
+    },
+    {
+      "epoch": 0.7848031059345535,
+      "grad_norm": 0.17541223764419556,
+      "learning_rate": 1.6965506059517988e-05,
+      "loss": 0.5348,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7850804215196894,
+      "grad_norm": 0.18578267097473145,
+      "learning_rate": 1.696034834298358e-05,
+      "loss": 0.5355,
+      "step": 2831
+    },
+    {
+      "epoch": 0.7853577371048253,
+      "grad_norm": 0.19077135622501373,
+      "learning_rate": 1.6955189756117028e-05,
+      "loss": 0.5387,
+      "step": 2832
+    },
+    {
+      "epoch": 0.7856350526899611,
+      "grad_norm": 0.19363151490688324,
+      "learning_rate": 1.6950030299924925e-05,
+      "loss": 0.54,
+      "step": 2833
+    },
+    {
+      "epoch": 0.785912368275097,
+      "grad_norm": 0.19038641452789307,
+      "learning_rate": 1.6944869975414e-05,
+      "loss": 0.5365,
+      "step": 2834
+    },
+    {
+      "epoch": 0.7861896838602329,
+      "grad_norm": 0.18864920735359192,
+      "learning_rate": 1.6939708783591184e-05,
+      "loss": 0.5272,
+      "step": 2835
+    },
+    {
+      "epoch": 0.7864669994453688,
+      "grad_norm": 0.195342555642128,
+      "learning_rate": 1.6934546725463558e-05,
+      "loss": 0.5459,
+      "step": 2836
+    },
+    {
+      "epoch": 0.7867443150305047,
+      "grad_norm": 0.1781705915927887,
+      "learning_rate": 1.6929383802038372e-05,
+      "loss": 0.525,
+      "step": 2837
+    },
+    {
+      "epoch": 0.7870216306156406,
+      "grad_norm": 0.18777886033058167,
+      "learning_rate": 1.6924220014323054e-05,
+      "loss": 0.5254,
+      "step": 2838
+    },
+    {
+      "epoch": 0.7872989462007765,
+      "grad_norm": 0.19651533663272858,
+      "learning_rate": 1.6919055363325193e-05,
+      "loss": 0.5388,
+      "step": 2839
+    },
+    {
+      "epoch": 0.7875762617859123,
+      "grad_norm": 0.19021400809288025,
+      "learning_rate": 1.6913889850052546e-05,
+      "loss": 0.5362,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7878535773710482,
+      "grad_norm": 0.18892168998718262,
+      "learning_rate": 1.690872347551305e-05,
+      "loss": 0.5446,
+      "step": 2841
+    },
+    {
+      "epoch": 0.7881308929561841,
+      "grad_norm": 0.1848224401473999,
+      "learning_rate": 1.6903556240714795e-05,
+      "loss": 0.5353,
+      "step": 2842
+    },
+    {
+      "epoch": 0.78840820854132,
+      "grad_norm": 0.19968490302562714,
+      "learning_rate": 1.6898388146666046e-05,
+      "loss": 0.5173,
+      "step": 2843
+    },
+    {
+      "epoch": 0.7886855241264559,
+      "grad_norm": 0.19547419250011444,
+      "learning_rate": 1.689321919437524e-05,
+      "loss": 0.5421,
+      "step": 2844
+    },
+    {
+      "epoch": 0.7889628397115918,
+      "grad_norm": 0.18620312213897705,
+      "learning_rate": 1.688804938485097e-05,
+      "loss": 0.5297,
+      "step": 2845
+    },
+    {
+      "epoch": 0.7892401552967276,
+      "grad_norm": 0.18695352971553802,
+      "learning_rate": 1.6882878719102007e-05,
+      "loss": 0.5386,
+      "step": 2846
+    },
+    {
+      "epoch": 0.7895174708818635,
+      "grad_norm": 0.18656794726848602,
+      "learning_rate": 1.6877707198137285e-05,
+      "loss": 0.557,
+      "step": 2847
+    },
+    {
+      "epoch": 0.7897947864669994,
+      "grad_norm": 0.18816913664340973,
+      "learning_rate": 1.6872534822965903e-05,
+      "loss": 0.5638,
+      "step": 2848
+    },
+    {
+      "epoch": 0.7900721020521353,
+      "grad_norm": 0.18742120265960693,
+      "learning_rate": 1.6867361594597126e-05,
+      "loss": 0.5482,
+      "step": 2849
+    },
+    {
+      "epoch": 0.7903494176372712,
+      "grad_norm": 0.1863267421722412,
+      "learning_rate": 1.6862187514040396e-05,
+      "loss": 0.5471,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7906267332224071,
+      "grad_norm": 0.24633349478244781,
+      "learning_rate": 1.6857012582305303e-05,
+      "loss": 0.5251,
+      "step": 2851
+    },
+    {
+      "epoch": 0.790904048807543,
+      "grad_norm": 0.1922018826007843,
+      "learning_rate": 1.6851836800401624e-05,
+      "loss": 0.5706,
+      "step": 2852
+    },
+    {
+      "epoch": 0.7911813643926788,
+      "grad_norm": 0.18678666651248932,
+      "learning_rate": 1.684666016933928e-05,
+      "loss": 0.55,
+      "step": 2853
+    },
+    {
+      "epoch": 0.7914586799778147,
+      "grad_norm": 0.1959305703639984,
+      "learning_rate": 1.6841482690128376e-05,
+      "loss": 0.5414,
+      "step": 2854
+    },
+    {
+      "epoch": 0.7917359955629506,
+      "grad_norm": 0.19318552315235138,
+      "learning_rate": 1.6836304363779178e-05,
+      "loss": 0.5383,
+      "step": 2855
+    },
+    {
+      "epoch": 0.7920133111480865,
+      "grad_norm": 0.18508724868297577,
+      "learning_rate": 1.6831125191302104e-05,
+      "loss": 0.5371,
+      "step": 2856
+    },
+    {
+      "epoch": 0.7922906267332224,
+      "grad_norm": 0.18992586433887482,
+      "learning_rate": 1.682594517370776e-05,
+      "loss": 0.499,
+      "step": 2857
+    },
+    {
+      "epoch": 0.7925679423183583,
+      "grad_norm": 0.19711445271968842,
+      "learning_rate": 1.682076431200689e-05,
+      "loss": 0.5777,
+      "step": 2858
+    },
+    {
+      "epoch": 0.7928452579034941,
+      "grad_norm": 0.19418618083000183,
+      "learning_rate": 1.6815582607210435e-05,
+      "loss": 0.5459,
+      "step": 2859
+    },
+    {
+      "epoch": 0.79312257348863,
+      "grad_norm": 0.19879227876663208,
+      "learning_rate": 1.6810400060329472e-05,
+      "loss": 0.5578,
+      "step": 2860
+    },
+    {
+      "epoch": 0.7933998890737659,
+      "grad_norm": 0.1837853342294693,
+      "learning_rate": 1.680521667237525e-05,
+      "loss": 0.534,
+      "step": 2861
+    },
+    {
+      "epoch": 0.7936772046589018,
+      "grad_norm": 0.18517963588237762,
+      "learning_rate": 1.68000324443592e-05,
+      "loss": 0.5368,
+      "step": 2862
+    },
+    {
+      "epoch": 0.7939545202440377,
+      "grad_norm": 0.18853336572647095,
+      "learning_rate": 1.6794847377292885e-05,
+      "loss": 0.5728,
+      "step": 2863
+    },
+    {
+      "epoch": 0.7942318358291736,
+      "grad_norm": 0.1775895059108734,
+      "learning_rate": 1.678966147218806e-05,
+      "loss": 0.5283,
+      "step": 2864
+    },
+    {
+      "epoch": 0.7945091514143094,
+      "grad_norm": 0.18305779993534088,
+      "learning_rate": 1.678447473005663e-05,
+      "loss": 0.531,
+      "step": 2865
+    },
+    {
+      "epoch": 0.7947864669994453,
+      "grad_norm": 0.19156721234321594,
+      "learning_rate": 1.6779287151910665e-05,
+      "loss": 0.5485,
+      "step": 2866
+    },
+    {
+      "epoch": 0.7950637825845812,
+      "grad_norm": 0.18802639842033386,
+      "learning_rate": 1.6774098738762398e-05,
+      "loss": 0.548,
+      "step": 2867
+    },
+    {
+      "epoch": 0.7953410981697171,
+      "grad_norm": 0.19210045039653778,
+      "learning_rate": 1.6768909491624224e-05,
+      "loss": 0.548,
+      "step": 2868
+    },
+    {
+      "epoch": 0.795618413754853,
+      "grad_norm": 0.18298813700675964,
+      "learning_rate": 1.6763719411508713e-05,
+      "loss": 0.5385,
+      "step": 2869
+    },
+    {
+      "epoch": 0.7958957293399889,
+      "grad_norm": 0.206680029630661,
+      "learning_rate": 1.675852849942857e-05,
+      "loss": 0.5443,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7961730449251248,
+      "grad_norm": 0.19752557575702667,
+      "learning_rate": 1.67533367563967e-05,
+      "loss": 0.541,
+      "step": 2871
+    },
+    {
+      "epoch": 0.7964503605102606,
+      "grad_norm": 0.1768447309732437,
+      "learning_rate": 1.674814418342613e-05,
+      "loss": 0.5125,
+      "step": 2872
+    },
+    {
+      "epoch": 0.7967276760953965,
+      "grad_norm": 0.16779829561710358,
+      "learning_rate": 1.6742950781530086e-05,
+      "loss": 0.5446,
+      "step": 2873
+    },
+    {
+      "epoch": 0.7970049916805324,
+      "grad_norm": 0.18453587591648102,
+      "learning_rate": 1.6737756551721924e-05,
+      "loss": 0.538,
+      "step": 2874
+    },
+    {
+      "epoch": 0.7972823072656683,
+      "grad_norm": 0.18118412792682648,
+      "learning_rate": 1.673256149501518e-05,
+      "loss": 0.538,
+      "step": 2875
+    },
+    {
+      "epoch": 0.7975596228508042,
+      "grad_norm": 0.18454188108444214,
+      "learning_rate": 1.672736561242355e-05,
+      "loss": 0.5208,
+      "step": 2876
+    },
+    {
+      "epoch": 0.7978369384359401,
+      "grad_norm": 0.18930627405643463,
+      "learning_rate": 1.672216890496089e-05,
+      "loss": 0.5545,
+      "step": 2877
+    },
+    {
+      "epoch": 0.7981142540210759,
+      "grad_norm": 0.18117345869541168,
+      "learning_rate": 1.6716971373641212e-05,
+      "loss": 0.5444,
+      "step": 2878
+    },
+    {
+      "epoch": 0.7983915696062118,
+      "grad_norm": 0.20891335606575012,
+      "learning_rate": 1.671177301947869e-05,
+      "loss": 0.5583,
+      "step": 2879
+    },
+    {
+      "epoch": 0.7986688851913477,
+      "grad_norm": 0.18599353730678558,
+      "learning_rate": 1.670657384348766e-05,
+      "loss": 0.56,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7989462007764836,
+      "grad_norm": 0.23091299831867218,
+      "learning_rate": 1.6701373846682626e-05,
+      "loss": 0.5332,
+      "step": 2881
+    },
+    {
+      "epoch": 0.7992235163616195,
+      "grad_norm": 0.18974825739860535,
+      "learning_rate": 1.6696173030078242e-05,
+      "loss": 0.56,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7995008319467554,
+      "grad_norm": 0.1794738471508026,
+      "learning_rate": 1.6690971394689324e-05,
+      "loss": 0.5671,
+      "step": 2883
+    },
+    {
+      "epoch": 0.7997781475318912,
+      "grad_norm": 0.24422964453697205,
+      "learning_rate": 1.6685768941530848e-05,
+      "loss": 0.5261,
+      "step": 2884
+    },
+    {
+      "epoch": 0.8000554631170271,
+      "grad_norm": 0.18428778648376465,
+      "learning_rate": 1.6680565671617955e-05,
+      "loss": 0.5315,
+      "step": 2885
+    },
+    {
+      "epoch": 0.800332778702163,
+      "grad_norm": 0.18910729885101318,
+      "learning_rate": 1.667536158596593e-05,
+      "loss": 0.5447,
+      "step": 2886
+    },
+    {
+      "epoch": 0.8006100942872989,
+      "grad_norm": 0.1889326423406601,
+      "learning_rate": 1.667015668559024e-05,
+      "loss": 0.5313,
+      "step": 2887
+    },
+    {
+      "epoch": 0.8008874098724348,
+      "grad_norm": 0.19478273391723633,
+      "learning_rate": 1.66649509715065e-05,
+      "loss": 0.5358,
+      "step": 2888
+    },
+    {
+      "epoch": 0.8011647254575707,
+      "grad_norm": 0.1966940313577652,
+      "learning_rate": 1.6659744444730467e-05,
+      "loss": 0.5367,
+      "step": 2889
+    },
+    {
+      "epoch": 0.8014420410427066,
+      "grad_norm": 0.19594216346740723,
+      "learning_rate": 1.665453710627809e-05,
+      "loss": 0.5571,
+      "step": 2890
+    },
+    {
+      "epoch": 0.8017193566278424,
+      "grad_norm": 0.19384315609931946,
+      "learning_rate": 1.6649328957165448e-05,
+      "loss": 0.5478,
+      "step": 2891
+    },
+    {
+      "epoch": 0.8019966722129783,
+      "grad_norm": 0.18817134201526642,
+      "learning_rate": 1.6644119998408795e-05,
+      "loss": 0.5533,
+      "step": 2892
+    },
+    {
+      "epoch": 0.8022739877981142,
+      "grad_norm": 0.19234129786491394,
+      "learning_rate": 1.6638910231024528e-05,
+      "loss": 0.55,
+      "step": 2893
+    },
+    {
+      "epoch": 0.8025513033832501,
+      "grad_norm": 0.19165417551994324,
+      "learning_rate": 1.6633699656029224e-05,
+      "loss": 0.5372,
+      "step": 2894
+    },
+    {
+      "epoch": 0.802828618968386,
+      "grad_norm": 0.20091277360916138,
+      "learning_rate": 1.6628488274439592e-05,
+      "loss": 0.5571,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8031059345535219,
+      "grad_norm": 0.1875421106815338,
+      "learning_rate": 1.6623276087272517e-05,
+      "loss": 0.5346,
+      "step": 2896
+    },
+    {
+      "epoch": 0.8033832501386577,
+      "grad_norm": 0.24504978954792023,
+      "learning_rate": 1.661806309554503e-05,
+      "loss": 0.5277,
+      "step": 2897
+    },
+    {
+      "epoch": 0.8036605657237936,
+      "grad_norm": 0.188764289021492,
+      "learning_rate": 1.661284930027433e-05,
+      "loss": 0.5354,
+      "step": 2898
+    },
+    {
+      "epoch": 0.8039378813089295,
+      "grad_norm": 0.1898432970046997,
+      "learning_rate": 1.6607634702477765e-05,
+      "loss": 0.5349,
+      "step": 2899
+    },
+    {
+      "epoch": 0.8042151968940654,
+      "grad_norm": 0.18189935386180878,
+      "learning_rate": 1.6602419303172835e-05,
+      "loss": 0.5146,
+      "step": 2900
+    },
+    {
+      "epoch": 0.8044925124792013,
+      "grad_norm": 0.19001318514347076,
+      "learning_rate": 1.659720310337721e-05,
+      "loss": 0.5438,
+      "step": 2901
+    },
+    {
+      "epoch": 0.8047698280643372,
+      "grad_norm": 0.19630925357341766,
+      "learning_rate": 1.6591986104108706e-05,
+      "loss": 0.5644,
+      "step": 2902
+    },
+    {
+      "epoch": 0.805047143649473,
+      "grad_norm": 0.1834658980369568,
+      "learning_rate": 1.65867683063853e-05,
+      "loss": 0.5588,
+      "step": 2903
+    },
+    {
+      "epoch": 0.8053244592346089,
+      "grad_norm": 0.198238343000412,
+      "learning_rate": 1.658154971122512e-05,
+      "loss": 0.5471,
+      "step": 2904
+    },
+    {
+      "epoch": 0.8056017748197448,
+      "grad_norm": 0.18793722987174988,
+      "learning_rate": 1.657633031964645e-05,
+      "loss": 0.5416,
+      "step": 2905
+    },
+    {
+      "epoch": 0.8058790904048807,
+      "grad_norm": 0.1919628083705902,
+      "learning_rate": 1.657111013266774e-05,
+      "loss": 0.5489,
+      "step": 2906
+    },
+    {
+      "epoch": 0.8061564059900166,
+      "grad_norm": 0.18970000743865967,
+      "learning_rate": 1.6565889151307576e-05,
+      "loss": 0.5374,
+      "step": 2907
+    },
+    {
+      "epoch": 0.8064337215751525,
+      "grad_norm": 0.19195859134197235,
+      "learning_rate": 1.656066737658471e-05,
+      "loss": 0.5696,
+      "step": 2908
+    },
+    {
+      "epoch": 0.8067110371602884,
+      "grad_norm": 0.18715813755989075,
+      "learning_rate": 1.6555444809518066e-05,
+      "loss": 0.5365,
+      "step": 2909
+    },
+    {
+      "epoch": 0.8069883527454242,
+      "grad_norm": 0.19223229587078094,
+      "learning_rate": 1.6550221451126682e-05,
+      "loss": 0.5148,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8072656683305601,
+      "grad_norm": 0.2136392742395401,
+      "learning_rate": 1.6544997302429794e-05,
+      "loss": 0.5269,
+      "step": 2911
+    },
+    {
+      "epoch": 0.807542983915696,
+      "grad_norm": 0.19737182557582855,
+      "learning_rate": 1.6539772364446755e-05,
+      "loss": 0.5592,
+      "step": 2912
+    },
+    {
+      "epoch": 0.8078202995008319,
+      "grad_norm": 0.20777487754821777,
+      "learning_rate": 1.6534546638197098e-05,
+      "loss": 0.5124,
+      "step": 2913
+    },
+    {
+      "epoch": 0.8080976150859678,
+      "grad_norm": 0.1888158619403839,
+      "learning_rate": 1.6529320124700495e-05,
+      "loss": 0.5433,
+      "step": 2914
+    },
+    {
+      "epoch": 0.8083749306711037,
+      "grad_norm": 0.24319934844970703,
+      "learning_rate": 1.6524092824976787e-05,
+      "loss": 0.5763,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8086522462562395,
+      "grad_norm": 0.19144728779792786,
+      "learning_rate": 1.6518864740045947e-05,
+      "loss": 0.5245,
+      "step": 2916
+    },
+    {
+      "epoch": 0.8089295618413754,
+      "grad_norm": 0.19189846515655518,
+      "learning_rate": 1.6513635870928122e-05,
+      "loss": 0.5435,
+      "step": 2917
+    },
+    {
+      "epoch": 0.8092068774265113,
+      "grad_norm": 0.24125546216964722,
+      "learning_rate": 1.6508406218643597e-05,
+      "loss": 0.5599,
+      "step": 2918
+    },
+    {
+      "epoch": 0.8094841930116472,
+      "grad_norm": 0.2077561616897583,
+      "learning_rate": 1.650317578421282e-05,
+      "loss": 0.5466,
+      "step": 2919
+    },
+    {
+      "epoch": 0.8097615085967831,
+      "grad_norm": 0.2178879827260971,
+      "learning_rate": 1.6497944568656383e-05,
+      "loss": 0.5306,
+      "step": 2920
+    },
+    {
+      "epoch": 0.810038824181919,
+      "grad_norm": 0.24422143399715424,
+      "learning_rate": 1.649271257299504e-05,
+      "loss": 0.5388,
+      "step": 2921
+    },
+    {
+      "epoch": 0.8103161397670549,
+      "grad_norm": 0.18929331004619598,
+      "learning_rate": 1.6487479798249687e-05,
+      "loss": 0.5236,
+      "step": 2922
+    },
+    {
+      "epoch": 0.8105934553521907,
+      "grad_norm": 0.17720143496990204,
+      "learning_rate": 1.648224624544138e-05,
+      "loss": 0.5172,
+      "step": 2923
+    },
+    {
+      "epoch": 0.8108707709373266,
+      "grad_norm": 0.18783891201019287,
+      "learning_rate": 1.6477011915591325e-05,
+      "loss": 0.5389,
+      "step": 2924
+    },
+    {
+      "epoch": 0.8111480865224625,
+      "grad_norm": 0.19297097623348236,
+      "learning_rate": 1.6471776809720873e-05,
+      "loss": 0.543,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8114254021075984,
+      "grad_norm": 0.18489378690719604,
+      "learning_rate": 1.6466540928851538e-05,
+      "loss": 0.5512,
+      "step": 2926
+    },
+    {
+      "epoch": 0.8117027176927343,
+      "grad_norm": 0.18271875381469727,
+      "learning_rate": 1.6461304274004972e-05,
+      "loss": 0.5417,
+      "step": 2927
+    },
+    {
+      "epoch": 0.8119800332778702,
+      "grad_norm": 0.1847040057182312,
+      "learning_rate": 1.6456066846202994e-05,
+      "loss": 0.5387,
+      "step": 2928
+    },
+    {
+      "epoch": 0.812257348863006,
+      "grad_norm": 0.20353619754314423,
+      "learning_rate": 1.6450828646467555e-05,
+      "loss": 0.5454,
+      "step": 2929
+    },
+    {
+      "epoch": 0.812534664448142,
+      "grad_norm": 0.2704160809516907,
+      "learning_rate": 1.644558967582078e-05,
+      "loss": 0.5598,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8128119800332779,
+      "grad_norm": 0.2357567995786667,
+      "learning_rate": 1.6440349935284917e-05,
+      "loss": 0.5724,
+      "step": 2931
+    },
+    {
+      "epoch": 0.8130892956184138,
+      "grad_norm": 0.18530336022377014,
+      "learning_rate": 1.6435109425882385e-05,
+      "loss": 0.5365,
+      "step": 2932
+    },
+    {
+      "epoch": 0.8133666112035497,
+      "grad_norm": 0.18962644040584564,
+      "learning_rate": 1.6429868148635745e-05,
+      "loss": 0.5429,
+      "step": 2933
+    },
+    {
+      "epoch": 0.8136439267886856,
+      "grad_norm": 0.1827327460050583,
+      "learning_rate": 1.6424626104567708e-05,
+      "loss": 0.5368,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8139212423738215,
+      "grad_norm": 0.19838617742061615,
+      "learning_rate": 1.641938329470114e-05,
+      "loss": 0.5366,
+      "step": 2935
+    },
+    {
+      "epoch": 0.8141985579589573,
+      "grad_norm": 0.18842142820358276,
+      "learning_rate": 1.6414139720059045e-05,
+      "loss": 0.546,
+      "step": 2936
+    },
+    {
+      "epoch": 0.8144758735440932,
+      "grad_norm": 0.1957085281610489,
+      "learning_rate": 1.6408895381664594e-05,
+      "loss": 0.5144,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8147531891292291,
+      "grad_norm": 0.18413354456424713,
+      "learning_rate": 1.6403650280541087e-05,
+      "loss": 0.5441,
+      "step": 2938
+    },
+    {
+      "epoch": 0.815030504714365,
+      "grad_norm": 0.18572686612606049,
+      "learning_rate": 1.6398404417711984e-05,
+      "loss": 0.543,
+      "step": 2939
+    },
+    {
+      "epoch": 0.8153078202995009,
+      "grad_norm": 0.18513022363185883,
+      "learning_rate": 1.639315779420089e-05,
+      "loss": 0.5416,
+      "step": 2940
+    },
+    {
+      "epoch": 0.8155851358846368,
+      "grad_norm": 0.1878947764635086,
+      "learning_rate": 1.6387910411031564e-05,
+      "loss": 0.5487,
+      "step": 2941
+    },
+    {
+      "epoch": 0.8158624514697727,
+      "grad_norm": 0.20760974287986755,
+      "learning_rate": 1.6382662269227912e-05,
+      "loss": 0.5861,
+      "step": 2942
+    },
+    {
+      "epoch": 0.8161397670549085,
+      "grad_norm": 0.1904260218143463,
+      "learning_rate": 1.637741336981398e-05,
+      "loss": 0.549,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8164170826400444,
+      "grad_norm": 0.1811297982931137,
+      "learning_rate": 1.637216371381397e-05,
+      "loss": 0.5556,
+      "step": 2944
+    },
+    {
+      "epoch": 0.8166943982251803,
+      "grad_norm": 0.18820282816886902,
+      "learning_rate": 1.6366913302252228e-05,
+      "loss": 0.5262,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8169717138103162,
+      "grad_norm": 0.18562357127666473,
+      "learning_rate": 1.636166213615325e-05,
+      "loss": 0.5353,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8172490293954521,
+      "grad_norm": 0.19570721685886383,
+      "learning_rate": 1.6356410216541675e-05,
+      "loss": 0.524,
+      "step": 2947
+    },
+    {
+      "epoch": 0.817526344980588,
+      "grad_norm": 0.19673757255077362,
+      "learning_rate": 1.635115754444229e-05,
+      "loss": 0.5532,
+      "step": 2948
+    },
+    {
+      "epoch": 0.8178036605657238,
+      "grad_norm": 0.1866602897644043,
+      "learning_rate": 1.6345904120880045e-05,
+      "loss": 0.5407,
+      "step": 2949
+    },
+    {
+      "epoch": 0.8180809761508597,
+      "grad_norm": 0.1872847080230713,
+      "learning_rate": 1.634064994688e-05,
+      "loss": 0.522,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8183582917359956,
+      "grad_norm": 0.18448910117149353,
+      "learning_rate": 1.63353950234674e-05,
+      "loss": 0.5506,
+      "step": 2951
+    },
+    {
+      "epoch": 0.8186356073211315,
+      "grad_norm": 0.19117306172847748,
+      "learning_rate": 1.6330139351667607e-05,
+      "loss": 0.538,
+      "step": 2952
+    },
+    {
+      "epoch": 0.8189129229062674,
+      "grad_norm": 0.19007255136966705,
+      "learning_rate": 1.6324882932506152e-05,
+      "loss": 0.545,
+      "step": 2953
+    },
+    {
+      "epoch": 0.8191902384914033,
+      "grad_norm": 0.1907234787940979,
+      "learning_rate": 1.63196257670087e-05,
+      "loss": 0.5216,
+      "step": 2954
+    },
+    {
+      "epoch": 0.8194675540765392,
+      "grad_norm": 0.19023658335208893,
+      "learning_rate": 1.6314367856201063e-05,
+      "loss": 0.528,
+      "step": 2955
+    },
+    {
+      "epoch": 0.819744869661675,
+      "grad_norm": 0.19689738750457764,
+      "learning_rate": 1.6309109201109197e-05,
+      "loss": 0.5579,
+      "step": 2956
+    },
+    {
+      "epoch": 0.8200221852468109,
+      "grad_norm": 0.18936869502067566,
+      "learning_rate": 1.63038498027592e-05,
+      "loss": 0.5171,
+      "step": 2957
+    },
+    {
+      "epoch": 0.8202995008319468,
+      "grad_norm": 0.18959283828735352,
+      "learning_rate": 1.6298589662177334e-05,
+      "loss": 0.5109,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8205768164170827,
+      "grad_norm": 0.18975764513015747,
+      "learning_rate": 1.6293328780389976e-05,
+      "loss": 0.5241,
+      "step": 2959
+    },
+    {
+      "epoch": 0.8208541320022186,
+      "grad_norm": 0.1963840126991272,
+      "learning_rate": 1.6288067158423676e-05,
+      "loss": 0.5311,
+      "step": 2960
+    },
+    {
+      "epoch": 0.8211314475873545,
+      "grad_norm": 0.1869080811738968,
+      "learning_rate": 1.6282804797305107e-05,
+      "loss": 0.5268,
+      "step": 2961
+    },
+    {
+      "epoch": 0.8214087631724903,
+      "grad_norm": 0.19208255410194397,
+      "learning_rate": 1.62775416980611e-05,
+      "loss": 0.5561,
+      "step": 2962
+    },
+    {
+      "epoch": 0.8216860787576262,
+      "grad_norm": 0.21065904200077057,
+      "learning_rate": 1.6272277861718622e-05,
+      "loss": 0.5612,
+      "step": 2963
+    },
+    {
+      "epoch": 0.8219633943427621,
+      "grad_norm": 0.19388340413570404,
+      "learning_rate": 1.626701328930479e-05,
+      "loss": 0.5448,
+      "step": 2964
+    },
+    {
+      "epoch": 0.822240709927898,
+      "grad_norm": 0.19354204833507538,
+      "learning_rate": 1.626174798184686e-05,
+      "loss": 0.5521,
+      "step": 2965
+    },
+    {
+      "epoch": 0.8225180255130339,
+      "grad_norm": 0.1918734908103943,
+      "learning_rate": 1.6256481940372235e-05,
+      "loss": 0.5737,
+      "step": 2966
+    },
+    {
+      "epoch": 0.8227953410981698,
+      "grad_norm": 0.19950683414936066,
+      "learning_rate": 1.625121516590845e-05,
+      "loss": 0.5251,
+      "step": 2967
+    },
+    {
+      "epoch": 0.8230726566833056,
+      "grad_norm": 0.1884751170873642,
+      "learning_rate": 1.624594765948321e-05,
+      "loss": 0.5487,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8233499722684415,
+      "grad_norm": 0.20038394629955292,
+      "learning_rate": 1.624067942212433e-05,
+      "loss": 0.5366,
+      "step": 2969
+    },
+    {
+      "epoch": 0.8236272878535774,
+      "grad_norm": 0.18730241060256958,
+      "learning_rate": 1.6235410454859784e-05,
+      "loss": 0.5574,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8239046034387133,
+      "grad_norm": 0.1982334703207016,
+      "learning_rate": 1.6230140758717692e-05,
+      "loss": 0.5472,
+      "step": 2971
+    },
+    {
+      "epoch": 0.8241819190238492,
+      "grad_norm": 0.19432491064071655,
+      "learning_rate": 1.6224870334726315e-05,
+      "loss": 0.54,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8244592346089851,
+      "grad_norm": 0.1818057745695114,
+      "learning_rate": 1.6219599183914038e-05,
+      "loss": 0.5594,
+      "step": 2973
+    },
+    {
+      "epoch": 0.824736550194121,
+      "grad_norm": 0.18658170104026794,
+      "learning_rate": 1.6214327307309417e-05,
+      "loss": 0.537,
+      "step": 2974
+    },
+    {
+      "epoch": 0.8250138657792568,
+      "grad_norm": 0.19775459170341492,
+      "learning_rate": 1.620905470594113e-05,
+      "loss": 0.5423,
+      "step": 2975
+    },
+    {
+      "epoch": 0.8252911813643927,
+      "grad_norm": 0.1877543330192566,
+      "learning_rate": 1.6203781380837997e-05,
+      "loss": 0.5337,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8255684969495286,
+      "grad_norm": 0.1873265951871872,
+      "learning_rate": 1.619850733302899e-05,
+      "loss": 0.5153,
+      "step": 2977
+    },
+    {
+      "epoch": 0.8258458125346645,
+      "grad_norm": 0.1832212209701538,
+      "learning_rate": 1.619323256354321e-05,
+      "loss": 0.5291,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8261231281198004,
+      "grad_norm": 0.1939922571182251,
+      "learning_rate": 1.6187957073409907e-05,
+      "loss": 0.5428,
+      "step": 2979
+    },
+    {
+      "epoch": 0.8264004437049363,
+      "grad_norm": 0.2593255639076233,
+      "learning_rate": 1.6182680863658468e-05,
+      "loss": 0.5448,
+      "step": 2980
+    },
+    {
+      "epoch": 0.8266777592900721,
+      "grad_norm": 0.18556508421897888,
+      "learning_rate": 1.6177403935318422e-05,
+      "loss": 0.529,
+      "step": 2981
+    },
+    {
+      "epoch": 0.826955074875208,
+      "grad_norm": 0.20643781125545502,
+      "learning_rate": 1.6172126289419437e-05,
+      "loss": 0.535,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8272323904603439,
+      "grad_norm": 0.192849263548851,
+      "learning_rate": 1.6166847926991324e-05,
+      "loss": 0.5457,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8275097060454798,
+      "grad_norm": 0.19120796024799347,
+      "learning_rate": 1.616156884906403e-05,
+      "loss": 0.542,
+      "step": 2984
+    },
+    {
+      "epoch": 0.8277870216306157,
+      "grad_norm": 0.20270869135856628,
+      "learning_rate": 1.615628905666764e-05,
+      "loss": 0.5621,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8280643372157516,
+      "grad_norm": 0.20787782967090607,
+      "learning_rate": 1.6151008550832377e-05,
+      "loss": 0.5007,
+      "step": 2986
+    },
+    {
+      "epoch": 0.8283416528008875,
+      "grad_norm": 0.23962002992630005,
+      "learning_rate": 1.6145727332588626e-05,
+      "loss": 0.5351,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8286189683860233,
+      "grad_norm": 0.19550423324108124,
+      "learning_rate": 1.614044540296687e-05,
+      "loss": 0.5599,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8288962839711592,
+      "grad_norm": 0.20781289041042328,
+      "learning_rate": 1.6135162762997776e-05,
+      "loss": 0.5381,
+      "step": 2989
+    },
+    {
+      "epoch": 0.8291735995562951,
+      "grad_norm": 0.1908697932958603,
+      "learning_rate": 1.61298794137121e-05,
+      "loss": 0.5109,
+      "step": 2990
+    },
+    {
+      "epoch": 0.829450915141431,
+      "grad_norm": 0.19248178601264954,
+      "learning_rate": 1.6124595356140794e-05,
+      "loss": 0.5061,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8297282307265669,
+      "grad_norm": 0.4288308918476105,
+      "learning_rate": 1.611931059131489e-05,
+      "loss": 0.5418,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8300055463117028,
+      "grad_norm": 0.2553388178348541,
+      "learning_rate": 1.6114025120265604e-05,
+      "loss": 0.5308,
+      "step": 2993
+    },
+    {
+      "epoch": 0.8302828618968386,
+      "grad_norm": 0.19662028551101685,
+      "learning_rate": 1.6108738944024265e-05,
+      "loss": 0.5647,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8305601774819745,
+      "grad_norm": 0.1986682265996933,
+      "learning_rate": 1.6103452063622343e-05,
+      "loss": 0.5594,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8308374930671104,
+      "grad_norm": 0.18815158307552338,
+      "learning_rate": 1.6098164480091454e-05,
+      "loss": 0.5347,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8311148086522463,
+      "grad_norm": 0.19016186892986298,
+      "learning_rate": 1.6092876194463343e-05,
+      "loss": 0.5022,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8313921242373822,
+      "grad_norm": 0.19305314123630524,
+      "learning_rate": 1.6087587207769897e-05,
+      "loss": 0.5561,
+      "step": 2998
+    },
+    {
+      "epoch": 0.8316694398225181,
+      "grad_norm": 0.19363488256931305,
+      "learning_rate": 1.6082297521043134e-05,
+      "loss": 0.5704,
+      "step": 2999
+    },
+    {
+      "epoch": 0.831946755407654,
+      "grad_norm": 0.2677402198314667,
+      "learning_rate": 1.6077007135315212e-05,
+      "loss": 0.5709,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8322240709927898,
+      "grad_norm": 0.18514001369476318,
+      "learning_rate": 1.6071716051618426e-05,
+      "loss": 0.5328,
+      "step": 3001
+    },
+    {
+      "epoch": 0.8325013865779257,
+      "grad_norm": 0.20008349418640137,
+      "learning_rate": 1.606642427098521e-05,
+      "loss": 0.5535,
+      "step": 3002
+    },
+    {
+      "epoch": 0.8327787021630616,
+      "grad_norm": 0.20995013415813446,
+      "learning_rate": 1.606113179444813e-05,
+      "loss": 0.5541,
+      "step": 3003
+    },
+    {
+      "epoch": 0.8330560177481975,
+      "grad_norm": 0.21254870295524597,
+      "learning_rate": 1.6055838623039886e-05,
+      "loss": 0.5221,
+      "step": 3004
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.1829461008310318,
+      "learning_rate": 1.6050544757793312e-05,
+      "loss": 0.5445,
+      "step": 3005
+    },
+    {
+      "epoch": 0.8336106489184693,
+      "grad_norm": 0.20135165750980377,
+      "learning_rate": 1.604525019974139e-05,
+      "loss": 0.5337,
+      "step": 3006
+    },
+    {
+      "epoch": 0.8338879645036051,
+      "grad_norm": 0.1872393935918808,
+      "learning_rate": 1.6039954949917218e-05,
+      "loss": 0.5506,
+      "step": 3007
+    },
+    {
+      "epoch": 0.834165280088741,
+      "grad_norm": 0.17982006072998047,
+      "learning_rate": 1.6034659009354055e-05,
+      "loss": 0.5252,
+      "step": 3008
+    },
+    {
+      "epoch": 0.8344425956738769,
+      "grad_norm": 0.19605563580989838,
+      "learning_rate": 1.6029362379085264e-05,
+      "loss": 0.527,
+      "step": 3009
+    },
+    {
+      "epoch": 0.8347199112590128,
+      "grad_norm": 0.20329341292381287,
+      "learning_rate": 1.602406506014437e-05,
+      "loss": 0.5729,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8349972268441487,
+      "grad_norm": 0.18132315576076508,
+      "learning_rate": 1.6018767053565008e-05,
+      "loss": 0.5391,
+      "step": 3011
+    },
+    {
+      "epoch": 0.8352745424292846,
+      "grad_norm": 0.19777770340442657,
+      "learning_rate": 1.6013468360380966e-05,
+      "loss": 0.5366,
+      "step": 3012
+    },
+    {
+      "epoch": 0.8355518580144204,
+      "grad_norm": 0.19389207661151886,
+      "learning_rate": 1.6008168981626164e-05,
+      "loss": 0.5569,
+      "step": 3013
+    },
+    {
+      "epoch": 0.8358291735995563,
+      "grad_norm": 0.18683470785617828,
+      "learning_rate": 1.6002868918334647e-05,
+      "loss": 0.5496,
+      "step": 3014
+    },
+    {
+      "epoch": 0.8361064891846922,
+      "grad_norm": 0.18466855585575104,
+      "learning_rate": 1.5997568171540594e-05,
+      "loss": 0.5547,
+      "step": 3015
+    },
+    {
+      "epoch": 0.8363838047698281,
+      "grad_norm": 0.19307468831539154,
+      "learning_rate": 1.5992266742278322e-05,
+      "loss": 0.5518,
+      "step": 3016
+    },
+    {
+      "epoch": 0.836661120354964,
+      "grad_norm": 0.19225814938545227,
+      "learning_rate": 1.5986964631582287e-05,
+      "loss": 0.548,
+      "step": 3017
+    },
+    {
+      "epoch": 0.8369384359400999,
+      "grad_norm": 0.19282346963882446,
+      "learning_rate": 1.5981661840487063e-05,
+      "loss": 0.5594,
+      "step": 3018
+    },
+    {
+      "epoch": 0.8372157515252358,
+      "grad_norm": 0.18779483437538147,
+      "learning_rate": 1.5976358370027373e-05,
+      "loss": 0.5191,
+      "step": 3019
+    },
+    {
+      "epoch": 0.8374930671103716,
+      "grad_norm": 0.18765175342559814,
+      "learning_rate": 1.597105422123806e-05,
+      "loss": 0.5401,
+      "step": 3020
+    },
+    {
+      "epoch": 0.8377703826955075,
+      "grad_norm": 0.1844259798526764,
+      "learning_rate": 1.5965749395154107e-05,
+      "loss": 0.5459,
+      "step": 3021
+    },
+    {
+      "epoch": 0.8380476982806434,
+      "grad_norm": 0.18752428889274597,
+      "learning_rate": 1.5960443892810617e-05,
+      "loss": 0.5339,
+      "step": 3022
+    },
+    {
+      "epoch": 0.8383250138657793,
+      "grad_norm": 0.19530834257602692,
+      "learning_rate": 1.5955137715242847e-05,
+      "loss": 0.5479,
+      "step": 3023
+    },
+    {
+      "epoch": 0.8386023294509152,
+      "grad_norm": 0.1960788518190384,
+      "learning_rate": 1.5949830863486166e-05,
+      "loss": 0.5511,
+      "step": 3024
+    },
+    {
+      "epoch": 0.8388796450360511,
+      "grad_norm": 0.2669583261013031,
+      "learning_rate": 1.594452333857608e-05,
+      "loss": 0.5637,
+      "step": 3025
+    },
+    {
+      "epoch": 0.8391569606211869,
+      "grad_norm": 0.18217816948890686,
+      "learning_rate": 1.5939215141548224e-05,
+      "loss": 0.4905,
+      "step": 3026
+    },
+    {
+      "epoch": 0.8394342762063228,
+      "grad_norm": 0.19247345626354218,
+      "learning_rate": 1.5933906273438383e-05,
+      "loss": 0.5234,
+      "step": 3027
+    },
+    {
+      "epoch": 0.8397115917914587,
+      "grad_norm": 0.1867692619562149,
+      "learning_rate": 1.592859673528244e-05,
+      "loss": 0.5166,
+      "step": 3028
+    },
+    {
+      "epoch": 0.8399889073765946,
+      "grad_norm": 0.18724261224269867,
+      "learning_rate": 1.5923286528116446e-05,
+      "loss": 0.5524,
+      "step": 3029
+    },
+    {
+      "epoch": 0.8402662229617305,
+      "grad_norm": 0.18874408304691315,
+      "learning_rate": 1.5917975652976544e-05,
+      "loss": 0.5283,
+      "step": 3030
+    },
+    {
+      "epoch": 0.8405435385468664,
+      "grad_norm": 0.2242184728384018,
+      "learning_rate": 1.5912664110899038e-05,
+      "loss": 0.532,
+      "step": 3031
+    },
+    {
+      "epoch": 0.8408208541320022,
+      "grad_norm": 0.20637090504169464,
+      "learning_rate": 1.5907351902920346e-05,
+      "loss": 0.5619,
+      "step": 3032
+    },
+    {
+      "epoch": 0.8410981697171381,
+      "grad_norm": 0.20153935253620148,
+      "learning_rate": 1.590203903007702e-05,
+      "loss": 0.5704,
+      "step": 3033
+    },
+    {
+      "epoch": 0.841375485302274,
+      "grad_norm": 0.1838448941707611,
+      "learning_rate": 1.5896725493405746e-05,
+      "loss": 0.5533,
+      "step": 3034
+    },
+    {
+      "epoch": 0.8416528008874099,
+      "grad_norm": 0.187151700258255,
+      "learning_rate": 1.589141129394333e-05,
+      "loss": 0.5276,
+      "step": 3035
+    },
+    {
+      "epoch": 0.8419301164725458,
+      "grad_norm": 0.19188909232616425,
+      "learning_rate": 1.5886096432726723e-05,
+      "loss": 0.5588,
+      "step": 3036
+    },
+    {
+      "epoch": 0.8422074320576817,
+      "grad_norm": 0.18532563745975494,
+      "learning_rate": 1.5880780910792984e-05,
+      "loss": 0.5363,
+      "step": 3037
+    },
+    {
+      "epoch": 0.8424847476428176,
+      "grad_norm": 0.18388155102729797,
+      "learning_rate": 1.587546472917932e-05,
+      "loss": 0.5544,
+      "step": 3038
+    },
+    {
+      "epoch": 0.8427620632279534,
+      "grad_norm": 0.18780486285686493,
+      "learning_rate": 1.5870147888923054e-05,
+      "loss": 0.5501,
+      "step": 3039
+    },
+    {
+      "epoch": 0.8430393788130893,
+      "grad_norm": 0.18718671798706055,
+      "learning_rate": 1.5864830391061644e-05,
+      "loss": 0.5491,
+      "step": 3040
+    },
+    {
+      "epoch": 0.8433166943982252,
+      "grad_norm": 0.1841050237417221,
+      "learning_rate": 1.585951223663268e-05,
+      "loss": 0.539,
+      "step": 3041
+    },
+    {
+      "epoch": 0.8435940099833611,
+      "grad_norm": 0.17960785329341888,
+      "learning_rate": 1.5854193426673862e-05,
+      "loss": 0.5134,
+      "step": 3042
+    },
+    {
+      "epoch": 0.843871325568497,
+      "grad_norm": 0.1937013566493988,
+      "learning_rate": 1.5848873962223044e-05,
+      "loss": 0.5083,
+      "step": 3043
+    },
+    {
+      "epoch": 0.8441486411536329,
+      "grad_norm": 0.18566550314426422,
+      "learning_rate": 1.5843553844318193e-05,
+      "loss": 0.5569,
+      "step": 3044
+    },
+    {
+      "epoch": 0.8444259567387687,
+      "grad_norm": 0.1879187971353531,
+      "learning_rate": 1.5838233073997395e-05,
+      "loss": 0.5611,
+      "step": 3045
+    },
+    {
+      "epoch": 0.8447032723239046,
+      "grad_norm": 0.1857539415359497,
+      "learning_rate": 1.5832911652298882e-05,
+      "loss": 0.5464,
+      "step": 3046
+    },
+    {
+      "epoch": 0.8449805879090405,
+      "grad_norm": 0.1914960891008377,
+      "learning_rate": 1.5827589580261e-05,
+      "loss": 0.5547,
+      "step": 3047
+    },
+    {
+      "epoch": 0.8452579034941764,
+      "grad_norm": 0.18924832344055176,
+      "learning_rate": 1.582226685892223e-05,
+      "loss": 0.5739,
+      "step": 3048
+    },
+    {
+      "epoch": 0.8455352190793123,
+      "grad_norm": 0.18020758032798767,
+      "learning_rate": 1.5816943489321174e-05,
+      "loss": 0.5355,
+      "step": 3049
+    },
+    {
+      "epoch": 0.8458125346644482,
+      "grad_norm": 0.22242438793182373,
+      "learning_rate": 1.5811619472496562e-05,
+      "loss": 0.5551,
+      "step": 3050
+    },
+    {
+      "epoch": 0.846089850249584,
+      "grad_norm": 0.18375588953495026,
+      "learning_rate": 1.5806294809487248e-05,
+      "loss": 0.5393,
+      "step": 3051
+    },
+    {
+      "epoch": 0.8463671658347199,
+      "grad_norm": 0.20009155571460724,
+      "learning_rate": 1.5800969501332223e-05,
+      "loss": 0.5555,
+      "step": 3052
+    },
+    {
+      "epoch": 0.8466444814198558,
+      "grad_norm": 0.1867419183254242,
+      "learning_rate": 1.5795643549070588e-05,
+      "loss": 0.5541,
+      "step": 3053
+    },
+    {
+      "epoch": 0.8469217970049917,
+      "grad_norm": 0.20315201580524445,
+      "learning_rate": 1.5790316953741583e-05,
+      "loss": 0.5302,
+      "step": 3054
+    },
+    {
+      "epoch": 0.8471991125901276,
+      "grad_norm": 0.19646741449832916,
+      "learning_rate": 1.578498971638456e-05,
+      "loss": 0.5491,
+      "step": 3055
+    },
+    {
+      "epoch": 0.8474764281752635,
+      "grad_norm": 0.1929592341184616,
+      "learning_rate": 1.5779661838039013e-05,
+      "loss": 0.554,
+      "step": 3056
+    },
+    {
+      "epoch": 0.8477537437603994,
+      "grad_norm": 0.18516919016838074,
+      "learning_rate": 1.577433331974455e-05,
+      "loss": 0.5474,
+      "step": 3057
+    },
+    {
+      "epoch": 0.8480310593455352,
+      "grad_norm": 0.17926859855651855,
+      "learning_rate": 1.57690041625409e-05,
+      "loss": 0.541,
+      "step": 3058
+    },
+    {
+      "epoch": 0.8483083749306711,
+      "grad_norm": 0.19980867207050323,
+      "learning_rate": 1.576367436746793e-05,
+      "loss": 0.581,
+      "step": 3059
+    },
+    {
+      "epoch": 0.848585690515807,
+      "grad_norm": 0.20209050178527832,
+      "learning_rate": 1.575834393556562e-05,
+      "loss": 0.5749,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8488630061009429,
+      "grad_norm": 0.19711333513259888,
+      "learning_rate": 1.575301286787408e-05,
+      "loss": 0.5537,
+      "step": 3061
+    },
+    {
+      "epoch": 0.8491403216860788,
+      "grad_norm": 0.185336172580719,
+      "learning_rate": 1.5747681165433544e-05,
+      "loss": 0.5378,
+      "step": 3062
+    },
+    {
+      "epoch": 0.8494176372712147,
+      "grad_norm": 0.2029864341020584,
+      "learning_rate": 1.5742348829284366e-05,
+      "loss": 0.5489,
+      "step": 3063
+    },
+    {
+      "epoch": 0.8496949528563505,
+      "grad_norm": 0.18520487844944,
+      "learning_rate": 1.5737015860467032e-05,
+      "loss": 0.5589,
+      "step": 3064
+    },
+    {
+      "epoch": 0.8499722684414864,
+      "grad_norm": 0.18894660472869873,
+      "learning_rate": 1.573168226002213e-05,
+      "loss": 0.5486,
+      "step": 3065
+    },
+    {
+      "epoch": 0.8502495840266223,
+      "grad_norm": 0.2093021273612976,
+      "learning_rate": 1.5726348028990404e-05,
+      "loss": 0.528,
+      "step": 3066
+    },
+    {
+      "epoch": 0.8505268996117582,
+      "grad_norm": 0.18530364334583282,
+      "learning_rate": 1.5721013168412698e-05,
+      "loss": 0.5244,
+      "step": 3067
+    },
+    {
+      "epoch": 0.8508042151968941,
+      "grad_norm": 0.18845012784004211,
+      "learning_rate": 1.5715677679329978e-05,
+      "loss": 0.5211,
+      "step": 3068
+    },
+    {
+      "epoch": 0.85108153078203,
+      "grad_norm": 0.20350565016269684,
+      "learning_rate": 1.571034156278335e-05,
+      "loss": 0.5711,
+      "step": 3069
+    },
+    {
+      "epoch": 0.8513588463671659,
+      "grad_norm": 0.18500900268554688,
+      "learning_rate": 1.5705004819814025e-05,
+      "loss": 0.5116,
+      "step": 3070
+    },
+    {
+      "epoch": 0.8516361619523017,
+      "grad_norm": 0.18262405693531036,
+      "learning_rate": 1.5699667451463344e-05,
+      "loss": 0.5485,
+      "step": 3071
+    },
+    {
+      "epoch": 0.8519134775374376,
+      "grad_norm": 0.19544097781181335,
+      "learning_rate": 1.5694329458772776e-05,
+      "loss": 0.5465,
+      "step": 3072
+    },
+    {
+      "epoch": 0.8521907931225735,
+      "grad_norm": 0.18556345999240875,
+      "learning_rate": 1.5688990842783892e-05,
+      "loss": 0.5409,
+      "step": 3073
+    },
+    {
+      "epoch": 0.8524681087077094,
+      "grad_norm": 0.23834922909736633,
+      "learning_rate": 1.5683651604538405e-05,
+      "loss": 0.5192,
+      "step": 3074
+    },
+    {
+      "epoch": 0.8527454242928453,
+      "grad_norm": 0.20461677014827728,
+      "learning_rate": 1.5678311745078138e-05,
+      "loss": 0.5674,
+      "step": 3075
+    },
+    {
+      "epoch": 0.8530227398779812,
+      "grad_norm": 0.19039872288703918,
+      "learning_rate": 1.5672971265445046e-05,
+      "loss": 0.5433,
+      "step": 3076
+    },
+    {
+      "epoch": 0.853300055463117,
+      "grad_norm": 0.19499389827251434,
+      "learning_rate": 1.566763016668119e-05,
+      "loss": 0.5577,
+      "step": 3077
+    },
+    {
+      "epoch": 0.8535773710482529,
+      "grad_norm": 0.185109481215477,
+      "learning_rate": 1.5662288449828767e-05,
+      "loss": 0.5192,
+      "step": 3078
+    },
+    {
+      "epoch": 0.8538546866333888,
+      "grad_norm": 0.2005089968442917,
+      "learning_rate": 1.5656946115930084e-05,
+      "loss": 0.5423,
+      "step": 3079
+    },
+    {
+      "epoch": 0.8541320022185247,
+      "grad_norm": 0.19989848136901855,
+      "learning_rate": 1.5651603166027574e-05,
+      "loss": 0.5307,
+      "step": 3080
+    },
+    {
+      "epoch": 0.8544093178036606,
+      "grad_norm": 0.18417179584503174,
+      "learning_rate": 1.5646259601163783e-05,
+      "loss": 0.5479,
+      "step": 3081
+    },
+    {
+      "epoch": 0.8546866333887965,
+      "grad_norm": 0.19350256025791168,
+      "learning_rate": 1.5640915422381387e-05,
+      "loss": 0.547,
+      "step": 3082
+    },
+    {
+      "epoch": 0.8549639489739324,
+      "grad_norm": 0.19599801301956177,
+      "learning_rate": 1.5635570630723173e-05,
+      "loss": 0.572,
+      "step": 3083
+    },
+    {
+      "epoch": 0.8552412645590682,
+      "grad_norm": 0.18424679338932037,
+      "learning_rate": 1.5630225227232055e-05,
+      "loss": 0.5421,
+      "step": 3084
+    },
+    {
+      "epoch": 0.8555185801442041,
+      "grad_norm": 0.18912889063358307,
+      "learning_rate": 1.562487921295106e-05,
+      "loss": 0.5503,
+      "step": 3085
+    },
+    {
+      "epoch": 0.85579589572934,
+      "grad_norm": 0.1971350610256195,
+      "learning_rate": 1.561953258892334e-05,
+      "loss": 0.5493,
+      "step": 3086
+    },
+    {
+      "epoch": 0.8560732113144759,
+      "grad_norm": 0.19339221715927124,
+      "learning_rate": 1.5614185356192156e-05,
+      "loss": 0.5249,
+      "step": 3087
+    },
+    {
+      "epoch": 0.8563505268996118,
+      "grad_norm": 0.22969551384449005,
+      "learning_rate": 1.5608837515800906e-05,
+      "loss": 0.5398,
+      "step": 3088
+    },
+    {
+      "epoch": 0.8566278424847477,
+      "grad_norm": 0.20485931634902954,
+      "learning_rate": 1.560348906879309e-05,
+      "loss": 0.5455,
+      "step": 3089
+    },
+    {
+      "epoch": 0.8569051580698835,
+      "grad_norm": 0.2660558521747589,
+      "learning_rate": 1.5598140016212328e-05,
+      "loss": 0.5636,
+      "step": 3090
+    },
+    {
+      "epoch": 0.8571824736550194,
+      "grad_norm": 0.19878970086574554,
+      "learning_rate": 1.559279035910237e-05,
+      "loss": 0.5513,
+      "step": 3091
+    },
+    {
+      "epoch": 0.8574597892401553,
+      "grad_norm": 0.19739742577075958,
+      "learning_rate": 1.5587440098507067e-05,
+      "loss": 0.5521,
+      "step": 3092
+    },
+    {
+      "epoch": 0.8577371048252912,
+      "grad_norm": 0.19214226305484772,
+      "learning_rate": 1.55820892354704e-05,
+      "loss": 0.5721,
+      "step": 3093
+    },
+    {
+      "epoch": 0.8580144204104271,
+      "grad_norm": 0.19705668091773987,
+      "learning_rate": 1.5576737771036464e-05,
+      "loss": 0.555,
+      "step": 3094
+    },
+    {
+      "epoch": 0.858291735995563,
+      "grad_norm": 0.17740927636623383,
+      "learning_rate": 1.557138570624948e-05,
+      "loss": 0.5478,
+      "step": 3095
+    },
+    {
+      "epoch": 0.8585690515806988,
+      "grad_norm": 0.19050319492816925,
+      "learning_rate": 1.556603304215376e-05,
+      "loss": 0.5511,
+      "step": 3096
+    },
+    {
+      "epoch": 0.8588463671658347,
+      "grad_norm": 0.1907372921705246,
+      "learning_rate": 1.556067977979377e-05,
+      "loss": 0.543,
+      "step": 3097
+    },
+    {
+      "epoch": 0.8591236827509706,
+      "grad_norm": 0.18947333097457886,
+      "learning_rate": 1.5555325920214055e-05,
+      "loss": 0.56,
+      "step": 3098
+    },
+    {
+      "epoch": 0.8594009983361065,
+      "grad_norm": 0.19369062781333923,
+      "learning_rate": 1.5549971464459308e-05,
+      "loss": 0.5506,
+      "step": 3099
+    },
+    {
+      "epoch": 0.8596783139212424,
+      "grad_norm": 0.18818899989128113,
+      "learning_rate": 1.554461641357432e-05,
+      "loss": 0.5434,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8599556295063783,
+      "grad_norm": 0.19132985174655914,
+      "learning_rate": 1.5539260768604e-05,
+      "loss": 0.5378,
+      "step": 3101
+    },
+    {
+      "epoch": 0.8602329450915142,
+      "grad_norm": 0.20558297634124756,
+      "learning_rate": 1.5533904530593386e-05,
+      "loss": 0.5428,
+      "step": 3102
+    },
+    {
+      "epoch": 0.86051026067665,
+      "grad_norm": 0.19788722693920135,
+      "learning_rate": 1.5528547700587616e-05,
+      "loss": 0.5417,
+      "step": 3103
+    },
+    {
+      "epoch": 0.8607875762617859,
+      "grad_norm": 0.20902171730995178,
+      "learning_rate": 1.552319027963195e-05,
+      "loss": 0.5426,
+      "step": 3104
+    },
+    {
+      "epoch": 0.8610648918469218,
+      "grad_norm": 0.18526656925678253,
+      "learning_rate": 1.5517832268771764e-05,
+      "loss": 0.5342,
+      "step": 3105
+    },
+    {
+      "epoch": 0.8613422074320577,
+      "grad_norm": 0.1860983669757843,
+      "learning_rate": 1.551247366905254e-05,
+      "loss": 0.5466,
+      "step": 3106
+    },
+    {
+      "epoch": 0.8616195230171936,
+      "grad_norm": 0.18875092267990112,
+      "learning_rate": 1.5507114481519895e-05,
+      "loss": 0.5417,
+      "step": 3107
+    },
+    {
+      "epoch": 0.8618968386023295,
+      "grad_norm": 0.1913326233625412,
+      "learning_rate": 1.5501754707219536e-05,
+      "loss": 0.5589,
+      "step": 3108
+    },
+    {
+      "epoch": 0.8621741541874653,
+      "grad_norm": 0.19133377075195312,
+      "learning_rate": 1.549639434719731e-05,
+      "loss": 0.5345,
+      "step": 3109
+    },
+    {
+      "epoch": 0.8624514697726012,
+      "grad_norm": 0.18826799094676971,
+      "learning_rate": 1.549103340249916e-05,
+      "loss": 0.527,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8627287853577371,
+      "grad_norm": 0.18173575401306152,
+      "learning_rate": 1.548567187417114e-05,
+      "loss": 0.5431,
+      "step": 3111
+    },
+    {
+      "epoch": 0.863006100942873,
+      "grad_norm": 0.19065049290657043,
+      "learning_rate": 1.548030976325944e-05,
+      "loss": 0.5147,
+      "step": 3112
+    },
+    {
+      "epoch": 0.8632834165280089,
+      "grad_norm": 0.2025771290063858,
+      "learning_rate": 1.547494707081034e-05,
+      "loss": 0.5635,
+      "step": 3113
+    },
+    {
+      "epoch": 0.8635607321131448,
+      "grad_norm": 0.18411415815353394,
+      "learning_rate": 1.546958379787025e-05,
+      "loss": 0.5246,
+      "step": 3114
+    },
+    {
+      "epoch": 0.8638380476982807,
+      "grad_norm": 0.19902820885181427,
+      "learning_rate": 1.546421994548568e-05,
+      "loss": 0.5564,
+      "step": 3115
+    },
+    {
+      "epoch": 0.8641153632834165,
+      "grad_norm": 0.18202657997608185,
+      "learning_rate": 1.5458855514703266e-05,
+      "loss": 0.5364,
+      "step": 3116
+    },
+    {
+      "epoch": 0.8643926788685524,
+      "grad_norm": 0.19231395423412323,
+      "learning_rate": 1.545349050656974e-05,
+      "loss": 0.5521,
+      "step": 3117
+    },
+    {
+      "epoch": 0.8646699944536883,
+      "grad_norm": 0.18620963394641876,
+      "learning_rate": 1.5448124922131974e-05,
+      "loss": 0.5308,
+      "step": 3118
+    },
+    {
+      "epoch": 0.8649473100388242,
+      "grad_norm": 0.18969541788101196,
+      "learning_rate": 1.5442758762436923e-05,
+      "loss": 0.5446,
+      "step": 3119
+    },
+    {
+      "epoch": 0.8652246256239601,
+      "grad_norm": 0.19268915057182312,
+      "learning_rate": 1.543739202853167e-05,
+      "loss": 0.531,
+      "step": 3120
+    },
+    {
+      "epoch": 0.865501941209096,
+      "grad_norm": 0.4672553837299347,
+      "learning_rate": 1.5432024721463413e-05,
+      "loss": 0.5564,
+      "step": 3121
+    },
+    {
+      "epoch": 0.8657792567942318,
+      "grad_norm": 0.19262655079364777,
+      "learning_rate": 1.5426656842279445e-05,
+      "loss": 0.558,
+      "step": 3122
+    },
+    {
+      "epoch": 0.8660565723793677,
+      "grad_norm": 0.20529362559318542,
+      "learning_rate": 1.5421288392027185e-05,
+      "loss": 0.5247,
+      "step": 3123
+    },
+    {
+      "epoch": 0.8663338879645036,
+      "grad_norm": 0.19508616626262665,
+      "learning_rate": 1.5415919371754166e-05,
+      "loss": 0.5307,
+      "step": 3124
+    },
+    {
+      "epoch": 0.8666112035496395,
+      "grad_norm": 0.18688172101974487,
+      "learning_rate": 1.541054978250802e-05,
+      "loss": 0.5422,
+      "step": 3125
+    },
+    {
+      "epoch": 0.8668885191347754,
+      "grad_norm": 0.19961917400360107,
+      "learning_rate": 1.5405179625336495e-05,
+      "loss": 0.5598,
+      "step": 3126
+    },
+    {
+      "epoch": 0.8671658347199113,
+      "grad_norm": 0.18617486953735352,
+      "learning_rate": 1.5399808901287457e-05,
+      "loss": 0.5221,
+      "step": 3127
+    },
+    {
+      "epoch": 0.8674431503050472,
+      "grad_norm": 0.1895580142736435,
+      "learning_rate": 1.5394437611408873e-05,
+      "loss": 0.5515,
+      "step": 3128
+    },
+    {
+      "epoch": 0.867720465890183,
+      "grad_norm": 0.19370147585868835,
+      "learning_rate": 1.5389065756748826e-05,
+      "loss": 0.5312,
+      "step": 3129
+    },
+    {
+      "epoch": 0.8679977814753189,
+      "grad_norm": 0.19625920057296753,
+      "learning_rate": 1.5383693338355504e-05,
+      "loss": 0.526,
+      "step": 3130
+    },
+    {
+      "epoch": 0.8682750970604548,
+      "grad_norm": 0.1878737509250641,
+      "learning_rate": 1.537832035727721e-05,
+      "loss": 0.5229,
+      "step": 3131
+    },
+    {
+      "epoch": 0.8685524126455907,
+      "grad_norm": 0.19030463695526123,
+      "learning_rate": 1.537294681456235e-05,
+      "loss": 0.5197,
+      "step": 3132
+    },
+    {
+      "epoch": 0.8688297282307266,
+      "grad_norm": 0.19420188665390015,
+      "learning_rate": 1.536757271125946e-05,
+      "loss": 0.5369,
+      "step": 3133
+    },
+    {
+      "epoch": 0.8691070438158625,
+      "grad_norm": 0.2041894942522049,
+      "learning_rate": 1.5362198048417147e-05,
+      "loss": 0.5315,
+      "step": 3134
+    },
+    {
+      "epoch": 0.8693843594009983,
+      "grad_norm": 0.1823713183403015,
+      "learning_rate": 1.535682282708417e-05,
+      "loss": 0.5358,
+      "step": 3135
+    },
+    {
+      "epoch": 0.8696616749861342,
+      "grad_norm": 0.18803556263446808,
+      "learning_rate": 1.5351447048309367e-05,
+      "loss": 0.5234,
+      "step": 3136
+    },
+    {
+      "epoch": 0.8699389905712701,
+      "grad_norm": 0.20315398275852203,
+      "learning_rate": 1.5346070713141697e-05,
+      "loss": 0.5523,
+      "step": 3137
+    },
+    {
+      "epoch": 0.870216306156406,
+      "grad_norm": 0.18089507520198822,
+      "learning_rate": 1.5340693822630224e-05,
+      "loss": 0.5425,
+      "step": 3138
+    },
+    {
+      "epoch": 0.8704936217415419,
+      "grad_norm": 0.19417926669120789,
+      "learning_rate": 1.5335316377824127e-05,
+      "loss": 0.5329,
+      "step": 3139
+    },
+    {
+      "epoch": 0.8707709373266778,
+      "grad_norm": 0.19868281483650208,
+      "learning_rate": 1.5329938379772685e-05,
+      "loss": 0.5564,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8710482529118136,
+      "grad_norm": 0.186373770236969,
+      "learning_rate": 1.5324559829525285e-05,
+      "loss": 0.5498,
+      "step": 3141
+    },
+    {
+      "epoch": 0.8713255684969495,
+      "grad_norm": 0.21897022426128387,
+      "learning_rate": 1.531918072813143e-05,
+      "loss": 0.5508,
+      "step": 3142
+    },
+    {
+      "epoch": 0.8716028840820854,
+      "grad_norm": 0.19098646938800812,
+      "learning_rate": 1.5313801076640715e-05,
+      "loss": 0.5481,
+      "step": 3143
+    },
+    {
+      "epoch": 0.8718801996672213,
+      "grad_norm": 0.1954183280467987,
+      "learning_rate": 1.5308420876102863e-05,
+      "loss": 0.5531,
+      "step": 3144
+    },
+    {
+      "epoch": 0.8721575152523572,
+      "grad_norm": 0.18958479166030884,
+      "learning_rate": 1.5303040127567694e-05,
+      "loss": 0.5437,
+      "step": 3145
+    },
+    {
+      "epoch": 0.8724348308374931,
+      "grad_norm": 0.17530956864356995,
+      "learning_rate": 1.5297658832085126e-05,
+      "loss": 0.5216,
+      "step": 3146
+    },
+    {
+      "epoch": 0.872712146422629,
+      "grad_norm": 0.19810372591018677,
+      "learning_rate": 1.5292276990705202e-05,
+      "loss": 0.5623,
+      "step": 3147
+    },
+    {
+      "epoch": 0.8729894620077648,
+      "grad_norm": 0.19809319078922272,
+      "learning_rate": 1.5286894604478054e-05,
+      "loss": 0.5226,
+      "step": 3148
+    },
+    {
+      "epoch": 0.8732667775929007,
+      "grad_norm": 0.20191790163516998,
+      "learning_rate": 1.528151167445393e-05,
+      "loss": 0.5612,
+      "step": 3149
+    },
+    {
+      "epoch": 0.8735440931780366,
+      "grad_norm": 0.18988439440727234,
+      "learning_rate": 1.5276128201683187e-05,
+      "loss": 0.5362,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8738214087631725,
+      "grad_norm": 0.181321382522583,
+      "learning_rate": 1.5270744187216277e-05,
+      "loss": 0.5082,
+      "step": 3151
+    },
+    {
+      "epoch": 0.8740987243483084,
+      "grad_norm": 0.18067540228366852,
+      "learning_rate": 1.526535963210377e-05,
+      "loss": 0.529,
+      "step": 3152
+    },
+    {
+      "epoch": 0.8743760399334443,
+      "grad_norm": 0.19349637627601624,
+      "learning_rate": 1.5259974537396325e-05,
+      "loss": 0.5495,
+      "step": 3153
+    },
+    {
+      "epoch": 0.8746533555185801,
+      "grad_norm": 0.17970281839370728,
+      "learning_rate": 1.5254588904144735e-05,
+      "loss": 0.529,
+      "step": 3154
+    },
+    {
+      "epoch": 0.874930671103716,
+      "grad_norm": 0.18023180961608887,
+      "learning_rate": 1.5249202733399859e-05,
+      "loss": 0.5638,
+      "step": 3155
+    },
+    {
+      "epoch": 0.8752079866888519,
+      "grad_norm": 0.20077262818813324,
+      "learning_rate": 1.5243816026212695e-05,
+      "loss": 0.5346,
+      "step": 3156
+    },
+    {
+      "epoch": 0.8754853022739878,
+      "grad_norm": 0.18828438222408295,
+      "learning_rate": 1.5238428783634326e-05,
+      "loss": 0.5721,
+      "step": 3157
+    },
+    {
+      "epoch": 0.8757626178591237,
+      "grad_norm": 0.19122296571731567,
+      "learning_rate": 1.5233041006715948e-05,
+      "loss": 0.5158,
+      "step": 3158
+    },
+    {
+      "epoch": 0.8760399334442596,
+      "grad_norm": 0.19371193647384644,
+      "learning_rate": 1.5227652696508859e-05,
+      "loss": 0.5758,
+      "step": 3159
+    },
+    {
+      "epoch": 0.8763172490293955,
+      "grad_norm": 0.19302338361740112,
+      "learning_rate": 1.5222263854064465e-05,
+      "loss": 0.5643,
+      "step": 3160
+    },
+    {
+      "epoch": 0.8765945646145313,
+      "grad_norm": 0.28244075179100037,
+      "learning_rate": 1.5216874480434264e-05,
+      "loss": 0.5462,
+      "step": 3161
+    },
+    {
+      "epoch": 0.8768718801996672,
+      "grad_norm": 0.1919064074754715,
+      "learning_rate": 1.521148457666987e-05,
+      "loss": 0.5317,
+      "step": 3162
+    },
+    {
+      "epoch": 0.8771491957848031,
+      "grad_norm": 0.18293742835521698,
+      "learning_rate": 1.5206094143823e-05,
+      "loss": 0.514,
+      "step": 3163
+    },
+    {
+      "epoch": 0.877426511369939,
+      "grad_norm": 0.193914532661438,
+      "learning_rate": 1.520070318294546e-05,
+      "loss": 0.5488,
+      "step": 3164
+    },
+    {
+      "epoch": 0.8777038269550749,
+      "grad_norm": 0.1869155466556549,
+      "learning_rate": 1.5195311695089175e-05,
+      "loss": 0.5591,
+      "step": 3165
+    },
+    {
+      "epoch": 0.8779811425402108,
+      "grad_norm": 0.19056154787540436,
+      "learning_rate": 1.5189919681306173e-05,
+      "loss": 0.5584,
+      "step": 3166
+    },
+    {
+      "epoch": 0.8782584581253466,
+      "grad_norm": 0.1936255842447281,
+      "learning_rate": 1.5184527142648569e-05,
+      "loss": 0.5477,
+      "step": 3167
+    },
+    {
+      "epoch": 0.8785357737104825,
+      "grad_norm": 0.19368582963943481,
+      "learning_rate": 1.5179134080168595e-05,
+      "loss": 0.5568,
+      "step": 3168
+    },
+    {
+      "epoch": 0.8788130892956184,
+      "grad_norm": 0.19679602980613708,
+      "learning_rate": 1.517374049491858e-05,
+      "loss": 0.5549,
+      "step": 3169
+    },
+    {
+      "epoch": 0.8790904048807543,
+      "grad_norm": 0.20116961002349854,
+      "learning_rate": 1.5168346387950955e-05,
+      "loss": 0.5565,
+      "step": 3170
+    },
+    {
+      "epoch": 0.8793677204658902,
+      "grad_norm": 0.1948871910572052,
+      "learning_rate": 1.5162951760318256e-05,
+      "loss": 0.5502,
+      "step": 3171
+    },
+    {
+      "epoch": 0.8796450360510261,
+      "grad_norm": 0.19126254320144653,
+      "learning_rate": 1.515755661307311e-05,
+      "loss": 0.5574,
+      "step": 3172
+    },
+    {
+      "epoch": 0.879922351636162,
+      "grad_norm": 0.18868835270404816,
+      "learning_rate": 1.5152160947268262e-05,
+      "loss": 0.5377,
+      "step": 3173
+    },
+    {
+      "epoch": 0.8801996672212978,
+      "grad_norm": 0.20093926787376404,
+      "learning_rate": 1.5146764763956542e-05,
+      "loss": 0.5486,
+      "step": 3174
+    },
+    {
+      "epoch": 0.8804769828064337,
+      "grad_norm": 0.20768702030181885,
+      "learning_rate": 1.5141368064190897e-05,
+      "loss": 0.5431,
+      "step": 3175
+    },
+    {
+      "epoch": 0.8807542983915696,
+      "grad_norm": 0.19229131937026978,
+      "learning_rate": 1.5135970849024356e-05,
+      "loss": 0.5298,
+      "step": 3176
+    },
+    {
+      "epoch": 0.8810316139767055,
+      "grad_norm": 0.18817956745624542,
+      "learning_rate": 1.5130573119510064e-05,
+      "loss": 0.5414,
+      "step": 3177
+    },
+    {
+      "epoch": 0.8813089295618414,
+      "grad_norm": 0.19554218649864197,
+      "learning_rate": 1.5125174876701262e-05,
+      "loss": 0.5381,
+      "step": 3178
+    },
+    {
+      "epoch": 0.8815862451469773,
+      "grad_norm": 0.1836453080177307,
+      "learning_rate": 1.5119776121651288e-05,
+      "loss": 0.508,
+      "step": 3179
+    },
+    {
+      "epoch": 0.8818635607321131,
+      "grad_norm": 0.2512352764606476,
+      "learning_rate": 1.5114376855413586e-05,
+      "loss": 0.5146,
+      "step": 3180
+    },
+    {
+      "epoch": 0.882140876317249,
+      "grad_norm": 0.19116099178791046,
+      "learning_rate": 1.5108977079041692e-05,
+      "loss": 0.5332,
+      "step": 3181
+    },
+    {
+      "epoch": 0.8824181919023849,
+      "grad_norm": 0.19376236200332642,
+      "learning_rate": 1.5103576793589244e-05,
+      "loss": 0.5188,
+      "step": 3182
+    },
+    {
+      "epoch": 0.8826955074875208,
+      "grad_norm": 0.2322767972946167,
+      "learning_rate": 1.5098176000109984e-05,
+      "loss": 0.5468,
+      "step": 3183
+    },
+    {
+      "epoch": 0.8829728230726567,
+      "grad_norm": 0.19449107348918915,
+      "learning_rate": 1.5092774699657747e-05,
+      "loss": 0.5223,
+      "step": 3184
+    },
+    {
+      "epoch": 0.8832501386577926,
+      "grad_norm": 0.19483071565628052,
+      "learning_rate": 1.5087372893286475e-05,
+      "loss": 0.5526,
+      "step": 3185
+    },
+    {
+      "epoch": 0.8835274542429284,
+      "grad_norm": 0.18979991972446442,
+      "learning_rate": 1.5081970582050201e-05,
+      "loss": 0.5145,
+      "step": 3186
+    },
+    {
+      "epoch": 0.8838047698280643,
+      "grad_norm": 0.18646521866321564,
+      "learning_rate": 1.5076567767003056e-05,
+      "loss": 0.5308,
+      "step": 3187
+    },
+    {
+      "epoch": 0.8840820854132002,
+      "grad_norm": 0.1952681541442871,
+      "learning_rate": 1.5071164449199277e-05,
+      "loss": 0.5384,
+      "step": 3188
+    },
+    {
+      "epoch": 0.8843594009983361,
+      "grad_norm": 0.1863500475883484,
+      "learning_rate": 1.506576062969319e-05,
+      "loss": 0.5721,
+      "step": 3189
+    },
+    {
+      "epoch": 0.884636716583472,
+      "grad_norm": 0.19182690978050232,
+      "learning_rate": 1.5060356309539226e-05,
+      "loss": 0.5171,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8849140321686079,
+      "grad_norm": 0.18404962122440338,
+      "learning_rate": 1.5054951489791908e-05,
+      "loss": 0.5362,
+      "step": 3191
+    },
+    {
+      "epoch": 0.8851913477537438,
+      "grad_norm": 0.19122952222824097,
+      "learning_rate": 1.5049546171505869e-05,
+      "loss": 0.5424,
+      "step": 3192
+    },
+    {
+      "epoch": 0.8854686633388796,
+      "grad_norm": 0.1863165944814682,
+      "learning_rate": 1.5044140355735816e-05,
+      "loss": 0.53,
+      "step": 3193
+    },
+    {
+      "epoch": 0.8857459789240155,
+      "grad_norm": 0.19876129925251007,
+      "learning_rate": 1.5038734043536582e-05,
+      "loss": 0.5484,
+      "step": 3194
+    },
+    {
+      "epoch": 0.8860232945091514,
+      "grad_norm": 0.18133728206157684,
+      "learning_rate": 1.5033327235963065e-05,
+      "loss": 0.5036,
+      "step": 3195
+    },
+    {
+      "epoch": 0.8863006100942873,
+      "grad_norm": 0.19382363557815552,
+      "learning_rate": 1.5027919934070291e-05,
+      "loss": 0.5558,
+      "step": 3196
+    },
+    {
+      "epoch": 0.8865779256794232,
+      "grad_norm": 0.2071426957845688,
+      "learning_rate": 1.5022512138913358e-05,
+      "loss": 0.539,
+      "step": 3197
+    },
+    {
+      "epoch": 0.8868552412645591,
+      "grad_norm": 0.19322386384010315,
+      "learning_rate": 1.5017103851547476e-05,
+      "loss": 0.551,
+      "step": 3198
+    },
+    {
+      "epoch": 0.8871325568496949,
+      "grad_norm": 0.18517249822616577,
+      "learning_rate": 1.5011695073027942e-05,
+      "loss": 0.5171,
+      "step": 3199
+    },
+    {
+      "epoch": 0.8874098724348308,
+      "grad_norm": 0.18865883350372314,
+      "learning_rate": 1.5006285804410156e-05,
+      "loss": 0.5206,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8876871880199667,
+      "grad_norm": 0.19007782638072968,
+      "learning_rate": 1.5000876046749603e-05,
+      "loss": 0.5194,
+      "step": 3201
+    },
+    {
+      "epoch": 0.8879645036051026,
+      "grad_norm": 0.18567977845668793,
+      "learning_rate": 1.4995465801101877e-05,
+      "loss": 0.535,
+      "step": 3202
+    },
+    {
+      "epoch": 0.8882418191902385,
+      "grad_norm": 0.19124850630760193,
+      "learning_rate": 1.4990055068522654e-05,
+      "loss": 0.5304,
+      "step": 3203
+    },
+    {
+      "epoch": 0.8885191347753744,
+      "grad_norm": 0.19194400310516357,
+      "learning_rate": 1.4984643850067717e-05,
+      "loss": 0.549,
+      "step": 3204
+    },
+    {
+      "epoch": 0.8887964503605102,
+      "grad_norm": 0.18528962135314941,
+      "learning_rate": 1.4979232146792936e-05,
+      "loss": 0.5053,
+      "step": 3205
+    },
+    {
+      "epoch": 0.8890737659456461,
+      "grad_norm": 0.1908206045627594,
+      "learning_rate": 1.4973819959754273e-05,
+      "loss": 0.5427,
+      "step": 3206
+    },
+    {
+      "epoch": 0.889351081530782,
+      "grad_norm": 0.18456414341926575,
+      "learning_rate": 1.4968407290007796e-05,
+      "loss": 0.567,
+      "step": 3207
+    },
+    {
+      "epoch": 0.8896283971159179,
+      "grad_norm": 0.19162026047706604,
+      "learning_rate": 1.496299413860966e-05,
+      "loss": 0.5609,
+      "step": 3208
+    },
+    {
+      "epoch": 0.8899057127010538,
+      "grad_norm": 0.18303340673446655,
+      "learning_rate": 1.4957580506616109e-05,
+      "loss": 0.5281,
+      "step": 3209
+    },
+    {
+      "epoch": 0.8901830282861897,
+      "grad_norm": 0.19368955492973328,
+      "learning_rate": 1.4952166395083486e-05,
+      "loss": 0.5348,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8904603438713256,
+      "grad_norm": 0.1844678521156311,
+      "learning_rate": 1.4946751805068238e-05,
+      "loss": 0.542,
+      "step": 3211
+    },
+    {
+      "epoch": 0.8907376594564614,
+      "grad_norm": 0.18950358033180237,
+      "learning_rate": 1.4941336737626879e-05,
+      "loss": 0.5558,
+      "step": 3212
+    },
+    {
+      "epoch": 0.8910149750415973,
+      "grad_norm": 0.19502434134483337,
+      "learning_rate": 1.4935921193816046e-05,
+      "loss": 0.5109,
+      "step": 3213
+    },
+    {
+      "epoch": 0.8912922906267332,
+      "grad_norm": 0.18396303057670593,
+      "learning_rate": 1.4930505174692447e-05,
+      "loss": 0.5152,
+      "step": 3214
+    },
+    {
+      "epoch": 0.8915696062118691,
+      "grad_norm": 0.20252270996570587,
+      "learning_rate": 1.4925088681312895e-05,
+      "loss": 0.5219,
+      "step": 3215
+    },
+    {
+      "epoch": 0.891846921797005,
+      "grad_norm": 0.18550090491771698,
+      "learning_rate": 1.4919671714734288e-05,
+      "loss": 0.5321,
+      "step": 3216
+    },
+    {
+      "epoch": 0.8921242373821409,
+      "grad_norm": 0.19548344612121582,
+      "learning_rate": 1.4914254276013622e-05,
+      "loss": 0.559,
+      "step": 3217
+    },
+    {
+      "epoch": 0.8924015529672767,
+      "grad_norm": 0.1747296005487442,
+      "learning_rate": 1.4908836366207985e-05,
+      "loss": 0.5218,
+      "step": 3218
+    },
+    {
+      "epoch": 0.8926788685524126,
+      "grad_norm": 0.18561501801013947,
+      "learning_rate": 1.4903417986374548e-05,
+      "loss": 0.5615,
+      "step": 3219
+    },
+    {
+      "epoch": 0.8929561841375485,
+      "grad_norm": 0.19387111067771912,
+      "learning_rate": 1.4897999137570586e-05,
+      "loss": 0.5437,
+      "step": 3220
+    },
+    {
+      "epoch": 0.8932334997226844,
+      "grad_norm": 0.19557684659957886,
+      "learning_rate": 1.4892579820853459e-05,
+      "loss": 0.5387,
+      "step": 3221
+    },
+    {
+      "epoch": 0.8935108153078203,
+      "grad_norm": 0.1842055469751358,
+      "learning_rate": 1.488716003728062e-05,
+      "loss": 0.5277,
+      "step": 3222
+    },
+    {
+      "epoch": 0.8937881308929562,
+      "grad_norm": 0.1835155189037323,
+      "learning_rate": 1.4881739787909607e-05,
+      "loss": 0.5342,
+      "step": 3223
+    },
+    {
+      "epoch": 0.894065446478092,
+      "grad_norm": 0.1914055347442627,
+      "learning_rate": 1.4876319073798061e-05,
+      "loss": 0.548,
+      "step": 3224
+    },
+    {
+      "epoch": 0.8943427620632279,
+      "grad_norm": 0.18737882375717163,
+      "learning_rate": 1.4870897896003705e-05,
+      "loss": 0.5281,
+      "step": 3225
+    },
+    {
+      "epoch": 0.8946200776483638,
+      "grad_norm": 0.20303812623023987,
+      "learning_rate": 1.4865476255584351e-05,
+      "loss": 0.5603,
+      "step": 3226
+    },
+    {
+      "epoch": 0.8948973932334997,
+      "grad_norm": 0.20020204782485962,
+      "learning_rate": 1.486005415359791e-05,
+      "loss": 0.5434,
+      "step": 3227
+    },
+    {
+      "epoch": 0.8951747088186356,
+      "grad_norm": 0.1958005726337433,
+      "learning_rate": 1.4854631591102374e-05,
+      "loss": 0.522,
+      "step": 3228
+    },
+    {
+      "epoch": 0.8954520244037715,
+      "grad_norm": 0.23392035067081451,
+      "learning_rate": 1.4849208569155829e-05,
+      "loss": 0.5574,
+      "step": 3229
+    },
+    {
+      "epoch": 0.8957293399889074,
+      "grad_norm": 0.1955118626356125,
+      "learning_rate": 1.4843785088816455e-05,
+      "loss": 0.5327,
+      "step": 3230
+    },
+    {
+      "epoch": 0.8960066555740432,
+      "grad_norm": 0.18367242813110352,
+      "learning_rate": 1.4838361151142511e-05,
+      "loss": 0.5166,
+      "step": 3231
+    },
+    {
+      "epoch": 0.8962839711591791,
+      "grad_norm": 0.19902175664901733,
+      "learning_rate": 1.4832936757192354e-05,
+      "loss": 0.5283,
+      "step": 3232
+    },
+    {
+      "epoch": 0.896561286744315,
+      "grad_norm": 0.19180850684642792,
+      "learning_rate": 1.4827511908024419e-05,
+      "loss": 0.5277,
+      "step": 3233
+    },
+    {
+      "epoch": 0.8968386023294509,
+      "grad_norm": 0.18553797900676727,
+      "learning_rate": 1.4822086604697253e-05,
+      "loss": 0.5325,
+      "step": 3234
+    },
+    {
+      "epoch": 0.8971159179145868,
+      "grad_norm": 0.1895224153995514,
+      "learning_rate": 1.4816660848269462e-05,
+      "loss": 0.552,
+      "step": 3235
+    },
+    {
+      "epoch": 0.8973932334997227,
+      "grad_norm": 0.18854071199893951,
+      "learning_rate": 1.4811234639799761e-05,
+      "loss": 0.5471,
+      "step": 3236
+    },
+    {
+      "epoch": 0.8976705490848585,
+      "grad_norm": 0.17953041195869446,
+      "learning_rate": 1.480580798034695e-05,
+      "loss": 0.5556,
+      "step": 3237
+    },
+    {
+      "epoch": 0.8979478646699944,
+      "grad_norm": 0.1898965984582901,
+      "learning_rate": 1.480038087096991e-05,
+      "loss": 0.5282,
+      "step": 3238
+    },
+    {
+      "epoch": 0.8982251802551303,
+      "grad_norm": 0.19018866121768951,
+      "learning_rate": 1.4794953312727613e-05,
+      "loss": 0.5301,
+      "step": 3239
+    },
+    {
+      "epoch": 0.8985024958402662,
+      "grad_norm": 0.1814391165971756,
+      "learning_rate": 1.4789525306679122e-05,
+      "loss": 0.5526,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8987798114254021,
+      "grad_norm": 0.19814236462116241,
+      "learning_rate": 1.4784096853883586e-05,
+      "loss": 0.5741,
+      "step": 3241
+    },
+    {
+      "epoch": 0.899057127010538,
+      "grad_norm": 0.19339510798454285,
+      "learning_rate": 1.4778667955400233e-05,
+      "loss": 0.5375,
+      "step": 3242
+    },
+    {
+      "epoch": 0.8993344425956739,
+      "grad_norm": 0.29308080673217773,
+      "learning_rate": 1.4773238612288393e-05,
+      "loss": 0.5239,
+      "step": 3243
+    },
+    {
+      "epoch": 0.8996117581808097,
+      "grad_norm": 0.1853429526090622,
+      "learning_rate": 1.476780882560747e-05,
+      "loss": 0.5436,
+      "step": 3244
+    },
+    {
+      "epoch": 0.8998890737659456,
+      "grad_norm": 0.18454696238040924,
+      "learning_rate": 1.4762378596416961e-05,
+      "loss": 0.5032,
+      "step": 3245
+    },
+    {
+      "epoch": 0.9001663893510815,
+      "grad_norm": 0.19088363647460938,
+      "learning_rate": 1.4756947925776448e-05,
+      "loss": 0.5369,
+      "step": 3246
+    },
+    {
+      "epoch": 0.9004437049362174,
+      "grad_norm": 0.18263565003871918,
+      "learning_rate": 1.4751516814745598e-05,
+      "loss": 0.5313,
+      "step": 3247
+    },
+    {
+      "epoch": 0.9007210205213533,
+      "grad_norm": 0.1870230883359909,
+      "learning_rate": 1.4746085264384165e-05,
+      "loss": 0.551,
+      "step": 3248
+    },
+    {
+      "epoch": 0.9009983361064892,
+      "grad_norm": 0.18567577004432678,
+      "learning_rate": 1.4740653275751987e-05,
+      "loss": 0.5585,
+      "step": 3249
+    },
+    {
+      "epoch": 0.901275651691625,
+      "grad_norm": 0.20058301091194153,
+      "learning_rate": 1.4735220849908987e-05,
+      "loss": 0.5031,
+      "step": 3250
+    },
+    {
+      "epoch": 0.9015529672767609,
+      "grad_norm": 0.18694059550762177,
+      "learning_rate": 1.4729787987915186e-05,
+      "loss": 0.5334,
+      "step": 3251
+    },
+    {
+      "epoch": 0.9018302828618968,
+      "grad_norm": 0.18202227354049683,
+      "learning_rate": 1.4724354690830663e-05,
+      "loss": 0.5553,
+      "step": 3252
+    },
+    {
+      "epoch": 0.9021075984470327,
+      "grad_norm": 0.19465142488479614,
+      "learning_rate": 1.4718920959715616e-05,
+      "loss": 0.5115,
+      "step": 3253
+    },
+    {
+      "epoch": 0.9023849140321686,
+      "grad_norm": 0.18764632940292358,
+      "learning_rate": 1.4713486795630291e-05,
+      "loss": 0.5546,
+      "step": 3254
+    },
+    {
+      "epoch": 0.9026622296173045,
+      "grad_norm": 0.2006525844335556,
+      "learning_rate": 1.4708052199635053e-05,
+      "loss": 0.5239,
+      "step": 3255
+    },
+    {
+      "epoch": 0.9029395452024404,
+      "grad_norm": 0.18893550336360931,
+      "learning_rate": 1.4702617172790325e-05,
+      "loss": 0.5246,
+      "step": 3256
+    },
+    {
+      "epoch": 0.9032168607875762,
+      "grad_norm": 0.2028273344039917,
+      "learning_rate": 1.4697181716156633e-05,
+      "loss": 0.5548,
+      "step": 3257
+    },
+    {
+      "epoch": 0.9034941763727121,
+      "grad_norm": 0.18957914412021637,
+      "learning_rate": 1.4691745830794574e-05,
+      "loss": 0.5261,
+      "step": 3258
+    },
+    {
+      "epoch": 0.903771491957848,
+      "grad_norm": 0.19126209616661072,
+      "learning_rate": 1.4686309517764835e-05,
+      "loss": 0.5479,
+      "step": 3259
+    },
+    {
+      "epoch": 0.9040488075429839,
+      "grad_norm": 0.1924603134393692,
+      "learning_rate": 1.4680872778128183e-05,
+      "loss": 0.5456,
+      "step": 3260
+    },
+    {
+      "epoch": 0.9043261231281198,
+      "grad_norm": 0.1945699006319046,
+      "learning_rate": 1.4675435612945468e-05,
+      "loss": 0.5437,
+      "step": 3261
+    },
+    {
+      "epoch": 0.9046034387132557,
+      "grad_norm": 0.18470460176467896,
+      "learning_rate": 1.466999802327763e-05,
+      "loss": 0.5342,
+      "step": 3262
+    },
+    {
+      "epoch": 0.9048807542983915,
+      "grad_norm": 0.18286101520061493,
+      "learning_rate": 1.4664560010185685e-05,
+      "loss": 0.5007,
+      "step": 3263
+    },
+    {
+      "epoch": 0.9051580698835274,
+      "grad_norm": 0.20376376807689667,
+      "learning_rate": 1.4659121574730736e-05,
+      "loss": 0.5517,
+      "step": 3264
+    },
+    {
+      "epoch": 0.9054353854686633,
+      "grad_norm": 0.18289178609848022,
+      "learning_rate": 1.465368271797396e-05,
+      "loss": 0.5254,
+      "step": 3265
+    },
+    {
+      "epoch": 0.9057127010537992,
+      "grad_norm": 0.18297207355499268,
+      "learning_rate": 1.4648243440976625e-05,
+      "loss": 0.5287,
+      "step": 3266
+    },
+    {
+      "epoch": 0.9059900166389351,
+      "grad_norm": 0.19243334233760834,
+      "learning_rate": 1.464280374480008e-05,
+      "loss": 0.5408,
+      "step": 3267
+    },
+    {
+      "epoch": 0.906267332224071,
+      "grad_norm": 0.19214653968811035,
+      "learning_rate": 1.463736363050575e-05,
+      "loss": 0.5455,
+      "step": 3268
+    },
+    {
+      "epoch": 0.9065446478092068,
+      "grad_norm": 0.1994084268808365,
+      "learning_rate": 1.4631923099155143e-05,
+      "loss": 0.546,
+      "step": 3269
+    },
+    {
+      "epoch": 0.9068219633943427,
+      "grad_norm": 0.19682733714580536,
+      "learning_rate": 1.4626482151809865e-05,
+      "loss": 0.5031,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9070992789794786,
+      "grad_norm": 0.3178056478500366,
+      "learning_rate": 1.462104078953157e-05,
+      "loss": 0.5183,
+      "step": 3271
+    },
+    {
+      "epoch": 0.9073765945646145,
+      "grad_norm": 0.1968078464269638,
+      "learning_rate": 1.4615599013382028e-05,
+      "loss": 0.5475,
+      "step": 3272
+    },
+    {
+      "epoch": 0.9076539101497504,
+      "grad_norm": 0.18579484522342682,
+      "learning_rate": 1.461015682442306e-05,
+      "loss": 0.5296,
+      "step": 3273
+    },
+    {
+      "epoch": 0.9079312257348863,
+      "grad_norm": 0.1962941437959671,
+      "learning_rate": 1.4604714223716595e-05,
+      "loss": 0.5565,
+      "step": 3274
+    },
+    {
+      "epoch": 0.9082085413200222,
+      "grad_norm": 0.19378679990768433,
+      "learning_rate": 1.4599271212324617e-05,
+      "loss": 0.5506,
+      "step": 3275
+    },
+    {
+      "epoch": 0.908485856905158,
+      "grad_norm": 0.22947950661182404,
+      "learning_rate": 1.4593827791309206e-05,
+      "loss": 0.5625,
+      "step": 3276
+    },
+    {
+      "epoch": 0.9087631724902939,
+      "grad_norm": 0.1949935257434845,
+      "learning_rate": 1.458838396173252e-05,
+      "loss": 0.5327,
+      "step": 3277
+    },
+    {
+      "epoch": 0.9090404880754298,
+      "grad_norm": 0.19015999138355255,
+      "learning_rate": 1.458293972465679e-05,
+      "loss": 0.5168,
+      "step": 3278
+    },
+    {
+      "epoch": 0.9093178036605657,
+      "grad_norm": 0.19812412559986115,
+      "learning_rate": 1.4577495081144337e-05,
+      "loss": 0.5389,
+      "step": 3279
+    },
+    {
+      "epoch": 0.9095951192457016,
+      "grad_norm": 0.19743192195892334,
+      "learning_rate": 1.4572050032257548e-05,
+      "loss": 0.5492,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9098724348308375,
+      "grad_norm": 0.17543548345565796,
+      "learning_rate": 1.4566604579058904e-05,
+      "loss": 0.5064,
+      "step": 3281
+    },
+    {
+      "epoch": 0.9101497504159733,
+      "grad_norm": 0.20693279802799225,
+      "learning_rate": 1.4561158722610948e-05,
+      "loss": 0.5312,
+      "step": 3282
+    },
+    {
+      "epoch": 0.9104270660011092,
+      "grad_norm": 0.19201938807964325,
+      "learning_rate": 1.4555712463976318e-05,
+      "loss": 0.518,
+      "step": 3283
+    },
+    {
+      "epoch": 0.9107043815862451,
+      "grad_norm": 0.1977842152118683,
+      "learning_rate": 1.4550265804217722e-05,
+      "loss": 0.5263,
+      "step": 3284
+    },
+    {
+      "epoch": 0.910981697171381,
+      "grad_norm": 0.18007364869117737,
+      "learning_rate": 1.4544818744397947e-05,
+      "loss": 0.521,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9112590127565169,
+      "grad_norm": 0.19621910154819489,
+      "learning_rate": 1.453937128557986e-05,
+      "loss": 0.5701,
+      "step": 3286
+    },
+    {
+      "epoch": 0.9115363283416528,
+      "grad_norm": 0.1956057995557785,
+      "learning_rate": 1.4533923428826399e-05,
+      "loss": 0.5542,
+      "step": 3287
+    },
+    {
+      "epoch": 0.9118136439267887,
+      "grad_norm": 0.20553717017173767,
+      "learning_rate": 1.452847517520059e-05,
+      "loss": 0.5379,
+      "step": 3288
+    },
+    {
+      "epoch": 0.9120909595119245,
+      "grad_norm": 0.19297057390213013,
+      "learning_rate": 1.4523026525765532e-05,
+      "loss": 0.5562,
+      "step": 3289
+    },
+    {
+      "epoch": 0.9123682750970604,
+      "grad_norm": 0.19896887242794037,
+      "learning_rate": 1.4517577481584399e-05,
+      "loss": 0.5502,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9126455906821963,
+      "grad_norm": 0.18490025401115417,
+      "learning_rate": 1.4512128043720447e-05,
+      "loss": 0.5426,
+      "step": 3291
+    },
+    {
+      "epoch": 0.9129229062673322,
+      "grad_norm": 0.19716937839984894,
+      "learning_rate": 1.4506678213236998e-05,
+      "loss": 0.5433,
+      "step": 3292
+    },
+    {
+      "epoch": 0.9132002218524681,
+      "grad_norm": 0.19106687605381012,
+      "learning_rate": 1.4501227991197472e-05,
+      "loss": 0.5437,
+      "step": 3293
+    },
+    {
+      "epoch": 0.913477537437604,
+      "grad_norm": 0.18353629112243652,
+      "learning_rate": 1.4495777378665337e-05,
+      "loss": 0.5374,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9137548530227398,
+      "grad_norm": 0.19178996980190277,
+      "learning_rate": 1.4490326376704161e-05,
+      "loss": 0.5471,
+      "step": 3295
+    },
+    {
+      "epoch": 0.9140321686078757,
+      "grad_norm": 0.18615961074829102,
+      "learning_rate": 1.4484874986377573e-05,
+      "loss": 0.5503,
+      "step": 3296
+    },
+    {
+      "epoch": 0.9143094841930116,
+      "grad_norm": 0.19387958943843842,
+      "learning_rate": 1.447942320874929e-05,
+      "loss": 0.5174,
+      "step": 3297
+    },
+    {
+      "epoch": 0.9145867997781475,
+      "grad_norm": 0.18303687870502472,
+      "learning_rate": 1.4473971044883095e-05,
+      "loss": 0.5335,
+      "step": 3298
+    },
+    {
+      "epoch": 0.9148641153632834,
+      "grad_norm": 0.9202612638473511,
+      "learning_rate": 1.4468518495842848e-05,
+      "loss": 0.547,
+      "step": 3299
+    },
+    {
+      "epoch": 0.9151414309484193,
+      "grad_norm": 0.1895207315683365,
+      "learning_rate": 1.446306556269249e-05,
+      "loss": 0.5561,
+      "step": 3300
+    },
+    {
+      "epoch": 0.9154187465335551,
+      "grad_norm": 0.18821243941783905,
+      "learning_rate": 1.4457612246496027e-05,
+      "loss": 0.5318,
+      "step": 3301
+    },
+    {
+      "epoch": 0.915696062118691,
+      "grad_norm": 0.1886141002178192,
+      "learning_rate": 1.4452158548317551e-05,
+      "loss": 0.5517,
+      "step": 3302
+    },
+    {
+      "epoch": 0.9159733777038269,
+      "grad_norm": 0.1940373033285141,
+      "learning_rate": 1.444670446922122e-05,
+      "loss": 0.5617,
+      "step": 3303
+    },
+    {
+      "epoch": 0.9162506932889628,
+      "grad_norm": 0.18342509865760803,
+      "learning_rate": 1.444125001027127e-05,
+      "loss": 0.5438,
+      "step": 3304
+    },
+    {
+      "epoch": 0.9165280088740987,
+      "grad_norm": 0.19425570964813232,
+      "learning_rate": 1.4435795172532014e-05,
+      "loss": 0.5413,
+      "step": 3305
+    },
+    {
+      "epoch": 0.9168053244592346,
+      "grad_norm": 0.19505858421325684,
+      "learning_rate": 1.4430339957067826e-05,
+      "loss": 0.5666,
+      "step": 3306
+    },
+    {
+      "epoch": 0.9170826400443705,
+      "grad_norm": 0.19822010397911072,
+      "learning_rate": 1.4424884364943172e-05,
+      "loss": 0.521,
+      "step": 3307
+    },
+    {
+      "epoch": 0.9173599556295063,
+      "grad_norm": 0.1916007101535797,
+      "learning_rate": 1.4419428397222582e-05,
+      "loss": 0.5649,
+      "step": 3308
+    },
+    {
+      "epoch": 0.9176372712146422,
+      "grad_norm": 0.19743779301643372,
+      "learning_rate": 1.441397205497065e-05,
+      "loss": 0.5337,
+      "step": 3309
+    },
+    {
+      "epoch": 0.9179145867997781,
+      "grad_norm": 0.19250613451004028,
+      "learning_rate": 1.4408515339252068e-05,
+      "loss": 0.5477,
+      "step": 3310
+    },
+    {
+      "epoch": 0.918191902384914,
+      "grad_norm": 0.18895046412944794,
+      "learning_rate": 1.4403058251131574e-05,
+      "loss": 0.5782,
+      "step": 3311
+    },
+    {
+      "epoch": 0.9184692179700499,
+      "grad_norm": 0.1895778477191925,
+      "learning_rate": 1.4397600791673999e-05,
+      "loss": 0.5344,
+      "step": 3312
+    },
+    {
+      "epoch": 0.9187465335551858,
+      "grad_norm": 0.1935378462076187,
+      "learning_rate": 1.4392142961944228e-05,
+      "loss": 0.5568,
+      "step": 3313
+    },
+    {
+      "epoch": 0.9190238491403216,
+      "grad_norm": 0.18773426115512848,
+      "learning_rate": 1.4386684763007235e-05,
+      "loss": 0.546,
+      "step": 3314
+    },
+    {
+      "epoch": 0.9193011647254575,
+      "grad_norm": 0.18107493221759796,
+      "learning_rate": 1.438122619592806e-05,
+      "loss": 0.5007,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9195784803105934,
+      "grad_norm": 0.19908292591571808,
+      "learning_rate": 1.4375767261771814e-05,
+      "loss": 0.545,
+      "step": 3316
+    },
+    {
+      "epoch": 0.9198557958957293,
+      "grad_norm": 0.19010509550571442,
+      "learning_rate": 1.4370307961603673e-05,
+      "loss": 0.5593,
+      "step": 3317
+    },
+    {
+      "epoch": 0.9201331114808652,
+      "grad_norm": 0.19176799058914185,
+      "learning_rate": 1.4364848296488897e-05,
+      "loss": 0.5676,
+      "step": 3318
+    },
+    {
+      "epoch": 0.9204104270660011,
+      "grad_norm": 0.1841881275177002,
+      "learning_rate": 1.4359388267492812e-05,
+      "loss": 0.5361,
+      "step": 3319
+    },
+    {
+      "epoch": 0.920687742651137,
+      "grad_norm": 0.19731801748275757,
+      "learning_rate": 1.4353927875680808e-05,
+      "loss": 0.5213,
+      "step": 3320
+    },
+    {
+      "epoch": 0.9209650582362728,
+      "grad_norm": 0.18500731885433197,
+      "learning_rate": 1.4348467122118364e-05,
+      "loss": 0.5417,
+      "step": 3321
+    },
+    {
+      "epoch": 0.9212423738214087,
+      "grad_norm": 0.2047429382801056,
+      "learning_rate": 1.4343006007871004e-05,
+      "loss": 0.5333,
+      "step": 3322
+    },
+    {
+      "epoch": 0.9215196894065446,
+      "grad_norm": 0.2538173198699951,
+      "learning_rate": 1.4337544534004346e-05,
+      "loss": 0.5529,
+      "step": 3323
+    },
+    {
+      "epoch": 0.9217970049916805,
+      "grad_norm": 0.1965561956167221,
+      "learning_rate": 1.4332082701584063e-05,
+      "loss": 0.5432,
+      "step": 3324
+    },
+    {
+      "epoch": 0.9220743205768164,
+      "grad_norm": 0.1974562108516693,
+      "learning_rate": 1.4326620511675906e-05,
+      "loss": 0.5159,
+      "step": 3325
+    },
+    {
+      "epoch": 0.9223516361619523,
+      "grad_norm": 0.19231897592544556,
+      "learning_rate": 1.4321157965345688e-05,
+      "loss": 0.5674,
+      "step": 3326
+    },
+    {
+      "epoch": 0.9226289517470881,
+      "grad_norm": 0.19817808270454407,
+      "learning_rate": 1.4315695063659304e-05,
+      "loss": 0.5496,
+      "step": 3327
+    },
+    {
+      "epoch": 0.922906267332224,
+      "grad_norm": 0.1902833878993988,
+      "learning_rate": 1.4310231807682706e-05,
+      "loss": 0.5342,
+      "step": 3328
+    },
+    {
+      "epoch": 0.9231835829173599,
+      "grad_norm": 0.19699136912822723,
+      "learning_rate": 1.4304768198481923e-05,
+      "loss": 0.5417,
+      "step": 3329
+    },
+    {
+      "epoch": 0.9234608985024958,
+      "grad_norm": 0.19054579734802246,
+      "learning_rate": 1.4299304237123043e-05,
+      "loss": 0.5556,
+      "step": 3330
+    },
+    {
+      "epoch": 0.9237382140876317,
+      "grad_norm": 0.191939577460289,
+      "learning_rate": 1.4293839924672242e-05,
+      "loss": 0.5304,
+      "step": 3331
+    },
+    {
+      "epoch": 0.9240155296727676,
+      "grad_norm": 0.22973594069480896,
+      "learning_rate": 1.4288375262195739e-05,
+      "loss": 0.5337,
+      "step": 3332
+    },
+    {
+      "epoch": 0.9242928452579035,
+      "grad_norm": 0.1892794668674469,
+      "learning_rate": 1.4282910250759843e-05,
+      "loss": 0.5527,
+      "step": 3333
+    },
+    {
+      "epoch": 0.9245701608430393,
+      "grad_norm": 0.19066397845745087,
+      "learning_rate": 1.427744489143092e-05,
+      "loss": 0.5446,
+      "step": 3334
+    },
+    {
+      "epoch": 0.9248474764281752,
+      "grad_norm": 0.18730609118938446,
+      "learning_rate": 1.4271979185275402e-05,
+      "loss": 0.5425,
+      "step": 3335
+    },
+    {
+      "epoch": 0.9251247920133111,
+      "grad_norm": 0.18019746243953705,
+      "learning_rate": 1.4266513133359802e-05,
+      "loss": 0.5263,
+      "step": 3336
+    },
+    {
+      "epoch": 0.925402107598447,
+      "grad_norm": 0.19604218006134033,
+      "learning_rate": 1.4261046736750686e-05,
+      "loss": 0.5552,
+      "step": 3337
+    },
+    {
+      "epoch": 0.9256794231835829,
+      "grad_norm": 0.1902369260787964,
+      "learning_rate": 1.4255579996514693e-05,
+      "loss": 0.5531,
+      "step": 3338
+    },
+    {
+      "epoch": 0.9259567387687188,
+      "grad_norm": 0.19522936642169952,
+      "learning_rate": 1.4250112913718525e-05,
+      "loss": 0.545,
+      "step": 3339
+    },
+    {
+      "epoch": 0.9262340543538546,
+      "grad_norm": 0.21855325996875763,
+      "learning_rate": 1.4244645489428968e-05,
+      "loss": 0.5133,
+      "step": 3340
+    },
+    {
+      "epoch": 0.9265113699389905,
+      "grad_norm": 0.20022931694984436,
+      "learning_rate": 1.4239177724712843e-05,
+      "loss": 0.5392,
+      "step": 3341
+    },
+    {
+      "epoch": 0.9267886855241264,
+      "grad_norm": 0.19183357059955597,
+      "learning_rate": 1.423370962063707e-05,
+      "loss": 0.5338,
+      "step": 3342
+    },
+    {
+      "epoch": 0.9270660011092623,
+      "grad_norm": 0.18884657323360443,
+      "learning_rate": 1.4228241178268617e-05,
+      "loss": 0.5143,
+      "step": 3343
+    },
+    {
+      "epoch": 0.9273433166943982,
+      "grad_norm": 0.19082637131214142,
+      "learning_rate": 1.4222772398674522e-05,
+      "loss": 0.5281,
+      "step": 3344
+    },
+    {
+      "epoch": 0.9276206322795341,
+      "grad_norm": 0.20209652185440063,
+      "learning_rate": 1.4217303282921888e-05,
+      "loss": 0.526,
+      "step": 3345
+    },
+    {
+      "epoch": 0.92789794786467,
+      "grad_norm": 0.18856458365917206,
+      "learning_rate": 1.4211833832077881e-05,
+      "loss": 0.52,
+      "step": 3346
+    },
+    {
+      "epoch": 0.9281752634498058,
+      "grad_norm": 0.1941954344511032,
+      "learning_rate": 1.4206364047209742e-05,
+      "loss": 0.5401,
+      "step": 3347
+    },
+    {
+      "epoch": 0.9284525790349417,
+      "grad_norm": 0.18749526143074036,
+      "learning_rate": 1.4200893929384767e-05,
+      "loss": 0.5449,
+      "step": 3348
+    },
+    {
+      "epoch": 0.9287298946200776,
+      "grad_norm": 0.18534240126609802,
+      "learning_rate": 1.4195423479670319e-05,
+      "loss": 0.5039,
+      "step": 3349
+    },
+    {
+      "epoch": 0.9290072102052135,
+      "grad_norm": 0.18905483186244965,
+      "learning_rate": 1.4189952699133837e-05,
+      "loss": 0.5632,
+      "step": 3350
+    },
+    {
+      "epoch": 0.9292845257903494,
+      "grad_norm": 0.20442554354667664,
+      "learning_rate": 1.4184481588842805e-05,
+      "loss": 0.5339,
+      "step": 3351
+    },
+    {
+      "epoch": 0.9295618413754853,
+      "grad_norm": 0.1810925155878067,
+      "learning_rate": 1.4179010149864785e-05,
+      "loss": 0.512,
+      "step": 3352
+    },
+    {
+      "epoch": 0.9298391569606211,
+      "grad_norm": 0.2010018676519394,
+      "learning_rate": 1.4173538383267404e-05,
+      "loss": 0.5514,
+      "step": 3353
+    },
+    {
+      "epoch": 0.930116472545757,
+      "grad_norm": 0.18728989362716675,
+      "learning_rate": 1.4168066290118342e-05,
+      "loss": 0.5161,
+      "step": 3354
+    },
+    {
+      "epoch": 0.9303937881308929,
+      "grad_norm": 0.18747897446155548,
+      "learning_rate": 1.4162593871485352e-05,
+      "loss": 0.5549,
+      "step": 3355
+    },
+    {
+      "epoch": 0.9306711037160288,
+      "grad_norm": 0.20961935818195343,
+      "learning_rate": 1.415712112843625e-05,
+      "loss": 0.5422,
+      "step": 3356
+    },
+    {
+      "epoch": 0.9309484193011647,
+      "grad_norm": 0.1834210753440857,
+      "learning_rate": 1.415164806203891e-05,
+      "loss": 0.5074,
+      "step": 3357
+    },
+    {
+      "epoch": 0.9312257348863006,
+      "grad_norm": 0.1954893171787262,
+      "learning_rate": 1.4146174673361273e-05,
+      "loss": 0.5722,
+      "step": 3358
+    },
+    {
+      "epoch": 0.9315030504714364,
+      "grad_norm": 0.1870308816432953,
+      "learning_rate": 1.4140700963471346e-05,
+      "loss": 0.5381,
+      "step": 3359
+    },
+    {
+      "epoch": 0.9317803660565723,
+      "grad_norm": 0.19276247918605804,
+      "learning_rate": 1.4135226933437185e-05,
+      "loss": 0.5517,
+      "step": 3360
+    },
+    {
+      "epoch": 0.9320576816417082,
+      "grad_norm": 0.18969060480594635,
+      "learning_rate": 1.412975258432693e-05,
+      "loss": 0.5347,
+      "step": 3361
+    },
+    {
+      "epoch": 0.9323349972268441,
+      "grad_norm": 0.19431856274604797,
+      "learning_rate": 1.4124277917208765e-05,
+      "loss": 0.5077,
+      "step": 3362
+    },
+    {
+      "epoch": 0.93261231281198,
+      "grad_norm": 0.1918652504682541,
+      "learning_rate": 1.4118802933150943e-05,
+      "loss": 0.526,
+      "step": 3363
+    },
+    {
+      "epoch": 0.9328896283971159,
+      "grad_norm": 0.19903786480426788,
+      "learning_rate": 1.4113327633221782e-05,
+      "loss": 0.5312,
+      "step": 3364
+    },
+    {
+      "epoch": 0.9331669439822518,
+      "grad_norm": 0.21233876049518585,
+      "learning_rate": 1.4107852018489653e-05,
+      "loss": 0.5256,
+      "step": 3365
+    },
+    {
+      "epoch": 0.9334442595673876,
+      "grad_norm": 0.1878969967365265,
+      "learning_rate": 1.4102376090022997e-05,
+      "loss": 0.5439,
+      "step": 3366
+    },
+    {
+      "epoch": 0.9337215751525235,
+      "grad_norm": 0.194586381316185,
+      "learning_rate": 1.409689984889031e-05,
+      "loss": 0.5392,
+      "step": 3367
+    },
+    {
+      "epoch": 0.9339988907376594,
+      "grad_norm": 0.1912042200565338,
+      "learning_rate": 1.4091423296160152e-05,
+      "loss": 0.5269,
+      "step": 3368
+    },
+    {
+      "epoch": 0.9342762063227953,
+      "grad_norm": 0.18847908079624176,
+      "learning_rate": 1.4085946432901154e-05,
+      "loss": 0.5175,
+      "step": 3369
+    },
+    {
+      "epoch": 0.9345535219079312,
+      "grad_norm": 0.19912898540496826,
+      "learning_rate": 1.4080469260181977e-05,
+      "loss": 0.5336,
+      "step": 3370
+    },
+    {
+      "epoch": 0.9348308374930671,
+      "grad_norm": 0.1928989291191101,
+      "learning_rate": 1.4074991779071378e-05,
+      "loss": 0.5847,
+      "step": 3371
+    },
+    {
+      "epoch": 0.9351081530782029,
+      "grad_norm": 0.20281600952148438,
+      "learning_rate": 1.4069513990638156e-05,
+      "loss": 0.5772,
+      "step": 3372
+    },
+    {
+      "epoch": 0.9353854686633388,
+      "grad_norm": 0.21058966219425201,
+      "learning_rate": 1.4064035895951169e-05,
+      "loss": 0.5201,
+      "step": 3373
+    },
+    {
+      "epoch": 0.9356627842484747,
+      "grad_norm": 0.190606027841568,
+      "learning_rate": 1.4058557496079342e-05,
+      "loss": 0.5226,
+      "step": 3374
+    },
+    {
+      "epoch": 0.9359400998336106,
+      "grad_norm": 0.1862584948539734,
+      "learning_rate": 1.4053078792091654e-05,
+      "loss": 0.5275,
+      "step": 3375
+    },
+    {
+      "epoch": 0.9362174154187465,
+      "grad_norm": 0.1864033192396164,
+      "learning_rate": 1.4047599785057144e-05,
+      "loss": 0.515,
+      "step": 3376
+    },
+    {
+      "epoch": 0.9364947310038824,
+      "grad_norm": 0.21961943805217743,
+      "learning_rate": 1.4042120476044912e-05,
+      "loss": 0.5344,
+      "step": 3377
+    },
+    {
+      "epoch": 0.9367720465890182,
+      "grad_norm": 0.18101632595062256,
+      "learning_rate": 1.4036640866124123e-05,
+      "loss": 0.547,
+      "step": 3378
+    },
+    {
+      "epoch": 0.9370493621741541,
+      "grad_norm": 0.21359553933143616,
+      "learning_rate": 1.4031160956363982e-05,
+      "loss": 0.5392,
+      "step": 3379
+    },
+    {
+      "epoch": 0.93732667775929,
+      "grad_norm": 0.1979496330022812,
+      "learning_rate": 1.4025680747833775e-05,
+      "loss": 0.5188,
+      "step": 3380
+    },
+    {
+      "epoch": 0.937603993344426,
+      "grad_norm": 0.18914289772510529,
+      "learning_rate": 1.402020024160283e-05,
+      "loss": 0.5291,
+      "step": 3381
+    },
+    {
+      "epoch": 0.9378813089295619,
+      "grad_norm": 0.18374024331569672,
+      "learning_rate": 1.4014719438740543e-05,
+      "loss": 0.5368,
+      "step": 3382
+    },
+    {
+      "epoch": 0.9381586245146978,
+      "grad_norm": 0.201919287443161,
+      "learning_rate": 1.4009238340316358e-05,
+      "loss": 0.5429,
+      "step": 3383
+    },
+    {
+      "epoch": 0.9384359400998337,
+      "grad_norm": 0.19644081592559814,
+      "learning_rate": 1.400375694739979e-05,
+      "loss": 0.5151,
+      "step": 3384
+    },
+    {
+      "epoch": 0.9387132556849695,
+      "grad_norm": 0.18970987200737,
+      "learning_rate": 1.39982752610604e-05,
+      "loss": 0.5852,
+      "step": 3385
+    },
+    {
+      "epoch": 0.9389905712701054,
+      "grad_norm": 0.19986121356487274,
+      "learning_rate": 1.3992793282367808e-05,
+      "loss": 0.5209,
+      "step": 3386
+    },
+    {
+      "epoch": 0.9392678868552413,
+      "grad_norm": 0.19541729986667633,
+      "learning_rate": 1.3987311012391698e-05,
+      "loss": 0.5315,
+      "step": 3387
+    },
+    {
+      "epoch": 0.9395452024403772,
+      "grad_norm": 0.1904657930135727,
+      "learning_rate": 1.3981828452201804e-05,
+      "loss": 0.5483,
+      "step": 3388
+    },
+    {
+      "epoch": 0.9398225180255131,
+      "grad_norm": 0.19447685778141022,
+      "learning_rate": 1.3976345602867916e-05,
+      "loss": 0.5446,
+      "step": 3389
+    },
+    {
+      "epoch": 0.940099833610649,
+      "grad_norm": 0.1892615556716919,
+      "learning_rate": 1.3970862465459891e-05,
+      "loss": 0.5366,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9403771491957849,
+      "grad_norm": 0.1949482262134552,
+      "learning_rate": 1.3965379041047624e-05,
+      "loss": 0.5577,
+      "step": 3391
+    },
+    {
+      "epoch": 0.9406544647809207,
+      "grad_norm": 0.19417253136634827,
+      "learning_rate": 1.3959895330701083e-05,
+      "loss": 0.5431,
+      "step": 3392
+    },
+    {
+      "epoch": 0.9409317803660566,
+      "grad_norm": 0.19106963276863098,
+      "learning_rate": 1.3954411335490284e-05,
+      "loss": 0.5308,
+      "step": 3393
+    },
+    {
+      "epoch": 0.9412090959511925,
+      "grad_norm": 0.19582101702690125,
+      "learning_rate": 1.39489270564853e-05,
+      "loss": 0.5221,
+      "step": 3394
+    },
+    {
+      "epoch": 0.9414864115363284,
+      "grad_norm": 0.1944214105606079,
+      "learning_rate": 1.3943442494756259e-05,
+      "loss": 0.5428,
+      "step": 3395
+    },
+    {
+      "epoch": 0.9417637271214643,
+      "grad_norm": 0.19008556008338928,
+      "learning_rate": 1.3937957651373342e-05,
+      "loss": 0.5458,
+      "step": 3396
+    },
+    {
+      "epoch": 0.9420410427066002,
+      "grad_norm": 0.186477929353714,
+      "learning_rate": 1.3932472527406792e-05,
+      "loss": 0.5315,
+      "step": 3397
+    },
+    {
+      "epoch": 0.942318358291736,
+      "grad_norm": 0.20386064052581787,
+      "learning_rate": 1.3926987123926897e-05,
+      "loss": 0.5309,
+      "step": 3398
+    },
+    {
+      "epoch": 0.9425956738768719,
+      "grad_norm": 0.18348954617977142,
+      "learning_rate": 1.3921501442004011e-05,
+      "loss": 0.5333,
+      "step": 3399
+    },
+    {
+      "epoch": 0.9428729894620078,
+      "grad_norm": 0.19923634827136993,
+      "learning_rate": 1.3916015482708528e-05,
+      "loss": 0.5301,
+      "step": 3400
+    },
+    {
+      "epoch": 0.9431503050471437,
+      "grad_norm": 0.19134169816970825,
+      "learning_rate": 1.3910529247110906e-05,
+      "loss": 0.519,
+      "step": 3401
+    },
+    {
+      "epoch": 0.9434276206322796,
+      "grad_norm": 0.18783889710903168,
+      "learning_rate": 1.390504273628166e-05,
+      "loss": 0.5207,
+      "step": 3402
+    },
+    {
+      "epoch": 0.9437049362174155,
+      "grad_norm": 0.18133123219013214,
+      "learning_rate": 1.3899555951291348e-05,
+      "loss": 0.5244,
+      "step": 3403
+    },
+    {
+      "epoch": 0.9439822518025514,
+      "grad_norm": 0.17967382073402405,
+      "learning_rate": 1.3894068893210594e-05,
+      "loss": 0.558,
+      "step": 3404
+    },
+    {
+      "epoch": 0.9442595673876872,
+      "grad_norm": 0.1942291408777237,
+      "learning_rate": 1.3888581563110059e-05,
+      "loss": 0.5307,
+      "step": 3405
+    },
+    {
+      "epoch": 0.9445368829728231,
+      "grad_norm": 0.22371730208396912,
+      "learning_rate": 1.3883093962060472e-05,
+      "loss": 0.5386,
+      "step": 3406
+    },
+    {
+      "epoch": 0.944814198557959,
+      "grad_norm": 0.18771930038928986,
+      "learning_rate": 1.387760609113261e-05,
+      "loss": 0.5267,
+      "step": 3407
+    },
+    {
+      "epoch": 0.9450915141430949,
+      "grad_norm": 0.19484716653823853,
+      "learning_rate": 1.3872117951397298e-05,
+      "loss": 0.5578,
+      "step": 3408
+    },
+    {
+      "epoch": 0.9453688297282308,
+      "grad_norm": 0.190118208527565,
+      "learning_rate": 1.3866629543925424e-05,
+      "loss": 0.4886,
+      "step": 3409
+    },
+    {
+      "epoch": 0.9456461453133667,
+      "grad_norm": 0.18805035948753357,
+      "learning_rate": 1.3861140869787914e-05,
+      "loss": 0.5393,
+      "step": 3410
+    },
+    {
+      "epoch": 0.9459234608985025,
+      "grad_norm": 0.1881994605064392,
+      "learning_rate": 1.385565193005576e-05,
+      "loss": 0.5368,
+      "step": 3411
+    },
+    {
+      "epoch": 0.9462007764836384,
+      "grad_norm": 0.18705572187900543,
+      "learning_rate": 1.3850162725799997e-05,
+      "loss": 0.5706,
+      "step": 3412
+    },
+    {
+      "epoch": 0.9464780920687743,
+      "grad_norm": 0.23857946693897247,
+      "learning_rate": 1.3844673258091714e-05,
+      "loss": 0.5151,
+      "step": 3413
+    },
+    {
+      "epoch": 0.9467554076539102,
+      "grad_norm": 0.190442755818367,
+      "learning_rate": 1.383918352800205e-05,
+      "loss": 0.5287,
+      "step": 3414
+    },
+    {
+      "epoch": 0.9470327232390461,
+      "grad_norm": 0.19661828875541687,
+      "learning_rate": 1.38336935366022e-05,
+      "loss": 0.5474,
+      "step": 3415
+    },
+    {
+      "epoch": 0.947310038824182,
+      "grad_norm": 0.1797836571931839,
+      "learning_rate": 1.3828203284963409e-05,
+      "loss": 0.5416,
+      "step": 3416
+    },
+    {
+      "epoch": 0.9475873544093179,
+      "grad_norm": 0.23815502226352692,
+      "learning_rate": 1.382271277415696e-05,
+      "loss": 0.5501,
+      "step": 3417
+    },
+    {
+      "epoch": 0.9478646699944537,
+      "grad_norm": 0.19180168211460114,
+      "learning_rate": 1.381722200525421e-05,
+      "loss": 0.5254,
+      "step": 3418
+    },
+    {
+      "epoch": 0.9481419855795896,
+      "grad_norm": 0.2042774260044098,
+      "learning_rate": 1.3811730979326545e-05,
+      "loss": 0.5437,
+      "step": 3419
+    },
+    {
+      "epoch": 0.9484193011647255,
+      "grad_norm": 0.1914202868938446,
+      "learning_rate": 1.3806239697445414e-05,
+      "loss": 0.5602,
+      "step": 3420
+    },
+    {
+      "epoch": 0.9486966167498614,
+      "grad_norm": 0.2009340077638626,
+      "learning_rate": 1.3800748160682309e-05,
+      "loss": 0.5497,
+      "step": 3421
+    },
+    {
+      "epoch": 0.9489739323349973,
+      "grad_norm": 0.19518032670021057,
+      "learning_rate": 1.3795256370108776e-05,
+      "loss": 0.5149,
+      "step": 3422
+    },
+    {
+      "epoch": 0.9492512479201332,
+      "grad_norm": 0.19419553875923157,
+      "learning_rate": 1.3789764326796407e-05,
+      "loss": 0.5407,
+      "step": 3423
+    },
+    {
+      "epoch": 0.949528563505269,
+      "grad_norm": 0.19669625163078308,
+      "learning_rate": 1.3784272031816844e-05,
+      "loss": 0.537,
+      "step": 3424
+    },
+    {
+      "epoch": 0.9498058790904049,
+      "grad_norm": 0.1969965547323227,
+      "learning_rate": 1.3778779486241786e-05,
+      "loss": 0.5324,
+      "step": 3425
+    },
+    {
+      "epoch": 0.9500831946755408,
+      "grad_norm": 0.1851070374250412,
+      "learning_rate": 1.3773286691142966e-05,
+      "loss": 0.5533,
+      "step": 3426
+    },
+    {
+      "epoch": 0.9503605102606767,
+      "grad_norm": 0.18586984276771545,
+      "learning_rate": 1.3767793647592175e-05,
+      "loss": 0.5347,
+      "step": 3427
+    },
+    {
+      "epoch": 0.9506378258458126,
+      "grad_norm": 0.1796058714389801,
+      "learning_rate": 1.3762300356661261e-05,
+      "loss": 0.5223,
+      "step": 3428
+    },
+    {
+      "epoch": 0.9509151414309485,
+      "grad_norm": 0.1968797892332077,
+      "learning_rate": 1.3756806819422097e-05,
+      "loss": 0.5296,
+      "step": 3429
+    },
+    {
+      "epoch": 0.9511924570160843,
+      "grad_norm": 0.19946354627609253,
+      "learning_rate": 1.3751313036946627e-05,
+      "loss": 0.555,
+      "step": 3430
+    },
+    {
+      "epoch": 0.9514697726012202,
+      "grad_norm": 0.19412177801132202,
+      "learning_rate": 1.3745819010306832e-05,
+      "loss": 0.5367,
+      "step": 3431
+    },
+    {
+      "epoch": 0.9517470881863561,
+      "grad_norm": 0.1967850923538208,
+      "learning_rate": 1.374032474057474e-05,
+      "loss": 0.5355,
+      "step": 3432
+    },
+    {
+      "epoch": 0.952024403771492,
+      "grad_norm": 0.26605215668678284,
+      "learning_rate": 1.3734830228822428e-05,
+      "loss": 0.5161,
+      "step": 3433
+    },
+    {
+      "epoch": 0.9523017193566279,
+      "grad_norm": 0.21650773286819458,
+      "learning_rate": 1.372933547612202e-05,
+      "loss": 0.5398,
+      "step": 3434
+    },
+    {
+      "epoch": 0.9525790349417638,
+      "grad_norm": 0.20562243461608887,
+      "learning_rate": 1.3723840483545697e-05,
+      "loss": 0.5509,
+      "step": 3435
+    },
+    {
+      "epoch": 0.9528563505268997,
+      "grad_norm": 0.19407188892364502,
+      "learning_rate": 1.3718345252165663e-05,
+      "loss": 0.5373,
+      "step": 3436
+    },
+    {
+      "epoch": 0.9531336661120355,
+      "grad_norm": 0.19095157086849213,
+      "learning_rate": 1.3712849783054197e-05,
+      "loss": 0.54,
+      "step": 3437
+    },
+    {
+      "epoch": 0.9534109816971714,
+      "grad_norm": 0.1831715852022171,
+      "learning_rate": 1.3707354077283599e-05,
+      "loss": 0.5451,
+      "step": 3438
+    },
+    {
+      "epoch": 0.9536882972823073,
+      "grad_norm": 0.19795599579811096,
+      "learning_rate": 1.3701858135926238e-05,
+      "loss": 0.5329,
+      "step": 3439
+    },
+    {
+      "epoch": 0.9539656128674432,
+      "grad_norm": 0.19685760140419006,
+      "learning_rate": 1.3696361960054506e-05,
+      "loss": 0.5525,
+      "step": 3440
+    },
+    {
+      "epoch": 0.9542429284525791,
+      "grad_norm": 0.1895620971918106,
+      "learning_rate": 1.3690865550740864e-05,
+      "loss": 0.539,
+      "step": 3441
+    },
+    {
+      "epoch": 0.954520244037715,
+      "grad_norm": 0.1906086504459381,
+      "learning_rate": 1.3685368909057799e-05,
+      "loss": 0.5449,
+      "step": 3442
+    },
+    {
+      "epoch": 0.9547975596228508,
+      "grad_norm": 0.19026583433151245,
+      "learning_rate": 1.3679872036077853e-05,
+      "loss": 0.5303,
+      "step": 3443
+    },
+    {
+      "epoch": 0.9550748752079867,
+      "grad_norm": 0.1858745813369751,
+      "learning_rate": 1.3674374932873615e-05,
+      "loss": 0.5364,
+      "step": 3444
+    },
+    {
+      "epoch": 0.9553521907931226,
+      "grad_norm": 0.19552922248840332,
+      "learning_rate": 1.3668877600517712e-05,
+      "loss": 0.5367,
+      "step": 3445
+    },
+    {
+      "epoch": 0.9556295063782585,
+      "grad_norm": 0.1855076402425766,
+      "learning_rate": 1.3663380040082821e-05,
+      "loss": 0.532,
+      "step": 3446
+    },
+    {
+      "epoch": 0.9559068219633944,
+      "grad_norm": 0.19864481687545776,
+      "learning_rate": 1.365788225264166e-05,
+      "loss": 0.5252,
+      "step": 3447
+    },
+    {
+      "epoch": 0.9561841375485303,
+      "grad_norm": 0.18544144928455353,
+      "learning_rate": 1.3652384239266993e-05,
+      "loss": 0.5359,
+      "step": 3448
+    },
+    {
+      "epoch": 0.9564614531336662,
+      "grad_norm": 0.18177370727062225,
+      "learning_rate": 1.364688600103163e-05,
+      "loss": 0.5516,
+      "step": 3449
+    },
+    {
+      "epoch": 0.956738768718802,
+      "grad_norm": 0.2049761265516281,
+      "learning_rate": 1.3641387539008424e-05,
+      "loss": 0.5446,
+      "step": 3450
+    },
+    {
+      "epoch": 0.9570160843039379,
+      "grad_norm": 0.18312260508537292,
+      "learning_rate": 1.3635888854270268e-05,
+      "loss": 0.5292,
+      "step": 3451
+    },
+    {
+      "epoch": 0.9572933998890738,
+      "grad_norm": 0.18809424340724945,
+      "learning_rate": 1.36303899478901e-05,
+      "loss": 0.5296,
+      "step": 3452
+    },
+    {
+      "epoch": 0.9575707154742097,
+      "grad_norm": 0.183831587433815,
+      "learning_rate": 1.3624890820940902e-05,
+      "loss": 0.5213,
+      "step": 3453
+    },
+    {
+      "epoch": 0.9578480310593456,
+      "grad_norm": 0.18347223103046417,
+      "learning_rate": 1.3619391474495708e-05,
+      "loss": 0.5516,
+      "step": 3454
+    },
+    {
+      "epoch": 0.9581253466444815,
+      "grad_norm": 0.1897597759962082,
+      "learning_rate": 1.3613891909627575e-05,
+      "loss": 0.5539,
+      "step": 3455
+    },
+    {
+      "epoch": 0.9584026622296173,
+      "grad_norm": 0.19610610604286194,
+      "learning_rate": 1.360839212740962e-05,
+      "loss": 0.5581,
+      "step": 3456
+    },
+    {
+      "epoch": 0.9586799778147532,
+      "grad_norm": 0.20589366555213928,
+      "learning_rate": 1.3602892128914992e-05,
+      "loss": 0.5266,
+      "step": 3457
+    },
+    {
+      "epoch": 0.9589572933998891,
+      "grad_norm": 0.18203216791152954,
+      "learning_rate": 1.3597391915216896e-05,
+      "loss": 0.5222,
+      "step": 3458
+    },
+    {
+      "epoch": 0.959234608985025,
+      "grad_norm": 0.19360937178134918,
+      "learning_rate": 1.3591891487388553e-05,
+      "loss": 0.5271,
+      "step": 3459
+    },
+    {
+      "epoch": 0.9595119245701609,
+      "grad_norm": 0.2998809814453125,
+      "learning_rate": 1.3586390846503259e-05,
+      "loss": 0.5281,
+      "step": 3460
+    },
+    {
+      "epoch": 0.9597892401552968,
+      "grad_norm": 0.18753725290298462,
+      "learning_rate": 1.3580889993634322e-05,
+      "loss": 0.5553,
+      "step": 3461
+    },
+    {
+      "epoch": 0.9600665557404326,
+      "grad_norm": 0.19862516224384308,
+      "learning_rate": 1.3575388929855112e-05,
+      "loss": 0.4975,
+      "step": 3462
+    },
+    {
+      "epoch": 0.9603438713255685,
+      "grad_norm": 0.18844319880008698,
+      "learning_rate": 1.3569887656239033e-05,
+      "loss": 0.4977,
+      "step": 3463
+    },
+    {
+      "epoch": 0.9606211869107044,
+      "grad_norm": 0.2365204095840454,
+      "learning_rate": 1.3564386173859523e-05,
+      "loss": 0.5243,
+      "step": 3464
+    },
+    {
+      "epoch": 0.9608985024958403,
+      "grad_norm": 0.19346970319747925,
+      "learning_rate": 1.3558884483790072e-05,
+      "loss": 0.5504,
+      "step": 3465
+    },
+    {
+      "epoch": 0.9611758180809762,
+      "grad_norm": 0.19397403299808502,
+      "learning_rate": 1.3553382587104201e-05,
+      "loss": 0.5448,
+      "step": 3466
+    },
+    {
+      "epoch": 0.9614531336661121,
+      "grad_norm": 0.19531095027923584,
+      "learning_rate": 1.3547880484875477e-05,
+      "loss": 0.5614,
+      "step": 3467
+    },
+    {
+      "epoch": 0.961730449251248,
+      "grad_norm": 0.20053645968437195,
+      "learning_rate": 1.354237817817751e-05,
+      "loss": 0.5412,
+      "step": 3468
+    },
+    {
+      "epoch": 0.9620077648363838,
+      "grad_norm": 0.19779855012893677,
+      "learning_rate": 1.3536875668083943e-05,
+      "loss": 0.5675,
+      "step": 3469
+    },
+    {
+      "epoch": 0.9622850804215197,
+      "grad_norm": 0.19257789850234985,
+      "learning_rate": 1.3531372955668462e-05,
+      "loss": 0.5128,
+      "step": 3470
+    },
+    {
+      "epoch": 0.9625623960066556,
+      "grad_norm": 0.19003498554229736,
+      "learning_rate": 1.352587004200479e-05,
+      "loss": 0.545,
+      "step": 3471
+    },
+    {
+      "epoch": 0.9628397115917915,
+      "grad_norm": 0.1935829222202301,
+      "learning_rate": 1.3520366928166695e-05,
+      "loss": 0.5568,
+      "step": 3472
+    },
+    {
+      "epoch": 0.9631170271769274,
+      "grad_norm": 0.19856667518615723,
+      "learning_rate": 1.3514863615227979e-05,
+      "loss": 0.5247,
+      "step": 3473
+    },
+    {
+      "epoch": 0.9633943427620633,
+      "grad_norm": 0.18320922553539276,
+      "learning_rate": 1.3509360104262478e-05,
+      "loss": 0.5358,
+      "step": 3474
+    },
+    {
+      "epoch": 0.9636716583471991,
+      "grad_norm": 0.20552265644073486,
+      "learning_rate": 1.3503856396344086e-05,
+      "loss": 0.5631,
+      "step": 3475
+    },
+    {
+      "epoch": 0.963948973932335,
+      "grad_norm": 0.19140848517417908,
+      "learning_rate": 1.3498352492546706e-05,
+      "loss": 0.5241,
+      "step": 3476
+    },
+    {
+      "epoch": 0.9642262895174709,
+      "grad_norm": 0.18474581837654114,
+      "learning_rate": 1.3492848393944312e-05,
+      "loss": 0.4961,
+      "step": 3477
+    },
+    {
+      "epoch": 0.9645036051026068,
+      "grad_norm": 0.1950991153717041,
+      "learning_rate": 1.3487344101610885e-05,
+      "loss": 0.5223,
+      "step": 3478
+    },
+    {
+      "epoch": 0.9647809206877427,
+      "grad_norm": 0.19947074353694916,
+      "learning_rate": 1.348183961662047e-05,
+      "loss": 0.5232,
+      "step": 3479
+    },
+    {
+      "epoch": 0.9650582362728786,
+      "grad_norm": 0.19454912841320038,
+      "learning_rate": 1.3476334940047127e-05,
+      "loss": 0.5611,
+      "step": 3480
+    },
+    {
+      "epoch": 0.9653355518580145,
+      "grad_norm": 0.19796130061149597,
+      "learning_rate": 1.3470830072964973e-05,
+      "loss": 0.5459,
+      "step": 3481
+    },
+    {
+      "epoch": 0.9656128674431503,
+      "grad_norm": 0.1934625804424286,
+      "learning_rate": 1.346532501644815e-05,
+      "loss": 0.5142,
+      "step": 3482
+    },
+    {
+      "epoch": 0.9658901830282862,
+      "grad_norm": 0.18265816569328308,
+      "learning_rate": 1.345981977157084e-05,
+      "loss": 0.5416,
+      "step": 3483
+    },
+    {
+      "epoch": 0.9661674986134221,
+      "grad_norm": 0.19438108801841736,
+      "learning_rate": 1.3454314339407262e-05,
+      "loss": 0.5264,
+      "step": 3484
+    },
+    {
+      "epoch": 0.966444814198558,
+      "grad_norm": 0.20060043036937714,
+      "learning_rate": 1.3448808721031673e-05,
+      "loss": 0.5472,
+      "step": 3485
+    },
+    {
+      "epoch": 0.9667221297836939,
+      "grad_norm": 0.20769962668418884,
+      "learning_rate": 1.3443302917518361e-05,
+      "loss": 0.5334,
+      "step": 3486
+    },
+    {
+      "epoch": 0.9669994453688298,
+      "grad_norm": 0.19313625991344452,
+      "learning_rate": 1.3437796929941661e-05,
+      "loss": 0.5565,
+      "step": 3487
+    },
+    {
+      "epoch": 0.9672767609539656,
+      "grad_norm": 0.20908023416996002,
+      "learning_rate": 1.3432290759375935e-05,
+      "loss": 0.5485,
+      "step": 3488
+    },
+    {
+      "epoch": 0.9675540765391015,
+      "grad_norm": 0.1994648575782776,
+      "learning_rate": 1.342678440689558e-05,
+      "loss": 0.5566,
+      "step": 3489
+    },
+    {
+      "epoch": 0.9678313921242374,
+      "grad_norm": 0.18501751124858856,
+      "learning_rate": 1.342127787357503e-05,
+      "loss": 0.5345,
+      "step": 3490
+    },
+    {
+      "epoch": 0.9681087077093733,
+      "grad_norm": 0.17410939931869507,
+      "learning_rate": 1.341577116048876e-05,
+      "loss": 0.5523,
+      "step": 3491
+    },
+    {
+      "epoch": 0.9683860232945092,
+      "grad_norm": 0.1948653906583786,
+      "learning_rate": 1.3410264268711276e-05,
+      "loss": 0.5419,
+      "step": 3492
+    },
+    {
+      "epoch": 0.9686633388796451,
+      "grad_norm": 0.1910163313150406,
+      "learning_rate": 1.3404757199317108e-05,
+      "loss": 0.5215,
+      "step": 3493
+    },
+    {
+      "epoch": 0.968940654464781,
+      "grad_norm": 0.21614839136600494,
+      "learning_rate": 1.3399249953380849e-05,
+      "loss": 0.5235,
+      "step": 3494
+    },
+    {
+      "epoch": 0.9692179700499168,
+      "grad_norm": 0.19417038559913635,
+      "learning_rate": 1.3393742531977094e-05,
+      "loss": 0.5504,
+      "step": 3495
+    },
+    {
+      "epoch": 0.9694952856350527,
+      "grad_norm": 0.18813247978687286,
+      "learning_rate": 1.3388234936180493e-05,
+      "loss": 0.5334,
+      "step": 3496
+    },
+    {
+      "epoch": 0.9697726012201886,
+      "grad_norm": 0.18560314178466797,
+      "learning_rate": 1.3382727167065723e-05,
+      "loss": 0.5295,
+      "step": 3497
+    },
+    {
+      "epoch": 0.9700499168053245,
+      "grad_norm": 0.19894826412200928,
+      "learning_rate": 1.3377219225707495e-05,
+      "loss": 0.5235,
+      "step": 3498
+    },
+    {
+      "epoch": 0.9703272323904604,
+      "grad_norm": 0.18726693093776703,
+      "learning_rate": 1.3371711113180552e-05,
+      "loss": 0.5032,
+      "step": 3499
+    },
+    {
+      "epoch": 0.9706045479755963,
+      "grad_norm": 0.20481140911579132,
+      "learning_rate": 1.3366202830559679e-05,
+      "loss": 0.5465,
+      "step": 3500
+    },
+    {
+      "epoch": 0.9708818635607321,
+      "grad_norm": 0.18740220367908478,
+      "learning_rate": 1.3360694378919683e-05,
+      "loss": 0.5052,
+      "step": 3501
+    },
+    {
+      "epoch": 0.971159179145868,
+      "grad_norm": 0.2005304992198944,
+      "learning_rate": 1.3355185759335409e-05,
+      "loss": 0.5565,
+      "step": 3502
+    },
+    {
+      "epoch": 0.9714364947310039,
+      "grad_norm": 0.1884390115737915,
+      "learning_rate": 1.3349676972881736e-05,
+      "loss": 0.5513,
+      "step": 3503
+    },
+    {
+      "epoch": 0.9717138103161398,
+      "grad_norm": 0.19259166717529297,
+      "learning_rate": 1.3344168020633574e-05,
+      "loss": 0.5136,
+      "step": 3504
+    },
+    {
+      "epoch": 0.9719911259012757,
+      "grad_norm": 0.18772351741790771,
+      "learning_rate": 1.3338658903665868e-05,
+      "loss": 0.5418,
+      "step": 3505
+    },
+    {
+      "epoch": 0.9722684414864116,
+      "grad_norm": 0.19562283158302307,
+      "learning_rate": 1.3333149623053584e-05,
+      "loss": 0.5291,
+      "step": 3506
+    },
+    {
+      "epoch": 0.9725457570715474,
+      "grad_norm": 0.18652015924453735,
+      "learning_rate": 1.332764017987174e-05,
+      "loss": 0.5406,
+      "step": 3507
+    },
+    {
+      "epoch": 0.9728230726566833,
+      "grad_norm": 0.19781257212162018,
+      "learning_rate": 1.3322130575195366e-05,
+      "loss": 0.5253,
+      "step": 3508
+    },
+    {
+      "epoch": 0.9731003882418192,
+      "grad_norm": 0.20048947632312775,
+      "learning_rate": 1.3316620810099536e-05,
+      "loss": 0.5339,
+      "step": 3509
+    },
+    {
+      "epoch": 0.9733777038269551,
+      "grad_norm": 0.1913914531469345,
+      "learning_rate": 1.331111088565935e-05,
+      "loss": 0.5399,
+      "step": 3510
+    },
+    {
+      "epoch": 0.973655019412091,
+      "grad_norm": 0.18670813739299774,
+      "learning_rate": 1.3305600802949941e-05,
+      "loss": 0.4995,
+      "step": 3511
+    },
+    {
+      "epoch": 0.9739323349972269,
+      "grad_norm": 0.1913759857416153,
+      "learning_rate": 1.3300090563046472e-05,
+      "loss": 0.5324,
+      "step": 3512
+    },
+    {
+      "epoch": 0.9742096505823628,
+      "grad_norm": 0.18199758231639862,
+      "learning_rate": 1.3294580167024135e-05,
+      "loss": 0.535,
+      "step": 3513
+    },
+    {
+      "epoch": 0.9744869661674986,
+      "grad_norm": 0.18627969920635223,
+      "learning_rate": 1.328906961595815e-05,
+      "loss": 0.5282,
+      "step": 3514
+    },
+    {
+      "epoch": 0.9747642817526345,
+      "grad_norm": 0.1850520372390747,
+      "learning_rate": 1.3283558910923785e-05,
+      "loss": 0.5126,
+      "step": 3515
+    },
+    {
+      "epoch": 0.9750415973377704,
+      "grad_norm": 0.19907443225383759,
+      "learning_rate": 1.327804805299631e-05,
+      "loss": 0.5228,
+      "step": 3516
+    },
+    {
+      "epoch": 0.9753189129229063,
+      "grad_norm": 0.19155220687389374,
+      "learning_rate": 1.3272537043251054e-05,
+      "loss": 0.5289,
+      "step": 3517
+    },
+    {
+      "epoch": 0.9755962285080422,
+      "grad_norm": 0.19737032055854797,
+      "learning_rate": 1.3267025882763345e-05,
+      "loss": 0.5342,
+      "step": 3518
+    },
+    {
+      "epoch": 0.9758735440931781,
+      "grad_norm": 0.186650350689888,
+      "learning_rate": 1.3261514572608569e-05,
+      "loss": 0.5542,
+      "step": 3519
+    },
+    {
+      "epoch": 0.9761508596783139,
+      "grad_norm": 0.1934208869934082,
+      "learning_rate": 1.3256003113862122e-05,
+      "loss": 0.53,
+      "step": 3520
+    },
+    {
+      "epoch": 0.9764281752634498,
+      "grad_norm": 0.1932929903268814,
+      "learning_rate": 1.3250491507599439e-05,
+      "loss": 0.5322,
+      "step": 3521
+    },
+    {
+      "epoch": 0.9767054908485857,
+      "grad_norm": 0.20356619358062744,
+      "learning_rate": 1.3244979754895978e-05,
+      "loss": 0.5443,
+      "step": 3522
+    },
+    {
+      "epoch": 0.9769828064337216,
+      "grad_norm": 0.19713589549064636,
+      "learning_rate": 1.3239467856827229e-05,
+      "loss": 0.5335,
+      "step": 3523
+    },
+    {
+      "epoch": 0.9772601220188575,
+      "grad_norm": 0.18841132521629333,
+      "learning_rate": 1.323395581446871e-05,
+      "loss": 0.5614,
+      "step": 3524
+    },
+    {
+      "epoch": 0.9775374376039934,
+      "grad_norm": 0.18878857791423798,
+      "learning_rate": 1.3228443628895962e-05,
+      "loss": 0.536,
+      "step": 3525
+    },
+    {
+      "epoch": 0.9778147531891292,
+      "grad_norm": 0.1925961673259735,
+      "learning_rate": 1.3222931301184565e-05,
+      "loss": 0.5743,
+      "step": 3526
+    },
+    {
+      "epoch": 0.9780920687742651,
+      "grad_norm": 0.2000453919172287,
+      "learning_rate": 1.321741883241012e-05,
+      "loss": 0.5144,
+      "step": 3527
+    },
+    {
+      "epoch": 0.978369384359401,
+      "grad_norm": 0.1939268261194229,
+      "learning_rate": 1.3211906223648251e-05,
+      "loss": 0.5379,
+      "step": 3528
+    },
+    {
+      "epoch": 0.9786466999445369,
+      "grad_norm": 0.19331765174865723,
+      "learning_rate": 1.3206393475974615e-05,
+      "loss": 0.5428,
+      "step": 3529
+    },
+    {
+      "epoch": 0.9789240155296728,
+      "grad_norm": 0.20399914681911469,
+      "learning_rate": 1.3200880590464898e-05,
+      "loss": 0.5323,
+      "step": 3530
+    },
+    {
+      "epoch": 0.9792013311148087,
+      "grad_norm": 0.2036525458097458,
+      "learning_rate": 1.3195367568194807e-05,
+      "loss": 0.541,
+      "step": 3531
+    },
+    {
+      "epoch": 0.9794786466999446,
+      "grad_norm": 0.19523896276950836,
+      "learning_rate": 1.3189854410240082e-05,
+      "loss": 0.5479,
+      "step": 3532
+    },
+    {
+      "epoch": 0.9797559622850804,
+      "grad_norm": 0.20129665732383728,
+      "learning_rate": 1.318434111767648e-05,
+      "loss": 0.5565,
+      "step": 3533
+    },
+    {
+      "epoch": 0.9800332778702163,
+      "grad_norm": 0.19564275443553925,
+      "learning_rate": 1.3178827691579801e-05,
+      "loss": 0.5781,
+      "step": 3534
+    },
+    {
+      "epoch": 0.9803105934553522,
+      "grad_norm": 0.2035827934741974,
+      "learning_rate": 1.317331413302585e-05,
+      "loss": 0.554,
+      "step": 3535
+    },
+    {
+      "epoch": 0.9805879090404881,
+      "grad_norm": 0.1899930238723755,
+      "learning_rate": 1.3167800443090475e-05,
+      "loss": 0.5187,
+      "step": 3536
+    },
+    {
+      "epoch": 0.980865224625624,
+      "grad_norm": 0.1863166242837906,
+      "learning_rate": 1.3162286622849538e-05,
+      "loss": 0.5199,
+      "step": 3537
+    },
+    {
+      "epoch": 0.9811425402107599,
+      "grad_norm": 0.20462384819984436,
+      "learning_rate": 1.3156772673378936e-05,
+      "loss": 0.5479,
+      "step": 3538
+    },
+    {
+      "epoch": 0.9814198557958957,
+      "grad_norm": 0.19956259429454803,
+      "learning_rate": 1.3151258595754581e-05,
+      "loss": 0.5492,
+      "step": 3539
+    },
+    {
+      "epoch": 0.9816971713810316,
+      "grad_norm": 0.2239493578672409,
+      "learning_rate": 1.3145744391052422e-05,
+      "loss": 0.563,
+      "step": 3540
+    },
+    {
+      "epoch": 0.9819744869661675,
+      "grad_norm": 0.20383387804031372,
+      "learning_rate": 1.3140230060348425e-05,
+      "loss": 0.5309,
+      "step": 3541
+    },
+    {
+      "epoch": 0.9822518025513034,
+      "grad_norm": 0.1931990385055542,
+      "learning_rate": 1.3134715604718579e-05,
+      "loss": 0.5233,
+      "step": 3542
+    },
+    {
+      "epoch": 0.9825291181364393,
+      "grad_norm": 0.19594894349575043,
+      "learning_rate": 1.3129201025238902e-05,
+      "loss": 0.5366,
+      "step": 3543
+    },
+    {
+      "epoch": 0.9828064337215752,
+      "grad_norm": 0.3933532238006592,
+      "learning_rate": 1.3123686322985434e-05,
+      "loss": 0.5762,
+      "step": 3544
+    },
+    {
+      "epoch": 0.983083749306711,
+      "grad_norm": 0.19356705248355865,
+      "learning_rate": 1.311817149903424e-05,
+      "loss": 0.5247,
+      "step": 3545
+    },
+    {
+      "epoch": 0.9833610648918469,
+      "grad_norm": 0.19310222566127777,
+      "learning_rate": 1.3112656554461405e-05,
+      "loss": 0.5501,
+      "step": 3546
+    },
+    {
+      "epoch": 0.9836383804769828,
+      "grad_norm": 0.1923639476299286,
+      "learning_rate": 1.310714149034305e-05,
+      "loss": 0.5265,
+      "step": 3547
+    },
+    {
+      "epoch": 0.9839156960621187,
+      "grad_norm": 0.19612760841846466,
+      "learning_rate": 1.3101626307755303e-05,
+      "loss": 0.5292,
+      "step": 3548
+    },
+    {
+      "epoch": 0.9841930116472546,
+      "grad_norm": 0.19919681549072266,
+      "learning_rate": 1.3096111007774322e-05,
+      "loss": 0.565,
+      "step": 3549
+    },
+    {
+      "epoch": 0.9844703272323905,
+      "grad_norm": 0.20053736865520477,
+      "learning_rate": 1.3090595591476293e-05,
+      "loss": 0.5669,
+      "step": 3550
+    },
+    {
+      "epoch": 0.9847476428175264,
+      "grad_norm": 0.18242114782333374,
+      "learning_rate": 1.3085080059937413e-05,
+      "loss": 0.531,
+      "step": 3551
+    },
+    {
+      "epoch": 0.9850249584026622,
+      "grad_norm": 0.18836815655231476,
+      "learning_rate": 1.3079564414233912e-05,
+      "loss": 0.5238,
+      "step": 3552
+    },
+    {
+      "epoch": 0.9853022739877981,
+      "grad_norm": 0.185248464345932,
+      "learning_rate": 1.3074048655442042e-05,
+      "loss": 0.5294,
+      "step": 3553
+    },
+    {
+      "epoch": 0.985579589572934,
+      "grad_norm": 0.18509770929813385,
+      "learning_rate": 1.3068532784638065e-05,
+      "loss": 0.5492,
+      "step": 3554
+    },
+    {
+      "epoch": 0.9858569051580699,
+      "grad_norm": 0.1903439611196518,
+      "learning_rate": 1.3063016802898288e-05,
+      "loss": 0.5463,
+      "step": 3555
+    },
+    {
+      "epoch": 0.9861342207432058,
+      "grad_norm": 0.19315816462039948,
+      "learning_rate": 1.3057500711299006e-05,
+      "loss": 0.533,
+      "step": 3556
+    },
+    {
+      "epoch": 0.9864115363283417,
+      "grad_norm": 0.19946327805519104,
+      "learning_rate": 1.305198451091657e-05,
+      "loss": 0.5282,
+      "step": 3557
+    },
+    {
+      "epoch": 0.9866888519134775,
+      "grad_norm": 0.1908363550901413,
+      "learning_rate": 1.3046468202827328e-05,
+      "loss": 0.5208,
+      "step": 3558
+    },
+    {
+      "epoch": 0.9869661674986134,
+      "grad_norm": 0.1873141974210739,
+      "learning_rate": 1.304095178810766e-05,
+      "loss": 0.5146,
+      "step": 3559
+    },
+    {
+      "epoch": 0.9872434830837493,
+      "grad_norm": 0.18219764530658722,
+      "learning_rate": 1.303543526783397e-05,
+      "loss": 0.5109,
+      "step": 3560
+    },
+    {
+      "epoch": 0.9875207986688852,
+      "grad_norm": 0.1835818737745285,
+      "learning_rate": 1.3029918643082673e-05,
+      "loss": 0.5352,
+      "step": 3561
+    },
+    {
+      "epoch": 0.9877981142540211,
+      "grad_norm": 0.1892389953136444,
+      "learning_rate": 1.3024401914930207e-05,
+      "loss": 0.5209,
+      "step": 3562
+    },
+    {
+      "epoch": 0.988075429839157,
+      "grad_norm": 0.18364709615707397,
+      "learning_rate": 1.3018885084453036e-05,
+      "loss": 0.5213,
+      "step": 3563
+    },
+    {
+      "epoch": 0.9883527454242929,
+      "grad_norm": 0.18862204253673553,
+      "learning_rate": 1.3013368152727634e-05,
+      "loss": 0.5151,
+      "step": 3564
+    },
+    {
+      "epoch": 0.9886300610094287,
+      "grad_norm": 0.1929624229669571,
+      "learning_rate": 1.3007851120830506e-05,
+      "loss": 0.5347,
+      "step": 3565
+    },
+    {
+      "epoch": 0.9889073765945646,
+      "grad_norm": 0.19009216129779816,
+      "learning_rate": 1.3002333989838167e-05,
+      "loss": 0.5589,
+      "step": 3566
+    },
+    {
+      "epoch": 0.9891846921797005,
+      "grad_norm": 0.1797810047864914,
+      "learning_rate": 1.299681676082716e-05,
+      "loss": 0.5186,
+      "step": 3567
+    },
+    {
+      "epoch": 0.9894620077648364,
+      "grad_norm": 0.1861737072467804,
+      "learning_rate": 1.2991299434874038e-05,
+      "loss": 0.5292,
+      "step": 3568
+    },
+    {
+      "epoch": 0.9897393233499723,
+      "grad_norm": 0.186946839094162,
+      "learning_rate": 1.298578201305538e-05,
+      "loss": 0.5393,
+      "step": 3569
+    },
+    {
+      "epoch": 0.9900166389351082,
+      "grad_norm": 0.20192833244800568,
+      "learning_rate": 1.2980264496447784e-05,
+      "loss": 0.5487,
+      "step": 3570
+    },
+    {
+      "epoch": 0.990293954520244,
+      "grad_norm": 0.1849672794342041,
+      "learning_rate": 1.2974746886127858e-05,
+      "loss": 0.5342,
+      "step": 3571
+    },
+    {
+      "epoch": 0.9905712701053799,
+      "grad_norm": 0.18184737861156464,
+      "learning_rate": 1.2969229183172236e-05,
+      "loss": 0.5387,
+      "step": 3572
+    },
+    {
+      "epoch": 0.9908485856905158,
+      "grad_norm": 0.19623394310474396,
+      "learning_rate": 1.2963711388657566e-05,
+      "loss": 0.5588,
+      "step": 3573
+    },
+    {
+      "epoch": 0.9911259012756517,
+      "grad_norm": 0.19016727805137634,
+      "learning_rate": 1.2958193503660524e-05,
+      "loss": 0.5393,
+      "step": 3574
+    },
+    {
+      "epoch": 0.9914032168607876,
+      "grad_norm": 0.19216102361679077,
+      "learning_rate": 1.2952675529257785e-05,
+      "loss": 0.5383,
+      "step": 3575
+    },
+    {
+      "epoch": 0.9916805324459235,
+      "grad_norm": 0.21044804155826569,
+      "learning_rate": 1.2947157466526062e-05,
+      "loss": 0.5453,
+      "step": 3576
+    },
+    {
+      "epoch": 0.9919578480310594,
+      "grad_norm": 0.19791410863399506,
+      "learning_rate": 1.2941639316542062e-05,
+      "loss": 0.5562,
+      "step": 3577
+    },
+    {
+      "epoch": 0.9922351636161952,
+      "grad_norm": 0.18991726636886597,
+      "learning_rate": 1.2936121080382534e-05,
+      "loss": 0.4977,
+      "step": 3578
+    },
+    {
+      "epoch": 0.9925124792013311,
+      "grad_norm": 0.19211836159229279,
+      "learning_rate": 1.293060275912423e-05,
+      "loss": 0.5227,
+      "step": 3579
+    },
+    {
+      "epoch": 0.992789794786467,
+      "grad_norm": 0.19635222852230072,
+      "learning_rate": 1.292508435384392e-05,
+      "loss": 0.5454,
+      "step": 3580
+    },
+    {
+      "epoch": 0.9930671103716029,
+      "grad_norm": 0.18526218831539154,
+      "learning_rate": 1.2919565865618388e-05,
+      "loss": 0.5429,
+      "step": 3581
+    },
+    {
+      "epoch": 0.9933444259567388,
+      "grad_norm": 0.1954220086336136,
+      "learning_rate": 1.291404729552444e-05,
+      "loss": 0.5479,
+      "step": 3582
+    },
+    {
+      "epoch": 0.9936217415418747,
+      "grad_norm": 0.1860000044107437,
+      "learning_rate": 1.2908528644638895e-05,
+      "loss": 0.5291,
+      "step": 3583
+    },
+    {
+      "epoch": 0.9938990571270105,
+      "grad_norm": 0.19651706516742706,
+      "learning_rate": 1.2903009914038586e-05,
+      "loss": 0.5345,
+      "step": 3584
+    },
+    {
+      "epoch": 0.9941763727121464,
+      "grad_norm": 0.18839724361896515,
+      "learning_rate": 1.2897491104800366e-05,
+      "loss": 0.5624,
+      "step": 3585
+    },
+    {
+      "epoch": 0.9944536882972823,
+      "grad_norm": 0.1895817518234253,
+      "learning_rate": 1.28919722180011e-05,
+      "loss": 0.5577,
+      "step": 3586
+    },
+    {
+      "epoch": 0.9947310038824182,
+      "grad_norm": 0.18223224580287933,
+      "learning_rate": 1.288645325471767e-05,
+      "loss": 0.5511,
+      "step": 3587
+    },
+    {
+      "epoch": 0.9950083194675541,
+      "grad_norm": 0.18843968212604523,
+      "learning_rate": 1.2880934216026971e-05,
+      "loss": 0.5416,
+      "step": 3588
+    },
+    {
+      "epoch": 0.99528563505269,
+      "grad_norm": 0.18690787255764008,
+      "learning_rate": 1.2875415103005915e-05,
+      "loss": 0.5348,
+      "step": 3589
+    },
+    {
+      "epoch": 0.9955629506378258,
+      "grad_norm": 0.18256263434886932,
+      "learning_rate": 1.2869895916731426e-05,
+      "loss": 0.5147,
+      "step": 3590
+    },
+    {
+      "epoch": 0.9958402662229617,
+      "grad_norm": 0.18710075318813324,
+      "learning_rate": 1.2864376658280441e-05,
+      "loss": 0.5245,
+      "step": 3591
+    },
+    {
+      "epoch": 0.9961175818080976,
+      "grad_norm": 0.20847736299037933,
+      "learning_rate": 1.2858857328729915e-05,
+      "loss": 0.5432,
+      "step": 3592
+    },
+    {
+      "epoch": 0.9963948973932335,
+      "grad_norm": 0.18452630937099457,
+      "learning_rate": 1.2853337929156822e-05,
+      "loss": 0.5279,
+      "step": 3593
+    },
+    {
+      "epoch": 0.9966722129783694,
+      "grad_norm": 0.19052648544311523,
+      "learning_rate": 1.2847818460638131e-05,
+      "loss": 0.5312,
+      "step": 3594
+    },
+    {
+      "epoch": 0.9969495285635053,
+      "grad_norm": 0.19351086020469666,
+      "learning_rate": 1.2842298924250848e-05,
+      "loss": 0.525,
+      "step": 3595
+    },
+    {
+      "epoch": 0.9972268441486412,
+      "grad_norm": 0.17867478728294373,
+      "learning_rate": 1.2836779321071974e-05,
+      "loss": 0.5493,
+      "step": 3596
+    },
+    {
+      "epoch": 0.997504159733777,
+      "grad_norm": 0.18988832831382751,
+      "learning_rate": 1.2831259652178532e-05,
+      "loss": 0.5449,
+      "step": 3597
+    },
+    {
+      "epoch": 0.9977814753189129,
+      "grad_norm": 0.18697360157966614,
+      "learning_rate": 1.2825739918647553e-05,
+      "loss": 0.5146,
+      "step": 3598
+    },
+    {
+      "epoch": 0.9980587909040488,
+      "grad_norm": 0.19430890679359436,
+      "learning_rate": 1.2820220121556087e-05,
+      "loss": 0.5423,
+      "step": 3599
+    },
+    {
+      "epoch": 0.9983361064891847,
+      "grad_norm": 0.19263558089733124,
+      "learning_rate": 1.2814700261981195e-05,
+      "loss": 0.5393,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9986134220743206,
+      "grad_norm": 0.1892475187778473,
+      "learning_rate": 1.2809180340999938e-05,
+      "loss": 0.5205,
+      "step": 3601
+    },
+    {
+      "epoch": 0.9988907376594565,
+      "grad_norm": 0.19061142206192017,
+      "learning_rate": 1.280366035968941e-05,
+      "loss": 0.5195,
+      "step": 3602
+    },
+    {
+      "epoch": 0.9991680532445923,
+      "grad_norm": 0.1829683780670166,
+      "learning_rate": 1.2798140319126695e-05,
+      "loss": 0.5111,
+      "step": 3603
+    },
+    {
+      "epoch": 0.9994453688297282,
+      "grad_norm": 0.18301549553871155,
+      "learning_rate": 1.279262022038891e-05,
+      "loss": 0.5393,
+      "step": 3604
+    },
+    {
+      "epoch": 0.9997226844148641,
+      "grad_norm": 0.18907521665096283,
+      "learning_rate": 1.2787100064553162e-05,
+      "loss": 0.534,
+      "step": 3605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.20034904778003693,
+      "learning_rate": 1.2781579852696588e-05,
+      "loss": 0.5388,
+      "step": 3606
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.8209076523780823,
+      "eval_runtime": 415.5978,
+      "eval_samples_per_second": 98.684,
+      "eval_steps_per_second": 1.542,
+      "step": 3606
+    },
+    {
+      "epoch": 1.0002773155851359,
+      "grad_norm": 0.19878089427947998,
+      "learning_rate": 1.2776059585896324e-05,
+      "loss": 0.537,
+      "step": 3607
+    },
+    {
+      "epoch": 1.0005546311702718,
+      "grad_norm": 0.18699733912944794,
+      "learning_rate": 1.2770539265229522e-05,
+      "loss": 0.5383,
+      "step": 3608
+    },
+    {
+      "epoch": 1.0008319467554077,
+      "grad_norm": 0.18291421234607697,
+      "learning_rate": 1.2765018891773343e-05,
+      "loss": 0.5536,
+      "step": 3609
+    },
+    {
+      "epoch": 1.0011092623405435,
+      "grad_norm": 0.19343863427639008,
+      "learning_rate": 1.2759498466604951e-05,
+      "loss": 0.5325,
+      "step": 3610
+    },
+    {
+      "epoch": 1.0013865779256794,
+      "grad_norm": 0.19165514409542084,
+      "learning_rate": 1.2753977990801536e-05,
+      "loss": 0.5405,
+      "step": 3611
+    },
+    {
+      "epoch": 1.0016638935108153,
+      "grad_norm": 0.1763203889131546,
+      "learning_rate": 1.2748457465440289e-05,
+      "loss": 0.5398,
+      "step": 3612
+    },
+    {
+      "epoch": 1.0019412090959512,
+      "grad_norm": 0.2186586856842041,
+      "learning_rate": 1.27429368915984e-05,
+      "loss": 0.5192,
+      "step": 3613
+    },
+    {
+      "epoch": 1.002218524681087,
+      "grad_norm": 0.19609209895133972,
+      "learning_rate": 1.2737416270353094e-05,
+      "loss": 0.5364,
+      "step": 3614
+    },
+    {
+      "epoch": 1.002495840266223,
+      "grad_norm": 0.20778897404670715,
+      "learning_rate": 1.273189560278158e-05,
+      "loss": 0.5505,
+      "step": 3615
+    },
+    {
+      "epoch": 1.0027731558513588,
+      "grad_norm": 0.2087172418832779,
+      "learning_rate": 1.2726374889961095e-05,
+      "loss": 0.5627,
+      "step": 3616
+    },
+    {
+      "epoch": 1.0030504714364947,
+      "grad_norm": 0.19748911261558533,
+      "learning_rate": 1.2720854132968865e-05,
+      "loss": 0.541,
+      "step": 3617
+    },
+    {
+      "epoch": 1.0033277870216306,
+      "grad_norm": 0.18848268687725067,
+      "learning_rate": 1.2715333332882146e-05,
+      "loss": 0.5373,
+      "step": 3618
+    },
+    {
+      "epoch": 1.0036051026067665,
+      "grad_norm": 0.19345615804195404,
+      "learning_rate": 1.2709812490778187e-05,
+      "loss": 0.5429,
+      "step": 3619
+    },
+    {
+      "epoch": 1.0038824181919024,
+      "grad_norm": 0.18660636246204376,
+      "learning_rate": 1.270429160773425e-05,
+      "loss": 0.5608,
+      "step": 3620
+    },
+    {
+      "epoch": 1.0041597337770383,
+      "grad_norm": 0.1913338303565979,
+      "learning_rate": 1.2698770684827612e-05,
+      "loss": 0.54,
+      "step": 3621
+    },
+    {
+      "epoch": 1.0044370493621742,
+      "grad_norm": 0.258350670337677,
+      "learning_rate": 1.2693249723135542e-05,
+      "loss": 0.4896,
+      "step": 3622
+    },
+    {
+      "epoch": 1.00471436494731,
+      "grad_norm": 0.18567179143428802,
+      "learning_rate": 1.2687728723735337e-05,
+      "loss": 0.5154,
+      "step": 3623
+    },
+    {
+      "epoch": 1.004991680532446,
+      "grad_norm": 0.1973087191581726,
+      "learning_rate": 1.2682207687704279e-05,
+      "loss": 0.557,
+      "step": 3624
+    },
+    {
+      "epoch": 1.0052689961175818,
+      "grad_norm": 0.22406698763370514,
+      "learning_rate": 1.2676686616119675e-05,
+      "loss": 0.5303,
+      "step": 3625
+    },
+    {
+      "epoch": 1.0055463117027177,
+      "grad_norm": 0.19072286784648895,
+      "learning_rate": 1.2671165510058834e-05,
+      "loss": 0.5455,
+      "step": 3626
+    },
+    {
+      "epoch": 1.0058236272878536,
+      "grad_norm": 0.1937733143568039,
+      "learning_rate": 1.2665644370599064e-05,
+      "loss": 0.5198,
+      "step": 3627
+    },
+    {
+      "epoch": 1.0061009428729895,
+      "grad_norm": 0.2021392285823822,
+      "learning_rate": 1.2660123198817692e-05,
+      "loss": 0.5313,
+      "step": 3628
+    },
+    {
+      "epoch": 1.0063782584581253,
+      "grad_norm": 0.18881459534168243,
+      "learning_rate": 1.2654601995792036e-05,
+      "loss": 0.5254,
+      "step": 3629
+    },
+    {
+      "epoch": 1.0066555740432612,
+      "grad_norm": 0.18779276311397552,
+      "learning_rate": 1.2649080762599442e-05,
+      "loss": 0.5202,
+      "step": 3630
+    },
+    {
+      "epoch": 1.006932889628397,
+      "grad_norm": 0.1880016028881073,
+      "learning_rate": 1.2643559500317234e-05,
+      "loss": 0.5177,
+      "step": 3631
+    },
+    {
+      "epoch": 1.007210205213533,
+      "grad_norm": 0.18926078081130981,
+      "learning_rate": 1.2638038210022765e-05,
+      "loss": 0.5349,
+      "step": 3632
+    },
+    {
+      "epoch": 1.0074875207986689,
+      "grad_norm": 0.18671317398548126,
+      "learning_rate": 1.2632516892793389e-05,
+      "loss": 0.5284,
+      "step": 3633
+    },
+    {
+      "epoch": 1.0077648363838048,
+      "grad_norm": 0.19339770078659058,
+      "learning_rate": 1.2626995549706452e-05,
+      "loss": 0.5238,
+      "step": 3634
+    },
+    {
+      "epoch": 1.0080421519689406,
+      "grad_norm": 0.18992464244365692,
+      "learning_rate": 1.2621474181839322e-05,
+      "loss": 0.5461,
+      "step": 3635
+    },
+    {
+      "epoch": 1.0083194675540765,
+      "grad_norm": 0.19679436087608337,
+      "learning_rate": 1.2615952790269356e-05,
+      "loss": 0.5174,
+      "step": 3636
+    },
+    {
+      "epoch": 1.0085967831392124,
+      "grad_norm": 0.1883174180984497,
+      "learning_rate": 1.2610431376073931e-05,
+      "loss": 0.5542,
+      "step": 3637
+    },
+    {
+      "epoch": 1.0088740987243483,
+      "grad_norm": 0.1904815286397934,
+      "learning_rate": 1.260490994033042e-05,
+      "loss": 0.5224,
+      "step": 3638
+    },
+    {
+      "epoch": 1.0091514143094842,
+      "grad_norm": 0.18984469771385193,
+      "learning_rate": 1.2599388484116198e-05,
+      "loss": 0.5237,
+      "step": 3639
+    },
+    {
+      "epoch": 1.00942872989462,
+      "grad_norm": 0.19481045007705688,
+      "learning_rate": 1.259386700850865e-05,
+      "loss": 0.5315,
+      "step": 3640
+    },
+    {
+      "epoch": 1.009706045479756,
+      "grad_norm": 0.17928001284599304,
+      "learning_rate": 1.2588345514585163e-05,
+      "loss": 0.5292,
+      "step": 3641
+    },
+    {
+      "epoch": 1.0099833610648918,
+      "grad_norm": 0.1879146546125412,
+      "learning_rate": 1.2582824003423124e-05,
+      "loss": 0.5269,
+      "step": 3642
+    },
+    {
+      "epoch": 1.0102606766500277,
+      "grad_norm": 0.18970663845539093,
+      "learning_rate": 1.2577302476099926e-05,
+      "loss": 0.5107,
+      "step": 3643
+    },
+    {
+      "epoch": 1.0105379922351636,
+      "grad_norm": 0.19013284146785736,
+      "learning_rate": 1.257178093369297e-05,
+      "loss": 0.5348,
+      "step": 3644
+    },
+    {
+      "epoch": 1.0108153078202995,
+      "grad_norm": 0.19010770320892334,
+      "learning_rate": 1.2566259377279652e-05,
+      "loss": 0.5264,
+      "step": 3645
+    },
+    {
+      "epoch": 1.0110926234054354,
+      "grad_norm": 0.18681344389915466,
+      "learning_rate": 1.2560737807937374e-05,
+      "loss": 0.497,
+      "step": 3646
+    },
+    {
+      "epoch": 1.0113699389905713,
+      "grad_norm": 0.1887647807598114,
+      "learning_rate": 1.2555216226743537e-05,
+      "loss": 0.5271,
+      "step": 3647
+    },
+    {
+      "epoch": 1.0116472545757071,
+      "grad_norm": 0.18254730105400085,
+      "learning_rate": 1.2549694634775555e-05,
+      "loss": 0.5283,
+      "step": 3648
+    },
+    {
+      "epoch": 1.011924570160843,
+      "grad_norm": 0.1893124133348465,
+      "learning_rate": 1.2544173033110832e-05,
+      "loss": 0.5273,
+      "step": 3649
+    },
+    {
+      "epoch": 1.012201885745979,
+      "grad_norm": 0.18291127681732178,
+      "learning_rate": 1.2538651422826777e-05,
+      "loss": 0.5425,
+      "step": 3650
+    },
+    {
+      "epoch": 1.0124792013311148,
+      "grad_norm": 0.200171560049057,
+      "learning_rate": 1.2533129805000807e-05,
+      "loss": 0.532,
+      "step": 3651
+    },
+    {
+      "epoch": 1.0127565169162507,
+      "grad_norm": 0.1930498629808426,
+      "learning_rate": 1.2527608180710338e-05,
+      "loss": 0.5295,
+      "step": 3652
+    },
+    {
+      "epoch": 1.0130338325013866,
+      "grad_norm": 0.18753856420516968,
+      "learning_rate": 1.2522086551032778e-05,
+      "loss": 0.5358,
+      "step": 3653
+    },
+    {
+      "epoch": 1.0133111480865225,
+      "grad_norm": 0.19522987306118011,
+      "learning_rate": 1.2516564917045548e-05,
+      "loss": 0.5332,
+      "step": 3654
+    },
+    {
+      "epoch": 1.0135884636716583,
+      "grad_norm": 0.17891767621040344,
+      "learning_rate": 1.2511043279826062e-05,
+      "loss": 0.5166,
+      "step": 3655
+    },
+    {
+      "epoch": 1.0138657792567942,
+      "grad_norm": 0.18819722533226013,
+      "learning_rate": 1.250552164045174e-05,
+      "loss": 0.5445,
+      "step": 3656
+    },
+    {
+      "epoch": 1.01414309484193,
+      "grad_norm": 0.1814819872379303,
+      "learning_rate": 1.25e-05,
+      "loss": 0.5191,
+      "step": 3657
+    },
+    {
+      "epoch": 1.014420410427066,
+      "grad_norm": 0.18296414613723755,
+      "learning_rate": 1.2494478359548261e-05,
+      "loss": 0.5173,
+      "step": 3658
+    },
+    {
+      "epoch": 1.0146977260122019,
+      "grad_norm": 0.20985326170921326,
+      "learning_rate": 1.2488956720173939e-05,
+      "loss": 0.5062,
+      "step": 3659
+    },
+    {
+      "epoch": 1.0149750415973378,
+      "grad_norm": 0.17893539369106293,
+      "learning_rate": 1.2483435082954453e-05,
+      "loss": 0.5458,
+      "step": 3660
+    },
+    {
+      "epoch": 1.0152523571824736,
+      "grad_norm": 0.19261109828948975,
+      "learning_rate": 1.2477913448967227e-05,
+      "loss": 0.5637,
+      "step": 3661
+    },
+    {
+      "epoch": 1.0155296727676095,
+      "grad_norm": 0.2319900244474411,
+      "learning_rate": 1.2472391819289667e-05,
+      "loss": 0.5379,
+      "step": 3662
+    },
+    {
+      "epoch": 1.0158069883527454,
+      "grad_norm": 0.1867554783821106,
+      "learning_rate": 1.2466870194999192e-05,
+      "loss": 0.5321,
+      "step": 3663
+    },
+    {
+      "epoch": 1.0160843039378813,
+      "grad_norm": 0.1840423047542572,
+      "learning_rate": 1.2461348577173224e-05,
+      "loss": 0.5228,
+      "step": 3664
+    },
+    {
+      "epoch": 1.0163616195230172,
+      "grad_norm": 0.18526844680309296,
+      "learning_rate": 1.2455826966889175e-05,
+      "loss": 0.5387,
+      "step": 3665
+    },
+    {
+      "epoch": 1.016638935108153,
+      "grad_norm": 0.19361324608325958,
+      "learning_rate": 1.2450305365224446e-05,
+      "loss": 0.5261,
+      "step": 3666
+    },
+    {
+      "epoch": 1.016916250693289,
+      "grad_norm": 0.19102880358695984,
+      "learning_rate": 1.2444783773256466e-05,
+      "loss": 0.5433,
+      "step": 3667
+    },
+    {
+      "epoch": 1.0171935662784248,
+      "grad_norm": 0.18817844986915588,
+      "learning_rate": 1.2439262192062631e-05,
+      "loss": 0.538,
+      "step": 3668
+    },
+    {
+      "epoch": 1.0174708818635607,
+      "grad_norm": 0.1901366263628006,
+      "learning_rate": 1.2433740622720353e-05,
+      "loss": 0.5453,
+      "step": 3669
+    },
+    {
+      "epoch": 1.0177481974486966,
+      "grad_norm": 0.22999554872512817,
+      "learning_rate": 1.242821906630703e-05,
+      "loss": 0.5164,
+      "step": 3670
+    },
+    {
+      "epoch": 1.0180255130338325,
+      "grad_norm": 0.1897146850824356,
+      "learning_rate": 1.2422697523900075e-05,
+      "loss": 0.5528,
+      "step": 3671
+    },
+    {
+      "epoch": 1.0183028286189684,
+      "grad_norm": 0.1865740269422531,
+      "learning_rate": 1.241717599657688e-05,
+      "loss": 0.53,
+      "step": 3672
+    },
+    {
+      "epoch": 1.0185801442041043,
+      "grad_norm": 0.18005676567554474,
+      "learning_rate": 1.2411654485414839e-05,
+      "loss": 0.5056,
+      "step": 3673
+    },
+    {
+      "epoch": 1.0188574597892401,
+      "grad_norm": 0.1854349970817566,
+      "learning_rate": 1.240613299149135e-05,
+      "loss": 0.549,
+      "step": 3674
+    },
+    {
+      "epoch": 1.019134775374376,
+      "grad_norm": 0.19649791717529297,
+      "learning_rate": 1.2400611515883805e-05,
+      "loss": 0.5321,
+      "step": 3675
+    },
+    {
+      "epoch": 1.019412090959512,
+      "grad_norm": 0.18856771290302277,
+      "learning_rate": 1.2395090059669585e-05,
+      "loss": 0.5304,
+      "step": 3676
+    },
+    {
+      "epoch": 1.0196894065446478,
+      "grad_norm": 0.18438786268234253,
+      "learning_rate": 1.238956862392607e-05,
+      "loss": 0.551,
+      "step": 3677
+    },
+    {
+      "epoch": 1.0199667221297837,
+      "grad_norm": 0.1798408478498459,
+      "learning_rate": 1.2384047209730647e-05,
+      "loss": 0.5237,
+      "step": 3678
+    },
+    {
+      "epoch": 1.0202440377149196,
+      "grad_norm": 0.17963095009326935,
+      "learning_rate": 1.2378525818160683e-05,
+      "loss": 0.5229,
+      "step": 3679
+    },
+    {
+      "epoch": 1.0205213533000554,
+      "grad_norm": 0.18897545337677002,
+      "learning_rate": 1.237300445029355e-05,
+      "loss": 0.519,
+      "step": 3680
+    },
+    {
+      "epoch": 1.0207986688851913,
+      "grad_norm": 0.18016240000724792,
+      "learning_rate": 1.2367483107206614e-05,
+      "loss": 0.5092,
+      "step": 3681
+    },
+    {
+      "epoch": 1.0210759844703272,
+      "grad_norm": 0.18622079491615295,
+      "learning_rate": 1.2361961789977238e-05,
+      "loss": 0.505,
+      "step": 3682
+    },
+    {
+      "epoch": 1.021353300055463,
+      "grad_norm": 0.1844959259033203,
+      "learning_rate": 1.2356440499682769e-05,
+      "loss": 0.5358,
+      "step": 3683
+    },
+    {
+      "epoch": 1.021630615640599,
+      "grad_norm": 0.18017494678497314,
+      "learning_rate": 1.2350919237400563e-05,
+      "loss": 0.521,
+      "step": 3684
+    },
+    {
+      "epoch": 1.0219079312257349,
+      "grad_norm": 0.18644961714744568,
+      "learning_rate": 1.2345398004207965e-05,
+      "loss": 0.5239,
+      "step": 3685
+    },
+    {
+      "epoch": 1.0221852468108708,
+      "grad_norm": 0.18692266941070557,
+      "learning_rate": 1.2339876801182315e-05,
+      "loss": 0.5055,
+      "step": 3686
+    },
+    {
+      "epoch": 1.0224625623960066,
+      "grad_norm": 0.18799254298210144,
+      "learning_rate": 1.2334355629400934e-05,
+      "loss": 0.5153,
+      "step": 3687
+    },
+    {
+      "epoch": 1.0227398779811425,
+      "grad_norm": 0.19313998520374298,
+      "learning_rate": 1.2328834489941168e-05,
+      "loss": 0.5274,
+      "step": 3688
+    },
+    {
+      "epoch": 1.0230171935662784,
+      "grad_norm": 0.24748623371124268,
+      "learning_rate": 1.2323313383880326e-05,
+      "loss": 0.5168,
+      "step": 3689
+    },
+    {
+      "epoch": 1.0232945091514143,
+      "grad_norm": 0.18509146571159363,
+      "learning_rate": 1.231779231229572e-05,
+      "loss": 0.508,
+      "step": 3690
+    },
+    {
+      "epoch": 1.0235718247365502,
+      "grad_norm": 0.2279065102338791,
+      "learning_rate": 1.2312271276264666e-05,
+      "loss": 0.5219,
+      "step": 3691
+    },
+    {
+      "epoch": 1.023849140321686,
+      "grad_norm": 0.18918190896511078,
+      "learning_rate": 1.230675027686446e-05,
+      "loss": 0.509,
+      "step": 3692
+    },
+    {
+      "epoch": 1.024126455906822,
+      "grad_norm": 0.19168592989444733,
+      "learning_rate": 1.2301229315172394e-05,
+      "loss": 0.5128,
+      "step": 3693
+    },
+    {
+      "epoch": 1.0244037714919578,
+      "grad_norm": 0.18460464477539062,
+      "learning_rate": 1.229570839226575e-05,
+      "loss": 0.4949,
+      "step": 3694
+    },
+    {
+      "epoch": 1.0246810870770937,
+      "grad_norm": 0.1955164670944214,
+      "learning_rate": 1.2290187509221816e-05,
+      "loss": 0.5084,
+      "step": 3695
+    },
+    {
+      "epoch": 1.0249584026622296,
+      "grad_norm": 0.21448062360286713,
+      "learning_rate": 1.2284666667117858e-05,
+      "loss": 0.5258,
+      "step": 3696
+    },
+    {
+      "epoch": 1.0252357182473655,
+      "grad_norm": 0.195621058344841,
+      "learning_rate": 1.2279145867031136e-05,
+      "loss": 0.5409,
+      "step": 3697
+    },
+    {
+      "epoch": 1.0255130338325014,
+      "grad_norm": 0.18348294496536255,
+      "learning_rate": 1.2273625110038908e-05,
+      "loss": 0.5288,
+      "step": 3698
+    },
+    {
+      "epoch": 1.0257903494176372,
+      "grad_norm": 0.2350398451089859,
+      "learning_rate": 1.2268104397218421e-05,
+      "loss": 0.5176,
+      "step": 3699
+    },
+    {
+      "epoch": 1.0260676650027731,
+      "grad_norm": 0.18816140294075012,
+      "learning_rate": 1.2262583729646909e-05,
+      "loss": 0.5174,
+      "step": 3700
+    },
+    {
+      "epoch": 1.026344980587909,
+      "grad_norm": 0.18344999849796295,
+      "learning_rate": 1.22570631084016e-05,
+      "loss": 0.5001,
+      "step": 3701
+    },
+    {
+      "epoch": 1.026622296173045,
+      "grad_norm": 0.19813844561576843,
+      "learning_rate": 1.2251542534559716e-05,
+      "loss": 0.5229,
+      "step": 3702
+    },
+    {
+      "epoch": 1.0268996117581808,
+      "grad_norm": 0.18872547149658203,
+      "learning_rate": 1.2246022009198469e-05,
+      "loss": 0.5265,
+      "step": 3703
+    },
+    {
+      "epoch": 1.0271769273433167,
+      "grad_norm": 0.19747066497802734,
+      "learning_rate": 1.2240501533395048e-05,
+      "loss": 0.5381,
+      "step": 3704
+    },
+    {
+      "epoch": 1.0274542429284526,
+      "grad_norm": 0.1966402530670166,
+      "learning_rate": 1.2234981108226662e-05,
+      "loss": 0.5447,
+      "step": 3705
+    },
+    {
+      "epoch": 1.0277315585135884,
+      "grad_norm": 0.1881251186132431,
+      "learning_rate": 1.222946073477048e-05,
+      "loss": 0.4944,
+      "step": 3706
+    },
+    {
+      "epoch": 1.0280088740987243,
+      "grad_norm": 0.19006265699863434,
+      "learning_rate": 1.222394041410368e-05,
+      "loss": 0.5279,
+      "step": 3707
+    },
+    {
+      "epoch": 1.0282861896838602,
+      "grad_norm": 0.17501787841320038,
+      "learning_rate": 1.2218420147303412e-05,
+      "loss": 0.4797,
+      "step": 3708
+    },
+    {
+      "epoch": 1.028563505268996,
+      "grad_norm": 0.17646637558937073,
+      "learning_rate": 1.2212899935446841e-05,
+      "loss": 0.4878,
+      "step": 3709
+    },
+    {
+      "epoch": 1.028840820854132,
+      "grad_norm": 0.19758492708206177,
+      "learning_rate": 1.2207379779611095e-05,
+      "loss": 0.5238,
+      "step": 3710
+    },
+    {
+      "epoch": 1.0291181364392679,
+      "grad_norm": 0.19391943514347076,
+      "learning_rate": 1.2201859680873305e-05,
+      "loss": 0.5086,
+      "step": 3711
+    },
+    {
+      "epoch": 1.0293954520244037,
+      "grad_norm": 0.18827463686466217,
+      "learning_rate": 1.2196339640310595e-05,
+      "loss": 0.5257,
+      "step": 3712
+    },
+    {
+      "epoch": 1.0296727676095396,
+      "grad_norm": 0.19285833835601807,
+      "learning_rate": 1.2190819659000063e-05,
+      "loss": 0.5179,
+      "step": 3713
+    },
+    {
+      "epoch": 1.0299500831946755,
+      "grad_norm": 0.18147997558116913,
+      "learning_rate": 1.2185299738018813e-05,
+      "loss": 0.4985,
+      "step": 3714
+    },
+    {
+      "epoch": 1.0302273987798114,
+      "grad_norm": 0.19085568189620972,
+      "learning_rate": 1.2179779878443915e-05,
+      "loss": 0.5131,
+      "step": 3715
+    },
+    {
+      "epoch": 1.0305047143649473,
+      "grad_norm": 0.20084667205810547,
+      "learning_rate": 1.217426008135245e-05,
+      "loss": 0.519,
+      "step": 3716
+    },
+    {
+      "epoch": 1.0307820299500832,
+      "grad_norm": 0.20030198991298676,
+      "learning_rate": 1.2168740347821473e-05,
+      "loss": 0.5277,
+      "step": 3717
+    },
+    {
+      "epoch": 1.031059345535219,
+      "grad_norm": 0.1940310150384903,
+      "learning_rate": 1.2163220678928028e-05,
+      "loss": 0.5001,
+      "step": 3718
+    },
+    {
+      "epoch": 1.031336661120355,
+      "grad_norm": 0.18660347163677216,
+      "learning_rate": 1.2157701075749153e-05,
+      "loss": 0.529,
+      "step": 3719
+    },
+    {
+      "epoch": 1.0316139767054908,
+      "grad_norm": 0.1930490881204605,
+      "learning_rate": 1.2152181539361871e-05,
+      "loss": 0.5056,
+      "step": 3720
+    },
+    {
+      "epoch": 1.0318912922906267,
+      "grad_norm": 0.1963385045528412,
+      "learning_rate": 1.2146662070843184e-05,
+      "loss": 0.5057,
+      "step": 3721
+    },
+    {
+      "epoch": 1.0321686078757626,
+      "grad_norm": 0.19734053313732147,
+      "learning_rate": 1.2141142671270085e-05,
+      "loss": 0.5287,
+      "step": 3722
+    },
+    {
+      "epoch": 1.0324459234608985,
+      "grad_norm": 0.18422985076904297,
+      "learning_rate": 1.2135623341719561e-05,
+      "loss": 0.5135,
+      "step": 3723
+    },
+    {
+      "epoch": 1.0327232390460344,
+      "grad_norm": 0.18918901681900024,
+      "learning_rate": 1.213010408326858e-05,
+      "loss": 0.5098,
+      "step": 3724
+    },
+    {
+      "epoch": 1.0330005546311702,
+      "grad_norm": 0.18276064097881317,
+      "learning_rate": 1.2124584896994085e-05,
+      "loss": 0.5098,
+      "step": 3725
+    },
+    {
+      "epoch": 1.0332778702163061,
+      "grad_norm": 0.2021724432706833,
+      "learning_rate": 1.2119065783973031e-05,
+      "loss": 0.5127,
+      "step": 3726
+    },
+    {
+      "epoch": 1.033555185801442,
+      "grad_norm": 0.18858903646469116,
+      "learning_rate": 1.2113546745282333e-05,
+      "loss": 0.5057,
+      "step": 3727
+    },
+    {
+      "epoch": 1.033832501386578,
+      "grad_norm": 0.18476137518882751,
+      "learning_rate": 1.2108027781998902e-05,
+      "loss": 0.5254,
+      "step": 3728
+    },
+    {
+      "epoch": 1.0341098169717138,
+      "grad_norm": 0.18318045139312744,
+      "learning_rate": 1.2102508895199633e-05,
+      "loss": 0.5248,
+      "step": 3729
+    },
+    {
+      "epoch": 1.0343871325568497,
+      "grad_norm": 0.18871383368968964,
+      "learning_rate": 1.2096990085961417e-05,
+      "loss": 0.5073,
+      "step": 3730
+    },
+    {
+      "epoch": 1.0346644481419855,
+      "grad_norm": 0.2086830884218216,
+      "learning_rate": 1.209147135536111e-05,
+      "loss": 0.5038,
+      "step": 3731
+    },
+    {
+      "epoch": 1.0349417637271214,
+      "grad_norm": 0.1914805769920349,
+      "learning_rate": 1.2085952704475562e-05,
+      "loss": 0.5322,
+      "step": 3732
+    },
+    {
+      "epoch": 1.0352190793122573,
+      "grad_norm": 0.18802616000175476,
+      "learning_rate": 1.2080434134381615e-05,
+      "loss": 0.5161,
+      "step": 3733
+    },
+    {
+      "epoch": 1.0354963948973932,
+      "grad_norm": 0.1888350397348404,
+      "learning_rate": 1.2074915646156083e-05,
+      "loss": 0.5178,
+      "step": 3734
+    },
+    {
+      "epoch": 1.035773710482529,
+      "grad_norm": 0.1935519278049469,
+      "learning_rate": 1.2069397240875774e-05,
+      "loss": 0.5151,
+      "step": 3735
+    },
+    {
+      "epoch": 1.036051026067665,
+      "grad_norm": 0.18621766567230225,
+      "learning_rate": 1.2063878919617467e-05,
+      "loss": 0.5034,
+      "step": 3736
+    },
+    {
+      "epoch": 1.0363283416528009,
+      "grad_norm": 0.20036379992961884,
+      "learning_rate": 1.2058360683457941e-05,
+      "loss": 0.518,
+      "step": 3737
+    },
+    {
+      "epoch": 1.0366056572379367,
+      "grad_norm": 0.19412393867969513,
+      "learning_rate": 1.2052842533473945e-05,
+      "loss": 0.5095,
+      "step": 3738
+    },
+    {
+      "epoch": 1.0368829728230726,
+      "grad_norm": 0.20189900696277618,
+      "learning_rate": 1.2047324470742216e-05,
+      "loss": 0.5,
+      "step": 3739
+    },
+    {
+      "epoch": 1.0371602884082085,
+      "grad_norm": 0.19598862528800964,
+      "learning_rate": 1.204180649633948e-05,
+      "loss": 0.5352,
+      "step": 3740
+    },
+    {
+      "epoch": 1.0374376039933444,
+      "grad_norm": 0.18859632313251495,
+      "learning_rate": 1.2036288611342436e-05,
+      "loss": 0.5148,
+      "step": 3741
+    },
+    {
+      "epoch": 1.0377149195784803,
+      "grad_norm": 0.18398115038871765,
+      "learning_rate": 1.2030770816827769e-05,
+      "loss": 0.519,
+      "step": 3742
+    },
+    {
+      "epoch": 1.0379922351636162,
+      "grad_norm": 0.19209997355937958,
+      "learning_rate": 1.2025253113872144e-05,
+      "loss": 0.535,
+      "step": 3743
+    },
+    {
+      "epoch": 1.038269550748752,
+      "grad_norm": 0.20428086817264557,
+      "learning_rate": 1.2019735503552219e-05,
+      "loss": 0.5346,
+      "step": 3744
+    },
+    {
+      "epoch": 1.038546866333888,
+      "grad_norm": 0.20181040465831757,
+      "learning_rate": 1.2014217986944624e-05,
+      "loss": 0.5171,
+      "step": 3745
+    },
+    {
+      "epoch": 1.0388241819190238,
+      "grad_norm": 0.18710096180438995,
+      "learning_rate": 1.200870056512596e-05,
+      "loss": 0.5059,
+      "step": 3746
+    },
+    {
+      "epoch": 1.0391014975041597,
+      "grad_norm": 0.18923644721508026,
+      "learning_rate": 1.2003183239172843e-05,
+      "loss": 0.5271,
+      "step": 3747
+    },
+    {
+      "epoch": 1.0393788130892956,
+      "grad_norm": 0.20316655933856964,
+      "learning_rate": 1.1997666010161836e-05,
+      "loss": 0.5138,
+      "step": 3748
+    },
+    {
+      "epoch": 1.0396561286744315,
+      "grad_norm": 0.18475180864334106,
+      "learning_rate": 1.1992148879169499e-05,
+      "loss": 0.4857,
+      "step": 3749
+    },
+    {
+      "epoch": 1.0399334442595674,
+      "grad_norm": 0.18830114603042603,
+      "learning_rate": 1.1986631847272367e-05,
+      "loss": 0.5201,
+      "step": 3750
+    },
+    {
+      "epoch": 1.0402107598447032,
+      "grad_norm": 0.1905670464038849,
+      "learning_rate": 1.1981114915546967e-05,
+      "loss": 0.4813,
+      "step": 3751
+    },
+    {
+      "epoch": 1.0404880754298391,
+      "grad_norm": 0.18634763360023499,
+      "learning_rate": 1.1975598085069798e-05,
+      "loss": 0.53,
+      "step": 3752
+    },
+    {
+      "epoch": 1.040765391014975,
+      "grad_norm": 0.19833336770534515,
+      "learning_rate": 1.197008135691733e-05,
+      "loss": 0.5381,
+      "step": 3753
+    },
+    {
+      "epoch": 1.041042706600111,
+      "grad_norm": 0.19526107609272003,
+      "learning_rate": 1.1964564732166032e-05,
+      "loss": 0.514,
+      "step": 3754
+    },
+    {
+      "epoch": 1.0413200221852468,
+      "grad_norm": 0.1905699223279953,
+      "learning_rate": 1.195904821189234e-05,
+      "loss": 0.5321,
+      "step": 3755
+    },
+    {
+      "epoch": 1.0415973377703827,
+      "grad_norm": 0.203719362616539,
+      "learning_rate": 1.1953531797172673e-05,
+      "loss": 0.5565,
+      "step": 3756
+    },
+    {
+      "epoch": 1.0418746533555185,
+      "grad_norm": 0.1842213273048401,
+      "learning_rate": 1.1948015489083433e-05,
+      "loss": 0.5008,
+      "step": 3757
+    },
+    {
+      "epoch": 1.0421519689406544,
+      "grad_norm": 0.2012074589729309,
+      "learning_rate": 1.1942499288700997e-05,
+      "loss": 0.5199,
+      "step": 3758
+    },
+    {
+      "epoch": 1.0424292845257903,
+      "grad_norm": 0.19444574415683746,
+      "learning_rate": 1.193698319710172e-05,
+      "loss": 0.5228,
+      "step": 3759
+    },
+    {
+      "epoch": 1.0427066001109262,
+      "grad_norm": 0.1847216635942459,
+      "learning_rate": 1.1931467215361934e-05,
+      "loss": 0.4973,
+      "step": 3760
+    },
+    {
+      "epoch": 1.042983915696062,
+      "grad_norm": 0.2011430710554123,
+      "learning_rate": 1.192595134455796e-05,
+      "loss": 0.5167,
+      "step": 3761
+    },
+    {
+      "epoch": 1.043261231281198,
+      "grad_norm": 0.1845628172159195,
+      "learning_rate": 1.192043558576609e-05,
+      "loss": 0.5279,
+      "step": 3762
+    },
+    {
+      "epoch": 1.0435385468663338,
+      "grad_norm": 0.19030199944972992,
+      "learning_rate": 1.1914919940062585e-05,
+      "loss": 0.5217,
+      "step": 3763
+    },
+    {
+      "epoch": 1.0438158624514697,
+      "grad_norm": 0.1975383758544922,
+      "learning_rate": 1.190940440852371e-05,
+      "loss": 0.5218,
+      "step": 3764
+    },
+    {
+      "epoch": 1.0440931780366056,
+      "grad_norm": 0.19011008739471436,
+      "learning_rate": 1.190388899222568e-05,
+      "loss": 0.5273,
+      "step": 3765
+    },
+    {
+      "epoch": 1.0443704936217415,
+      "grad_norm": 0.18708908557891846,
+      "learning_rate": 1.1898373692244699e-05,
+      "loss": 0.5336,
+      "step": 3766
+    },
+    {
+      "epoch": 1.0446478092068774,
+      "grad_norm": 0.18800389766693115,
+      "learning_rate": 1.189285850965695e-05,
+      "loss": 0.5131,
+      "step": 3767
+    },
+    {
+      "epoch": 1.0449251247920133,
+      "grad_norm": 0.187372624874115,
+      "learning_rate": 1.1887343445538597e-05,
+      "loss": 0.5171,
+      "step": 3768
+    },
+    {
+      "epoch": 1.0452024403771492,
+      "grad_norm": 0.19775499403476715,
+      "learning_rate": 1.1881828500965765e-05,
+      "loss": 0.5291,
+      "step": 3769
+    },
+    {
+      "epoch": 1.045479755962285,
+      "grad_norm": 0.19341982901096344,
+      "learning_rate": 1.1876313677014569e-05,
+      "loss": 0.5253,
+      "step": 3770
+    },
+    {
+      "epoch": 1.045757071547421,
+      "grad_norm": 0.19665475189685822,
+      "learning_rate": 1.1870798974761102e-05,
+      "loss": 0.5123,
+      "step": 3771
+    },
+    {
+      "epoch": 1.0460343871325568,
+      "grad_norm": 0.19843867421150208,
+      "learning_rate": 1.1865284395281426e-05,
+      "loss": 0.4966,
+      "step": 3772
+    },
+    {
+      "epoch": 1.0463117027176927,
+      "grad_norm": 0.18300753831863403,
+      "learning_rate": 1.1859769939651582e-05,
+      "loss": 0.5091,
+      "step": 3773
+    },
+    {
+      "epoch": 1.0465890183028286,
+      "grad_norm": 0.1935308873653412,
+      "learning_rate": 1.1854255608947581e-05,
+      "loss": 0.5112,
+      "step": 3774
+    },
+    {
+      "epoch": 1.0468663338879645,
+      "grad_norm": 0.19752877950668335,
+      "learning_rate": 1.1848741404245421e-05,
+      "loss": 0.5062,
+      "step": 3775
+    },
+    {
+      "epoch": 1.0471436494731003,
+      "grad_norm": 0.18157391250133514,
+      "learning_rate": 1.1843227326621069e-05,
+      "loss": 0.5018,
+      "step": 3776
+    },
+    {
+      "epoch": 1.0474209650582362,
+      "grad_norm": 0.20472866296768188,
+      "learning_rate": 1.1837713377150463e-05,
+      "loss": 0.5358,
+      "step": 3777
+    },
+    {
+      "epoch": 1.0476982806433721,
+      "grad_norm": 0.1965123564004898,
+      "learning_rate": 1.1832199556909528e-05,
+      "loss": 0.5269,
+      "step": 3778
+    },
+    {
+      "epoch": 1.047975596228508,
+      "grad_norm": 0.19199901819229126,
+      "learning_rate": 1.1826685866974153e-05,
+      "loss": 0.4876,
+      "step": 3779
+    },
+    {
+      "epoch": 1.0482529118136439,
+      "grad_norm": 0.1838652491569519,
+      "learning_rate": 1.1821172308420203e-05,
+      "loss": 0.542,
+      "step": 3780
+    },
+    {
+      "epoch": 1.0485302273987798,
+      "grad_norm": 0.19129469990730286,
+      "learning_rate": 1.1815658882323519e-05,
+      "loss": 0.4951,
+      "step": 3781
+    },
+    {
+      "epoch": 1.0488075429839157,
+      "grad_norm": 0.18997079133987427,
+      "learning_rate": 1.181014558975992e-05,
+      "loss": 0.4899,
+      "step": 3782
+    },
+    {
+      "epoch": 1.0490848585690515,
+      "grad_norm": 0.19084186851978302,
+      "learning_rate": 1.1804632431805197e-05,
+      "loss": 0.5315,
+      "step": 3783
+    },
+    {
+      "epoch": 1.0493621741541874,
+      "grad_norm": 0.18761594593524933,
+      "learning_rate": 1.1799119409535101e-05,
+      "loss": 0.5063,
+      "step": 3784
+    },
+    {
+      "epoch": 1.0496394897393233,
+      "grad_norm": 0.1877257525920868,
+      "learning_rate": 1.1793606524025388e-05,
+      "loss": 0.506,
+      "step": 3785
+    },
+    {
+      "epoch": 1.0499168053244592,
+      "grad_norm": 0.19573919475078583,
+      "learning_rate": 1.1788093776351752e-05,
+      "loss": 0.5218,
+      "step": 3786
+    },
+    {
+      "epoch": 1.050194120909595,
+      "grad_norm": 0.19596537947654724,
+      "learning_rate": 1.1782581167589883e-05,
+      "loss": 0.5252,
+      "step": 3787
+    },
+    {
+      "epoch": 1.050471436494731,
+      "grad_norm": 0.19336049258708954,
+      "learning_rate": 1.1777068698815434e-05,
+      "loss": 0.5247,
+      "step": 3788
+    },
+    {
+      "epoch": 1.0507487520798668,
+      "grad_norm": 0.18527399003505707,
+      "learning_rate": 1.1771556371104039e-05,
+      "loss": 0.5034,
+      "step": 3789
+    },
+    {
+      "epoch": 1.0510260676650027,
+      "grad_norm": 0.18503272533416748,
+      "learning_rate": 1.1766044185531296e-05,
+      "loss": 0.4683,
+      "step": 3790
+    },
+    {
+      "epoch": 1.0513033832501386,
+      "grad_norm": 0.1880425363779068,
+      "learning_rate": 1.1760532143172772e-05,
+      "loss": 0.5317,
+      "step": 3791
+    },
+    {
+      "epoch": 1.0515806988352745,
+      "grad_norm": 0.18435950577259064,
+      "learning_rate": 1.1755020245104025e-05,
+      "loss": 0.5474,
+      "step": 3792
+    },
+    {
+      "epoch": 1.0518580144204104,
+      "grad_norm": 0.18995283544063568,
+      "learning_rate": 1.1749508492400564e-05,
+      "loss": 0.5066,
+      "step": 3793
+    },
+    {
+      "epoch": 1.0521353300055463,
+      "grad_norm": 0.18789511919021606,
+      "learning_rate": 1.1743996886137882e-05,
+      "loss": 0.5064,
+      "step": 3794
+    },
+    {
+      "epoch": 1.0524126455906821,
+      "grad_norm": 0.17998188734054565,
+      "learning_rate": 1.1738485427391431e-05,
+      "loss": 0.5077,
+      "step": 3795
+    },
+    {
+      "epoch": 1.052689961175818,
+      "grad_norm": 0.21583081781864166,
+      "learning_rate": 1.1732974117236656e-05,
+      "loss": 0.5048,
+      "step": 3796
+    },
+    {
+      "epoch": 1.052967276760954,
+      "grad_norm": 0.21214094758033752,
+      "learning_rate": 1.172746295674895e-05,
+      "loss": 0.5244,
+      "step": 3797
+    },
+    {
+      "epoch": 1.0532445923460898,
+      "grad_norm": 0.1979261040687561,
+      "learning_rate": 1.1721951947003689e-05,
+      "loss": 0.5102,
+      "step": 3798
+    },
+    {
+      "epoch": 1.0535219079312257,
+      "grad_norm": 0.1923878788948059,
+      "learning_rate": 1.1716441089076216e-05,
+      "loss": 0.5165,
+      "step": 3799
+    },
+    {
+      "epoch": 1.0537992235163616,
+      "grad_norm": 0.19521905481815338,
+      "learning_rate": 1.1710930384041852e-05,
+      "loss": 0.5411,
+      "step": 3800
+    },
+    {
+      "epoch": 1.0540765391014975,
+      "grad_norm": 0.1916571408510208,
+      "learning_rate": 1.1705419832975873e-05,
+      "loss": 0.5001,
+      "step": 3801
+    },
+    {
+      "epoch": 1.0543538546866333,
+      "grad_norm": 0.19345450401306152,
+      "learning_rate": 1.1699909436953532e-05,
+      "loss": 0.5371,
+      "step": 3802
+    },
+    {
+      "epoch": 1.0546311702717692,
+      "grad_norm": 0.2184910923242569,
+      "learning_rate": 1.1694399197050062e-05,
+      "loss": 0.5138,
+      "step": 3803
+    },
+    {
+      "epoch": 1.054908485856905,
+      "grad_norm": 0.19563588500022888,
+      "learning_rate": 1.1688889114340653e-05,
+      "loss": 0.51,
+      "step": 3804
+    },
+    {
+      "epoch": 1.055185801442041,
+      "grad_norm": 0.19763770699501038,
+      "learning_rate": 1.1683379189900465e-05,
+      "loss": 0.5183,
+      "step": 3805
+    },
+    {
+      "epoch": 1.0554631170271769,
+      "grad_norm": 0.19711549580097198,
+      "learning_rate": 1.1677869424804637e-05,
+      "loss": 0.4926,
+      "step": 3806
+    },
+    {
+      "epoch": 1.0557404326123128,
+      "grad_norm": 0.2019127458333969,
+      "learning_rate": 1.1672359820128265e-05,
+      "loss": 0.5459,
+      "step": 3807
+    },
+    {
+      "epoch": 1.0560177481974486,
+      "grad_norm": 0.2181474268436432,
+      "learning_rate": 1.166685037694642e-05,
+      "loss": 0.5211,
+      "step": 3808
+    },
+    {
+      "epoch": 1.0562950637825845,
+      "grad_norm": 0.20076914131641388,
+      "learning_rate": 1.1661341096334136e-05,
+      "loss": 0.5064,
+      "step": 3809
+    },
+    {
+      "epoch": 1.0565723793677204,
+      "grad_norm": 0.19240239262580872,
+      "learning_rate": 1.1655831979366427e-05,
+      "loss": 0.5198,
+      "step": 3810
+    },
+    {
+      "epoch": 1.0568496949528563,
+      "grad_norm": 0.19020958244800568,
+      "learning_rate": 1.1650323027118269e-05,
+      "loss": 0.5131,
+      "step": 3811
+    },
+    {
+      "epoch": 1.0571270105379922,
+      "grad_norm": 0.19757165014743805,
+      "learning_rate": 1.1644814240664594e-05,
+      "loss": 0.5075,
+      "step": 3812
+    },
+    {
+      "epoch": 1.057404326123128,
+      "grad_norm": 0.199832484126091,
+      "learning_rate": 1.1639305621080321e-05,
+      "loss": 0.4984,
+      "step": 3813
+    },
+    {
+      "epoch": 1.057681641708264,
+      "grad_norm": 0.19308264553546906,
+      "learning_rate": 1.1633797169440326e-05,
+      "loss": 0.5108,
+      "step": 3814
+    },
+    {
+      "epoch": 1.0579589572933998,
+      "grad_norm": 0.19443538784980774,
+      "learning_rate": 1.1628288886819453e-05,
+      "loss": 0.4896,
+      "step": 3815
+    },
+    {
+      "epoch": 1.0582362728785357,
+      "grad_norm": 0.21316394209861755,
+      "learning_rate": 1.1622780774292506e-05,
+      "loss": 0.5378,
+      "step": 3816
+    },
+    {
+      "epoch": 1.0585135884636716,
+      "grad_norm": 0.1971164047718048,
+      "learning_rate": 1.1617272832934282e-05,
+      "loss": 0.5282,
+      "step": 3817
+    },
+    {
+      "epoch": 1.0587909040488075,
+      "grad_norm": 0.19226108491420746,
+      "learning_rate": 1.161176506381951e-05,
+      "loss": 0.5409,
+      "step": 3818
+    },
+    {
+      "epoch": 1.0590682196339434,
+      "grad_norm": 0.17965470254421234,
+      "learning_rate": 1.1606257468022907e-05,
+      "loss": 0.5022,
+      "step": 3819
+    },
+    {
+      "epoch": 1.0593455352190793,
+      "grad_norm": 0.191642165184021,
+      "learning_rate": 1.1600750046619154e-05,
+      "loss": 0.5034,
+      "step": 3820
+    },
+    {
+      "epoch": 1.0596228508042151,
+      "grad_norm": 0.19330790638923645,
+      "learning_rate": 1.1595242800682893e-05,
+      "loss": 0.5155,
+      "step": 3821
+    },
+    {
+      "epoch": 1.059900166389351,
+      "grad_norm": 0.20066289603710175,
+      "learning_rate": 1.1589735731288725e-05,
+      "loss": 0.5337,
+      "step": 3822
+    },
+    {
+      "epoch": 1.060177481974487,
+      "grad_norm": 0.1891581416130066,
+      "learning_rate": 1.1584228839511242e-05,
+      "loss": 0.5264,
+      "step": 3823
+    },
+    {
+      "epoch": 1.0604547975596228,
+      "grad_norm": 0.19226865470409393,
+      "learning_rate": 1.1578722126424971e-05,
+      "loss": 0.4739,
+      "step": 3824
+    },
+    {
+      "epoch": 1.0607321131447587,
+      "grad_norm": 0.1854773312807083,
+      "learning_rate": 1.1573215593104425e-05,
+      "loss": 0.527,
+      "step": 3825
+    },
+    {
+      "epoch": 1.0610094287298946,
+      "grad_norm": 0.21180889010429382,
+      "learning_rate": 1.1567709240624067e-05,
+      "loss": 0.4695,
+      "step": 3826
+    },
+    {
+      "epoch": 1.0612867443150305,
+      "grad_norm": 0.1897820234298706,
+      "learning_rate": 1.1562203070058341e-05,
+      "loss": 0.4995,
+      "step": 3827
+    },
+    {
+      "epoch": 1.0615640599001663,
+      "grad_norm": 0.1939866989850998,
+      "learning_rate": 1.155669708248164e-05,
+      "loss": 0.5126,
+      "step": 3828
+    },
+    {
+      "epoch": 1.0618413754853022,
+      "grad_norm": 0.22312824428081512,
+      "learning_rate": 1.1551191278968328e-05,
+      "loss": 0.5291,
+      "step": 3829
+    },
+    {
+      "epoch": 1.062118691070438,
+      "grad_norm": 0.19337856769561768,
+      "learning_rate": 1.1545685660592741e-05,
+      "loss": 0.5188,
+      "step": 3830
+    },
+    {
+      "epoch": 1.062396006655574,
+      "grad_norm": 0.19467610120773315,
+      "learning_rate": 1.1540180228429164e-05,
+      "loss": 0.5388,
+      "step": 3831
+    },
+    {
+      "epoch": 1.0626733222407099,
+      "grad_norm": 0.19284315407276154,
+      "learning_rate": 1.1534674983551857e-05,
+      "loss": 0.5011,
+      "step": 3832
+    },
+    {
+      "epoch": 1.0629506378258458,
+      "grad_norm": 0.22472496330738068,
+      "learning_rate": 1.1529169927035028e-05,
+      "loss": 0.5105,
+      "step": 3833
+    },
+    {
+      "epoch": 1.0632279534109816,
+      "grad_norm": 0.18030276894569397,
+      "learning_rate": 1.1523665059952876e-05,
+      "loss": 0.515,
+      "step": 3834
+    },
+    {
+      "epoch": 1.0635052689961175,
+      "grad_norm": 0.19472011923789978,
+      "learning_rate": 1.1518160383379534e-05,
+      "loss": 0.5005,
+      "step": 3835
+    },
+    {
+      "epoch": 1.0637825845812534,
+      "grad_norm": 0.1929817497730255,
+      "learning_rate": 1.1512655898389115e-05,
+      "loss": 0.5114,
+      "step": 3836
+    },
+    {
+      "epoch": 1.0640599001663893,
+      "grad_norm": 0.1856139898300171,
+      "learning_rate": 1.150715160605569e-05,
+      "loss": 0.4847,
+      "step": 3837
+    },
+    {
+      "epoch": 1.0643372157515252,
+      "grad_norm": 0.1880098432302475,
+      "learning_rate": 1.1501647507453295e-05,
+      "loss": 0.5346,
+      "step": 3838
+    },
+    {
+      "epoch": 1.064614531336661,
+      "grad_norm": 0.18800011277198792,
+      "learning_rate": 1.149614360365592e-05,
+      "loss": 0.5113,
+      "step": 3839
+    },
+    {
+      "epoch": 1.064891846921797,
+      "grad_norm": 0.19980834424495697,
+      "learning_rate": 1.1490639895737523e-05,
+      "loss": 0.5214,
+      "step": 3840
+    },
+    {
+      "epoch": 1.0651691625069328,
+      "grad_norm": 0.18565940856933594,
+      "learning_rate": 1.1485136384772024e-05,
+      "loss": 0.5108,
+      "step": 3841
+    },
+    {
+      "epoch": 1.0654464780920687,
+      "grad_norm": 0.18709343671798706,
+      "learning_rate": 1.1479633071833306e-05,
+      "loss": 0.5071,
+      "step": 3842
+    },
+    {
+      "epoch": 1.0657237936772046,
+      "grad_norm": 0.19222643971443176,
+      "learning_rate": 1.1474129957995209e-05,
+      "loss": 0.5084,
+      "step": 3843
+    },
+    {
+      "epoch": 1.0660011092623405,
+      "grad_norm": 0.18951910734176636,
+      "learning_rate": 1.146862704433154e-05,
+      "loss": 0.4997,
+      "step": 3844
+    },
+    {
+      "epoch": 1.0662784248474764,
+      "grad_norm": 0.19320812821388245,
+      "learning_rate": 1.146312433191606e-05,
+      "loss": 0.5397,
+      "step": 3845
+    },
+    {
+      "epoch": 1.0665557404326123,
+      "grad_norm": 0.19874915480613708,
+      "learning_rate": 1.1457621821822492e-05,
+      "loss": 0.4976,
+      "step": 3846
+    },
+    {
+      "epoch": 1.0668330560177481,
+      "grad_norm": 0.19041554629802704,
+      "learning_rate": 1.1452119515124524e-05,
+      "loss": 0.5293,
+      "step": 3847
+    },
+    {
+      "epoch": 1.067110371602884,
+      "grad_norm": 0.18619757890701294,
+      "learning_rate": 1.1446617412895802e-05,
+      "loss": 0.5008,
+      "step": 3848
+    },
+    {
+      "epoch": 1.06738768718802,
+      "grad_norm": 0.18963642418384552,
+      "learning_rate": 1.1441115516209936e-05,
+      "loss": 0.5099,
+      "step": 3849
+    },
+    {
+      "epoch": 1.0676650027731558,
+      "grad_norm": 0.1950794905424118,
+      "learning_rate": 1.143561382614048e-05,
+      "loss": 0.5103,
+      "step": 3850
+    },
+    {
+      "epoch": 1.0679423183582917,
+      "grad_norm": 0.1896001100540161,
+      "learning_rate": 1.1430112343760971e-05,
+      "loss": 0.529,
+      "step": 3851
+    },
+    {
+      "epoch": 1.0682196339434276,
+      "grad_norm": 0.20062585175037384,
+      "learning_rate": 1.142461107014489e-05,
+      "loss": 0.5222,
+      "step": 3852
+    },
+    {
+      "epoch": 1.0684969495285634,
+      "grad_norm": 0.19482149183750153,
+      "learning_rate": 1.1419110006365682e-05,
+      "loss": 0.497,
+      "step": 3853
+    },
+    {
+      "epoch": 1.0687742651136993,
+      "grad_norm": 0.1913536936044693,
+      "learning_rate": 1.1413609153496742e-05,
+      "loss": 0.5164,
+      "step": 3854
+    },
+    {
+      "epoch": 1.0690515806988352,
+      "grad_norm": 0.20729845762252808,
+      "learning_rate": 1.140810851261145e-05,
+      "loss": 0.5594,
+      "step": 3855
+    },
+    {
+      "epoch": 1.069328896283971,
+      "grad_norm": 0.20025323331356049,
+      "learning_rate": 1.1402608084783112e-05,
+      "loss": 0.4986,
+      "step": 3856
+    },
+    {
+      "epoch": 1.069606211869107,
+      "grad_norm": 0.21821513772010803,
+      "learning_rate": 1.1397107871085009e-05,
+      "loss": 0.5312,
+      "step": 3857
+    },
+    {
+      "epoch": 1.0698835274542429,
+      "grad_norm": 0.19112320244312286,
+      "learning_rate": 1.1391607872590381e-05,
+      "loss": 0.4957,
+      "step": 3858
+    },
+    {
+      "epoch": 1.0701608430393788,
+      "grad_norm": 0.18980076909065247,
+      "learning_rate": 1.138610809037243e-05,
+      "loss": 0.5235,
+      "step": 3859
+    },
+    {
+      "epoch": 1.0704381586245146,
+      "grad_norm": 0.18163329362869263,
+      "learning_rate": 1.1380608525504298e-05,
+      "loss": 0.4966,
+      "step": 3860
+    },
+    {
+      "epoch": 1.0707154742096505,
+      "grad_norm": 0.19191473722457886,
+      "learning_rate": 1.1375109179059098e-05,
+      "loss": 0.5233,
+      "step": 3861
+    },
+    {
+      "epoch": 1.0709927897947864,
+      "grad_norm": 0.18972338736057281,
+      "learning_rate": 1.1369610052109902e-05,
+      "loss": 0.516,
+      "step": 3862
+    },
+    {
+      "epoch": 1.0712701053799223,
+      "grad_norm": 0.1891545057296753,
+      "learning_rate": 1.1364111145729737e-05,
+      "loss": 0.5029,
+      "step": 3863
+    },
+    {
+      "epoch": 1.0715474209650582,
+      "grad_norm": 0.1886187046766281,
+      "learning_rate": 1.1358612460991577e-05,
+      "loss": 0.5157,
+      "step": 3864
+    },
+    {
+      "epoch": 1.071824736550194,
+      "grad_norm": 0.2039126604795456,
+      "learning_rate": 1.1353113998968371e-05,
+      "loss": 0.535,
+      "step": 3865
+    },
+    {
+      "epoch": 1.07210205213533,
+      "grad_norm": 0.19889415800571442,
+      "learning_rate": 1.134761576073301e-05,
+      "loss": 0.5235,
+      "step": 3866
+    },
+    {
+      "epoch": 1.0723793677204658,
+      "grad_norm": 0.18273714184761047,
+      "learning_rate": 1.1342117747358344e-05,
+      "loss": 0.5083,
+      "step": 3867
+    },
+    {
+      "epoch": 1.0726566833056017,
+      "grad_norm": 0.19936375319957733,
+      "learning_rate": 1.1336619959917182e-05,
+      "loss": 0.508,
+      "step": 3868
+    },
+    {
+      "epoch": 1.0729339988907376,
+      "grad_norm": 0.19671058654785156,
+      "learning_rate": 1.133112239948229e-05,
+      "loss": 0.5239,
+      "step": 3869
+    },
+    {
+      "epoch": 1.0732113144758735,
+      "grad_norm": 0.18869206309318542,
+      "learning_rate": 1.132562506712639e-05,
+      "loss": 0.5204,
+      "step": 3870
+    },
+    {
+      "epoch": 1.0734886300610094,
+      "grad_norm": 0.18764182925224304,
+      "learning_rate": 1.132012796392215e-05,
+      "loss": 0.4993,
+      "step": 3871
+    },
+    {
+      "epoch": 1.0737659456461452,
+      "grad_norm": 0.21141819655895233,
+      "learning_rate": 1.1314631090942204e-05,
+      "loss": 0.4947,
+      "step": 3872
+    },
+    {
+      "epoch": 1.0740432612312811,
+      "grad_norm": 0.18649768829345703,
+      "learning_rate": 1.130913444925914e-05,
+      "loss": 0.5258,
+      "step": 3873
+    },
+    {
+      "epoch": 1.074320576816417,
+      "grad_norm": 0.18479731678962708,
+      "learning_rate": 1.1303638039945498e-05,
+      "loss": 0.4958,
+      "step": 3874
+    },
+    {
+      "epoch": 1.074597892401553,
+      "grad_norm": 0.20172236859798431,
+      "learning_rate": 1.1298141864073763e-05,
+      "loss": 0.5181,
+      "step": 3875
+    },
+    {
+      "epoch": 1.0748752079866888,
+      "grad_norm": 0.194131001830101,
+      "learning_rate": 1.1292645922716404e-05,
+      "loss": 0.538,
+      "step": 3876
+    },
+    {
+      "epoch": 1.0751525235718247,
+      "grad_norm": 0.18898829817771912,
+      "learning_rate": 1.1287150216945808e-05,
+      "loss": 0.4733,
+      "step": 3877
+    },
+    {
+      "epoch": 1.0754298391569606,
+      "grad_norm": 0.19593171775341034,
+      "learning_rate": 1.1281654747834337e-05,
+      "loss": 0.5304,
+      "step": 3878
+    },
+    {
+      "epoch": 1.0757071547420964,
+      "grad_norm": 0.1899636685848236,
+      "learning_rate": 1.1276159516454308e-05,
+      "loss": 0.5191,
+      "step": 3879
+    },
+    {
+      "epoch": 1.0759844703272323,
+      "grad_norm": 0.1919141411781311,
+      "learning_rate": 1.1270664523877982e-05,
+      "loss": 0.5044,
+      "step": 3880
+    },
+    {
+      "epoch": 1.0762617859123682,
+      "grad_norm": 0.1983211487531662,
+      "learning_rate": 1.1265169771177573e-05,
+      "loss": 0.4931,
+      "step": 3881
+    },
+    {
+      "epoch": 1.076539101497504,
+      "grad_norm": 0.1884281188249588,
+      "learning_rate": 1.1259675259425263e-05,
+      "loss": 0.5212,
+      "step": 3882
+    },
+    {
+      "epoch": 1.07681641708264,
+      "grad_norm": 0.18257342278957367,
+      "learning_rate": 1.125418098969317e-05,
+      "loss": 0.4964,
+      "step": 3883
+    },
+    {
+      "epoch": 1.0770937326677759,
+      "grad_norm": 0.19225247204303741,
+      "learning_rate": 1.1248686963053374e-05,
+      "loss": 0.5216,
+      "step": 3884
+    },
+    {
+      "epoch": 1.0773710482529117,
+      "grad_norm": 0.21373282372951508,
+      "learning_rate": 1.1243193180577902e-05,
+      "loss": 0.5106,
+      "step": 3885
+    },
+    {
+      "epoch": 1.0776483638380476,
+      "grad_norm": 0.21381765604019165,
+      "learning_rate": 1.123769964333874e-05,
+      "loss": 0.4909,
+      "step": 3886
+    },
+    {
+      "epoch": 1.0779256794231835,
+      "grad_norm": 0.18304717540740967,
+      "learning_rate": 1.1232206352407828e-05,
+      "loss": 0.5087,
+      "step": 3887
+    },
+    {
+      "epoch": 1.0782029950083194,
+      "grad_norm": 0.18388831615447998,
+      "learning_rate": 1.1226713308857036e-05,
+      "loss": 0.4925,
+      "step": 3888
+    },
+    {
+      "epoch": 1.0784803105934553,
+      "grad_norm": 0.19412866234779358,
+      "learning_rate": 1.1221220513758219e-05,
+      "loss": 0.5099,
+      "step": 3889
+    },
+    {
+      "epoch": 1.0787576261785912,
+      "grad_norm": 0.19013476371765137,
+      "learning_rate": 1.1215727968183159e-05,
+      "loss": 0.5227,
+      "step": 3890
+    },
+    {
+      "epoch": 1.079034941763727,
+      "grad_norm": 0.19106021523475647,
+      "learning_rate": 1.1210235673203601e-05,
+      "loss": 0.5123,
+      "step": 3891
+    },
+    {
+      "epoch": 1.079312257348863,
+      "grad_norm": 0.18381288647651672,
+      "learning_rate": 1.1204743629891225e-05,
+      "loss": 0.5182,
+      "step": 3892
+    },
+    {
+      "epoch": 1.0795895729339988,
+      "grad_norm": 0.19413931667804718,
+      "learning_rate": 1.1199251839317696e-05,
+      "loss": 0.5014,
+      "step": 3893
+    },
+    {
+      "epoch": 1.0798668885191347,
+      "grad_norm": 0.18920141458511353,
+      "learning_rate": 1.119376030255459e-05,
+      "loss": 0.5136,
+      "step": 3894
+    },
+    {
+      "epoch": 1.0801442041042706,
+      "grad_norm": 0.19880147278308868,
+      "learning_rate": 1.1188269020673456e-05,
+      "loss": 0.5539,
+      "step": 3895
+    },
+    {
+      "epoch": 1.0804215196894065,
+      "grad_norm": 0.1854148656129837,
+      "learning_rate": 1.118277799474579e-05,
+      "loss": 0.4911,
+      "step": 3896
+    },
+    {
+      "epoch": 1.0806988352745424,
+      "grad_norm": 0.18824362754821777,
+      "learning_rate": 1.1177287225843041e-05,
+      "loss": 0.4769,
+      "step": 3897
+    },
+    {
+      "epoch": 1.0809761508596782,
+      "grad_norm": 0.1937967985868454,
+      "learning_rate": 1.1171796715036597e-05,
+      "loss": 0.5147,
+      "step": 3898
+    },
+    {
+      "epoch": 1.0812534664448141,
+      "grad_norm": 0.18958862125873566,
+      "learning_rate": 1.11663064633978e-05,
+      "loss": 0.5197,
+      "step": 3899
+    },
+    {
+      "epoch": 1.08153078202995,
+      "grad_norm": 0.1861356496810913,
+      "learning_rate": 1.1160816471997951e-05,
+      "loss": 0.5102,
+      "step": 3900
+    },
+    {
+      "epoch": 1.081808097615086,
+      "grad_norm": 0.19085288047790527,
+      "learning_rate": 1.115532674190829e-05,
+      "loss": 0.4957,
+      "step": 3901
+    },
+    {
+      "epoch": 1.0820854132002218,
+      "grad_norm": 0.19016428291797638,
+      "learning_rate": 1.1149837274200004e-05,
+      "loss": 0.523,
+      "step": 3902
+    },
+    {
+      "epoch": 1.0823627287853577,
+      "grad_norm": 0.18524803221225739,
+      "learning_rate": 1.1144348069944244e-05,
+      "loss": 0.4999,
+      "step": 3903
+    },
+    {
+      "epoch": 1.0826400443704935,
+      "grad_norm": 0.18277236819267273,
+      "learning_rate": 1.1138859130212089e-05,
+      "loss": 0.5139,
+      "step": 3904
+    },
+    {
+      "epoch": 1.0829173599556294,
+      "grad_norm": 0.18841886520385742,
+      "learning_rate": 1.113337045607458e-05,
+      "loss": 0.5124,
+      "step": 3905
+    },
+    {
+      "epoch": 1.0831946755407653,
+      "grad_norm": 0.19280953705310822,
+      "learning_rate": 1.1127882048602703e-05,
+      "loss": 0.4953,
+      "step": 3906
+    },
+    {
+      "epoch": 1.0834719911259012,
+      "grad_norm": 0.19617757201194763,
+      "learning_rate": 1.1122393908867392e-05,
+      "loss": 0.5144,
+      "step": 3907
+    },
+    {
+      "epoch": 1.083749306711037,
+      "grad_norm": 0.19670793414115906,
+      "learning_rate": 1.1116906037939532e-05,
+      "loss": 0.5048,
+      "step": 3908
+    },
+    {
+      "epoch": 1.084026622296173,
+      "grad_norm": 0.18718746304512024,
+      "learning_rate": 1.1111418436889944e-05,
+      "loss": 0.5164,
+      "step": 3909
+    },
+    {
+      "epoch": 1.0843039378813089,
+      "grad_norm": 0.18242870271205902,
+      "learning_rate": 1.110593110678941e-05,
+      "loss": 0.5263,
+      "step": 3910
+    },
+    {
+      "epoch": 1.0845812534664447,
+      "grad_norm": 0.1876215934753418,
+      "learning_rate": 1.1100444048708653e-05,
+      "loss": 0.504,
+      "step": 3911
+    },
+    {
+      "epoch": 1.0848585690515806,
+      "grad_norm": 0.19080670177936554,
+      "learning_rate": 1.1094957263718345e-05,
+      "loss": 0.5236,
+      "step": 3912
+    },
+    {
+      "epoch": 1.0851358846367165,
+      "grad_norm": 0.194438174366951,
+      "learning_rate": 1.1089470752889093e-05,
+      "loss": 0.5175,
+      "step": 3913
+    },
+    {
+      "epoch": 1.0854132002218524,
+      "grad_norm": 0.19020043313503265,
+      "learning_rate": 1.1083984517291476e-05,
+      "loss": 0.5164,
+      "step": 3914
+    },
+    {
+      "epoch": 1.0856905158069883,
+      "grad_norm": 0.1921057403087616,
+      "learning_rate": 1.1078498557995995e-05,
+      "loss": 0.5049,
+      "step": 3915
+    },
+    {
+      "epoch": 1.0859678313921242,
+      "grad_norm": 0.19786351919174194,
+      "learning_rate": 1.1073012876073103e-05,
+      "loss": 0.5317,
+      "step": 3916
+    },
+    {
+      "epoch": 1.08624514697726,
+      "grad_norm": 0.1904228925704956,
+      "learning_rate": 1.106752747259321e-05,
+      "loss": 0.5635,
+      "step": 3917
+    },
+    {
+      "epoch": 1.086522462562396,
+      "grad_norm": 0.1912548840045929,
+      "learning_rate": 1.106204234862666e-05,
+      "loss": 0.5369,
+      "step": 3918
+    },
+    {
+      "epoch": 1.0867997781475318,
+      "grad_norm": 0.1982937604188919,
+      "learning_rate": 1.1056557505243746e-05,
+      "loss": 0.5137,
+      "step": 3919
+    },
+    {
+      "epoch": 1.0870770937326677,
+      "grad_norm": 0.19235102832317352,
+      "learning_rate": 1.1051072943514703e-05,
+      "loss": 0.5317,
+      "step": 3920
+    },
+    {
+      "epoch": 1.0873544093178036,
+      "grad_norm": 0.2086210548877716,
+      "learning_rate": 1.1045588664509717e-05,
+      "loss": 0.4916,
+      "step": 3921
+    },
+    {
+      "epoch": 1.0876317249029395,
+      "grad_norm": 0.19181272387504578,
+      "learning_rate": 1.104010466929892e-05,
+      "loss": 0.5142,
+      "step": 3922
+    },
+    {
+      "epoch": 1.0879090404880754,
+      "grad_norm": 0.18342861533164978,
+      "learning_rate": 1.1034620958952377e-05,
+      "loss": 0.5328,
+      "step": 3923
+    },
+    {
+      "epoch": 1.0881863560732112,
+      "grad_norm": 0.18724067509174347,
+      "learning_rate": 1.1029137534540113e-05,
+      "loss": 0.4951,
+      "step": 3924
+    },
+    {
+      "epoch": 1.0884636716583471,
+      "grad_norm": 0.19535782933235168,
+      "learning_rate": 1.1023654397132087e-05,
+      "loss": 0.498,
+      "step": 3925
+    },
+    {
+      "epoch": 1.088740987243483,
+      "grad_norm": 0.24757403135299683,
+      "learning_rate": 1.10181715477982e-05,
+      "loss": 0.4967,
+      "step": 3926
+    },
+    {
+      "epoch": 1.089018302828619,
+      "grad_norm": 0.1959121972322464,
+      "learning_rate": 1.1012688987608303e-05,
+      "loss": 0.5174,
+      "step": 3927
+    },
+    {
+      "epoch": 1.0892956184137548,
+      "grad_norm": 0.20022380352020264,
+      "learning_rate": 1.1007206717632193e-05,
+      "loss": 0.5224,
+      "step": 3928
+    },
+    {
+      "epoch": 1.0895729339988907,
+      "grad_norm": 0.18869644403457642,
+      "learning_rate": 1.1001724738939606e-05,
+      "loss": 0.5269,
+      "step": 3929
+    },
+    {
+      "epoch": 1.0898502495840265,
+      "grad_norm": 0.19060127437114716,
+      "learning_rate": 1.099624305260021e-05,
+      "loss": 0.517,
+      "step": 3930
+    },
+    {
+      "epoch": 1.0901275651691624,
+      "grad_norm": 0.18450330197811127,
+      "learning_rate": 1.0990761659683643e-05,
+      "loss": 0.5134,
+      "step": 3931
+    },
+    {
+      "epoch": 1.0904048807542983,
+      "grad_norm": 0.20092949271202087,
+      "learning_rate": 1.0985280561259462e-05,
+      "loss": 0.5229,
+      "step": 3932
+    },
+    {
+      "epoch": 1.0906821963394342,
+      "grad_norm": 0.187063530087471,
+      "learning_rate": 1.0979799758397173e-05,
+      "loss": 0.5074,
+      "step": 3933
+    },
+    {
+      "epoch": 1.09095951192457,
+      "grad_norm": 0.21675853431224823,
+      "learning_rate": 1.0974319252166226e-05,
+      "loss": 0.5443,
+      "step": 3934
+    },
+    {
+      "epoch": 1.091236827509706,
+      "grad_norm": 0.19182507693767548,
+      "learning_rate": 1.0968839043636021e-05,
+      "loss": 0.5165,
+      "step": 3935
+    },
+    {
+      "epoch": 1.0915141430948418,
+      "grad_norm": 0.19931842386722565,
+      "learning_rate": 1.0963359133875884e-05,
+      "loss": 0.5362,
+      "step": 3936
+    },
+    {
+      "epoch": 1.0917914586799777,
+      "grad_norm": 0.19970446825027466,
+      "learning_rate": 1.0957879523955087e-05,
+      "loss": 0.5175,
+      "step": 3937
+    },
+    {
+      "epoch": 1.0920687742651136,
+      "grad_norm": 0.18823319673538208,
+      "learning_rate": 1.0952400214942857e-05,
+      "loss": 0.5184,
+      "step": 3938
+    },
+    {
+      "epoch": 1.0923460898502495,
+      "grad_norm": 0.19610384106636047,
+      "learning_rate": 1.094692120790835e-05,
+      "loss": 0.5194,
+      "step": 3939
+    },
+    {
+      "epoch": 1.0926234054353854,
+      "grad_norm": 0.2017711102962494,
+      "learning_rate": 1.0941442503920664e-05,
+      "loss": 0.5254,
+      "step": 3940
+    },
+    {
+      "epoch": 1.0929007210205213,
+      "grad_norm": 0.18085281550884247,
+      "learning_rate": 1.0935964104048834e-05,
+      "loss": 0.4865,
+      "step": 3941
+    },
+    {
+      "epoch": 1.0931780366056572,
+      "grad_norm": 0.18639199435710907,
+      "learning_rate": 1.0930486009361847e-05,
+      "loss": 0.4902,
+      "step": 3942
+    },
+    {
+      "epoch": 1.093455352190793,
+      "grad_norm": 0.1943678855895996,
+      "learning_rate": 1.0925008220928624e-05,
+      "loss": 0.5216,
+      "step": 3943
+    },
+    {
+      "epoch": 1.093732667775929,
+      "grad_norm": 0.19227345287799835,
+      "learning_rate": 1.0919530739818022e-05,
+      "loss": 0.4983,
+      "step": 3944
+    },
+    {
+      "epoch": 1.0940099833610648,
+      "grad_norm": 0.18952839076519012,
+      "learning_rate": 1.091405356709885e-05,
+      "loss": 0.496,
+      "step": 3945
+    },
+    {
+      "epoch": 1.0942872989462007,
+      "grad_norm": 0.20736488699913025,
+      "learning_rate": 1.090857670383985e-05,
+      "loss": 0.5296,
+      "step": 3946
+    },
+    {
+      "epoch": 1.0945646145313366,
+      "grad_norm": 0.1944408416748047,
+      "learning_rate": 1.090310015110969e-05,
+      "loss": 0.5024,
+      "step": 3947
+    },
+    {
+      "epoch": 1.0948419301164725,
+      "grad_norm": 0.2053939253091812,
+      "learning_rate": 1.0897623909977006e-05,
+      "loss": 0.4962,
+      "step": 3948
+    },
+    {
+      "epoch": 1.0951192457016083,
+      "grad_norm": 0.200760155916214,
+      "learning_rate": 1.089214798151035e-05,
+      "loss": 0.5123,
+      "step": 3949
+    },
+    {
+      "epoch": 1.0953965612867442,
+      "grad_norm": 0.19097808003425598,
+      "learning_rate": 1.0886672366778224e-05,
+      "loss": 0.4831,
+      "step": 3950
+    },
+    {
+      "epoch": 1.0956738768718801,
+      "grad_norm": 0.1900891810655594,
+      "learning_rate": 1.0881197066849055e-05,
+      "loss": 0.5154,
+      "step": 3951
+    },
+    {
+      "epoch": 1.095951192457016,
+      "grad_norm": 0.19628465175628662,
+      "learning_rate": 1.0875722082791237e-05,
+      "loss": 0.4804,
+      "step": 3952
+    },
+    {
+      "epoch": 1.0962285080421519,
+      "grad_norm": 0.18627870082855225,
+      "learning_rate": 1.0870247415673072e-05,
+      "loss": 0.4819,
+      "step": 3953
+    },
+    {
+      "epoch": 1.0965058236272878,
+      "grad_norm": 0.1861443817615509,
+      "learning_rate": 1.0864773066562814e-05,
+      "loss": 0.4924,
+      "step": 3954
+    },
+    {
+      "epoch": 1.0967831392124237,
+      "grad_norm": 0.2025613784790039,
+      "learning_rate": 1.0859299036528657e-05,
+      "loss": 0.5246,
+      "step": 3955
+    },
+    {
+      "epoch": 1.0970604547975595,
+      "grad_norm": 0.18845801055431366,
+      "learning_rate": 1.0853825326638731e-05,
+      "loss": 0.499,
+      "step": 3956
+    },
+    {
+      "epoch": 1.0973377703826954,
+      "grad_norm": 0.1874900460243225,
+      "learning_rate": 1.0848351937961094e-05,
+      "loss": 0.5166,
+      "step": 3957
+    },
+    {
+      "epoch": 1.0976150859678313,
+      "grad_norm": 0.20147672295570374,
+      "learning_rate": 1.0842878871563752e-05,
+      "loss": 0.5236,
+      "step": 3958
+    },
+    {
+      "epoch": 1.0978924015529672,
+      "grad_norm": 0.2169303447008133,
+      "learning_rate": 1.083740612851465e-05,
+      "loss": 0.4983,
+      "step": 3959
+    },
+    {
+      "epoch": 1.098169717138103,
+      "grad_norm": 0.183817520737648,
+      "learning_rate": 1.083193370988166e-05,
+      "loss": 0.5103,
+      "step": 3960
+    },
+    {
+      "epoch": 1.098447032723239,
+      "grad_norm": 0.2015179991722107,
+      "learning_rate": 1.0826461616732596e-05,
+      "loss": 0.4943,
+      "step": 3961
+    },
+    {
+      "epoch": 1.0987243483083748,
+      "grad_norm": 0.19928953051567078,
+      "learning_rate": 1.0820989850135216e-05,
+      "loss": 0.5246,
+      "step": 3962
+    },
+    {
+      "epoch": 1.0990016638935107,
+      "grad_norm": 0.18814896047115326,
+      "learning_rate": 1.0815518411157198e-05,
+      "loss": 0.5284,
+      "step": 3963
+    },
+    {
+      "epoch": 1.0992789794786466,
+      "grad_norm": 0.20342199504375458,
+      "learning_rate": 1.0810047300866166e-05,
+      "loss": 0.5014,
+      "step": 3964
+    },
+    {
+      "epoch": 1.0995562950637825,
+      "grad_norm": 0.19590109586715698,
+      "learning_rate": 1.0804576520329679e-05,
+      "loss": 0.5148,
+      "step": 3965
+    },
+    {
+      "epoch": 1.0998336106489184,
+      "grad_norm": 0.1944301873445511,
+      "learning_rate": 1.0799106070615235e-05,
+      "loss": 0.4809,
+      "step": 3966
+    },
+    {
+      "epoch": 1.1001109262340543,
+      "grad_norm": 0.19030778110027313,
+      "learning_rate": 1.0793635952790264e-05,
+      "loss": 0.5093,
+      "step": 3967
+    },
+    {
+      "epoch": 1.1003882418191901,
+      "grad_norm": 0.1883806735277176,
+      "learning_rate": 1.0788166167922118e-05,
+      "loss": 0.5018,
+      "step": 3968
+    },
+    {
+      "epoch": 1.100665557404326,
+      "grad_norm": 0.21474680304527283,
+      "learning_rate": 1.0782696717078117e-05,
+      "loss": 0.5091,
+      "step": 3969
+    },
+    {
+      "epoch": 1.100942872989462,
+      "grad_norm": 0.24823465943336487,
+      "learning_rate": 1.0777227601325482e-05,
+      "loss": 0.5073,
+      "step": 3970
+    },
+    {
+      "epoch": 1.1012201885745978,
+      "grad_norm": 0.18175634741783142,
+      "learning_rate": 1.0771758821731386e-05,
+      "loss": 0.4993,
+      "step": 3971
+    },
+    {
+      "epoch": 1.1014975041597337,
+      "grad_norm": 0.18903425335884094,
+      "learning_rate": 1.0766290379362928e-05,
+      "loss": 0.5283,
+      "step": 3972
+    },
+    {
+      "epoch": 1.1017748197448696,
+      "grad_norm": 0.19946777820587158,
+      "learning_rate": 1.0760822275287159e-05,
+      "loss": 0.5284,
+      "step": 3973
+    },
+    {
+      "epoch": 1.1020521353300055,
+      "grad_norm": 0.19179320335388184,
+      "learning_rate": 1.075535451057104e-05,
+      "loss": 0.5264,
+      "step": 3974
+    },
+    {
+      "epoch": 1.1023294509151413,
+      "grad_norm": 0.18837237358093262,
+      "learning_rate": 1.0749887086281474e-05,
+      "loss": 0.5051,
+      "step": 3975
+    },
+    {
+      "epoch": 1.1026067665002772,
+      "grad_norm": 0.19744771718978882,
+      "learning_rate": 1.0744420003485312e-05,
+      "loss": 0.5313,
+      "step": 3976
+    },
+    {
+      "epoch": 1.102884082085413,
+      "grad_norm": 0.20925208926200867,
+      "learning_rate": 1.0738953263249319e-05,
+      "loss": 0.5033,
+      "step": 3977
+    },
+    {
+      "epoch": 1.103161397670549,
+      "grad_norm": 0.19691763818264008,
+      "learning_rate": 1.0733486866640203e-05,
+      "loss": 0.5205,
+      "step": 3978
+    },
+    {
+      "epoch": 1.1034387132556849,
+      "grad_norm": 0.19739267230033875,
+      "learning_rate": 1.07280208147246e-05,
+      "loss": 0.4742,
+      "step": 3979
+    },
+    {
+      "epoch": 1.1037160288408208,
+      "grad_norm": 0.1945018321275711,
+      "learning_rate": 1.0722555108569085e-05,
+      "loss": 0.4977,
+      "step": 3980
+    },
+    {
+      "epoch": 1.1039933444259566,
+      "grad_norm": 0.19838854670524597,
+      "learning_rate": 1.071708974924016e-05,
+      "loss": 0.5137,
+      "step": 3981
+    },
+    {
+      "epoch": 1.1042706600110925,
+      "grad_norm": 0.21607692539691925,
+      "learning_rate": 1.071162473780426e-05,
+      "loss": 0.5166,
+      "step": 3982
+    },
+    {
+      "epoch": 1.1045479755962284,
+      "grad_norm": 0.17799584567546844,
+      "learning_rate": 1.0706160075327761e-05,
+      "loss": 0.4938,
+      "step": 3983
+    },
+    {
+      "epoch": 1.1048252911813643,
+      "grad_norm": 0.1956464648246765,
+      "learning_rate": 1.0700695762876958e-05,
+      "loss": 0.5244,
+      "step": 3984
+    },
+    {
+      "epoch": 1.1051026067665002,
+      "grad_norm": 0.18592742085456848,
+      "learning_rate": 1.0695231801518083e-05,
+      "loss": 0.466,
+      "step": 3985
+    },
+    {
+      "epoch": 1.105379922351636,
+      "grad_norm": 0.19382350146770477,
+      "learning_rate": 1.0689768192317296e-05,
+      "loss": 0.4924,
+      "step": 3986
+    },
+    {
+      "epoch": 1.105657237936772,
+      "grad_norm": 0.2073841392993927,
+      "learning_rate": 1.0684304936340697e-05,
+      "loss": 0.5109,
+      "step": 3987
+    },
+    {
+      "epoch": 1.1059345535219078,
+      "grad_norm": 0.19084088504314423,
+      "learning_rate": 1.0678842034654315e-05,
+      "loss": 0.5137,
+      "step": 3988
+    },
+    {
+      "epoch": 1.1062118691070437,
+      "grad_norm": 0.1962527185678482,
+      "learning_rate": 1.0673379488324095e-05,
+      "loss": 0.5245,
+      "step": 3989
+    },
+    {
+      "epoch": 1.1064891846921796,
+      "grad_norm": 0.18985575437545776,
+      "learning_rate": 1.066791729841594e-05,
+      "loss": 0.4747,
+      "step": 3990
+    },
+    {
+      "epoch": 1.1067665002773155,
+      "grad_norm": 0.2049971967935562,
+      "learning_rate": 1.0662455465995657e-05,
+      "loss": 0.4954,
+      "step": 3991
+    },
+    {
+      "epoch": 1.1070438158624514,
+      "grad_norm": 0.19408412277698517,
+      "learning_rate": 1.0656993992128999e-05,
+      "loss": 0.5252,
+      "step": 3992
+    },
+    {
+      "epoch": 1.1073211314475873,
+      "grad_norm": 0.19346264004707336,
+      "learning_rate": 1.0651532877881639e-05,
+      "loss": 0.4995,
+      "step": 3993
+    },
+    {
+      "epoch": 1.1075984470327231,
+      "grad_norm": 0.19205859303474426,
+      "learning_rate": 1.0646072124319193e-05,
+      "loss": 0.5091,
+      "step": 3994
+    },
+    {
+      "epoch": 1.107875762617859,
+      "grad_norm": 0.20613279938697815,
+      "learning_rate": 1.0640611732507192e-05,
+      "loss": 0.5191,
+      "step": 3995
+    },
+    {
+      "epoch": 1.108153078202995,
+      "grad_norm": 0.1821645051240921,
+      "learning_rate": 1.0635151703511104e-05,
+      "loss": 0.493,
+      "step": 3996
+    },
+    {
+      "epoch": 1.1084303937881308,
+      "grad_norm": 0.20445798337459564,
+      "learning_rate": 1.062969203839633e-05,
+      "loss": 0.5365,
+      "step": 3997
+    },
+    {
+      "epoch": 1.1087077093732667,
+      "grad_norm": 0.19484387338161469,
+      "learning_rate": 1.062423273822819e-05,
+      "loss": 0.5099,
+      "step": 3998
+    },
+    {
+      "epoch": 1.1089850249584026,
+      "grad_norm": 0.19972245395183563,
+      "learning_rate": 1.0618773804071943e-05,
+      "loss": 0.5233,
+      "step": 3999
+    },
+    {
+      "epoch": 1.1092623405435384,
+      "grad_norm": 0.20230787992477417,
+      "learning_rate": 1.0613315236992766e-05,
+      "loss": 0.524,
+      "step": 4000
+    },
+    {
+      "epoch": 1.1095396561286743,
+      "grad_norm": 0.18941400945186615,
+      "learning_rate": 1.0607857038055774e-05,
+      "loss": 0.4988,
+      "step": 4001
+    },
+    {
+      "epoch": 1.1098169717138102,
+      "grad_norm": 0.1921200454235077,
+      "learning_rate": 1.0602399208326006e-05,
+      "loss": 0.5355,
+      "step": 4002
+    },
+    {
+      "epoch": 1.110094287298946,
+      "grad_norm": 0.1832018345594406,
+      "learning_rate": 1.0596941748868426e-05,
+      "loss": 0.499,
+      "step": 4003
+    },
+    {
+      "epoch": 1.110371602884082,
+      "grad_norm": 0.19228579103946686,
+      "learning_rate": 1.0591484660747933e-05,
+      "loss": 0.5189,
+      "step": 4004
+    },
+    {
+      "epoch": 1.1106489184692179,
+      "grad_norm": 0.19055628776550293,
+      "learning_rate": 1.0586027945029352e-05,
+      "loss": 0.5012,
+      "step": 4005
+    },
+    {
+      "epoch": 1.1109262340543538,
+      "grad_norm": 0.19514985382556915,
+      "learning_rate": 1.0580571602777425e-05,
+      "loss": 0.4883,
+      "step": 4006
+    },
+    {
+      "epoch": 1.1112035496394896,
+      "grad_norm": 0.1857844740152359,
+      "learning_rate": 1.057511563505683e-05,
+      "loss": 0.4863,
+      "step": 4007
+    },
+    {
+      "epoch": 1.1114808652246255,
+      "grad_norm": 0.19990986585617065,
+      "learning_rate": 1.0569660042932177e-05,
+      "loss": 0.5079,
+      "step": 4008
+    },
+    {
+      "epoch": 1.1117581808097614,
+      "grad_norm": 0.1960282176733017,
+      "learning_rate": 1.0564204827467994e-05,
+      "loss": 0.5277,
+      "step": 4009
+    },
+    {
+      "epoch": 1.1120354963948973,
+      "grad_norm": 0.20044219493865967,
+      "learning_rate": 1.0558749989728729e-05,
+      "loss": 0.5121,
+      "step": 4010
+    },
+    {
+      "epoch": 1.1123128119800332,
+      "grad_norm": 0.19683614373207092,
+      "learning_rate": 1.0553295530778784e-05,
+      "loss": 0.4932,
+      "step": 4011
+    },
+    {
+      "epoch": 1.112590127565169,
+      "grad_norm": 0.1851675808429718,
+      "learning_rate": 1.0547841451682453e-05,
+      "loss": 0.4994,
+      "step": 4012
+    },
+    {
+      "epoch": 1.112867443150305,
+      "grad_norm": 0.20108520984649658,
+      "learning_rate": 1.0542387753503974e-05,
+      "loss": 0.5352,
+      "step": 4013
+    },
+    {
+      "epoch": 1.1131447587354408,
+      "grad_norm": 0.18748925626277924,
+      "learning_rate": 1.0536934437307514e-05,
+      "loss": 0.4929,
+      "step": 4014
+    },
+    {
+      "epoch": 1.1134220743205767,
+      "grad_norm": 0.183696448802948,
+      "learning_rate": 1.0531481504157153e-05,
+      "loss": 0.5099,
+      "step": 4015
+    },
+    {
+      "epoch": 1.1136993899057126,
+      "grad_norm": 0.20530031621456146,
+      "learning_rate": 1.0526028955116912e-05,
+      "loss": 0.5101,
+      "step": 4016
+    },
+    {
+      "epoch": 1.1139767054908485,
+      "grad_norm": 0.20076102018356323,
+      "learning_rate": 1.0520576791250711e-05,
+      "loss": 0.533,
+      "step": 4017
+    },
+    {
+      "epoch": 1.1142540210759844,
+      "grad_norm": 0.2018887996673584,
+      "learning_rate": 1.0515125013622428e-05,
+      "loss": 0.5015,
+      "step": 4018
+    },
+    {
+      "epoch": 1.1145313366611203,
+      "grad_norm": 0.1974845677614212,
+      "learning_rate": 1.0509673623295843e-05,
+      "loss": 0.4778,
+      "step": 4019
+    },
+    {
+      "epoch": 1.1148086522462561,
+      "grad_norm": 0.18530985713005066,
+      "learning_rate": 1.0504222621334664e-05,
+      "loss": 0.5003,
+      "step": 4020
+    },
+    {
+      "epoch": 1.115085967831392,
+      "grad_norm": 0.1943405568599701,
+      "learning_rate": 1.0498772008802531e-05,
+      "loss": 0.5041,
+      "step": 4021
+    },
+    {
+      "epoch": 1.115363283416528,
+      "grad_norm": 0.19365082681179047,
+      "learning_rate": 1.0493321786763003e-05,
+      "loss": 0.5156,
+      "step": 4022
+    },
+    {
+      "epoch": 1.1156405990016638,
+      "grad_norm": 0.20058946311473846,
+      "learning_rate": 1.0487871956279558e-05,
+      "loss": 0.4925,
+      "step": 4023
+    },
+    {
+      "epoch": 1.1159179145867997,
+      "grad_norm": 0.1888214498758316,
+      "learning_rate": 1.0482422518415602e-05,
+      "loss": 0.5321,
+      "step": 4024
+    },
+    {
+      "epoch": 1.1161952301719356,
+      "grad_norm": 0.19549217820167542,
+      "learning_rate": 1.047697347423447e-05,
+      "loss": 0.5257,
+      "step": 4025
+    },
+    {
+      "epoch": 1.1164725457570714,
+      "grad_norm": 0.19175288081169128,
+      "learning_rate": 1.0471524824799413e-05,
+      "loss": 0.4774,
+      "step": 4026
+    },
+    {
+      "epoch": 1.1167498613422073,
+      "grad_norm": 0.19548679888248444,
+      "learning_rate": 1.04660765711736e-05,
+      "loss": 0.4782,
+      "step": 4027
+    },
+    {
+      "epoch": 1.1170271769273432,
+      "grad_norm": 0.20252208411693573,
+      "learning_rate": 1.0460628714420145e-05,
+      "loss": 0.5024,
+      "step": 4028
+    },
+    {
+      "epoch": 1.117304492512479,
+      "grad_norm": 0.18733076751232147,
+      "learning_rate": 1.0455181255602056e-05,
+      "loss": 0.4894,
+      "step": 4029
+    },
+    {
+      "epoch": 1.117581808097615,
+      "grad_norm": 0.19177298247814178,
+      "learning_rate": 1.0449734195782281e-05,
+      "loss": 0.5067,
+      "step": 4030
+    },
+    {
+      "epoch": 1.1178591236827509,
+      "grad_norm": 0.18974518775939941,
+      "learning_rate": 1.0444287536023681e-05,
+      "loss": 0.4863,
+      "step": 4031
+    },
+    {
+      "epoch": 1.1181364392678868,
+      "grad_norm": 0.20173045992851257,
+      "learning_rate": 1.0438841277389055e-05,
+      "loss": 0.507,
+      "step": 4032
+    },
+    {
+      "epoch": 1.1184137548530226,
+      "grad_norm": 0.189706489443779,
+      "learning_rate": 1.0433395420941101e-05,
+      "loss": 0.495,
+      "step": 4033
+    },
+    {
+      "epoch": 1.1186910704381585,
+      "grad_norm": 0.20018784701824188,
+      "learning_rate": 1.0427949967742452e-05,
+      "loss": 0.5267,
+      "step": 4034
+    },
+    {
+      "epoch": 1.1189683860232944,
+      "grad_norm": 0.18162201344966888,
+      "learning_rate": 1.0422504918855664e-05,
+      "loss": 0.5028,
+      "step": 4035
+    },
+    {
+      "epoch": 1.1192457016084303,
+      "grad_norm": 0.2000206708908081,
+      "learning_rate": 1.041706027534321e-05,
+      "loss": 0.5162,
+      "step": 4036
+    },
+    {
+      "epoch": 1.1195230171935662,
+      "grad_norm": 0.19805558025836945,
+      "learning_rate": 1.0411616038267486e-05,
+      "loss": 0.4861,
+      "step": 4037
+    },
+    {
+      "epoch": 1.119800332778702,
+      "grad_norm": 0.19431713223457336,
+      "learning_rate": 1.0406172208690797e-05,
+      "loss": 0.4991,
+      "step": 4038
+    },
+    {
+      "epoch": 1.120077648363838,
+      "grad_norm": 0.2039175033569336,
+      "learning_rate": 1.0400728787675387e-05,
+      "loss": 0.539,
+      "step": 4039
+    },
+    {
+      "epoch": 1.1203549639489738,
+      "grad_norm": 0.19760684669017792,
+      "learning_rate": 1.039528577628341e-05,
+      "loss": 0.5149,
+      "step": 4040
+    },
+    {
+      "epoch": 1.1206322795341097,
+      "grad_norm": 0.18955372273921967,
+      "learning_rate": 1.038984317557694e-05,
+      "loss": 0.4901,
+      "step": 4041
+    },
+    {
+      "epoch": 1.1209095951192456,
+      "grad_norm": 0.19011299312114716,
+      "learning_rate": 1.0384400986617977e-05,
+      "loss": 0.518,
+      "step": 4042
+    },
+    {
+      "epoch": 1.1211869107043815,
+      "grad_norm": 0.18622393906116486,
+      "learning_rate": 1.0378959210468434e-05,
+      "loss": 0.4912,
+      "step": 4043
+    },
+    {
+      "epoch": 1.1214642262895174,
+      "grad_norm": 0.19130538403987885,
+      "learning_rate": 1.0373517848190143e-05,
+      "loss": 0.4952,
+      "step": 4044
+    },
+    {
+      "epoch": 1.1217415418746532,
+      "grad_norm": 0.19988033175468445,
+      "learning_rate": 1.0368076900844856e-05,
+      "loss": 0.4964,
+      "step": 4045
+    },
+    {
+      "epoch": 1.1220188574597891,
+      "grad_norm": 0.20225760340690613,
+      "learning_rate": 1.0362636369494254e-05,
+      "loss": 0.5338,
+      "step": 4046
+    },
+    {
+      "epoch": 1.122296173044925,
+      "grad_norm": 0.2051144242286682,
+      "learning_rate": 1.0357196255199928e-05,
+      "loss": 0.5251,
+      "step": 4047
+    },
+    {
+      "epoch": 1.122573488630061,
+      "grad_norm": 0.18915054202079773,
+      "learning_rate": 1.0351756559023374e-05,
+      "loss": 0.4812,
+      "step": 4048
+    },
+    {
+      "epoch": 1.1228508042151968,
+      "grad_norm": 0.20346355438232422,
+      "learning_rate": 1.0346317282026045e-05,
+      "loss": 0.5366,
+      "step": 4049
+    },
+    {
+      "epoch": 1.1231281198003327,
+      "grad_norm": 0.1956186592578888,
+      "learning_rate": 1.0340878425269269e-05,
+      "loss": 0.5396,
+      "step": 4050
+    },
+    {
+      "epoch": 1.1234054353854686,
+      "grad_norm": 0.18970191478729248,
+      "learning_rate": 1.0335439989814316e-05,
+      "loss": 0.4687,
+      "step": 4051
+    },
+    {
+      "epoch": 1.1236827509706044,
+      "grad_norm": 0.19612254202365875,
+      "learning_rate": 1.033000197672237e-05,
+      "loss": 0.5025,
+      "step": 4052
+    },
+    {
+      "epoch": 1.1239600665557403,
+      "grad_norm": 0.19224813580513,
+      "learning_rate": 1.0324564387054535e-05,
+      "loss": 0.494,
+      "step": 4053
+    },
+    {
+      "epoch": 1.1242373821408762,
+      "grad_norm": 0.1851874589920044,
+      "learning_rate": 1.0319127221871823e-05,
+      "loss": 0.521,
+      "step": 4054
+    },
+    {
+      "epoch": 1.124514697726012,
+      "grad_norm": 0.19579973816871643,
+      "learning_rate": 1.0313690482235168e-05,
+      "loss": 0.5012,
+      "step": 4055
+    },
+    {
+      "epoch": 1.124792013311148,
+      "grad_norm": 0.21103787422180176,
+      "learning_rate": 1.0308254169205428e-05,
+      "loss": 0.4969,
+      "step": 4056
+    },
+    {
+      "epoch": 1.1250693288962839,
+      "grad_norm": 0.1894204467535019,
+      "learning_rate": 1.030281828384337e-05,
+      "loss": 0.494,
+      "step": 4057
+    },
+    {
+      "epoch": 1.1253466444814197,
+      "grad_norm": 0.1914178431034088,
+      "learning_rate": 1.0297382827209679e-05,
+      "loss": 0.5094,
+      "step": 4058
+    },
+    {
+      "epoch": 1.1256239600665556,
+      "grad_norm": 0.18707357347011566,
+      "learning_rate": 1.0291947800364948e-05,
+      "loss": 0.5023,
+      "step": 4059
+    },
+    {
+      "epoch": 1.1259012756516915,
+      "grad_norm": 0.19480562210083008,
+      "learning_rate": 1.0286513204369712e-05,
+      "loss": 0.5154,
+      "step": 4060
+    },
+    {
+      "epoch": 1.1261785912368274,
+      "grad_norm": 0.18776677548885345,
+      "learning_rate": 1.0281079040284392e-05,
+      "loss": 0.5143,
+      "step": 4061
+    },
+    {
+      "epoch": 1.1264559068219633,
+      "grad_norm": 0.18544964492321014,
+      "learning_rate": 1.0275645309169337e-05,
+      "loss": 0.4906,
+      "step": 4062
+    },
+    {
+      "epoch": 1.1267332224070992,
+      "grad_norm": 0.18632815778255463,
+      "learning_rate": 1.0270212012084817e-05,
+      "loss": 0.5276,
+      "step": 4063
+    },
+    {
+      "epoch": 1.127010537992235,
+      "grad_norm": 0.206988126039505,
+      "learning_rate": 1.0264779150091014e-05,
+      "loss": 0.5146,
+      "step": 4064
+    },
+    {
+      "epoch": 1.127287853577371,
+      "grad_norm": 0.19988800585269928,
+      "learning_rate": 1.0259346724248018e-05,
+      "loss": 0.4973,
+      "step": 4065
+    },
+    {
+      "epoch": 1.1275651691625068,
+      "grad_norm": 0.18873746693134308,
+      "learning_rate": 1.0253914735615838e-05,
+      "loss": 0.5136,
+      "step": 4066
+    },
+    {
+      "epoch": 1.1278424847476427,
+      "grad_norm": 0.19661124050617218,
+      "learning_rate": 1.0248483185254403e-05,
+      "loss": 0.5418,
+      "step": 4067
+    },
+    {
+      "epoch": 1.1281198003327786,
+      "grad_norm": 0.2008582353591919,
+      "learning_rate": 1.0243052074223555e-05,
+      "loss": 0.508,
+      "step": 4068
+    },
+    {
+      "epoch": 1.1283971159179145,
+      "grad_norm": 0.19019687175750732,
+      "learning_rate": 1.023762140358304e-05,
+      "loss": 0.5072,
+      "step": 4069
+    },
+    {
+      "epoch": 1.1286744315030504,
+      "grad_norm": 0.19475378096103668,
+      "learning_rate": 1.0232191174392532e-05,
+      "loss": 0.5282,
+      "step": 4070
+    },
+    {
+      "epoch": 1.1289517470881862,
+      "grad_norm": 0.18883496522903442,
+      "learning_rate": 1.0226761387711612e-05,
+      "loss": 0.5092,
+      "step": 4071
+    },
+    {
+      "epoch": 1.1292290626733221,
+      "grad_norm": 0.1904006004333496,
+      "learning_rate": 1.0221332044599768e-05,
+      "loss": 0.5302,
+      "step": 4072
+    },
+    {
+      "epoch": 1.129506378258458,
+      "grad_norm": 0.19037893414497375,
+      "learning_rate": 1.0215903146116417e-05,
+      "loss": 0.4916,
+      "step": 4073
+    },
+    {
+      "epoch": 1.129783693843594,
+      "grad_norm": 0.2003484070301056,
+      "learning_rate": 1.021047469332088e-05,
+      "loss": 0.5259,
+      "step": 4074
+    },
+    {
+      "epoch": 1.1300610094287298,
+      "grad_norm": 0.18021680414676666,
+      "learning_rate": 1.0205046687272392e-05,
+      "loss": 0.5038,
+      "step": 4075
+    },
+    {
+      "epoch": 1.1303383250138657,
+      "grad_norm": 0.19366797804832458,
+      "learning_rate": 1.0199619129030093e-05,
+      "loss": 0.5091,
+      "step": 4076
+    },
+    {
+      "epoch": 1.1306156405990015,
+      "grad_norm": 0.19460560381412506,
+      "learning_rate": 1.0194192019653053e-05,
+      "loss": 0.5071,
+      "step": 4077
+    },
+    {
+      "epoch": 1.1308929561841374,
+      "grad_norm": 0.18521569669246674,
+      "learning_rate": 1.018876536020024e-05,
+      "loss": 0.5031,
+      "step": 4078
+    },
+    {
+      "epoch": 1.1311702717692733,
+      "grad_norm": 0.18586260080337524,
+      "learning_rate": 1.018333915173054e-05,
+      "loss": 0.4893,
+      "step": 4079
+    },
+    {
+      "epoch": 1.1314475873544092,
+      "grad_norm": 0.20619140565395355,
+      "learning_rate": 1.0177913395302748e-05,
+      "loss": 0.5124,
+      "step": 4080
+    },
+    {
+      "epoch": 1.131724902939545,
+      "grad_norm": 0.2014811784029007,
+      "learning_rate": 1.0172488091975583e-05,
+      "loss": 0.5045,
+      "step": 4081
+    },
+    {
+      "epoch": 1.132002218524681,
+      "grad_norm": 0.2058635950088501,
+      "learning_rate": 1.0167063242807654e-05,
+      "loss": 0.4832,
+      "step": 4082
+    },
+    {
+      "epoch": 1.1322795341098169,
+      "grad_norm": 0.19578316807746887,
+      "learning_rate": 1.016163884885749e-05,
+      "loss": 0.501,
+      "step": 4083
+    },
+    {
+      "epoch": 1.1325568496949527,
+      "grad_norm": 0.1896321028470993,
+      "learning_rate": 1.0156214911183546e-05,
+      "loss": 0.5113,
+      "step": 4084
+    },
+    {
+      "epoch": 1.1328341652800886,
+      "grad_norm": 0.19131551682949066,
+      "learning_rate": 1.0150791430844172e-05,
+      "loss": 0.502,
+      "step": 4085
+    },
+    {
+      "epoch": 1.1331114808652245,
+      "grad_norm": 0.20685172080993652,
+      "learning_rate": 1.0145368408897624e-05,
+      "loss": 0.4907,
+      "step": 4086
+    },
+    {
+      "epoch": 1.1333887964503604,
+      "grad_norm": 0.21301031112670898,
+      "learning_rate": 1.0139945846402091e-05,
+      "loss": 0.5013,
+      "step": 4087
+    },
+    {
+      "epoch": 1.1336661120354963,
+      "grad_norm": 0.1879834532737732,
+      "learning_rate": 1.013452374441565e-05,
+      "loss": 0.5197,
+      "step": 4088
+    },
+    {
+      "epoch": 1.1339434276206322,
+      "grad_norm": 0.19364705681800842,
+      "learning_rate": 1.01291021039963e-05,
+      "loss": 0.5064,
+      "step": 4089
+    },
+    {
+      "epoch": 1.134220743205768,
+      "grad_norm": 0.18613065779209137,
+      "learning_rate": 1.012368092620194e-05,
+      "loss": 0.4874,
+      "step": 4090
+    },
+    {
+      "epoch": 1.134498058790904,
+      "grad_norm": 0.1905667632818222,
+      "learning_rate": 1.0118260212090397e-05,
+      "loss": 0.5038,
+      "step": 4091
+    },
+    {
+      "epoch": 1.1347753743760398,
+      "grad_norm": 0.19485412538051605,
+      "learning_rate": 1.0112839962719387e-05,
+      "loss": 0.4862,
+      "step": 4092
+    },
+    {
+      "epoch": 1.1350526899611757,
+      "grad_norm": 0.1890053004026413,
+      "learning_rate": 1.0107420179146542e-05,
+      "loss": 0.5268,
+      "step": 4093
+    },
+    {
+      "epoch": 1.1353300055463116,
+      "grad_norm": 0.19711565971374512,
+      "learning_rate": 1.0102000862429415e-05,
+      "loss": 0.4994,
+      "step": 4094
+    },
+    {
+      "epoch": 1.1356073211314475,
+      "grad_norm": 0.18928836286067963,
+      "learning_rate": 1.0096582013625455e-05,
+      "loss": 0.5112,
+      "step": 4095
+    },
+    {
+      "epoch": 1.1358846367165834,
+      "grad_norm": 0.1987476795911789,
+      "learning_rate": 1.0091163633792023e-05,
+      "loss": 0.4815,
+      "step": 4096
+    },
+    {
+      "epoch": 1.1361619523017192,
+      "grad_norm": 0.20385397970676422,
+      "learning_rate": 1.0085745723986379e-05,
+      "loss": 0.5387,
+      "step": 4097
+    },
+    {
+      "epoch": 1.1364392678868551,
+      "grad_norm": 0.1833743155002594,
+      "learning_rate": 1.0080328285265715e-05,
+      "loss": 0.5216,
+      "step": 4098
+    },
+    {
+      "epoch": 1.136716583471991,
+      "grad_norm": 0.19815048575401306,
+      "learning_rate": 1.007491131868711e-05,
+      "loss": 0.5257,
+      "step": 4099
+    },
+    {
+      "epoch": 1.1369938990571269,
+      "grad_norm": 0.19522307813167572,
+      "learning_rate": 1.0069494825307554e-05,
+      "loss": 0.5109,
+      "step": 4100
+    },
+    {
+      "epoch": 1.1372712146422628,
+      "grad_norm": 0.20600605010986328,
+      "learning_rate": 1.0064078806183956e-05,
+      "loss": 0.5222,
+      "step": 4101
+    },
+    {
+      "epoch": 1.1375485302273987,
+      "grad_norm": 0.19125322997570038,
+      "learning_rate": 1.0058663262373125e-05,
+      "loss": 0.5145,
+      "step": 4102
+    },
+    {
+      "epoch": 1.1378258458125345,
+      "grad_norm": 0.19589276611804962,
+      "learning_rate": 1.005324819493177e-05,
+      "loss": 0.5181,
+      "step": 4103
+    },
+    {
+      "epoch": 1.1381031613976704,
+      "grad_norm": 0.182989239692688,
+      "learning_rate": 1.0047833604916515e-05,
+      "loss": 0.5209,
+      "step": 4104
+    },
+    {
+      "epoch": 1.1383804769828063,
+      "grad_norm": 0.18455688655376434,
+      "learning_rate": 1.0042419493383896e-05,
+      "loss": 0.4888,
+      "step": 4105
+    },
+    {
+      "epoch": 1.1386577925679422,
+      "grad_norm": 0.19933606684207916,
+      "learning_rate": 1.0037005861390346e-05,
+      "loss": 0.5306,
+      "step": 4106
+    },
+    {
+      "epoch": 1.138935108153078,
+      "grad_norm": 0.19362375140190125,
+      "learning_rate": 1.0031592709992204e-05,
+      "loss": 0.4934,
+      "step": 4107
+    },
+    {
+      "epoch": 1.139212423738214,
+      "grad_norm": 0.1925627440214157,
+      "learning_rate": 1.0026180040245728e-05,
+      "loss": 0.5044,
+      "step": 4108
+    },
+    {
+      "epoch": 1.1394897393233498,
+      "grad_norm": 0.19093094766139984,
+      "learning_rate": 1.0020767853207069e-05,
+      "loss": 0.5083,
+      "step": 4109
+    },
+    {
+      "epoch": 1.1397670549084857,
+      "grad_norm": 0.18980608880519867,
+      "learning_rate": 1.0015356149932288e-05,
+      "loss": 0.5099,
+      "step": 4110
+    },
+    {
+      "epoch": 1.1400443704936216,
+      "grad_norm": 0.18821600079536438,
+      "learning_rate": 1.0009944931477346e-05,
+      "loss": 0.5194,
+      "step": 4111
+    },
+    {
+      "epoch": 1.1403216860787575,
+      "grad_norm": 0.1851465255022049,
+      "learning_rate": 1.0004534198898124e-05,
+      "loss": 0.503,
+      "step": 4112
+    },
+    {
+      "epoch": 1.1405990016638934,
+      "grad_norm": 0.19353441894054413,
+      "learning_rate": 9.9991239532504e-06,
+      "loss": 0.511,
+      "step": 4113
+    },
+    {
+      "epoch": 1.1408763172490295,
+      "grad_norm": 0.19810713827610016,
+      "learning_rate": 9.993714195589847e-06,
+      "loss": 0.5106,
+      "step": 4114
+    },
+    {
+      "epoch": 1.1411536328341654,
+      "grad_norm": 0.19608542323112488,
+      "learning_rate": 9.98830492697206e-06,
+      "loss": 0.5384,
+      "step": 4115
+    },
+    {
+      "epoch": 1.1414309484193013,
+      "grad_norm": 0.43584343791007996,
+      "learning_rate": 9.982896148452527e-06,
+      "loss": 0.5011,
+      "step": 4116
+    },
+    {
+      "epoch": 1.1417082640044371,
+      "grad_norm": 0.19775094091892242,
+      "learning_rate": 9.977487861086647e-06,
+      "loss": 0.5184,
+      "step": 4117
+    },
+    {
+      "epoch": 1.141985579589573,
+      "grad_norm": 0.2092062532901764,
+      "learning_rate": 9.97208006592971e-06,
+      "loss": 0.4917,
+      "step": 4118
+    },
+    {
+      "epoch": 1.142262895174709,
+      "grad_norm": 0.18967856466770172,
+      "learning_rate": 9.966672764036936e-06,
+      "loss": 0.4851,
+      "step": 4119
+    },
+    {
+      "epoch": 1.1425402107598448,
+      "grad_norm": 0.1816510409116745,
+      "learning_rate": 9.961265956463424e-06,
+      "loss": 0.4985,
+      "step": 4120
+    },
+    {
+      "epoch": 1.1428175263449807,
+      "grad_norm": 0.18744854629039764,
+      "learning_rate": 9.955859644264183e-06,
+      "loss": 0.4988,
+      "step": 4121
+    },
+    {
+      "epoch": 1.1430948419301166,
+      "grad_norm": 0.19459229707717896,
+      "learning_rate": 9.950453828494132e-06,
+      "loss": 0.5171,
+      "step": 4122
+    },
+    {
+      "epoch": 1.1433721575152525,
+      "grad_norm": 0.19057153165340424,
+      "learning_rate": 9.945048510208094e-06,
+      "loss": 0.5126,
+      "step": 4123
+    },
+    {
+      "epoch": 1.1436494731003883,
+      "grad_norm": 0.19819004833698273,
+      "learning_rate": 9.939643690460779e-06,
+      "loss": 0.5449,
+      "step": 4124
+    },
+    {
+      "epoch": 1.1439267886855242,
+      "grad_norm": 0.19458059966564178,
+      "learning_rate": 9.934239370306813e-06,
+      "loss": 0.4927,
+      "step": 4125
+    },
+    {
+      "epoch": 1.14420410427066,
+      "grad_norm": 0.20325776934623718,
+      "learning_rate": 9.928835550800727e-06,
+      "loss": 0.5221,
+      "step": 4126
+    },
+    {
+      "epoch": 1.144481419855796,
+      "grad_norm": 0.19814957678318024,
+      "learning_rate": 9.923432232996947e-06,
+      "loss": 0.5417,
+      "step": 4127
+    },
+    {
+      "epoch": 1.1447587354409319,
+      "grad_norm": 0.1916591227054596,
+      "learning_rate": 9.9180294179498e-06,
+      "loss": 0.4876,
+      "step": 4128
+    },
+    {
+      "epoch": 1.1450360510260678,
+      "grad_norm": 0.19258858263492584,
+      "learning_rate": 9.912627106713528e-06,
+      "loss": 0.4838,
+      "step": 4129
+    },
+    {
+      "epoch": 1.1453133666112036,
+      "grad_norm": 0.18850946426391602,
+      "learning_rate": 9.907225300342256e-06,
+      "loss": 0.4996,
+      "step": 4130
+    },
+    {
+      "epoch": 1.1455906821963395,
+      "grad_norm": 0.21922184526920319,
+      "learning_rate": 9.901823999890021e-06,
+      "loss": 0.5001,
+      "step": 4131
+    },
+    {
+      "epoch": 1.1458679977814754,
+      "grad_norm": 0.1909458041191101,
+      "learning_rate": 9.896423206410759e-06,
+      "loss": 0.5119,
+      "step": 4132
+    },
+    {
+      "epoch": 1.1461453133666113,
+      "grad_norm": 0.18811345100402832,
+      "learning_rate": 9.891022920958313e-06,
+      "loss": 0.5086,
+      "step": 4133
+    },
+    {
+      "epoch": 1.1464226289517472,
+      "grad_norm": 0.20187973976135254,
+      "learning_rate": 9.88562314458642e-06,
+      "loss": 0.5163,
+      "step": 4134
+    },
+    {
+      "epoch": 1.146699944536883,
+      "grad_norm": 0.1905764937400818,
+      "learning_rate": 9.880223878348713e-06,
+      "loss": 0.5092,
+      "step": 4135
+    },
+    {
+      "epoch": 1.146977260122019,
+      "grad_norm": 0.19595111906528473,
+      "learning_rate": 9.874825123298741e-06,
+      "loss": 0.5066,
+      "step": 4136
+    },
+    {
+      "epoch": 1.1472545757071548,
+      "grad_norm": 0.19632695615291595,
+      "learning_rate": 9.869426880489939e-06,
+      "loss": 0.5007,
+      "step": 4137
+    },
+    {
+      "epoch": 1.1475318912922907,
+      "grad_norm": 0.19036737084388733,
+      "learning_rate": 9.864029150975646e-06,
+      "loss": 0.4899,
+      "step": 4138
+    },
+    {
+      "epoch": 1.1478092068774266,
+      "grad_norm": 0.18314070999622345,
+      "learning_rate": 9.858631935809107e-06,
+      "loss": 0.4764,
+      "step": 4139
+    },
+    {
+      "epoch": 1.1480865224625625,
+      "grad_norm": 0.25677627325057983,
+      "learning_rate": 9.85323523604346e-06,
+      "loss": 0.5034,
+      "step": 4140
+    },
+    {
+      "epoch": 1.1483638380476984,
+      "grad_norm": 0.18910054862499237,
+      "learning_rate": 9.847839052731742e-06,
+      "loss": 0.5234,
+      "step": 4141
+    },
+    {
+      "epoch": 1.1486411536328343,
+      "grad_norm": 0.19723178446292877,
+      "learning_rate": 9.842443386926892e-06,
+      "loss": 0.5099,
+      "step": 4142
+    },
+    {
+      "epoch": 1.1489184692179701,
+      "grad_norm": 0.1976155936717987,
+      "learning_rate": 9.837048239681747e-06,
+      "loss": 0.498,
+      "step": 4143
+    },
+    {
+      "epoch": 1.149195784803106,
+      "grad_norm": 0.20631377398967743,
+      "learning_rate": 9.83165361204905e-06,
+      "loss": 0.4832,
+      "step": 4144
+    },
+    {
+      "epoch": 1.149473100388242,
+      "grad_norm": 0.204478919506073,
+      "learning_rate": 9.826259505081419e-06,
+      "loss": 0.494,
+      "step": 4145
+    },
+    {
+      "epoch": 1.1497504159733778,
+      "grad_norm": 0.19692255556583405,
+      "learning_rate": 9.820865919831406e-06,
+      "loss": 0.5219,
+      "step": 4146
+    },
+    {
+      "epoch": 1.1500277315585137,
+      "grad_norm": 0.1929834634065628,
+      "learning_rate": 9.815472857351433e-06,
+      "loss": 0.4878,
+      "step": 4147
+    },
+    {
+      "epoch": 1.1503050471436496,
+      "grad_norm": 0.1798853874206543,
+      "learning_rate": 9.810080318693832e-06,
+      "loss": 0.4949,
+      "step": 4148
+    },
+    {
+      "epoch": 1.1505823627287854,
+      "grad_norm": 0.19789017736911774,
+      "learning_rate": 9.804688304910824e-06,
+      "loss": 0.5227,
+      "step": 4149
+    },
+    {
+      "epoch": 1.1508596783139213,
+      "grad_norm": 0.19549022614955902,
+      "learning_rate": 9.799296817054542e-06,
+      "loss": 0.5191,
+      "step": 4150
+    },
+    {
+      "epoch": 1.1511369938990572,
+      "grad_norm": 0.19911247491836548,
+      "learning_rate": 9.793905856177008e-06,
+      "loss": 0.5172,
+      "step": 4151
+    },
+    {
+      "epoch": 1.151414309484193,
+      "grad_norm": 0.29837673902511597,
+      "learning_rate": 9.78851542333013e-06,
+      "loss": 0.5081,
+      "step": 4152
+    },
+    {
+      "epoch": 1.151691625069329,
+      "grad_norm": 0.1918451339006424,
+      "learning_rate": 9.783125519565737e-06,
+      "loss": 0.5116,
+      "step": 4153
+    },
+    {
+      "epoch": 1.1519689406544649,
+      "grad_norm": 0.19545188546180725,
+      "learning_rate": 9.777736145935538e-06,
+      "loss": 0.5042,
+      "step": 4154
+    },
+    {
+      "epoch": 1.1522462562396008,
+      "grad_norm": 0.19828790426254272,
+      "learning_rate": 9.772347303491144e-06,
+      "loss": 0.4944,
+      "step": 4155
+    },
+    {
+      "epoch": 1.1525235718247366,
+      "grad_norm": 0.19536136090755463,
+      "learning_rate": 9.766958993284051e-06,
+      "loss": 0.511,
+      "step": 4156
+    },
+    {
+      "epoch": 1.1528008874098725,
+      "grad_norm": 0.19933415949344635,
+      "learning_rate": 9.761571216365678e-06,
+      "loss": 0.5035,
+      "step": 4157
+    },
+    {
+      "epoch": 1.1530782029950084,
+      "grad_norm": 0.20315343141555786,
+      "learning_rate": 9.75618397378731e-06,
+      "loss": 0.5105,
+      "step": 4158
+    },
+    {
+      "epoch": 1.1533555185801443,
+      "grad_norm": 0.19544823467731476,
+      "learning_rate": 9.750797266600142e-06,
+      "loss": 0.4965,
+      "step": 4159
+    },
+    {
+      "epoch": 1.1536328341652802,
+      "grad_norm": 0.1914399266242981,
+      "learning_rate": 9.74541109585527e-06,
+      "loss": 0.4964,
+      "step": 4160
+    },
+    {
+      "epoch": 1.153910149750416,
+      "grad_norm": 0.19952736794948578,
+      "learning_rate": 9.740025462603675e-06,
+      "loss": 0.4937,
+      "step": 4161
+    },
+    {
+      "epoch": 1.154187465335552,
+      "grad_norm": 0.19488787651062012,
+      "learning_rate": 9.734640367896236e-06,
+      "loss": 0.5137,
+      "step": 4162
+    },
+    {
+      "epoch": 1.1544647809206878,
+      "grad_norm": 0.18852956593036652,
+      "learning_rate": 9.729255812783724e-06,
+      "loss": 0.5133,
+      "step": 4163
+    },
+    {
+      "epoch": 1.1547420965058237,
+      "grad_norm": 0.23788957297801971,
+      "learning_rate": 9.723871798316815e-06,
+      "loss": 0.5007,
+      "step": 4164
+    },
+    {
+      "epoch": 1.1550194120909596,
+      "grad_norm": 0.19701410830020905,
+      "learning_rate": 9.718488325546072e-06,
+      "loss": 0.4874,
+      "step": 4165
+    },
+    {
+      "epoch": 1.1552967276760955,
+      "grad_norm": 0.1924983710050583,
+      "learning_rate": 9.713105395521947e-06,
+      "loss": 0.491,
+      "step": 4166
+    },
+    {
+      "epoch": 1.1555740432612314,
+      "grad_norm": 0.19456344842910767,
+      "learning_rate": 9.707723009294802e-06,
+      "loss": 0.5032,
+      "step": 4167
+    },
+    {
+      "epoch": 1.1558513588463672,
+      "grad_norm": 0.2932334840297699,
+      "learning_rate": 9.702341167914875e-06,
+      "loss": 0.5158,
+      "step": 4168
+    },
+    {
+      "epoch": 1.1561286744315031,
+      "grad_norm": 0.19522738456726074,
+      "learning_rate": 9.696959872432311e-06,
+      "loss": 0.486,
+      "step": 4169
+    },
+    {
+      "epoch": 1.156405990016639,
+      "grad_norm": 0.19203148782253265,
+      "learning_rate": 9.691579123897137e-06,
+      "loss": 0.5187,
+      "step": 4170
+    },
+    {
+      "epoch": 1.156683305601775,
+      "grad_norm": 0.19738537073135376,
+      "learning_rate": 9.686198923359286e-06,
+      "loss": 0.4969,
+      "step": 4171
+    },
+    {
+      "epoch": 1.1569606211869108,
+      "grad_norm": 0.198639377951622,
+      "learning_rate": 9.680819271868578e-06,
+      "loss": 0.4974,
+      "step": 4172
+    },
+    {
+      "epoch": 1.1572379367720467,
+      "grad_norm": 0.19397403299808502,
+      "learning_rate": 9.675440170474717e-06,
+      "loss": 0.513,
+      "step": 4173
+    },
+    {
+      "epoch": 1.1575152523571826,
+      "grad_norm": 0.19262705743312836,
+      "learning_rate": 9.670061620227318e-06,
+      "loss": 0.5025,
+      "step": 4174
+    },
+    {
+      "epoch": 1.1577925679423184,
+      "grad_norm": 0.20281995832920074,
+      "learning_rate": 9.664683622175874e-06,
+      "loss": 0.5014,
+      "step": 4175
+    },
+    {
+      "epoch": 1.1580698835274543,
+      "grad_norm": 0.18004655838012695,
+      "learning_rate": 9.659306177369779e-06,
+      "loss": 0.5118,
+      "step": 4176
+    },
+    {
+      "epoch": 1.1583471991125902,
+      "grad_norm": 0.21135984361171722,
+      "learning_rate": 9.653929286858302e-06,
+      "loss": 0.4946,
+      "step": 4177
+    },
+    {
+      "epoch": 1.158624514697726,
+      "grad_norm": 0.2009570449590683,
+      "learning_rate": 9.648552951690635e-06,
+      "loss": 0.5502,
+      "step": 4178
+    },
+    {
+      "epoch": 1.158901830282862,
+      "grad_norm": 0.19548632204532623,
+      "learning_rate": 9.643177172915833e-06,
+      "loss": 0.5107,
+      "step": 4179
+    },
+    {
+      "epoch": 1.1591791458679979,
+      "grad_norm": 0.2423223853111267,
+      "learning_rate": 9.637801951582851e-06,
+      "loss": 0.4754,
+      "step": 4180
+    },
+    {
+      "epoch": 1.1594564614531337,
+      "grad_norm": 0.18860936164855957,
+      "learning_rate": 9.632427288740545e-06,
+      "loss": 0.4951,
+      "step": 4181
+    },
+    {
+      "epoch": 1.1597337770382696,
+      "grad_norm": 0.19608083367347717,
+      "learning_rate": 9.627053185437651e-06,
+      "loss": 0.4936,
+      "step": 4182
+    },
+    {
+      "epoch": 1.1600110926234055,
+      "grad_norm": 0.18897056579589844,
+      "learning_rate": 9.621679642722794e-06,
+      "loss": 0.507,
+      "step": 4183
+    },
+    {
+      "epoch": 1.1602884082085414,
+      "grad_norm": 0.18415631353855133,
+      "learning_rate": 9.616306661644497e-06,
+      "loss": 0.5157,
+      "step": 4184
+    },
+    {
+      "epoch": 1.1605657237936773,
+      "grad_norm": 0.300893098115921,
+      "learning_rate": 9.610934243251177e-06,
+      "loss": 0.5148,
+      "step": 4185
+    },
+    {
+      "epoch": 1.1608430393788132,
+      "grad_norm": 0.19000142812728882,
+      "learning_rate": 9.60556238859113e-06,
+      "loss": 0.5156,
+      "step": 4186
+    },
+    {
+      "epoch": 1.161120354963949,
+      "grad_norm": 0.20633526146411896,
+      "learning_rate": 9.60019109871254e-06,
+      "loss": 0.4992,
+      "step": 4187
+    },
+    {
+      "epoch": 1.161397670549085,
+      "grad_norm": 0.20525197684764862,
+      "learning_rate": 9.594820374663506e-06,
+      "loss": 0.4937,
+      "step": 4188
+    },
+    {
+      "epoch": 1.1616749861342208,
+      "grad_norm": 0.19740667939186096,
+      "learning_rate": 9.589450217491984e-06,
+      "loss": 0.4853,
+      "step": 4189
+    },
+    {
+      "epoch": 1.1619523017193567,
+      "grad_norm": 0.20025213062763214,
+      "learning_rate": 9.58408062824584e-06,
+      "loss": 0.5031,
+      "step": 4190
+    },
+    {
+      "epoch": 1.1622296173044926,
+      "grad_norm": 0.20170235633850098,
+      "learning_rate": 9.578711607972815e-06,
+      "loss": 0.5283,
+      "step": 4191
+    },
+    {
+      "epoch": 1.1625069328896285,
+      "grad_norm": 0.19781838357448578,
+      "learning_rate": 9.573343157720558e-06,
+      "loss": 0.5155,
+      "step": 4192
+    },
+    {
+      "epoch": 1.1627842484747644,
+      "grad_norm": 0.19260545074939728,
+      "learning_rate": 9.567975278536595e-06,
+      "loss": 0.4877,
+      "step": 4193
+    },
+    {
+      "epoch": 1.1630615640599002,
+      "grad_norm": 0.1921887993812561,
+      "learning_rate": 9.562607971468328e-06,
+      "loss": 0.5139,
+      "step": 4194
+    },
+    {
+      "epoch": 1.1633388796450361,
+      "grad_norm": 0.19691529870033264,
+      "learning_rate": 9.55724123756308e-06,
+      "loss": 0.5187,
+      "step": 4195
+    },
+    {
+      "epoch": 1.163616195230172,
+      "grad_norm": 0.1975255310535431,
+      "learning_rate": 9.551875077868028e-06,
+      "loss": 0.5256,
+      "step": 4196
+    },
+    {
+      "epoch": 1.163893510815308,
+      "grad_norm": 0.20154882967472076,
+      "learning_rate": 9.546509493430257e-06,
+      "loss": 0.5213,
+      "step": 4197
+    },
+    {
+      "epoch": 1.1641708264004438,
+      "grad_norm": 0.18560223281383514,
+      "learning_rate": 9.541144485296737e-06,
+      "loss": 0.501,
+      "step": 4198
+    },
+    {
+      "epoch": 1.1644481419855797,
+      "grad_norm": 0.18955029547214508,
+      "learning_rate": 9.535780054514325e-06,
+      "loss": 0.5273,
+      "step": 4199
+    },
+    {
+      "epoch": 1.1647254575707155,
+      "grad_norm": 0.1917003095149994,
+      "learning_rate": 9.530416202129756e-06,
+      "loss": 0.4887,
+      "step": 4200
+    },
+    {
+      "epoch": 1.1650027731558514,
+      "grad_norm": 0.20266690850257874,
+      "learning_rate": 9.525052929189661e-06,
+      "loss": 0.5112,
+      "step": 4201
+    },
+    {
+      "epoch": 1.1652800887409873,
+      "grad_norm": 0.18932059407234192,
+      "learning_rate": 9.519690236740563e-06,
+      "loss": 0.4984,
+      "step": 4202
+    },
+    {
+      "epoch": 1.1655574043261232,
+      "grad_norm": 0.19648674130439758,
+      "learning_rate": 9.51432812582886e-06,
+      "loss": 0.5004,
+      "step": 4203
+    },
+    {
+      "epoch": 1.165834719911259,
+      "grad_norm": 0.19848087430000305,
+      "learning_rate": 9.508966597500843e-06,
+      "loss": 0.5153,
+      "step": 4204
+    },
+    {
+      "epoch": 1.166112035496395,
+      "grad_norm": 0.2016671746969223,
+      "learning_rate": 9.50360565280269e-06,
+      "loss": 0.5134,
+      "step": 4205
+    },
+    {
+      "epoch": 1.1663893510815309,
+      "grad_norm": 0.22022029757499695,
+      "learning_rate": 9.498245292780463e-06,
+      "loss": 0.5048,
+      "step": 4206
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 0.19555556774139404,
+      "learning_rate": 9.49288551848011e-06,
+      "loss": 0.511,
+      "step": 4207
+    },
+    {
+      "epoch": 1.1669439822518026,
+      "grad_norm": 0.21261551976203918,
+      "learning_rate": 9.487526330947461e-06,
+      "loss": 0.5123,
+      "step": 4208
+    },
+    {
+      "epoch": 1.1672212978369385,
+      "grad_norm": 0.20261640846729279,
+      "learning_rate": 9.482167731228241e-06,
+      "loss": 0.5132,
+      "step": 4209
+    },
+    {
+      "epoch": 1.1674986134220744,
+      "grad_norm": 0.20545309782028198,
+      "learning_rate": 9.476809720368054e-06,
+      "loss": 0.5108,
+      "step": 4210
+    },
+    {
+      "epoch": 1.1677759290072103,
+      "grad_norm": 0.18970844149589539,
+      "learning_rate": 9.471452299412384e-06,
+      "loss": 0.5205,
+      "step": 4211
+    },
+    {
+      "epoch": 1.1680532445923462,
+      "grad_norm": 0.19278816878795624,
+      "learning_rate": 9.466095469406613e-06,
+      "loss": 0.5095,
+      "step": 4212
+    },
+    {
+      "epoch": 1.168330560177482,
+      "grad_norm": 0.19161520898342133,
+      "learning_rate": 9.460739231395999e-06,
+      "loss": 0.5151,
+      "step": 4213
+    },
+    {
+      "epoch": 1.168607875762618,
+      "grad_norm": 0.19239401817321777,
+      "learning_rate": 9.455383586425685e-06,
+      "loss": 0.4932,
+      "step": 4214
+    },
+    {
+      "epoch": 1.1688851913477538,
+      "grad_norm": 0.18713194131851196,
+      "learning_rate": 9.450028535540692e-06,
+      "loss": 0.4842,
+      "step": 4215
+    },
+    {
+      "epoch": 1.1691625069328897,
+      "grad_norm": 0.1931910663843155,
+      "learning_rate": 9.444674079785948e-06,
+      "loss": 0.4839,
+      "step": 4216
+    },
+    {
+      "epoch": 1.1694398225180256,
+      "grad_norm": 0.2027042806148529,
+      "learning_rate": 9.439320220206236e-06,
+      "loss": 0.4839,
+      "step": 4217
+    },
+    {
+      "epoch": 1.1697171381031615,
+      "grad_norm": 0.19624637067317963,
+      "learning_rate": 9.43396695784624e-06,
+      "loss": 0.5165,
+      "step": 4218
+    },
+    {
+      "epoch": 1.1699944536882974,
+      "grad_norm": 0.19271767139434814,
+      "learning_rate": 9.428614293750523e-06,
+      "loss": 0.5085,
+      "step": 4219
+    },
+    {
+      "epoch": 1.1702717692734332,
+      "grad_norm": 0.19581159949302673,
+      "learning_rate": 9.423262228963537e-06,
+      "loss": 0.5038,
+      "step": 4220
+    },
+    {
+      "epoch": 1.1705490848585691,
+      "grad_norm": 0.18047882616519928,
+      "learning_rate": 9.417910764529605e-06,
+      "loss": 0.5177,
+      "step": 4221
+    },
+    {
+      "epoch": 1.170826400443705,
+      "grad_norm": 0.19481384754180908,
+      "learning_rate": 9.412559901492935e-06,
+      "loss": 0.528,
+      "step": 4222
+    },
+    {
+      "epoch": 1.171103716028841,
+      "grad_norm": 0.2017330676317215,
+      "learning_rate": 9.407209640897635e-06,
+      "loss": 0.4888,
+      "step": 4223
+    },
+    {
+      "epoch": 1.1713810316139768,
+      "grad_norm": 0.2008255124092102,
+      "learning_rate": 9.401859983787674e-06,
+      "loss": 0.5369,
+      "step": 4224
+    },
+    {
+      "epoch": 1.1716583471991127,
+      "grad_norm": 0.19789865612983704,
+      "learning_rate": 9.396510931206912e-06,
+      "loss": 0.5095,
+      "step": 4225
+    },
+    {
+      "epoch": 1.1719356627842485,
+      "grad_norm": 0.18955931067466736,
+      "learning_rate": 9.391162484199097e-06,
+      "loss": 0.4998,
+      "step": 4226
+    },
+    {
+      "epoch": 1.1722129783693844,
+      "grad_norm": 0.19305095076560974,
+      "learning_rate": 9.385814643807845e-06,
+      "loss": 0.5198,
+      "step": 4227
+    },
+    {
+      "epoch": 1.1724902939545203,
+      "grad_norm": 0.19664254784584045,
+      "learning_rate": 9.380467411076667e-06,
+      "loss": 0.5123,
+      "step": 4228
+    },
+    {
+      "epoch": 1.1727676095396562,
+      "grad_norm": 0.1945103108882904,
+      "learning_rate": 9.375120787048944e-06,
+      "loss": 0.5063,
+      "step": 4229
+    },
+    {
+      "epoch": 1.173044925124792,
+      "grad_norm": 0.19993437826633453,
+      "learning_rate": 9.36977477276795e-06,
+      "loss": 0.502,
+      "step": 4230
+    },
+    {
+      "epoch": 1.173322240709928,
+      "grad_norm": 0.19845424592494965,
+      "learning_rate": 9.364429369276833e-06,
+      "loss": 0.4798,
+      "step": 4231
+    },
+    {
+      "epoch": 1.1735995562950639,
+      "grad_norm": 0.21802715957164764,
+      "learning_rate": 9.359084577618615e-06,
+      "loss": 0.5008,
+      "step": 4232
+    },
+    {
+      "epoch": 1.1738768718801997,
+      "grad_norm": 0.20177268981933594,
+      "learning_rate": 9.353740398836222e-06,
+      "loss": 0.5126,
+      "step": 4233
+    },
+    {
+      "epoch": 1.1741541874653356,
+      "grad_norm": 0.21453998982906342,
+      "learning_rate": 9.34839683397243e-06,
+      "loss": 0.5087,
+      "step": 4234
+    },
+    {
+      "epoch": 1.1744315030504715,
+      "grad_norm": 0.19861598312854767,
+      "learning_rate": 9.343053884069922e-06,
+      "loss": 0.5108,
+      "step": 4235
+    },
+    {
+      "epoch": 1.1747088186356074,
+      "grad_norm": 0.19748757779598236,
+      "learning_rate": 9.337711550171232e-06,
+      "loss": 0.5079,
+      "step": 4236
+    },
+    {
+      "epoch": 1.1749861342207433,
+      "grad_norm": 0.19762682914733887,
+      "learning_rate": 9.33236983331881e-06,
+      "loss": 0.4911,
+      "step": 4237
+    },
+    {
+      "epoch": 1.1752634498058792,
+      "grad_norm": 0.19971820712089539,
+      "learning_rate": 9.327028734554957e-06,
+      "loss": 0.5251,
+      "step": 4238
+    },
+    {
+      "epoch": 1.175540765391015,
+      "grad_norm": 0.20210981369018555,
+      "learning_rate": 9.321688254921862e-06,
+      "loss": 0.5057,
+      "step": 4239
+    },
+    {
+      "epoch": 1.175818080976151,
+      "grad_norm": 0.1958346962928772,
+      "learning_rate": 9.316348395461598e-06,
+      "loss": 0.4868,
+      "step": 4240
+    },
+    {
+      "epoch": 1.1760953965612868,
+      "grad_norm": 0.20058873295783997,
+      "learning_rate": 9.31100915721611e-06,
+      "loss": 0.5178,
+      "step": 4241
+    },
+    {
+      "epoch": 1.1763727121464227,
+      "grad_norm": 0.19912679493427277,
+      "learning_rate": 9.305670541227232e-06,
+      "loss": 0.5216,
+      "step": 4242
+    },
+    {
+      "epoch": 1.1766500277315586,
+      "grad_norm": 0.21662193536758423,
+      "learning_rate": 9.300332548536655e-06,
+      "loss": 0.5162,
+      "step": 4243
+    },
+    {
+      "epoch": 1.1769273433166945,
+      "grad_norm": 0.19769549369812012,
+      "learning_rate": 9.294995180185976e-06,
+      "loss": 0.5196,
+      "step": 4244
+    },
+    {
+      "epoch": 1.1772046589018303,
+      "grad_norm": 0.19544221460819244,
+      "learning_rate": 9.289658437216652e-06,
+      "loss": 0.5208,
+      "step": 4245
+    },
+    {
+      "epoch": 1.1774819744869662,
+      "grad_norm": 0.19446879625320435,
+      "learning_rate": 9.28432232067002e-06,
+      "loss": 0.5124,
+      "step": 4246
+    },
+    {
+      "epoch": 1.1777592900721021,
+      "grad_norm": 0.27411291003227234,
+      "learning_rate": 9.278986831587305e-06,
+      "loss": 0.4815,
+      "step": 4247
+    },
+    {
+      "epoch": 1.178036605657238,
+      "grad_norm": 0.19502323865890503,
+      "learning_rate": 9.273651971009599e-06,
+      "loss": 0.4944,
+      "step": 4248
+    },
+    {
+      "epoch": 1.1783139212423739,
+      "grad_norm": 0.2004517912864685,
+      "learning_rate": 9.268317739977872e-06,
+      "loss": 0.491,
+      "step": 4249
+    },
+    {
+      "epoch": 1.1785912368275098,
+      "grad_norm": 0.21437488496303558,
+      "learning_rate": 9.262984139532973e-06,
+      "loss": 0.4984,
+      "step": 4250
+    },
+    {
+      "epoch": 1.1788685524126457,
+      "grad_norm": 0.19150178134441376,
+      "learning_rate": 9.257651170715635e-06,
+      "loss": 0.4888,
+      "step": 4251
+    },
+    {
+      "epoch": 1.1791458679977815,
+      "grad_norm": 0.21565623581409454,
+      "learning_rate": 9.25231883456646e-06,
+      "loss": 0.5025,
+      "step": 4252
+    },
+    {
+      "epoch": 1.1794231835829174,
+      "grad_norm": 0.20012259483337402,
+      "learning_rate": 9.246987132125919e-06,
+      "loss": 0.533,
+      "step": 4253
+    },
+    {
+      "epoch": 1.1797004991680533,
+      "grad_norm": 0.1878737509250641,
+      "learning_rate": 9.241656064434382e-06,
+      "loss": 0.5027,
+      "step": 4254
+    },
+    {
+      "epoch": 1.1799778147531892,
+      "grad_norm": 0.19911964237689972,
+      "learning_rate": 9.236325632532074e-06,
+      "loss": 0.494,
+      "step": 4255
+    },
+    {
+      "epoch": 1.180255130338325,
+      "grad_norm": 0.208717480301857,
+      "learning_rate": 9.230995837459103e-06,
+      "loss": 0.5262,
+      "step": 4256
+    },
+    {
+      "epoch": 1.180532445923461,
+      "grad_norm": 0.19873370230197906,
+      "learning_rate": 9.225666680255452e-06,
+      "loss": 0.5128,
+      "step": 4257
+    },
+    {
+      "epoch": 1.1808097615085968,
+      "grad_norm": 0.19655872881412506,
+      "learning_rate": 9.22033816196099e-06,
+      "loss": 0.4987,
+      "step": 4258
+    },
+    {
+      "epoch": 1.1810870770937327,
+      "grad_norm": 0.2015899121761322,
+      "learning_rate": 9.215010283615443e-06,
+      "loss": 0.5177,
+      "step": 4259
+    },
+    {
+      "epoch": 1.1813643926788686,
+      "grad_norm": 0.19780929386615753,
+      "learning_rate": 9.20968304625842e-06,
+      "loss": 0.4959,
+      "step": 4260
+    },
+    {
+      "epoch": 1.1816417082640045,
+      "grad_norm": 0.18897369503974915,
+      "learning_rate": 9.204356450929413e-06,
+      "loss": 0.4847,
+      "step": 4261
+    },
+    {
+      "epoch": 1.1819190238491404,
+      "grad_norm": 0.19352437555789948,
+      "learning_rate": 9.19903049866778e-06,
+      "loss": 0.4902,
+      "step": 4262
+    },
+    {
+      "epoch": 1.1821963394342763,
+      "grad_norm": 0.2053721845149994,
+      "learning_rate": 9.19370519051275e-06,
+      "loss": 0.5054,
+      "step": 4263
+    },
+    {
+      "epoch": 1.1824736550194122,
+      "grad_norm": 0.19881947338581085,
+      "learning_rate": 9.188380527503443e-06,
+      "loss": 0.4888,
+      "step": 4264
+    },
+    {
+      "epoch": 1.182750970604548,
+      "grad_norm": 0.19415348768234253,
+      "learning_rate": 9.18305651067883e-06,
+      "loss": 0.4889,
+      "step": 4265
+    },
+    {
+      "epoch": 1.183028286189684,
+      "grad_norm": 0.1971830129623413,
+      "learning_rate": 9.177733141077775e-06,
+      "loss": 0.5026,
+      "step": 4266
+    },
+    {
+      "epoch": 1.1833056017748198,
+      "grad_norm": 0.19726091623306274,
+      "learning_rate": 9.172410419739e-06,
+      "loss": 0.5027,
+      "step": 4267
+    },
+    {
+      "epoch": 1.1835829173599557,
+      "grad_norm": 0.19601614773273468,
+      "learning_rate": 9.167088347701119e-06,
+      "loss": 0.495,
+      "step": 4268
+    },
+    {
+      "epoch": 1.1838602329450916,
+      "grad_norm": 0.18969157338142395,
+      "learning_rate": 9.16176692600261e-06,
+      "loss": 0.4859,
+      "step": 4269
+    },
+    {
+      "epoch": 1.1841375485302275,
+      "grad_norm": 0.18396854400634766,
+      "learning_rate": 9.156446155681811e-06,
+      "loss": 0.4981,
+      "step": 4270
+    },
+    {
+      "epoch": 1.1844148641153633,
+      "grad_norm": 0.19818167388439178,
+      "learning_rate": 9.151126037776955e-06,
+      "loss": 0.5199,
+      "step": 4271
+    },
+    {
+      "epoch": 1.1846921797004992,
+      "grad_norm": 0.20459376275539398,
+      "learning_rate": 9.145806573326137e-06,
+      "loss": 0.5215,
+      "step": 4272
+    },
+    {
+      "epoch": 1.184969495285635,
+      "grad_norm": 0.1927737444639206,
+      "learning_rate": 9.140487763367328e-06,
+      "loss": 0.5089,
+      "step": 4273
+    },
+    {
+      "epoch": 1.185246810870771,
+      "grad_norm": 0.18682962656021118,
+      "learning_rate": 9.135169608938354e-06,
+      "loss": 0.4812,
+      "step": 4274
+    },
+    {
+      "epoch": 1.1855241264559069,
+      "grad_norm": 0.2066759467124939,
+      "learning_rate": 9.12985211107695e-06,
+      "loss": 0.5111,
+      "step": 4275
+    },
+    {
+      "epoch": 1.1858014420410428,
+      "grad_norm": 0.20103536546230316,
+      "learning_rate": 9.124535270820685e-06,
+      "loss": 0.4976,
+      "step": 4276
+    },
+    {
+      "epoch": 1.1860787576261786,
+      "grad_norm": 0.1959027498960495,
+      "learning_rate": 9.119219089207017e-06,
+      "loss": 0.5396,
+      "step": 4277
+    },
+    {
+      "epoch": 1.1863560732113145,
+      "grad_norm": 0.1914219707250595,
+      "learning_rate": 9.11390356727328e-06,
+      "loss": 0.4949,
+      "step": 4278
+    },
+    {
+      "epoch": 1.1866333887964504,
+      "grad_norm": 0.20492805540561676,
+      "learning_rate": 9.108588706056673e-06,
+      "loss": 0.5225,
+      "step": 4279
+    },
+    {
+      "epoch": 1.1869107043815863,
+      "grad_norm": 0.19187003374099731,
+      "learning_rate": 9.10327450659426e-06,
+      "loss": 0.4893,
+      "step": 4280
+    },
+    {
+      "epoch": 1.1871880199667222,
+      "grad_norm": 0.2073322832584381,
+      "learning_rate": 9.097960969922983e-06,
+      "loss": 0.5182,
+      "step": 4281
+    },
+    {
+      "epoch": 1.187465335551858,
+      "grad_norm": 0.1949753612279892,
+      "learning_rate": 9.092648097079659e-06,
+      "loss": 0.5037,
+      "step": 4282
+    },
+    {
+      "epoch": 1.187742651136994,
+      "grad_norm": 0.20522314310073853,
+      "learning_rate": 9.087335889100967e-06,
+      "loss": 0.5164,
+      "step": 4283
+    },
+    {
+      "epoch": 1.1880199667221298,
+      "grad_norm": 0.19784744083881378,
+      "learning_rate": 9.082024347023457e-06,
+      "loss": 0.5207,
+      "step": 4284
+    },
+    {
+      "epoch": 1.1882972823072657,
+      "grad_norm": 0.20166823267936707,
+      "learning_rate": 9.076713471883557e-06,
+      "loss": 0.5184,
+      "step": 4285
+    },
+    {
+      "epoch": 1.1885745978924016,
+      "grad_norm": 0.20087933540344238,
+      "learning_rate": 9.07140326471756e-06,
+      "loss": 0.4829,
+      "step": 4286
+    },
+    {
+      "epoch": 1.1888519134775375,
+      "grad_norm": 0.2303885966539383,
+      "learning_rate": 9.066093726561622e-06,
+      "loss": 0.5022,
+      "step": 4287
+    },
+    {
+      "epoch": 1.1891292290626734,
+      "grad_norm": 0.19996869564056396,
+      "learning_rate": 9.060784858451774e-06,
+      "loss": 0.4958,
+      "step": 4288
+    },
+    {
+      "epoch": 1.1894065446478093,
+      "grad_norm": 0.19278928637504578,
+      "learning_rate": 9.055476661423925e-06,
+      "loss": 0.5143,
+      "step": 4289
+    },
+    {
+      "epoch": 1.1896838602329451,
+      "grad_norm": 0.19539318978786469,
+      "learning_rate": 9.050169136513842e-06,
+      "loss": 0.4935,
+      "step": 4290
+    },
+    {
+      "epoch": 1.189961175818081,
+      "grad_norm": 0.20165032148361206,
+      "learning_rate": 9.044862284757154e-06,
+      "loss": 0.484,
+      "step": 4291
+    },
+    {
+      "epoch": 1.190238491403217,
+      "grad_norm": 0.2109079658985138,
+      "learning_rate": 9.039556107189384e-06,
+      "loss": 0.5046,
+      "step": 4292
+    },
+    {
+      "epoch": 1.1905158069883528,
+      "grad_norm": 0.19363267719745636,
+      "learning_rate": 9.034250604845898e-06,
+      "loss": 0.4955,
+      "step": 4293
+    },
+    {
+      "epoch": 1.1907931225734887,
+      "grad_norm": 0.190843403339386,
+      "learning_rate": 9.028945778761942e-06,
+      "loss": 0.4936,
+      "step": 4294
+    },
+    {
+      "epoch": 1.1910704381586246,
+      "grad_norm": 0.1952875405550003,
+      "learning_rate": 9.023641629972626e-06,
+      "loss": 0.4903,
+      "step": 4295
+    },
+    {
+      "epoch": 1.1913477537437605,
+      "grad_norm": 0.1969594806432724,
+      "learning_rate": 9.018338159512937e-06,
+      "loss": 0.4951,
+      "step": 4296
+    },
+    {
+      "epoch": 1.1916250693288963,
+      "grad_norm": 0.19602110981941223,
+      "learning_rate": 9.013035368417716e-06,
+      "loss": 0.5064,
+      "step": 4297
+    },
+    {
+      "epoch": 1.1919023849140322,
+      "grad_norm": 0.1936318725347519,
+      "learning_rate": 9.007733257721679e-06,
+      "loss": 0.5272,
+      "step": 4298
+    },
+    {
+      "epoch": 1.192179700499168,
+      "grad_norm": 0.19315633177757263,
+      "learning_rate": 9.002431828459409e-06,
+      "loss": 0.5085,
+      "step": 4299
+    },
+    {
+      "epoch": 1.192457016084304,
+      "grad_norm": 0.20358605682849884,
+      "learning_rate": 8.997131081665357e-06,
+      "loss": 0.5105,
+      "step": 4300
+    },
+    {
+      "epoch": 1.1927343316694399,
+      "grad_norm": 0.194684237241745,
+      "learning_rate": 8.991831018373841e-06,
+      "loss": 0.5007,
+      "step": 4301
+    },
+    {
+      "epoch": 1.1930116472545758,
+      "grad_norm": 0.19725412130355835,
+      "learning_rate": 8.986531639619033e-06,
+      "loss": 0.4953,
+      "step": 4302
+    },
+    {
+      "epoch": 1.1932889628397116,
+      "grad_norm": 0.1992274820804596,
+      "learning_rate": 8.981232946434995e-06,
+      "loss": 0.4917,
+      "step": 4303
+    },
+    {
+      "epoch": 1.1935662784248475,
+      "grad_norm": 0.1960635781288147,
+      "learning_rate": 8.975934939855637e-06,
+      "loss": 0.4854,
+      "step": 4304
+    },
+    {
+      "epoch": 1.1938435940099834,
+      "grad_norm": 0.2065868228673935,
+      "learning_rate": 8.970637620914735e-06,
+      "loss": 0.497,
+      "step": 4305
+    },
+    {
+      "epoch": 1.1941209095951193,
+      "grad_norm": 0.19648823142051697,
+      "learning_rate": 8.965340990645947e-06,
+      "loss": 0.4887,
+      "step": 4306
+    },
+    {
+      "epoch": 1.1943982251802552,
+      "grad_norm": 0.20130853354930878,
+      "learning_rate": 8.960045050082783e-06,
+      "loss": 0.5177,
+      "step": 4307
+    },
+    {
+      "epoch": 1.194675540765391,
+      "grad_norm": 0.20241306722164154,
+      "learning_rate": 8.954749800258615e-06,
+      "loss": 0.5014,
+      "step": 4308
+    },
+    {
+      "epoch": 1.194952856350527,
+      "grad_norm": 0.1977449208498001,
+      "learning_rate": 8.94945524220669e-06,
+      "loss": 0.5068,
+      "step": 4309
+    },
+    {
+      "epoch": 1.1952301719356628,
+      "grad_norm": 0.19154322147369385,
+      "learning_rate": 8.944161376960119e-06,
+      "loss": 0.516,
+      "step": 4310
+    },
+    {
+      "epoch": 1.1955074875207987,
+      "grad_norm": 0.19326725602149963,
+      "learning_rate": 8.938868205551877e-06,
+      "loss": 0.493,
+      "step": 4311
+    },
+    {
+      "epoch": 1.1957848031059346,
+      "grad_norm": 0.1888340413570404,
+      "learning_rate": 8.933575729014788e-06,
+      "loss": 0.4899,
+      "step": 4312
+    },
+    {
+      "epoch": 1.1960621186910705,
+      "grad_norm": 0.19526329636573792,
+      "learning_rate": 8.928283948381575e-06,
+      "loss": 0.4789,
+      "step": 4313
+    },
+    {
+      "epoch": 1.1963394342762064,
+      "grad_norm": 0.1940467208623886,
+      "learning_rate": 8.922992864684791e-06,
+      "loss": 0.5252,
+      "step": 4314
+    },
+    {
+      "epoch": 1.1966167498613423,
+      "grad_norm": 0.20158347487449646,
+      "learning_rate": 8.917702478956872e-06,
+      "loss": 0.5069,
+      "step": 4315
+    },
+    {
+      "epoch": 1.1968940654464781,
+      "grad_norm": 0.20539873838424683,
+      "learning_rate": 8.912412792230104e-06,
+      "loss": 0.4972,
+      "step": 4316
+    },
+    {
+      "epoch": 1.197171381031614,
+      "grad_norm": 0.1870296597480774,
+      "learning_rate": 8.90712380553666e-06,
+      "loss": 0.4905,
+      "step": 4317
+    },
+    {
+      "epoch": 1.19744869661675,
+      "grad_norm": 0.1965525895357132,
+      "learning_rate": 8.90183551990855e-06,
+      "loss": 0.4966,
+      "step": 4318
+    },
+    {
+      "epoch": 1.1977260122018858,
+      "grad_norm": 0.20157171785831451,
+      "learning_rate": 8.896547936377658e-06,
+      "loss": 0.5305,
+      "step": 4319
+    },
+    {
+      "epoch": 1.1980033277870217,
+      "grad_norm": 0.2052687555551529,
+      "learning_rate": 8.89126105597574e-06,
+      "loss": 0.5178,
+      "step": 4320
+    },
+    {
+      "epoch": 1.1982806433721576,
+      "grad_norm": 0.19930841028690338,
+      "learning_rate": 8.885974879734399e-06,
+      "loss": 0.5181,
+      "step": 4321
+    },
+    {
+      "epoch": 1.1985579589572934,
+      "grad_norm": 0.19460836052894592,
+      "learning_rate": 8.880689408685114e-06,
+      "loss": 0.5114,
+      "step": 4322
+    },
+    {
+      "epoch": 1.1988352745424293,
+      "grad_norm": 0.19336189329624176,
+      "learning_rate": 8.87540464385921e-06,
+      "loss": 0.4967,
+      "step": 4323
+    },
+    {
+      "epoch": 1.1991125901275652,
+      "grad_norm": 0.19514434039592743,
+      "learning_rate": 8.8701205862879e-06,
+      "loss": 0.4995,
+      "step": 4324
+    },
+    {
+      "epoch": 1.199389905712701,
+      "grad_norm": 0.19089075922966003,
+      "learning_rate": 8.864837237002232e-06,
+      "loss": 0.4795,
+      "step": 4325
+    },
+    {
+      "epoch": 1.199667221297837,
+      "grad_norm": 0.2026674598455429,
+      "learning_rate": 8.85955459703313e-06,
+      "loss": 0.5033,
+      "step": 4326
+    },
+    {
+      "epoch": 1.1999445368829729,
+      "grad_norm": 0.2059757262468338,
+      "learning_rate": 8.854272667411379e-06,
+      "loss": 0.4985,
+      "step": 4327
+    },
+    {
+      "epoch": 1.2002218524681088,
+      "grad_norm": 0.20040294528007507,
+      "learning_rate": 8.848991449167623e-06,
+      "loss": 0.493,
+      "step": 4328
+    },
+    {
+      "epoch": 1.2004991680532446,
+      "grad_norm": 0.19929878413677216,
+      "learning_rate": 8.843710943332362e-06,
+      "loss": 0.495,
+      "step": 4329
+    },
+    {
+      "epoch": 1.2007764836383805,
+      "grad_norm": 0.1958535760641098,
+      "learning_rate": 8.838431150935975e-06,
+      "loss": 0.4971,
+      "step": 4330
+    },
+    {
+      "epoch": 1.2010537992235164,
+      "grad_norm": 0.18605239689350128,
+      "learning_rate": 8.83315207300868e-06,
+      "loss": 0.4861,
+      "step": 4331
+    },
+    {
+      "epoch": 1.2013311148086523,
+      "grad_norm": 0.43775662779808044,
+      "learning_rate": 8.827873710580564e-06,
+      "loss": 0.4962,
+      "step": 4332
+    },
+    {
+      "epoch": 1.2016084303937882,
+      "grad_norm": 0.1940368264913559,
+      "learning_rate": 8.822596064681577e-06,
+      "loss": 0.5247,
+      "step": 4333
+    },
+    {
+      "epoch": 1.201885745978924,
+      "grad_norm": 0.2012210190296173,
+      "learning_rate": 8.817319136341535e-06,
+      "loss": 0.5178,
+      "step": 4334
+    },
+    {
+      "epoch": 1.20216306156406,
+      "grad_norm": 0.18737319111824036,
+      "learning_rate": 8.812042926590098e-06,
+      "loss": 0.5166,
+      "step": 4335
+    },
+    {
+      "epoch": 1.2024403771491958,
+      "grad_norm": 0.19788531959056854,
+      "learning_rate": 8.806767436456792e-06,
+      "loss": 0.4959,
+      "step": 4336
+    },
+    {
+      "epoch": 1.2027176927343317,
+      "grad_norm": 0.1986195296049118,
+      "learning_rate": 8.801492666971012e-06,
+      "loss": 0.4669,
+      "step": 4337
+    },
+    {
+      "epoch": 1.2029950083194676,
+      "grad_norm": 0.21253184974193573,
+      "learning_rate": 8.796218619162004e-06,
+      "loss": 0.5126,
+      "step": 4338
+    },
+    {
+      "epoch": 1.2032723239046035,
+      "grad_norm": 0.19485290348529816,
+      "learning_rate": 8.790945294058876e-06,
+      "loss": 0.475,
+      "step": 4339
+    },
+    {
+      "epoch": 1.2035496394897394,
+      "grad_norm": 0.19246895611286163,
+      "learning_rate": 8.785672692690584e-06,
+      "loss": 0.5123,
+      "step": 4340
+    },
+    {
+      "epoch": 1.2038269550748752,
+      "grad_norm": 0.21028102934360504,
+      "learning_rate": 8.780400816085963e-06,
+      "loss": 0.5072,
+      "step": 4341
+    },
+    {
+      "epoch": 1.2041042706600111,
+      "grad_norm": 0.20249207317829132,
+      "learning_rate": 8.775129665273691e-06,
+      "loss": 0.4979,
+      "step": 4342
+    },
+    {
+      "epoch": 1.204381586245147,
+      "grad_norm": 0.19815127551555634,
+      "learning_rate": 8.769859241282307e-06,
+      "loss": 0.5116,
+      "step": 4343
+    },
+    {
+      "epoch": 1.204658901830283,
+      "grad_norm": 0.20030280947685242,
+      "learning_rate": 8.764589545140217e-06,
+      "loss": 0.5121,
+      "step": 4344
+    },
+    {
+      "epoch": 1.2049362174154188,
+      "grad_norm": 0.2132977396249771,
+      "learning_rate": 8.759320577875676e-06,
+      "loss": 0.5192,
+      "step": 4345
+    },
+    {
+      "epoch": 1.2052135330005547,
+      "grad_norm": 0.19407188892364502,
+      "learning_rate": 8.754052340516796e-06,
+      "loss": 0.4912,
+      "step": 4346
+    },
+    {
+      "epoch": 1.2054908485856906,
+      "grad_norm": 0.2085913121700287,
+      "learning_rate": 8.748784834091549e-06,
+      "loss": 0.4861,
+      "step": 4347
+    },
+    {
+      "epoch": 1.2057681641708264,
+      "grad_norm": 0.20071199536323547,
+      "learning_rate": 8.74351805962777e-06,
+      "loss": 0.5451,
+      "step": 4348
+    },
+    {
+      "epoch": 1.2060454797559623,
+      "grad_norm": 0.20174618065357208,
+      "learning_rate": 8.738252018153145e-06,
+      "loss": 0.5094,
+      "step": 4349
+    },
+    {
+      "epoch": 1.2063227953410982,
+      "grad_norm": 0.2093082219362259,
+      "learning_rate": 8.73298671069521e-06,
+      "loss": 0.5161,
+      "step": 4350
+    },
+    {
+      "epoch": 1.206600110926234,
+      "grad_norm": 0.21680103242397308,
+      "learning_rate": 8.727722138281381e-06,
+      "loss": 0.5058,
+      "step": 4351
+    },
+    {
+      "epoch": 1.20687742651137,
+      "grad_norm": 0.19285671412944794,
+      "learning_rate": 8.722458301938904e-06,
+      "loss": 0.4775,
+      "step": 4352
+    },
+    {
+      "epoch": 1.2071547420965059,
+      "grad_norm": 0.20253880321979523,
+      "learning_rate": 8.717195202694898e-06,
+      "loss": 0.501,
+      "step": 4353
+    },
+    {
+      "epoch": 1.2074320576816417,
+      "grad_norm": 0.19431787729263306,
+      "learning_rate": 8.711932841576325e-06,
+      "loss": 0.4927,
+      "step": 4354
+    },
+    {
+      "epoch": 1.2077093732667776,
+      "grad_norm": 0.20108050107955933,
+      "learning_rate": 8.706671219610027e-06,
+      "loss": 0.5135,
+      "step": 4355
+    },
+    {
+      "epoch": 1.2079866888519135,
+      "grad_norm": 0.20072422921657562,
+      "learning_rate": 8.70141033782267e-06,
+      "loss": 0.507,
+      "step": 4356
+    },
+    {
+      "epoch": 1.2082640044370494,
+      "grad_norm": 0.21369212865829468,
+      "learning_rate": 8.696150197240798e-06,
+      "loss": 0.5259,
+      "step": 4357
+    },
+    {
+      "epoch": 1.2085413200221853,
+      "grad_norm": 0.1956164538860321,
+      "learning_rate": 8.690890798890806e-06,
+      "loss": 0.5293,
+      "step": 4358
+    },
+    {
+      "epoch": 1.2088186356073212,
+      "grad_norm": 0.19677644968032837,
+      "learning_rate": 8.685632143798938e-06,
+      "loss": 0.5185,
+      "step": 4359
+    },
+    {
+      "epoch": 1.209095951192457,
+      "grad_norm": 0.19301322102546692,
+      "learning_rate": 8.680374232991304e-06,
+      "loss": 0.5092,
+      "step": 4360
+    },
+    {
+      "epoch": 1.209373266777593,
+      "grad_norm": 0.19476044178009033,
+      "learning_rate": 8.67511706749385e-06,
+      "loss": 0.5197,
+      "step": 4361
+    },
+    {
+      "epoch": 1.2096505823627288,
+      "grad_norm": 0.20564420521259308,
+      "learning_rate": 8.669860648332395e-06,
+      "loss": 0.519,
+      "step": 4362
+    },
+    {
+      "epoch": 1.2099278979478647,
+      "grad_norm": 0.20334792137145996,
+      "learning_rate": 8.664604976532605e-06,
+      "loss": 0.5407,
+      "step": 4363
+    },
+    {
+      "epoch": 1.2102052135330006,
+      "grad_norm": 0.1910853236913681,
+      "learning_rate": 8.659350053120003e-06,
+      "loss": 0.5174,
+      "step": 4364
+    },
+    {
+      "epoch": 1.2104825291181365,
+      "grad_norm": 0.1837315857410431,
+      "learning_rate": 8.65409587911996e-06,
+      "loss": 0.5198,
+      "step": 4365
+    },
+    {
+      "epoch": 1.2107598447032724,
+      "grad_norm": 0.19404159486293793,
+      "learning_rate": 8.64884245555771e-06,
+      "loss": 0.4944,
+      "step": 4366
+    },
+    {
+      "epoch": 1.2110371602884082,
+      "grad_norm": 0.18698063492774963,
+      "learning_rate": 8.643589783458328e-06,
+      "loss": 0.5151,
+      "step": 4367
+    },
+    {
+      "epoch": 1.2113144758735441,
+      "grad_norm": 0.2053699940443039,
+      "learning_rate": 8.638337863846752e-06,
+      "loss": 0.5182,
+      "step": 4368
+    },
+    {
+      "epoch": 1.21159179145868,
+      "grad_norm": 0.18703065812587738,
+      "learning_rate": 8.633086697747773e-06,
+      "loss": 0.4803,
+      "step": 4369
+    },
+    {
+      "epoch": 1.211869107043816,
+      "grad_norm": 0.19493581354618073,
+      "learning_rate": 8.627836286186035e-06,
+      "loss": 0.5313,
+      "step": 4370
+    },
+    {
+      "epoch": 1.2121464226289518,
+      "grad_norm": 0.1981756091117859,
+      "learning_rate": 8.622586630186019e-06,
+      "loss": 0.5012,
+      "step": 4371
+    },
+    {
+      "epoch": 1.2124237382140877,
+      "grad_norm": 0.18528394401073456,
+      "learning_rate": 8.61733773077209e-06,
+      "loss": 0.4845,
+      "step": 4372
+    },
+    {
+      "epoch": 1.2127010537992235,
+      "grad_norm": 0.20212890207767487,
+      "learning_rate": 8.612089588968437e-06,
+      "loss": 0.498,
+      "step": 4373
+    },
+    {
+      "epoch": 1.2129783693843594,
+      "grad_norm": 0.19095052778720856,
+      "learning_rate": 8.606842205799113e-06,
+      "loss": 0.4707,
+      "step": 4374
+    },
+    {
+      "epoch": 1.2132556849694953,
+      "grad_norm": 0.18967384099960327,
+      "learning_rate": 8.60159558228802e-06,
+      "loss": 0.5123,
+      "step": 4375
+    },
+    {
+      "epoch": 1.2135330005546312,
+      "grad_norm": 0.20997461676597595,
+      "learning_rate": 8.596349719458916e-06,
+      "loss": 0.4683,
+      "step": 4376
+    },
+    {
+      "epoch": 1.213810316139767,
+      "grad_norm": 0.19758398830890656,
+      "learning_rate": 8.591104618335413e-06,
+      "loss": 0.5227,
+      "step": 4377
+    },
+    {
+      "epoch": 1.214087631724903,
+      "grad_norm": 0.19900359213352203,
+      "learning_rate": 8.585860279940954e-06,
+      "loss": 0.4995,
+      "step": 4378
+    },
+    {
+      "epoch": 1.2143649473100389,
+      "grad_norm": 0.19332247972488403,
+      "learning_rate": 8.580616705298864e-06,
+      "loss": 0.5178,
+      "step": 4379
+    },
+    {
+      "epoch": 1.2146422628951747,
+      "grad_norm": 0.18881361186504364,
+      "learning_rate": 8.575373895432293e-06,
+      "loss": 0.493,
+      "step": 4380
+    },
+    {
+      "epoch": 1.2149195784803106,
+      "grad_norm": 0.222880020737648,
+      "learning_rate": 8.57013185136426e-06,
+      "loss": 0.4933,
+      "step": 4381
+    },
+    {
+      "epoch": 1.2151968940654465,
+      "grad_norm": 0.19780275225639343,
+      "learning_rate": 8.564890574117616e-06,
+      "loss": 0.4966,
+      "step": 4382
+    },
+    {
+      "epoch": 1.2154742096505824,
+      "grad_norm": 0.1969325840473175,
+      "learning_rate": 8.559650064715088e-06,
+      "loss": 0.5132,
+      "step": 4383
+    },
+    {
+      "epoch": 1.2157515252357183,
+      "grad_norm": 0.2127065658569336,
+      "learning_rate": 8.554410324179226e-06,
+      "loss": 0.496,
+      "step": 4384
+    },
+    {
+      "epoch": 1.2160288408208542,
+      "grad_norm": 0.20196539163589478,
+      "learning_rate": 8.549171353532443e-06,
+      "loss": 0.4894,
+      "step": 4385
+    },
+    {
+      "epoch": 1.21630615640599,
+      "grad_norm": 0.19963231682777405,
+      "learning_rate": 8.543933153797007e-06,
+      "loss": 0.5355,
+      "step": 4386
+    },
+    {
+      "epoch": 1.216583471991126,
+      "grad_norm": 0.20684166252613068,
+      "learning_rate": 8.53869572599503e-06,
+      "loss": 0.497,
+      "step": 4387
+    },
+    {
+      "epoch": 1.2168607875762618,
+      "grad_norm": 0.20197083055973053,
+      "learning_rate": 8.533459071148462e-06,
+      "loss": 0.4998,
+      "step": 4388
+    },
+    {
+      "epoch": 1.2171381031613977,
+      "grad_norm": 0.19906753301620483,
+      "learning_rate": 8.528223190279128e-06,
+      "loss": 0.4932,
+      "step": 4389
+    },
+    {
+      "epoch": 1.2174154187465336,
+      "grad_norm": 0.1864069402217865,
+      "learning_rate": 8.522988084408678e-06,
+      "loss": 0.4894,
+      "step": 4390
+    },
+    {
+      "epoch": 1.2176927343316695,
+      "grad_norm": 0.1982210874557495,
+      "learning_rate": 8.517753754558621e-06,
+      "loss": 0.4868,
+      "step": 4391
+    },
+    {
+      "epoch": 1.2179700499168054,
+      "grad_norm": 0.19526433944702148,
+      "learning_rate": 8.512520201750312e-06,
+      "loss": 0.5154,
+      "step": 4392
+    },
+    {
+      "epoch": 1.2182473655019412,
+      "grad_norm": 0.19613106548786163,
+      "learning_rate": 8.507287427004962e-06,
+      "loss": 0.5095,
+      "step": 4393
+    },
+    {
+      "epoch": 1.2185246810870771,
+      "grad_norm": 0.18780362606048584,
+      "learning_rate": 8.502055431343618e-06,
+      "loss": 0.5064,
+      "step": 4394
+    },
+    {
+      "epoch": 1.218801996672213,
+      "grad_norm": 0.20160475373268127,
+      "learning_rate": 8.49682421578718e-06,
+      "loss": 0.5269,
+      "step": 4395
+    },
+    {
+      "epoch": 1.219079312257349,
+      "grad_norm": 0.22338464856147766,
+      "learning_rate": 8.491593781356404e-06,
+      "loss": 0.5086,
+      "step": 4396
+    },
+    {
+      "epoch": 1.2193566278424848,
+      "grad_norm": 0.18976663053035736,
+      "learning_rate": 8.48636412907188e-06,
+      "loss": 0.5017,
+      "step": 4397
+    },
+    {
+      "epoch": 1.2196339434276207,
+      "grad_norm": 0.18721407651901245,
+      "learning_rate": 8.481135259954057e-06,
+      "loss": 0.4921,
+      "step": 4398
+    },
+    {
+      "epoch": 1.2199112590127565,
+      "grad_norm": 0.19557468593120575,
+      "learning_rate": 8.475907175023218e-06,
+      "loss": 0.495,
+      "step": 4399
+    },
+    {
+      "epoch": 1.2201885745978924,
+      "grad_norm": 0.1959930807352066,
+      "learning_rate": 8.470679875299507e-06,
+      "loss": 0.5033,
+      "step": 4400
+    },
+    {
+      "epoch": 1.2204658901830283,
+      "grad_norm": 0.19299939274787903,
+      "learning_rate": 8.465453361802907e-06,
+      "loss": 0.5275,
+      "step": 4401
+    },
+    {
+      "epoch": 1.2207432057681642,
+      "grad_norm": 0.19271641969680786,
+      "learning_rate": 8.460227635553247e-06,
+      "loss": 0.4899,
+      "step": 4402
+    },
+    {
+      "epoch": 1.2210205213533,
+      "grad_norm": 0.19292891025543213,
+      "learning_rate": 8.455002697570211e-06,
+      "loss": 0.4917,
+      "step": 4403
+    },
+    {
+      "epoch": 1.221297836938436,
+      "grad_norm": 0.19001443684101105,
+      "learning_rate": 8.44977854887332e-06,
+      "loss": 0.4953,
+      "step": 4404
+    },
+    {
+      "epoch": 1.2215751525235718,
+      "grad_norm": 0.2002885341644287,
+      "learning_rate": 8.44455519048194e-06,
+      "loss": 0.5099,
+      "step": 4405
+    },
+    {
+      "epoch": 1.2218524681087077,
+      "grad_norm": 0.1973315179347992,
+      "learning_rate": 8.439332623415287e-06,
+      "loss": 0.5097,
+      "step": 4406
+    },
+    {
+      "epoch": 1.2221297836938436,
+      "grad_norm": 0.1916041374206543,
+      "learning_rate": 8.434110848692427e-06,
+      "loss": 0.5177,
+      "step": 4407
+    },
+    {
+      "epoch": 1.2224070992789795,
+      "grad_norm": 0.2049477994441986,
+      "learning_rate": 8.428889867332268e-06,
+      "loss": 0.4952,
+      "step": 4408
+    },
+    {
+      "epoch": 1.2226844148641154,
+      "grad_norm": 0.2035469114780426,
+      "learning_rate": 8.423669680353549e-06,
+      "loss": 0.5438,
+      "step": 4409
+    },
+    {
+      "epoch": 1.2229617304492513,
+      "grad_norm": 0.1949811726808548,
+      "learning_rate": 8.418450288774884e-06,
+      "loss": 0.5214,
+      "step": 4410
+    },
+    {
+      "epoch": 1.2232390460343872,
+      "grad_norm": 0.22771196067333221,
+      "learning_rate": 8.413231693614704e-06,
+      "loss": 0.5108,
+      "step": 4411
+    },
+    {
+      "epoch": 1.223516361619523,
+      "grad_norm": 0.19894933700561523,
+      "learning_rate": 8.408013895891295e-06,
+      "loss": 0.4941,
+      "step": 4412
+    },
+    {
+      "epoch": 1.223793677204659,
+      "grad_norm": 0.19783443212509155,
+      "learning_rate": 8.40279689662279e-06,
+      "loss": 0.4934,
+      "step": 4413
+    },
+    {
+      "epoch": 1.2240709927897948,
+      "grad_norm": 0.1914057582616806,
+      "learning_rate": 8.397580696827166e-06,
+      "loss": 0.5151,
+      "step": 4414
+    },
+    {
+      "epoch": 1.2243483083749307,
+      "grad_norm": 0.19515813887119293,
+      "learning_rate": 8.392365297522243e-06,
+      "loss": 0.4888,
+      "step": 4415
+    },
+    {
+      "epoch": 1.2246256239600666,
+      "grad_norm": 0.19242678582668304,
+      "learning_rate": 8.387150699725673e-06,
+      "loss": 0.5101,
+      "step": 4416
+    },
+    {
+      "epoch": 1.2249029395452025,
+      "grad_norm": 0.2145782858133316,
+      "learning_rate": 8.381936904454973e-06,
+      "loss": 0.5157,
+      "step": 4417
+    },
+    {
+      "epoch": 1.2251802551303383,
+      "grad_norm": 0.19011950492858887,
+      "learning_rate": 8.376723912727488e-06,
+      "loss": 0.4891,
+      "step": 4418
+    },
+    {
+      "epoch": 1.2254575707154742,
+      "grad_norm": 0.19757391512393951,
+      "learning_rate": 8.371511725560416e-06,
+      "loss": 0.5173,
+      "step": 4419
+    },
+    {
+      "epoch": 1.2257348863006101,
+      "grad_norm": 0.2027006596326828,
+      "learning_rate": 8.36630034397078e-06,
+      "loss": 0.5116,
+      "step": 4420
+    },
+    {
+      "epoch": 1.226012201885746,
+      "grad_norm": 0.2004849761724472,
+      "learning_rate": 8.361089768975475e-06,
+      "loss": 0.5077,
+      "step": 4421
+    },
+    {
+      "epoch": 1.2262895174708819,
+      "grad_norm": 0.20179691910743713,
+      "learning_rate": 8.355880001591212e-06,
+      "loss": 0.4969,
+      "step": 4422
+    },
+    {
+      "epoch": 1.2265668330560178,
+      "grad_norm": 0.20845246315002441,
+      "learning_rate": 8.350671042834555e-06,
+      "loss": 0.4789,
+      "step": 4423
+    },
+    {
+      "epoch": 1.2268441486411537,
+      "grad_norm": 0.19270509481430054,
+      "learning_rate": 8.345462893721911e-06,
+      "loss": 0.4874,
+      "step": 4424
+    },
+    {
+      "epoch": 1.2271214642262895,
+      "grad_norm": 0.19383633136749268,
+      "learning_rate": 8.340255555269535e-06,
+      "loss": 0.5006,
+      "step": 4425
+    },
+    {
+      "epoch": 1.2273987798114254,
+      "grad_norm": 0.18180836737155914,
+      "learning_rate": 8.335049028493509e-06,
+      "loss": 0.5117,
+      "step": 4426
+    },
+    {
+      "epoch": 1.2276760953965613,
+      "grad_norm": 0.196406289935112,
+      "learning_rate": 8.32984331440976e-06,
+      "loss": 0.5224,
+      "step": 4427
+    },
+    {
+      "epoch": 1.2279534109816972,
+      "grad_norm": 0.19745458662509918,
+      "learning_rate": 8.324638414034069e-06,
+      "loss": 0.4721,
+      "step": 4428
+    },
+    {
+      "epoch": 1.228230726566833,
+      "grad_norm": 0.2035657912492752,
+      "learning_rate": 8.31943432838205e-06,
+      "loss": 0.4984,
+      "step": 4429
+    },
+    {
+      "epoch": 1.228508042151969,
+      "grad_norm": 0.204306960105896,
+      "learning_rate": 8.314231058469152e-06,
+      "loss": 0.4978,
+      "step": 4430
+    },
+    {
+      "epoch": 1.2287853577371048,
+      "grad_norm": 0.20982758700847626,
+      "learning_rate": 8.309028605310679e-06,
+      "loss": 0.4996,
+      "step": 4431
+    },
+    {
+      "epoch": 1.2290626733222407,
+      "grad_norm": 0.19134745001792908,
+      "learning_rate": 8.30382696992176e-06,
+      "loss": 0.4947,
+      "step": 4432
+    },
+    {
+      "epoch": 1.2293399889073766,
+      "grad_norm": 0.19805335998535156,
+      "learning_rate": 8.298626153317376e-06,
+      "loss": 0.5233,
+      "step": 4433
+    },
+    {
+      "epoch": 1.2296173044925125,
+      "grad_norm": 0.19368603825569153,
+      "learning_rate": 8.293426156512341e-06,
+      "loss": 0.5168,
+      "step": 4434
+    },
+    {
+      "epoch": 1.2298946200776484,
+      "grad_norm": 0.19627097249031067,
+      "learning_rate": 8.288226980521314e-06,
+      "loss": 0.4882,
+      "step": 4435
+    },
+    {
+      "epoch": 1.2301719356627843,
+      "grad_norm": 0.19731606543064117,
+      "learning_rate": 8.283028626358796e-06,
+      "loss": 0.5079,
+      "step": 4436
+    },
+    {
+      "epoch": 1.2304492512479202,
+      "grad_norm": 0.22327442467212677,
+      "learning_rate": 8.277831095039113e-06,
+      "loss": 0.5173,
+      "step": 4437
+    },
+    {
+      "epoch": 1.230726566833056,
+      "grad_norm": 0.2100764513015747,
+      "learning_rate": 8.272634387576453e-06,
+      "loss": 0.501,
+      "step": 4438
+    },
+    {
+      "epoch": 1.231003882418192,
+      "grad_norm": 0.19981728494167328,
+      "learning_rate": 8.267438504984823e-06,
+      "loss": 0.477,
+      "step": 4439
+    },
+    {
+      "epoch": 1.2312811980033278,
+      "grad_norm": 0.19193001091480255,
+      "learning_rate": 8.262243448278084e-06,
+      "loss": 0.5078,
+      "step": 4440
+    },
+    {
+      "epoch": 1.2315585135884637,
+      "grad_norm": 0.2083725780248642,
+      "learning_rate": 8.257049218469917e-06,
+      "loss": 0.5017,
+      "step": 4441
+    },
+    {
+      "epoch": 1.2318358291735996,
+      "grad_norm": 0.19924737513065338,
+      "learning_rate": 8.251855816573873e-06,
+      "loss": 0.5036,
+      "step": 4442
+    },
+    {
+      "epoch": 1.2321131447587355,
+      "grad_norm": 0.19425547122955322,
+      "learning_rate": 8.246663243603305e-06,
+      "loss": 0.5207,
+      "step": 4443
+    },
+    {
+      "epoch": 1.2323904603438713,
+      "grad_norm": 0.20682436227798462,
+      "learning_rate": 8.241471500571428e-06,
+      "loss": 0.5119,
+      "step": 4444
+    },
+    {
+      "epoch": 1.2326677759290072,
+      "grad_norm": 0.1909773349761963,
+      "learning_rate": 8.236280588491292e-06,
+      "loss": 0.4723,
+      "step": 4445
+    },
+    {
+      "epoch": 1.232945091514143,
+      "grad_norm": 0.2076360434293747,
+      "learning_rate": 8.231090508375777e-06,
+      "loss": 0.4959,
+      "step": 4446
+    },
+    {
+      "epoch": 1.233222407099279,
+      "grad_norm": 0.19352662563323975,
+      "learning_rate": 8.225901261237609e-06,
+      "loss": 0.5002,
+      "step": 4447
+    },
+    {
+      "epoch": 1.2334997226844149,
+      "grad_norm": 0.21294668316841125,
+      "learning_rate": 8.220712848089338e-06,
+      "loss": 0.5099,
+      "step": 4448
+    },
+    {
+      "epoch": 1.2337770382695508,
+      "grad_norm": 0.20197489857673645,
+      "learning_rate": 8.215525269943374e-06,
+      "loss": 0.4944,
+      "step": 4449
+    },
+    {
+      "epoch": 1.2340543538546866,
+      "grad_norm": 0.20826977491378784,
+      "learning_rate": 8.210338527811943e-06,
+      "loss": 0.5321,
+      "step": 4450
+    },
+    {
+      "epoch": 1.2343316694398225,
+      "grad_norm": 0.20179514586925507,
+      "learning_rate": 8.205152622707116e-06,
+      "loss": 0.5053,
+      "step": 4451
+    },
+    {
+      "epoch": 1.2346089850249584,
+      "grad_norm": 0.21204441785812378,
+      "learning_rate": 8.199967555640805e-06,
+      "loss": 0.4856,
+      "step": 4452
+    },
+    {
+      "epoch": 1.2348863006100943,
+      "grad_norm": 0.20789694786071777,
+      "learning_rate": 8.194783327624751e-06,
+      "loss": 0.5174,
+      "step": 4453
+    },
+    {
+      "epoch": 1.2351636161952302,
+      "grad_norm": 0.207809716463089,
+      "learning_rate": 8.189599939670531e-06,
+      "loss": 0.5028,
+      "step": 4454
+    },
+    {
+      "epoch": 1.235440931780366,
+      "grad_norm": 0.3477715849876404,
+      "learning_rate": 8.184417392789568e-06,
+      "loss": 0.4812,
+      "step": 4455
+    },
+    {
+      "epoch": 1.235718247365502,
+      "grad_norm": 0.20378254354000092,
+      "learning_rate": 8.179235687993108e-06,
+      "loss": 0.499,
+      "step": 4456
+    },
+    {
+      "epoch": 1.2359955629506378,
+      "grad_norm": 0.19878922402858734,
+      "learning_rate": 8.174054826292249e-06,
+      "loss": 0.508,
+      "step": 4457
+    },
+    {
+      "epoch": 1.2362728785357737,
+      "grad_norm": 0.21171467006206512,
+      "learning_rate": 8.168874808697896e-06,
+      "loss": 0.5037,
+      "step": 4458
+    },
+    {
+      "epoch": 1.2365501941209096,
+      "grad_norm": 0.1954033523797989,
+      "learning_rate": 8.163695636220828e-06,
+      "loss": 0.5014,
+      "step": 4459
+    },
+    {
+      "epoch": 1.2368275097060455,
+      "grad_norm": 0.19483618438243866,
+      "learning_rate": 8.158517309871626e-06,
+      "loss": 0.4929,
+      "step": 4460
+    },
+    {
+      "epoch": 1.2371048252911814,
+      "grad_norm": 0.19928301870822906,
+      "learning_rate": 8.153339830660719e-06,
+      "loss": 0.5078,
+      "step": 4461
+    },
+    {
+      "epoch": 1.2373821408763173,
+      "grad_norm": 0.20188100636005402,
+      "learning_rate": 8.148163199598379e-06,
+      "loss": 0.4913,
+      "step": 4462
+    },
+    {
+      "epoch": 1.2376594564614531,
+      "grad_norm": 0.21561436355113983,
+      "learning_rate": 8.142987417694699e-06,
+      "loss": 0.4676,
+      "step": 4463
+    },
+    {
+      "epoch": 1.237936772046589,
+      "grad_norm": 0.21500158309936523,
+      "learning_rate": 8.137812485959608e-06,
+      "loss": 0.4817,
+      "step": 4464
+    },
+    {
+      "epoch": 1.238214087631725,
+      "grad_norm": 0.20233996212482452,
+      "learning_rate": 8.132638405402874e-06,
+      "loss": 0.5101,
+      "step": 4465
+    },
+    {
+      "epoch": 1.2384914032168608,
+      "grad_norm": 0.20174197852611542,
+      "learning_rate": 8.1274651770341e-06,
+      "loss": 0.4972,
+      "step": 4466
+    },
+    {
+      "epoch": 1.2387687188019967,
+      "grad_norm": 0.190217986702919,
+      "learning_rate": 8.122292801862716e-06,
+      "loss": 0.498,
+      "step": 4467
+    },
+    {
+      "epoch": 1.2390460343871326,
+      "grad_norm": 0.20103545486927032,
+      "learning_rate": 8.11712128089799e-06,
+      "loss": 0.5011,
+      "step": 4468
+    },
+    {
+      "epoch": 1.2393233499722685,
+      "grad_norm": 0.20134317874908447,
+      "learning_rate": 8.111950615149031e-06,
+      "loss": 0.5089,
+      "step": 4469
+    },
+    {
+      "epoch": 1.2396006655574043,
+      "grad_norm": 0.20507292449474335,
+      "learning_rate": 8.106780805624764e-06,
+      "loss": 0.5092,
+      "step": 4470
+    },
+    {
+      "epoch": 1.2398779811425402,
+      "grad_norm": 0.19532889127731323,
+      "learning_rate": 8.101611853333955e-06,
+      "loss": 0.4975,
+      "step": 4471
+    },
+    {
+      "epoch": 1.240155296727676,
+      "grad_norm": 0.20147058367729187,
+      "learning_rate": 8.096443759285206e-06,
+      "loss": 0.5055,
+      "step": 4472
+    },
+    {
+      "epoch": 1.240432612312812,
+      "grad_norm": 0.20608854293823242,
+      "learning_rate": 8.09127652448695e-06,
+      "loss": 0.5094,
+      "step": 4473
+    },
+    {
+      "epoch": 1.2407099278979479,
+      "grad_norm": 0.20161250233650208,
+      "learning_rate": 8.086110149947457e-06,
+      "loss": 0.5182,
+      "step": 4474
+    },
+    {
+      "epoch": 1.2409872434830838,
+      "grad_norm": 0.19690173864364624,
+      "learning_rate": 8.080944636674811e-06,
+      "loss": 0.5211,
+      "step": 4475
+    },
+    {
+      "epoch": 1.2412645590682196,
+      "grad_norm": 0.1985633373260498,
+      "learning_rate": 8.075779985676949e-06,
+      "loss": 0.4994,
+      "step": 4476
+    },
+    {
+      "epoch": 1.2415418746533555,
+      "grad_norm": 0.1929730325937271,
+      "learning_rate": 8.070616197961631e-06,
+      "loss": 0.4815,
+      "step": 4477
+    },
+    {
+      "epoch": 1.2418191902384914,
+      "grad_norm": 0.19387488067150116,
+      "learning_rate": 8.065453274536447e-06,
+      "loss": 0.5091,
+      "step": 4478
+    },
+    {
+      "epoch": 1.2420965058236273,
+      "grad_norm": 0.19065174460411072,
+      "learning_rate": 8.060291216408814e-06,
+      "loss": 0.4868,
+      "step": 4479
+    },
+    {
+      "epoch": 1.2423738214087632,
+      "grad_norm": 0.20745143294334412,
+      "learning_rate": 8.055130024586e-06,
+      "loss": 0.4938,
+      "step": 4480
+    },
+    {
+      "epoch": 1.242651136993899,
+      "grad_norm": 0.19908511638641357,
+      "learning_rate": 8.04996970007508e-06,
+      "loss": 0.4939,
+      "step": 4481
+    },
+    {
+      "epoch": 1.242928452579035,
+      "grad_norm": 0.2046736627817154,
+      "learning_rate": 8.044810243882971e-06,
+      "loss": 0.5194,
+      "step": 4482
+    },
+    {
+      "epoch": 1.2432057681641708,
+      "grad_norm": 0.20767585933208466,
+      "learning_rate": 8.039651657016423e-06,
+      "loss": 0.5176,
+      "step": 4483
+    },
+    {
+      "epoch": 1.2434830837493067,
+      "grad_norm": 0.20827417075634003,
+      "learning_rate": 8.034493940482016e-06,
+      "loss": 0.4997,
+      "step": 4484
+    },
+    {
+      "epoch": 1.2437603993344426,
+      "grad_norm": 0.21914151310920715,
+      "learning_rate": 8.029337095286147e-06,
+      "loss": 0.4901,
+      "step": 4485
+    },
+    {
+      "epoch": 1.2440377149195785,
+      "grad_norm": 0.21400579810142517,
+      "learning_rate": 8.024181122435058e-06,
+      "loss": 0.4926,
+      "step": 4486
+    },
+    {
+      "epoch": 1.2443150305047144,
+      "grad_norm": 0.19826753437519073,
+      "learning_rate": 8.01902602293482e-06,
+      "loss": 0.5113,
+      "step": 4487
+    },
+    {
+      "epoch": 1.2445923460898503,
+      "grad_norm": 0.20881658792495728,
+      "learning_rate": 8.013871797791324e-06,
+      "loss": 0.4921,
+      "step": 4488
+    },
+    {
+      "epoch": 1.2448696616749861,
+      "grad_norm": 0.18899375200271606,
+      "learning_rate": 8.008718448010297e-06,
+      "loss": 0.5139,
+      "step": 4489
+    },
+    {
+      "epoch": 1.245146977260122,
+      "grad_norm": 0.1959080994129181,
+      "learning_rate": 8.003565974597298e-06,
+      "loss": 0.4864,
+      "step": 4490
+    },
+    {
+      "epoch": 1.245424292845258,
+      "grad_norm": 0.20414955914020538,
+      "learning_rate": 7.99841437855771e-06,
+      "loss": 0.4847,
+      "step": 4491
+    },
+    {
+      "epoch": 1.2457016084303938,
+      "grad_norm": 0.1957169771194458,
+      "learning_rate": 7.993263660896738e-06,
+      "loss": 0.5117,
+      "step": 4492
+    },
+    {
+      "epoch": 1.2459789240155297,
+      "grad_norm": 0.2063983678817749,
+      "learning_rate": 7.988113822619431e-06,
+      "loss": 0.4969,
+      "step": 4493
+    },
+    {
+      "epoch": 1.2462562396006656,
+      "grad_norm": 0.1924324780702591,
+      "learning_rate": 7.982964864730658e-06,
+      "loss": 0.4815,
+      "step": 4494
+    },
+    {
+      "epoch": 1.2465335551858014,
+      "grad_norm": 0.19338028132915497,
+      "learning_rate": 7.97781678823512e-06,
+      "loss": 0.4859,
+      "step": 4495
+    },
+    {
+      "epoch": 1.2468108707709373,
+      "grad_norm": 0.19696307182312012,
+      "learning_rate": 7.972669594137333e-06,
+      "loss": 0.5102,
+      "step": 4496
+    },
+    {
+      "epoch": 1.2470881863560732,
+      "grad_norm": 0.19288606941699982,
+      "learning_rate": 7.967523283441664e-06,
+      "loss": 0.4892,
+      "step": 4497
+    },
+    {
+      "epoch": 1.247365501941209,
+      "grad_norm": 0.19867482781410217,
+      "learning_rate": 7.962377857152284e-06,
+      "loss": 0.4955,
+      "step": 4498
+    },
+    {
+      "epoch": 1.247642817526345,
+      "grad_norm": 0.20236186683177948,
+      "learning_rate": 7.957233316273211e-06,
+      "loss": 0.5143,
+      "step": 4499
+    },
+    {
+      "epoch": 1.2479201331114809,
+      "grad_norm": 0.2065366804599762,
+      "learning_rate": 7.952089661808268e-06,
+      "loss": 0.4996,
+      "step": 4500
+    },
+    {
+      "epoch": 1.2481974486966168,
+      "grad_norm": 0.1983073204755783,
+      "learning_rate": 7.946946894761134e-06,
+      "loss": 0.4892,
+      "step": 4501
+    },
+    {
+      "epoch": 1.2484747642817526,
+      "grad_norm": 0.19757206737995148,
+      "learning_rate": 7.94180501613529e-06,
+      "loss": 0.5111,
+      "step": 4502
+    },
+    {
+      "epoch": 1.2487520798668885,
+      "grad_norm": 0.20565056800842285,
+      "learning_rate": 7.936664026934052e-06,
+      "loss": 0.4943,
+      "step": 4503
+    },
+    {
+      "epoch": 1.2490293954520244,
+      "grad_norm": 0.19800062477588654,
+      "learning_rate": 7.931523928160567e-06,
+      "loss": 0.4838,
+      "step": 4504
+    },
+    {
+      "epoch": 1.2493067110371603,
+      "grad_norm": 0.2019055336713791,
+      "learning_rate": 7.926384720817807e-06,
+      "loss": 0.5413,
+      "step": 4505
+    },
+    {
+      "epoch": 1.2495840266222962,
+      "grad_norm": 0.19655373692512512,
+      "learning_rate": 7.921246405908558e-06,
+      "loss": 0.5043,
+      "step": 4506
+    },
+    {
+      "epoch": 1.249861342207432,
+      "grad_norm": 0.1993308812379837,
+      "learning_rate": 7.916108984435448e-06,
+      "loss": 0.5122,
+      "step": 4507
+    },
+    {
+      "epoch": 1.250138657792568,
+      "grad_norm": 0.1926482617855072,
+      "learning_rate": 7.910972457400923e-06,
+      "loss": 0.5103,
+      "step": 4508
+    },
+    {
+      "epoch": 1.2504159733777038,
+      "grad_norm": 0.20536184310913086,
+      "learning_rate": 7.905836825807257e-06,
+      "loss": 0.5124,
+      "step": 4509
+    },
+    {
+      "epoch": 1.2506932889628397,
+      "grad_norm": 0.19252263009548187,
+      "learning_rate": 7.900702090656545e-06,
+      "loss": 0.5069,
+      "step": 4510
+    },
+    {
+      "epoch": 1.2509706045479756,
+      "grad_norm": 0.20172010362148285,
+      "learning_rate": 7.895568252950711e-06,
+      "loss": 0.5081,
+      "step": 4511
+    },
+    {
+      "epoch": 1.2512479201331115,
+      "grad_norm": 0.1947304904460907,
+      "learning_rate": 7.890435313691507e-06,
+      "loss": 0.5038,
+      "step": 4512
+    },
+    {
+      "epoch": 1.2515252357182474,
+      "grad_norm": 0.20323634147644043,
+      "learning_rate": 7.885303273880498e-06,
+      "loss": 0.5116,
+      "step": 4513
+    },
+    {
+      "epoch": 1.2518025513033832,
+      "grad_norm": 0.21063730120658875,
+      "learning_rate": 7.880172134519082e-06,
+      "loss": 0.5288,
+      "step": 4514
+    },
+    {
+      "epoch": 1.2520798668885191,
+      "grad_norm": 0.19582238793373108,
+      "learning_rate": 7.875041896608487e-06,
+      "loss": 0.507,
+      "step": 4515
+    },
+    {
+      "epoch": 1.252357182473655,
+      "grad_norm": 0.21959121525287628,
+      "learning_rate": 7.869912561149755e-06,
+      "loss": 0.5066,
+      "step": 4516
+    },
+    {
+      "epoch": 1.252634498058791,
+      "grad_norm": 0.21660704910755157,
+      "learning_rate": 7.864784129143747e-06,
+      "loss": 0.5058,
+      "step": 4517
+    },
+    {
+      "epoch": 1.2529118136439268,
+      "grad_norm": 0.20365749299526215,
+      "learning_rate": 7.85965660159117e-06,
+      "loss": 0.4966,
+      "step": 4518
+    },
+    {
+      "epoch": 1.2531891292290627,
+      "grad_norm": 0.21940533816814423,
+      "learning_rate": 7.85452997949253e-06,
+      "loss": 0.4925,
+      "step": 4519
+    },
+    {
+      "epoch": 1.2534664448141986,
+      "grad_norm": 0.2422044575214386,
+      "learning_rate": 7.84940426384817e-06,
+      "loss": 0.4982,
+      "step": 4520
+    },
+    {
+      "epoch": 1.2537437603993344,
+      "grad_norm": 0.1977468580007553,
+      "learning_rate": 7.844279455658257e-06,
+      "loss": 0.5266,
+      "step": 4521
+    },
+    {
+      "epoch": 1.2540210759844703,
+      "grad_norm": 0.20944328606128693,
+      "learning_rate": 7.839155555922773e-06,
+      "loss": 0.4822,
+      "step": 4522
+    },
+    {
+      "epoch": 1.2542983915696062,
+      "grad_norm": 0.1976504623889923,
+      "learning_rate": 7.834032565641525e-06,
+      "loss": 0.4964,
+      "step": 4523
+    },
+    {
+      "epoch": 1.254575707154742,
+      "grad_norm": 0.20402562618255615,
+      "learning_rate": 7.828910485814142e-06,
+      "loss": 0.5108,
+      "step": 4524
+    },
+    {
+      "epoch": 1.254853022739878,
+      "grad_norm": 0.19931097328662872,
+      "learning_rate": 7.823789317440086e-06,
+      "loss": 0.4945,
+      "step": 4525
+    },
+    {
+      "epoch": 1.2551303383250139,
+      "grad_norm": 0.22738957405090332,
+      "learning_rate": 7.818669061518628e-06,
+      "loss": 0.5012,
+      "step": 4526
+    },
+    {
+      "epoch": 1.2554076539101497,
+      "grad_norm": 0.20075024664402008,
+      "learning_rate": 7.813549719048862e-06,
+      "loss": 0.4728,
+      "step": 4527
+    },
+    {
+      "epoch": 1.2556849694952856,
+      "grad_norm": 0.18941983580589294,
+      "learning_rate": 7.808431291029717e-06,
+      "loss": 0.5031,
+      "step": 4528
+    },
+    {
+      "epoch": 1.2559622850804215,
+      "grad_norm": 0.1993735283613205,
+      "learning_rate": 7.803313778459925e-06,
+      "loss": 0.4867,
+      "step": 4529
+    },
+    {
+      "epoch": 1.2562396006655574,
+      "grad_norm": 0.20413918793201447,
+      "learning_rate": 7.798197182338051e-06,
+      "loss": 0.4906,
+      "step": 4530
+    },
+    {
+      "epoch": 1.2565169162506933,
+      "grad_norm": 0.19167618453502655,
+      "learning_rate": 7.793081503662477e-06,
+      "loss": 0.4974,
+      "step": 4531
+    },
+    {
+      "epoch": 1.2567942318358292,
+      "grad_norm": 0.20251625776290894,
+      "learning_rate": 7.78796674343141e-06,
+      "loss": 0.5196,
+      "step": 4532
+    },
+    {
+      "epoch": 1.257071547420965,
+      "grad_norm": 0.19917502999305725,
+      "learning_rate": 7.78285290264288e-06,
+      "loss": 0.5262,
+      "step": 4533
+    },
+    {
+      "epoch": 1.257348863006101,
+      "grad_norm": 0.23531167209148407,
+      "learning_rate": 7.777739982294719e-06,
+      "loss": 0.5052,
+      "step": 4534
+    },
+    {
+      "epoch": 1.2576261785912368,
+      "grad_norm": 0.20423223078250885,
+      "learning_rate": 7.772627983384604e-06,
+      "loss": 0.5154,
+      "step": 4535
+    },
+    {
+      "epoch": 1.2579034941763727,
+      "grad_norm": 0.19194969534873962,
+      "learning_rate": 7.767516906910018e-06,
+      "loss": 0.4841,
+      "step": 4536
+    },
+    {
+      "epoch": 1.2581808097615086,
+      "grad_norm": 0.20146818459033966,
+      "learning_rate": 7.762406753868273e-06,
+      "loss": 0.5213,
+      "step": 4537
+    },
+    {
+      "epoch": 1.2584581253466445,
+      "grad_norm": 0.2293848693370819,
+      "learning_rate": 7.757297525256482e-06,
+      "loss": 0.5098,
+      "step": 4538
+    },
+    {
+      "epoch": 1.2587354409317804,
+      "grad_norm": 0.201215922832489,
+      "learning_rate": 7.752189222071607e-06,
+      "loss": 0.5166,
+      "step": 4539
+    },
+    {
+      "epoch": 1.2590127565169162,
+      "grad_norm": 0.20330609381198883,
+      "learning_rate": 7.747081845310403e-06,
+      "loss": 0.478,
+      "step": 4540
+    },
+    {
+      "epoch": 1.2592900721020521,
+      "grad_norm": 0.19179701805114746,
+      "learning_rate": 7.741975395969456e-06,
+      "loss": 0.5067,
+      "step": 4541
+    },
+    {
+      "epoch": 1.259567387687188,
+      "grad_norm": 0.19922393560409546,
+      "learning_rate": 7.736869875045171e-06,
+      "loss": 0.519,
+      "step": 4542
+    },
+    {
+      "epoch": 1.259844703272324,
+      "grad_norm": 0.19848206639289856,
+      "learning_rate": 7.731765283533773e-06,
+      "loss": 0.4716,
+      "step": 4543
+    },
+    {
+      "epoch": 1.2601220188574598,
+      "grad_norm": 0.2069179117679596,
+      "learning_rate": 7.7266616224313e-06,
+      "loss": 0.5092,
+      "step": 4544
+    },
+    {
+      "epoch": 1.2603993344425957,
+      "grad_norm": 0.20515765249729156,
+      "learning_rate": 7.721558892733608e-06,
+      "loss": 0.5013,
+      "step": 4545
+    },
+    {
+      "epoch": 1.2606766500277315,
+      "grad_norm": 0.195294588804245,
+      "learning_rate": 7.716457095436378e-06,
+      "loss": 0.5055,
+      "step": 4546
+    },
+    {
+      "epoch": 1.2609539656128674,
+      "grad_norm": 0.1927175670862198,
+      "learning_rate": 7.711356231535111e-06,
+      "loss": 0.5105,
+      "step": 4547
+    },
+    {
+      "epoch": 1.2612312811980033,
+      "grad_norm": 0.19795742630958557,
+      "learning_rate": 7.706256302025109e-06,
+      "loss": 0.5169,
+      "step": 4548
+    },
+    {
+      "epoch": 1.2615085967831392,
+      "grad_norm": 0.2066584676504135,
+      "learning_rate": 7.701157307901515e-06,
+      "loss": 0.5266,
+      "step": 4549
+    },
+    {
+      "epoch": 1.261785912368275,
+      "grad_norm": 0.20272813737392426,
+      "learning_rate": 7.696059250159277e-06,
+      "loss": 0.4815,
+      "step": 4550
+    },
+    {
+      "epoch": 1.262063227953411,
+      "grad_norm": 0.20374014973640442,
+      "learning_rate": 7.690962129793153e-06,
+      "loss": 0.4925,
+      "step": 4551
+    },
+    {
+      "epoch": 1.2623405435385469,
+      "grad_norm": 0.19825245440006256,
+      "learning_rate": 7.68586594779773e-06,
+      "loss": 0.5236,
+      "step": 4552
+    },
+    {
+      "epoch": 1.2626178591236827,
+      "grad_norm": 0.19967573881149292,
+      "learning_rate": 7.68077070516741e-06,
+      "loss": 0.4796,
+      "step": 4553
+    },
+    {
+      "epoch": 1.2628951747088186,
+      "grad_norm": 0.19628724455833435,
+      "learning_rate": 7.67567640289641e-06,
+      "loss": 0.5064,
+      "step": 4554
+    },
+    {
+      "epoch": 1.2631724902939545,
+      "grad_norm": 0.20510771870613098,
+      "learning_rate": 7.670583041978754e-06,
+      "loss": 0.5365,
+      "step": 4555
+    },
+    {
+      "epoch": 1.2634498058790904,
+      "grad_norm": 0.1963818073272705,
+      "learning_rate": 7.665490623408308e-06,
+      "loss": 0.4989,
+      "step": 4556
+    },
+    {
+      "epoch": 1.2637271214642263,
+      "grad_norm": 0.2057606726884842,
+      "learning_rate": 7.660399148178727e-06,
+      "loss": 0.4998,
+      "step": 4557
+    },
+    {
+      "epoch": 1.2640044370493622,
+      "grad_norm": 0.18892644345760345,
+      "learning_rate": 7.655308617283493e-06,
+      "loss": 0.492,
+      "step": 4558
+    },
+    {
+      "epoch": 1.264281752634498,
+      "grad_norm": 0.1982104480266571,
+      "learning_rate": 7.650219031715906e-06,
+      "loss": 0.4747,
+      "step": 4559
+    },
+    {
+      "epoch": 1.264559068219634,
+      "grad_norm": 0.19365094602108002,
+      "learning_rate": 7.645130392469082e-06,
+      "loss": 0.491,
+      "step": 4560
+    },
+    {
+      "epoch": 1.2648363838047698,
+      "grad_norm": 0.1940053105354309,
+      "learning_rate": 7.640042700535944e-06,
+      "loss": 0.5134,
+      "step": 4561
+    },
+    {
+      "epoch": 1.2651136993899057,
+      "grad_norm": 0.19653920829296112,
+      "learning_rate": 7.634955956909234e-06,
+      "loss": 0.4746,
+      "step": 4562
+    },
+    {
+      "epoch": 1.2653910149750416,
+      "grad_norm": 0.19754603505134583,
+      "learning_rate": 7.629870162581516e-06,
+      "loss": 0.4839,
+      "step": 4563
+    },
+    {
+      "epoch": 1.2656683305601775,
+      "grad_norm": 0.1982675939798355,
+      "learning_rate": 7.62478531854516e-06,
+      "loss": 0.4898,
+      "step": 4564
+    },
+    {
+      "epoch": 1.2659456461453134,
+      "grad_norm": 0.20128501951694489,
+      "learning_rate": 7.61970142579236e-06,
+      "loss": 0.5171,
+      "step": 4565
+    },
+    {
+      "epoch": 1.2662229617304492,
+      "grad_norm": 0.20672714710235596,
+      "learning_rate": 7.6146184853151055e-06,
+      "loss": 0.51,
+      "step": 4566
+    },
+    {
+      "epoch": 1.2665002773155851,
+      "grad_norm": 0.1908549964427948,
+      "learning_rate": 7.609536498105224e-06,
+      "loss": 0.4792,
+      "step": 4567
+    },
+    {
+      "epoch": 1.266777592900721,
+      "grad_norm": 0.19964845478534698,
+      "learning_rate": 7.6044554651543424e-06,
+      "loss": 0.5005,
+      "step": 4568
+    },
+    {
+      "epoch": 1.267054908485857,
+      "grad_norm": 0.2022092193365097,
+      "learning_rate": 7.5993753874539015e-06,
+      "loss": 0.4815,
+      "step": 4569
+    },
+    {
+      "epoch": 1.2673322240709928,
+      "grad_norm": 0.20571796596050262,
+      "learning_rate": 7.594296265995164e-06,
+      "loss": 0.5162,
+      "step": 4570
+    },
+    {
+      "epoch": 1.2676095396561287,
+      "grad_norm": 0.2025289386510849,
+      "learning_rate": 7.589218101769202e-06,
+      "loss": 0.5145,
+      "step": 4571
+    },
+    {
+      "epoch": 1.2678868552412645,
+      "grad_norm": 0.18278613686561584,
+      "learning_rate": 7.584140895766895e-06,
+      "loss": 0.4959,
+      "step": 4572
+    },
+    {
+      "epoch": 1.2681641708264004,
+      "grad_norm": 0.20485979318618774,
+      "learning_rate": 7.579064648978939e-06,
+      "loss": 0.4931,
+      "step": 4573
+    },
+    {
+      "epoch": 1.2684414864115363,
+      "grad_norm": 0.20338453352451324,
+      "learning_rate": 7.5739893623958515e-06,
+      "loss": 0.501,
+      "step": 4574
+    },
+    {
+      "epoch": 1.2687188019966722,
+      "grad_norm": 0.1984669268131256,
+      "learning_rate": 7.5689150370079535e-06,
+      "loss": 0.5088,
+      "step": 4575
+    },
+    {
+      "epoch": 1.268996117581808,
+      "grad_norm": 0.19437134265899658,
+      "learning_rate": 7.563841673805372e-06,
+      "loss": 0.4721,
+      "step": 4576
+    },
+    {
+      "epoch": 1.269273433166944,
+      "grad_norm": 0.2029186487197876,
+      "learning_rate": 7.558769273778066e-06,
+      "loss": 0.5095,
+      "step": 4577
+    },
+    {
+      "epoch": 1.2695507487520798,
+      "grad_norm": 0.19555479288101196,
+      "learning_rate": 7.553697837915791e-06,
+      "loss": 0.4698,
+      "step": 4578
+    },
+    {
+      "epoch": 1.2698280643372157,
+      "grad_norm": 0.20628714561462402,
+      "learning_rate": 7.548627367208111e-06,
+      "loss": 0.4781,
+      "step": 4579
+    },
+    {
+      "epoch": 1.2701053799223516,
+      "grad_norm": 0.19298192858695984,
+      "learning_rate": 7.543557862644421e-06,
+      "loss": 0.4903,
+      "step": 4580
+    },
+    {
+      "epoch": 1.2703826955074875,
+      "grad_norm": 0.19302377104759216,
+      "learning_rate": 7.538489325213913e-06,
+      "loss": 0.5233,
+      "step": 4581
+    },
+    {
+      "epoch": 1.2706600110926234,
+      "grad_norm": 0.20301836729049683,
+      "learning_rate": 7.533421755905587e-06,
+      "loss": 0.5101,
+      "step": 4582
+    },
+    {
+      "epoch": 1.2709373266777593,
+      "grad_norm": 0.19993318617343903,
+      "learning_rate": 7.528355155708261e-06,
+      "loss": 0.4907,
+      "step": 4583
+    },
+    {
+      "epoch": 1.2712146422628952,
+      "grad_norm": 0.19807255268096924,
+      "learning_rate": 7.523289525610569e-06,
+      "loss": 0.4911,
+      "step": 4584
+    },
+    {
+      "epoch": 1.271491957848031,
+      "grad_norm": 0.20224188268184662,
+      "learning_rate": 7.518224866600945e-06,
+      "loss": 0.4961,
+      "step": 4585
+    },
+    {
+      "epoch": 1.271769273433167,
+      "grad_norm": 0.194349467754364,
+      "learning_rate": 7.513161179667636e-06,
+      "loss": 0.4848,
+      "step": 4586
+    },
+    {
+      "epoch": 1.2720465890183028,
+      "grad_norm": 0.19369667768478394,
+      "learning_rate": 7.508098465798707e-06,
+      "loss": 0.5079,
+      "step": 4587
+    },
+    {
+      "epoch": 1.2723239046034387,
+      "grad_norm": 0.19323889911174774,
+      "learning_rate": 7.50303672598203e-06,
+      "loss": 0.4986,
+      "step": 4588
+    },
+    {
+      "epoch": 1.2726012201885746,
+      "grad_norm": 0.19939887523651123,
+      "learning_rate": 7.4979759612052754e-06,
+      "loss": 0.5052,
+      "step": 4589
+    },
+    {
+      "epoch": 1.2728785357737105,
+      "grad_norm": 0.19758670032024384,
+      "learning_rate": 7.4929161724559355e-06,
+      "loss": 0.5028,
+      "step": 4590
+    },
+    {
+      "epoch": 1.2731558513588463,
+      "grad_norm": 0.19541890919208527,
+      "learning_rate": 7.487857360721312e-06,
+      "loss": 0.471,
+      "step": 4591
+    },
+    {
+      "epoch": 1.2734331669439822,
+      "grad_norm": 0.2024282068014145,
+      "learning_rate": 7.482799526988515e-06,
+      "loss": 0.5235,
+      "step": 4592
+    },
+    {
+      "epoch": 1.2737104825291181,
+      "grad_norm": 0.19754981994628906,
+      "learning_rate": 7.4777426722444505e-06,
+      "loss": 0.4934,
+      "step": 4593
+    },
+    {
+      "epoch": 1.273987798114254,
+      "grad_norm": 0.20427443087100983,
+      "learning_rate": 7.472686797475861e-06,
+      "loss": 0.4731,
+      "step": 4594
+    },
+    {
+      "epoch": 1.2742651136993899,
+      "grad_norm": 0.18958698213100433,
+      "learning_rate": 7.46763190366927e-06,
+      "loss": 0.4863,
+      "step": 4595
+    },
+    {
+      "epoch": 1.2745424292845258,
+      "grad_norm": 0.19229987263679504,
+      "learning_rate": 7.462577991811028e-06,
+      "loss": 0.4881,
+      "step": 4596
+    },
+    {
+      "epoch": 1.2748197448696617,
+      "grad_norm": 0.2041274905204773,
+      "learning_rate": 7.4575250628872745e-06,
+      "loss": 0.4933,
+      "step": 4597
+    },
+    {
+      "epoch": 1.2750970604547975,
+      "grad_norm": 0.1980563998222351,
+      "learning_rate": 7.452473117883989e-06,
+      "loss": 0.5078,
+      "step": 4598
+    },
+    {
+      "epoch": 1.2753743760399334,
+      "grad_norm": 0.19617661833763123,
+      "learning_rate": 7.4474221577869265e-06,
+      "loss": 0.5231,
+      "step": 4599
+    },
+    {
+      "epoch": 1.2756516916250693,
+      "grad_norm": 0.19393262267112732,
+      "learning_rate": 7.442372183581664e-06,
+      "loss": 0.4959,
+      "step": 4600
+    },
+    {
+      "epoch": 1.2759290072102052,
+      "grad_norm": 0.2189975529909134,
+      "learning_rate": 7.43732319625359e-06,
+      "loss": 0.501,
+      "step": 4601
+    },
+    {
+      "epoch": 1.276206322795341,
+      "grad_norm": 0.21371406316757202,
+      "learning_rate": 7.432275196787894e-06,
+      "loss": 0.5012,
+      "step": 4602
+    },
+    {
+      "epoch": 1.276483638380477,
+      "grad_norm": 0.20440706610679626,
+      "learning_rate": 7.427228186169575e-06,
+      "loss": 0.4698,
+      "step": 4603
+    },
+    {
+      "epoch": 1.2767609539656128,
+      "grad_norm": 0.19952230155467987,
+      "learning_rate": 7.422182165383434e-06,
+      "loss": 0.5198,
+      "step": 4604
+    },
+    {
+      "epoch": 1.2770382695507487,
+      "grad_norm": 0.20963895320892334,
+      "learning_rate": 7.417137135414088e-06,
+      "loss": 0.4953,
+      "step": 4605
+    },
+    {
+      "epoch": 1.2773155851358846,
+      "grad_norm": 0.210064098238945,
+      "learning_rate": 7.412093097245956e-06,
+      "loss": 0.5,
+      "step": 4606
+    },
+    {
+      "epoch": 1.2775929007210205,
+      "grad_norm": 0.20027229189872742,
+      "learning_rate": 7.4070500518632595e-06,
+      "loss": 0.499,
+      "step": 4607
+    },
+    {
+      "epoch": 1.2778702163061564,
+      "grad_norm": 0.19522671401500702,
+      "learning_rate": 7.4020080002500355e-06,
+      "loss": 0.4978,
+      "step": 4608
+    },
+    {
+      "epoch": 1.2781475318912923,
+      "grad_norm": 0.1936059147119522,
+      "learning_rate": 7.396966943390121e-06,
+      "loss": 0.4872,
+      "step": 4609
+    },
+    {
+      "epoch": 1.2784248474764282,
+      "grad_norm": 0.2015063762664795,
+      "learning_rate": 7.391926882267159e-06,
+      "loss": 0.5015,
+      "step": 4610
+    },
+    {
+      "epoch": 1.278702163061564,
+      "grad_norm": 0.20017270743846893,
+      "learning_rate": 7.386887817864592e-06,
+      "loss": 0.5063,
+      "step": 4611
+    },
+    {
+      "epoch": 1.2789794786467,
+      "grad_norm": 0.20706325769424438,
+      "learning_rate": 7.381849751165684e-06,
+      "loss": 0.5311,
+      "step": 4612
+    },
+    {
+      "epoch": 1.2792567942318358,
+      "grad_norm": 0.19370242953300476,
+      "learning_rate": 7.376812683153496e-06,
+      "loss": 0.4901,
+      "step": 4613
+    },
+    {
+      "epoch": 1.2795341098169717,
+      "grad_norm": 0.20481263101100922,
+      "learning_rate": 7.371776614810883e-06,
+      "loss": 0.5185,
+      "step": 4614
+    },
+    {
+      "epoch": 1.2798114254021076,
+      "grad_norm": 0.20902620255947113,
+      "learning_rate": 7.366741547120527e-06,
+      "loss": 0.5268,
+      "step": 4615
+    },
+    {
+      "epoch": 1.2800887409872435,
+      "grad_norm": 0.2147327959537506,
+      "learning_rate": 7.361707481064898e-06,
+      "loss": 0.5251,
+      "step": 4616
+    },
+    {
+      "epoch": 1.2803660565723793,
+      "grad_norm": 0.20069043338298798,
+      "learning_rate": 7.356674417626275e-06,
+      "loss": 0.5176,
+      "step": 4617
+    },
+    {
+      "epoch": 1.2806433721575152,
+      "grad_norm": 0.2038661241531372,
+      "learning_rate": 7.351642357786741e-06,
+      "loss": 0.5001,
+      "step": 4618
+    },
+    {
+      "epoch": 1.280920687742651,
+      "grad_norm": 0.20059525966644287,
+      "learning_rate": 7.34661130252819e-06,
+      "loss": 0.5081,
+      "step": 4619
+    },
+    {
+      "epoch": 1.281198003327787,
+      "grad_norm": 0.2055322527885437,
+      "learning_rate": 7.341581252832309e-06,
+      "loss": 0.5234,
+      "step": 4620
+    },
+    {
+      "epoch": 1.2814753189129229,
+      "grad_norm": 0.1974288672208786,
+      "learning_rate": 7.336552209680592e-06,
+      "loss": 0.4774,
+      "step": 4621
+    },
+    {
+      "epoch": 1.2817526344980588,
+      "grad_norm": 0.19128572940826416,
+      "learning_rate": 7.3315241740543455e-06,
+      "loss": 0.4954,
+      "step": 4622
+    },
+    {
+      "epoch": 1.2820299500831946,
+      "grad_norm": 0.3327876925468445,
+      "learning_rate": 7.326497146934669e-06,
+      "loss": 0.5184,
+      "step": 4623
+    },
+    {
+      "epoch": 1.2823072656683305,
+      "grad_norm": 0.19013381004333496,
+      "learning_rate": 7.3214711293024694e-06,
+      "loss": 0.4776,
+      "step": 4624
+    },
+    {
+      "epoch": 1.2825845812534664,
+      "grad_norm": 0.20403680205345154,
+      "learning_rate": 7.316446122138451e-06,
+      "loss": 0.522,
+      "step": 4625
+    },
+    {
+      "epoch": 1.2828618968386023,
+      "grad_norm": 0.19423431158065796,
+      "learning_rate": 7.311422126423131e-06,
+      "loss": 0.4802,
+      "step": 4626
+    },
+    {
+      "epoch": 1.2831392124237382,
+      "grad_norm": 0.20686981081962585,
+      "learning_rate": 7.306399143136825e-06,
+      "loss": 0.4908,
+      "step": 4627
+    },
+    {
+      "epoch": 1.283416528008874,
+      "grad_norm": 0.19685044884681702,
+      "learning_rate": 7.301377173259644e-06,
+      "loss": 0.4766,
+      "step": 4628
+    },
+    {
+      "epoch": 1.28369384359401,
+      "grad_norm": 0.20665518939495087,
+      "learning_rate": 7.296356217771515e-06,
+      "loss": 0.4733,
+      "step": 4629
+    },
+    {
+      "epoch": 1.2839711591791458,
+      "grad_norm": 0.20772451162338257,
+      "learning_rate": 7.291336277652158e-06,
+      "loss": 0.4944,
+      "step": 4630
+    },
+    {
+      "epoch": 1.2842484747642817,
+      "grad_norm": 0.20459213852882385,
+      "learning_rate": 7.286317353881094e-06,
+      "loss": 0.5046,
+      "step": 4631
+    },
+    {
+      "epoch": 1.2845257903494176,
+      "grad_norm": 0.198153555393219,
+      "learning_rate": 7.281299447437637e-06,
+      "loss": 0.5134,
+      "step": 4632
+    },
+    {
+      "epoch": 1.2848031059345535,
+      "grad_norm": 0.2004951536655426,
+      "learning_rate": 7.276282559300937e-06,
+      "loss": 0.4975,
+      "step": 4633
+    },
+    {
+      "epoch": 1.2850804215196894,
+      "grad_norm": 0.19988703727722168,
+      "learning_rate": 7.271266690449907e-06,
+      "loss": 0.5216,
+      "step": 4634
+    },
+    {
+      "epoch": 1.2853577371048253,
+      "grad_norm": 0.20041632652282715,
+      "learning_rate": 7.266251841863275e-06,
+      "loss": 0.4996,
+      "step": 4635
+    },
+    {
+      "epoch": 1.2856350526899611,
+      "grad_norm": 0.19913561642169952,
+      "learning_rate": 7.2612380145195735e-06,
+      "loss": 0.4907,
+      "step": 4636
+    },
+    {
+      "epoch": 1.285912368275097,
+      "grad_norm": 0.21043738722801208,
+      "learning_rate": 7.256225209397139e-06,
+      "loss": 0.5161,
+      "step": 4637
+    },
+    {
+      "epoch": 1.286189683860233,
+      "grad_norm": 0.2178940773010254,
+      "learning_rate": 7.2512134274740986e-06,
+      "loss": 0.5072,
+      "step": 4638
+    },
+    {
+      "epoch": 1.2864669994453688,
+      "grad_norm": 0.2040819674730301,
+      "learning_rate": 7.246202669728375e-06,
+      "loss": 0.522,
+      "step": 4639
+    },
+    {
+      "epoch": 1.2867443150305047,
+      "grad_norm": 0.20290911197662354,
+      "learning_rate": 7.241192937137708e-06,
+      "loss": 0.5249,
+      "step": 4640
+    },
+    {
+      "epoch": 1.2870216306156406,
+      "grad_norm": 0.2622935473918915,
+      "learning_rate": 7.2361842306796356e-06,
+      "loss": 0.5143,
+      "step": 4641
+    },
+    {
+      "epoch": 1.2872989462007765,
+      "grad_norm": 0.20093023777008057,
+      "learning_rate": 7.231176551331476e-06,
+      "loss": 0.5218,
+      "step": 4642
+    },
+    {
+      "epoch": 1.2875762617859123,
+      "grad_norm": 0.19977106153964996,
+      "learning_rate": 7.226169900070365e-06,
+      "loss": 0.5273,
+      "step": 4643
+    },
+    {
+      "epoch": 1.2878535773710482,
+      "grad_norm": 0.20094482600688934,
+      "learning_rate": 7.221164277873238e-06,
+      "loss": 0.5079,
+      "step": 4644
+    },
+    {
+      "epoch": 1.288130892956184,
+      "grad_norm": 0.2162494659423828,
+      "learning_rate": 7.216159685716817e-06,
+      "loss": 0.4924,
+      "step": 4645
+    },
+    {
+      "epoch": 1.28840820854132,
+      "grad_norm": 0.19907453656196594,
+      "learning_rate": 7.211156124577639e-06,
+      "loss": 0.4912,
+      "step": 4646
+    },
+    {
+      "epoch": 1.2886855241264559,
+      "grad_norm": 0.20266936719417572,
+      "learning_rate": 7.206153595432022e-06,
+      "loss": 0.4964,
+      "step": 4647
+    },
+    {
+      "epoch": 1.2889628397115918,
+      "grad_norm": 0.19707994163036346,
+      "learning_rate": 7.2011520992561e-06,
+      "loss": 0.5026,
+      "step": 4648
+    },
+    {
+      "epoch": 1.2892401552967276,
+      "grad_norm": 0.1953379213809967,
+      "learning_rate": 7.196151637025788e-06,
+      "loss": 0.5141,
+      "step": 4649
+    },
+    {
+      "epoch": 1.2895174708818635,
+      "grad_norm": 0.1917579025030136,
+      "learning_rate": 7.191152209716822e-06,
+      "loss": 0.4888,
+      "step": 4650
+    },
+    {
+      "epoch": 1.2897947864669994,
+      "grad_norm": 0.20420075953006744,
+      "learning_rate": 7.186153818304708e-06,
+      "loss": 0.5009,
+      "step": 4651
+    },
+    {
+      "epoch": 1.2900721020521353,
+      "grad_norm": 0.19985781610012054,
+      "learning_rate": 7.1811564637647734e-06,
+      "loss": 0.5,
+      "step": 4652
+    },
+    {
+      "epoch": 1.2903494176372712,
+      "grad_norm": 0.2002001255750656,
+      "learning_rate": 7.176160147072138e-06,
+      "loss": 0.5068,
+      "step": 4653
+    },
+    {
+      "epoch": 1.290626733222407,
+      "grad_norm": 0.2185979187488556,
+      "learning_rate": 7.171164869201709e-06,
+      "loss": 0.5064,
+      "step": 4654
+    },
+    {
+      "epoch": 1.290904048807543,
+      "grad_norm": 0.20650553703308105,
+      "learning_rate": 7.166170631128194e-06,
+      "loss": 0.5152,
+      "step": 4655
+    },
+    {
+      "epoch": 1.2911813643926788,
+      "grad_norm": 0.20761191844940186,
+      "learning_rate": 7.161177433826108e-06,
+      "loss": 0.4963,
+      "step": 4656
+    },
+    {
+      "epoch": 1.2914586799778147,
+      "grad_norm": 0.20723947882652283,
+      "learning_rate": 7.156185278269756e-06,
+      "loss": 0.5188,
+      "step": 4657
+    },
+    {
+      "epoch": 1.2917359955629506,
+      "grad_norm": 0.21793848276138306,
+      "learning_rate": 7.151194165433234e-06,
+      "loss": 0.4931,
+      "step": 4658
+    },
+    {
+      "epoch": 1.2920133111480865,
+      "grad_norm": 0.19249577820301056,
+      "learning_rate": 7.146204096290446e-06,
+      "loss": 0.5235,
+      "step": 4659
+    },
+    {
+      "epoch": 1.2922906267332224,
+      "grad_norm": 0.20929144322872162,
+      "learning_rate": 7.1412150718150884e-06,
+      "loss": 0.5141,
+      "step": 4660
+    },
+    {
+      "epoch": 1.2925679423183583,
+      "grad_norm": 0.3471876382827759,
+      "learning_rate": 7.136227092980649e-06,
+      "loss": 0.5095,
+      "step": 4661
+    },
+    {
+      "epoch": 1.2928452579034941,
+      "grad_norm": 0.1970401108264923,
+      "learning_rate": 7.131240160760408e-06,
+      "loss": 0.4993,
+      "step": 4662
+    },
+    {
+      "epoch": 1.29312257348863,
+      "grad_norm": 0.2073030173778534,
+      "learning_rate": 7.126254276127456e-06,
+      "loss": 0.5111,
+      "step": 4663
+    },
+    {
+      "epoch": 1.293399889073766,
+      "grad_norm": 0.19547441601753235,
+      "learning_rate": 7.1212694400546734e-06,
+      "loss": 0.5045,
+      "step": 4664
+    },
+    {
+      "epoch": 1.2936772046589018,
+      "grad_norm": 0.1951918601989746,
+      "learning_rate": 7.116285653514729e-06,
+      "loss": 0.4969,
+      "step": 4665
+    },
+    {
+      "epoch": 1.2939545202440377,
+      "grad_norm": 0.1991628259420395,
+      "learning_rate": 7.111302917480089e-06,
+      "loss": 0.4958,
+      "step": 4666
+    },
+    {
+      "epoch": 1.2942318358291736,
+      "grad_norm": 0.19393709301948547,
+      "learning_rate": 7.10632123292302e-06,
+      "loss": 0.4918,
+      "step": 4667
+    },
+    {
+      "epoch": 1.2945091514143094,
+      "grad_norm": 0.19460806250572205,
+      "learning_rate": 7.101340600815587e-06,
+      "loss": 0.4921,
+      "step": 4668
+    },
+    {
+      "epoch": 1.2947864669994453,
+      "grad_norm": 0.1952556073665619,
+      "learning_rate": 7.096361022129637e-06,
+      "loss": 0.4937,
+      "step": 4669
+    },
+    {
+      "epoch": 1.2950637825845812,
+      "grad_norm": 0.20183762907981873,
+      "learning_rate": 7.0913824978368075e-06,
+      "loss": 0.4777,
+      "step": 4670
+    },
+    {
+      "epoch": 1.295341098169717,
+      "grad_norm": 0.2849336564540863,
+      "learning_rate": 7.086405028908563e-06,
+      "loss": 0.4967,
+      "step": 4671
+    },
+    {
+      "epoch": 1.295618413754853,
+      "grad_norm": 0.19828931987285614,
+      "learning_rate": 7.081428616316127e-06,
+      "loss": 0.523,
+      "step": 4672
+    },
+    {
+      "epoch": 1.2958957293399889,
+      "grad_norm": 0.20423446595668793,
+      "learning_rate": 7.076453261030524e-06,
+      "loss": 0.4865,
+      "step": 4673
+    },
+    {
+      "epoch": 1.2961730449251248,
+      "grad_norm": 0.18961788713932037,
+      "learning_rate": 7.0714789640225865e-06,
+      "loss": 0.4901,
+      "step": 4674
+    },
+    {
+      "epoch": 1.2964503605102606,
+      "grad_norm": 0.2092195451259613,
+      "learning_rate": 7.0665057262629316e-06,
+      "loss": 0.5204,
+      "step": 4675
+    },
+    {
+      "epoch": 1.2967276760953965,
+      "grad_norm": 0.2078777551651001,
+      "learning_rate": 7.061533548721969e-06,
+      "loss": 0.5013,
+      "step": 4676
+    },
+    {
+      "epoch": 1.2970049916805324,
+      "grad_norm": 0.2117914855480194,
+      "learning_rate": 7.0565624323698955e-06,
+      "loss": 0.5268,
+      "step": 4677
+    },
+    {
+      "epoch": 1.2972823072656683,
+      "grad_norm": 0.19487378001213074,
+      "learning_rate": 7.051592378176711e-06,
+      "loss": 0.5368,
+      "step": 4678
+    },
+    {
+      "epoch": 1.2975596228508042,
+      "grad_norm": 0.19526413083076477,
+      "learning_rate": 7.046623387112212e-06,
+      "loss": 0.502,
+      "step": 4679
+    },
+    {
+      "epoch": 1.29783693843594,
+      "grad_norm": 0.19563087821006775,
+      "learning_rate": 7.041655460145971e-06,
+      "loss": 0.5248,
+      "step": 4680
+    },
+    {
+      "epoch": 1.298114254021076,
+      "grad_norm": 0.1995537430047989,
+      "learning_rate": 7.0366885982473635e-06,
+      "loss": 0.5115,
+      "step": 4681
+    },
+    {
+      "epoch": 1.2983915696062118,
+      "grad_norm": 0.20692096650600433,
+      "learning_rate": 7.0317228023855654e-06,
+      "loss": 0.4994,
+      "step": 4682
+    },
+    {
+      "epoch": 1.2986688851913477,
+      "grad_norm": 0.20720021426677704,
+      "learning_rate": 7.026758073529527e-06,
+      "loss": 0.4766,
+      "step": 4683
+    },
+    {
+      "epoch": 1.2989462007764836,
+      "grad_norm": 0.20226465165615082,
+      "learning_rate": 7.021794412647993e-06,
+      "loss": 0.5035,
+      "step": 4684
+    },
+    {
+      "epoch": 1.2992235163616195,
+      "grad_norm": 0.21326394379138947,
+      "learning_rate": 7.016831820709513e-06,
+      "loss": 0.4912,
+      "step": 4685
+    },
+    {
+      "epoch": 1.2995008319467554,
+      "grad_norm": 0.23130756616592407,
+      "learning_rate": 7.0118702986824225e-06,
+      "loss": 0.5091,
+      "step": 4686
+    },
+    {
+      "epoch": 1.2997781475318912,
+      "grad_norm": 0.19730432331562042,
+      "learning_rate": 7.006909847534837e-06,
+      "loss": 0.5001,
+      "step": 4687
+    },
+    {
+      "epoch": 1.3000554631170271,
+      "grad_norm": 0.21078179776668549,
+      "learning_rate": 7.0019504682346835e-06,
+      "loss": 0.4988,
+      "step": 4688
+    },
+    {
+      "epoch": 1.300332778702163,
+      "grad_norm": 0.2023860216140747,
+      "learning_rate": 6.996992161749656e-06,
+      "loss": 0.4996,
+      "step": 4689
+    },
+    {
+      "epoch": 1.300610094287299,
+      "grad_norm": 0.19739584624767303,
+      "learning_rate": 6.992034929047261e-06,
+      "loss": 0.4843,
+      "step": 4690
+    },
+    {
+      "epoch": 1.3008874098724348,
+      "grad_norm": 0.2066974639892578,
+      "learning_rate": 6.987078771094779e-06,
+      "loss": 0.5207,
+      "step": 4691
+    },
+    {
+      "epoch": 1.3011647254575707,
+      "grad_norm": 0.20626220107078552,
+      "learning_rate": 6.982123688859295e-06,
+      "loss": 0.5055,
+      "step": 4692
+    },
+    {
+      "epoch": 1.3014420410427066,
+      "grad_norm": 0.19193756580352783,
+      "learning_rate": 6.977169683307667e-06,
+      "loss": 0.5372,
+      "step": 4693
+    },
+    {
+      "epoch": 1.3017193566278424,
+      "grad_norm": 0.19858404994010925,
+      "learning_rate": 6.972216755406559e-06,
+      "loss": 0.4943,
+      "step": 4694
+    },
+    {
+      "epoch": 1.3019966722129783,
+      "grad_norm": 0.19214889407157898,
+      "learning_rate": 6.967264906122422e-06,
+      "loss": 0.4917,
+      "step": 4695
+    },
+    {
+      "epoch": 1.3022739877981142,
+      "grad_norm": 0.24965661764144897,
+      "learning_rate": 6.962314136421485e-06,
+      "loss": 0.5006,
+      "step": 4696
+    },
+    {
+      "epoch": 1.30255130338325,
+      "grad_norm": 0.20609916746616364,
+      "learning_rate": 6.957364447269785e-06,
+      "loss": 0.4954,
+      "step": 4697
+    },
+    {
+      "epoch": 1.302828618968386,
+      "grad_norm": 0.18501780927181244,
+      "learning_rate": 6.9524158396331225e-06,
+      "loss": 0.4895,
+      "step": 4698
+    },
+    {
+      "epoch": 1.3031059345535219,
+      "grad_norm": 0.19009456038475037,
+      "learning_rate": 6.947468314477115e-06,
+      "loss": 0.4812,
+      "step": 4699
+    },
+    {
+      "epoch": 1.3033832501386577,
+      "grad_norm": 0.20104162395000458,
+      "learning_rate": 6.942521872767148e-06,
+      "loss": 0.5101,
+      "step": 4700
+    },
+    {
+      "epoch": 1.3036605657237936,
+      "grad_norm": 0.21302615106105804,
+      "learning_rate": 6.937576515468405e-06,
+      "loss": 0.4959,
+      "step": 4701
+    },
+    {
+      "epoch": 1.3039378813089295,
+      "grad_norm": 0.19924134016036987,
+      "learning_rate": 6.932632243545864e-06,
+      "loss": 0.5155,
+      "step": 4702
+    },
+    {
+      "epoch": 1.3042151968940654,
+      "grad_norm": 0.2015244960784912,
+      "learning_rate": 6.927689057964274e-06,
+      "loss": 0.5192,
+      "step": 4703
+    },
+    {
+      "epoch": 1.3044925124792013,
+      "grad_norm": 0.21109920740127563,
+      "learning_rate": 6.9227469596881825e-06,
+      "loss": 0.5109,
+      "step": 4704
+    },
+    {
+      "epoch": 1.3047698280643372,
+      "grad_norm": 0.19857889413833618,
+      "learning_rate": 6.9178059496819246e-06,
+      "loss": 0.4955,
+      "step": 4705
+    },
+    {
+      "epoch": 1.305047143649473,
+      "grad_norm": 0.20074202120304108,
+      "learning_rate": 6.912866028909627e-06,
+      "loss": 0.4898,
+      "step": 4706
+    },
+    {
+      "epoch": 1.305324459234609,
+      "grad_norm": 0.20996864140033722,
+      "learning_rate": 6.907927198335197e-06,
+      "loss": 0.5067,
+      "step": 4707
+    },
+    {
+      "epoch": 1.3056017748197448,
+      "grad_norm": 0.20864242315292358,
+      "learning_rate": 6.902989458922319e-06,
+      "loss": 0.5017,
+      "step": 4708
+    },
+    {
+      "epoch": 1.3058790904048807,
+      "grad_norm": 0.2010875642299652,
+      "learning_rate": 6.898052811634498e-06,
+      "loss": 0.5122,
+      "step": 4709
+    },
+    {
+      "epoch": 1.3061564059900166,
+      "grad_norm": 0.1969454139471054,
+      "learning_rate": 6.893117257434994e-06,
+      "loss": 0.5032,
+      "step": 4710
+    },
+    {
+      "epoch": 1.3064337215751525,
+      "grad_norm": 0.2069738507270813,
+      "learning_rate": 6.88818279728686e-06,
+      "loss": 0.501,
+      "step": 4711
+    },
+    {
+      "epoch": 1.3067110371602884,
+      "grad_norm": 0.1900450736284256,
+      "learning_rate": 6.883249432152944e-06,
+      "loss": 0.5205,
+      "step": 4712
+    },
+    {
+      "epoch": 1.3069883527454242,
+      "grad_norm": 0.22067134082317352,
+      "learning_rate": 6.878317162995881e-06,
+      "loss": 0.4896,
+      "step": 4713
+    },
+    {
+      "epoch": 1.3072656683305601,
+      "grad_norm": 0.20373979210853577,
+      "learning_rate": 6.8733859907780865e-06,
+      "loss": 0.5217,
+      "step": 4714
+    },
+    {
+      "epoch": 1.307542983915696,
+      "grad_norm": 0.20570501685142517,
+      "learning_rate": 6.8684559164617525e-06,
+      "loss": 0.5074,
+      "step": 4715
+    },
+    {
+      "epoch": 1.307820299500832,
+      "grad_norm": 0.2102414071559906,
+      "learning_rate": 6.8635269410088725e-06,
+      "loss": 0.5142,
+      "step": 4716
+    },
+    {
+      "epoch": 1.3080976150859678,
+      "grad_norm": 0.20017775893211365,
+      "learning_rate": 6.8585990653812285e-06,
+      "loss": 0.5078,
+      "step": 4717
+    },
+    {
+      "epoch": 1.3083749306711037,
+      "grad_norm": 0.22755825519561768,
+      "learning_rate": 6.8536722905403666e-06,
+      "loss": 0.5103,
+      "step": 4718
+    },
+    {
+      "epoch": 1.3086522462562395,
+      "grad_norm": 0.20453383028507233,
+      "learning_rate": 6.848746617447644e-06,
+      "loss": 0.4962,
+      "step": 4719
+    },
+    {
+      "epoch": 1.3089295618413754,
+      "grad_norm": 0.21632753312587738,
+      "learning_rate": 6.8438220470641785e-06,
+      "loss": 0.5031,
+      "step": 4720
+    },
+    {
+      "epoch": 1.3092068774265113,
+      "grad_norm": 0.19868405163288116,
+      "learning_rate": 6.838898580350895e-06,
+      "loss": 0.4784,
+      "step": 4721
+    },
+    {
+      "epoch": 1.3094841930116472,
+      "grad_norm": 0.20900887250900269,
+      "learning_rate": 6.833976218268478e-06,
+      "loss": 0.5007,
+      "step": 4722
+    },
+    {
+      "epoch": 1.309761508596783,
+      "grad_norm": 0.20499762892723083,
+      "learning_rate": 6.829054961777423e-06,
+      "loss": 0.5004,
+      "step": 4723
+    },
+    {
+      "epoch": 1.310038824181919,
+      "grad_norm": 0.19857996702194214,
+      "learning_rate": 6.8241348118379966e-06,
+      "loss": 0.5087,
+      "step": 4724
+    },
+    {
+      "epoch": 1.3103161397670549,
+      "grad_norm": 0.1966305822134018,
+      "learning_rate": 6.819215769410243e-06,
+      "loss": 0.4808,
+      "step": 4725
+    },
+    {
+      "epoch": 1.3105934553521907,
+      "grad_norm": 0.2024124413728714,
+      "learning_rate": 6.814297835454009e-06,
+      "loss": 0.4972,
+      "step": 4726
+    },
+    {
+      "epoch": 1.3108707709373266,
+      "grad_norm": 0.21003521978855133,
+      "learning_rate": 6.8093810109289e-06,
+      "loss": 0.4985,
+      "step": 4727
+    },
+    {
+      "epoch": 1.3111480865224625,
+      "grad_norm": 0.19540292024612427,
+      "learning_rate": 6.804465296794332e-06,
+      "loss": 0.4898,
+      "step": 4728
+    },
+    {
+      "epoch": 1.3114254021075984,
+      "grad_norm": 0.2004072666168213,
+      "learning_rate": 6.799550694009479e-06,
+      "loss": 0.4926,
+      "step": 4729
+    },
+    {
+      "epoch": 1.3117027176927343,
+      "grad_norm": 0.19552730023860931,
+      "learning_rate": 6.794637203533321e-06,
+      "loss": 0.4841,
+      "step": 4730
+    },
+    {
+      "epoch": 1.3119800332778702,
+      "grad_norm": 0.20091231167316437,
+      "learning_rate": 6.789724826324602e-06,
+      "loss": 0.4953,
+      "step": 4731
+    },
+    {
+      "epoch": 1.312257348863006,
+      "grad_norm": 0.20696038007736206,
+      "learning_rate": 6.78481356334186e-06,
+      "loss": 0.5117,
+      "step": 4732
+    },
+    {
+      "epoch": 1.312534664448142,
+      "grad_norm": 0.19968093931674957,
+      "learning_rate": 6.779903415543418e-06,
+      "loss": 0.4942,
+      "step": 4733
+    },
+    {
+      "epoch": 1.3128119800332778,
+      "grad_norm": 0.19521543383598328,
+      "learning_rate": 6.7749943838873636e-06,
+      "loss": 0.4901,
+      "step": 4734
+    },
+    {
+      "epoch": 1.3130892956184137,
+      "grad_norm": 0.24511419236660004,
+      "learning_rate": 6.770086469331592e-06,
+      "loss": 0.5023,
+      "step": 4735
+    },
+    {
+      "epoch": 1.3133666112035496,
+      "grad_norm": 0.29729798436164856,
+      "learning_rate": 6.765179672833757e-06,
+      "loss": 0.4767,
+      "step": 4736
+    },
+    {
+      "epoch": 1.3136439267886855,
+      "grad_norm": 0.19325131177902222,
+      "learning_rate": 6.760273995351313e-06,
+      "loss": 0.4878,
+      "step": 4737
+    },
+    {
+      "epoch": 1.3139212423738214,
+      "grad_norm": 0.1989395171403885,
+      "learning_rate": 6.75536943784148e-06,
+      "loss": 0.4873,
+      "step": 4738
+    },
+    {
+      "epoch": 1.3141985579589572,
+      "grad_norm": 0.20601066946983337,
+      "learning_rate": 6.750466001261271e-06,
+      "loss": 0.4805,
+      "step": 4739
+    },
+    {
+      "epoch": 1.3144758735440931,
+      "grad_norm": 0.19977515935897827,
+      "learning_rate": 6.74556368656748e-06,
+      "loss": 0.5138,
+      "step": 4740
+    },
+    {
+      "epoch": 1.314753189129229,
+      "grad_norm": 0.19737781584262848,
+      "learning_rate": 6.740662494716675e-06,
+      "loss": 0.4954,
+      "step": 4741
+    },
+    {
+      "epoch": 1.315030504714365,
+      "grad_norm": 0.19452109932899475,
+      "learning_rate": 6.7357624266652044e-06,
+      "loss": 0.4723,
+      "step": 4742
+    },
+    {
+      "epoch": 1.3153078202995008,
+      "grad_norm": 0.19652196764945984,
+      "learning_rate": 6.730863483369203e-06,
+      "loss": 0.483,
+      "step": 4743
+    },
+    {
+      "epoch": 1.3155851358846367,
+      "grad_norm": 0.21024899184703827,
+      "learning_rate": 6.725965665784592e-06,
+      "loss": 0.5212,
+      "step": 4744
+    },
+    {
+      "epoch": 1.3158624514697725,
+      "grad_norm": 0.20697370171546936,
+      "learning_rate": 6.721068974867059e-06,
+      "loss": 0.4693,
+      "step": 4745
+    },
+    {
+      "epoch": 1.3161397670549084,
+      "grad_norm": 0.20207835733890533,
+      "learning_rate": 6.71617341157207e-06,
+      "loss": 0.5116,
+      "step": 4746
+    },
+    {
+      "epoch": 1.3164170826400443,
+      "grad_norm": 0.19856862723827362,
+      "learning_rate": 6.711278976854898e-06,
+      "loss": 0.5203,
+      "step": 4747
+    },
+    {
+      "epoch": 1.3166943982251802,
+      "grad_norm": 0.20049835741519928,
+      "learning_rate": 6.706385671670566e-06,
+      "loss": 0.485,
+      "step": 4748
+    },
+    {
+      "epoch": 1.316971713810316,
+      "grad_norm": 0.21266470849514008,
+      "learning_rate": 6.701493496973885e-06,
+      "loss": 0.5069,
+      "step": 4749
+    },
+    {
+      "epoch": 1.317249029395452,
+      "grad_norm": 0.20254302024841309,
+      "learning_rate": 6.69660245371945e-06,
+      "loss": 0.5294,
+      "step": 4750
+    },
+    {
+      "epoch": 1.3175263449805878,
+      "grad_norm": 0.19671931862831116,
+      "learning_rate": 6.691712542861639e-06,
+      "loss": 0.4901,
+      "step": 4751
+    },
+    {
+      "epoch": 1.3178036605657237,
+      "grad_norm": 0.2424955815076828,
+      "learning_rate": 6.686823765354599e-06,
+      "loss": 0.518,
+      "step": 4752
+    },
+    {
+      "epoch": 1.3180809761508596,
+      "grad_norm": 0.2099234014749527,
+      "learning_rate": 6.681936122152255e-06,
+      "loss": 0.5131,
+      "step": 4753
+    },
+    {
+      "epoch": 1.3183582917359955,
+      "grad_norm": 0.1978161633014679,
+      "learning_rate": 6.67704961420832e-06,
+      "loss": 0.4835,
+      "step": 4754
+    },
+    {
+      "epoch": 1.3186356073211314,
+      "grad_norm": 0.197959765791893,
+      "learning_rate": 6.6721642424762866e-06,
+      "loss": 0.4862,
+      "step": 4755
+    },
+    {
+      "epoch": 1.3189129229062673,
+      "grad_norm": 0.19919045269489288,
+      "learning_rate": 6.667280007909416e-06,
+      "loss": 0.4858,
+      "step": 4756
+    },
+    {
+      "epoch": 1.3191902384914032,
+      "grad_norm": 0.2116900086402893,
+      "learning_rate": 6.662396911460745e-06,
+      "loss": 0.5011,
+      "step": 4757
+    },
+    {
+      "epoch": 1.319467554076539,
+      "grad_norm": 0.20922990143299103,
+      "learning_rate": 6.657514954083099e-06,
+      "loss": 0.5122,
+      "step": 4758
+    },
+    {
+      "epoch": 1.319744869661675,
+      "grad_norm": 0.24156232178211212,
+      "learning_rate": 6.652634136729086e-06,
+      "loss": 0.4908,
+      "step": 4759
+    },
+    {
+      "epoch": 1.3200221852468108,
+      "grad_norm": 0.1963520497083664,
+      "learning_rate": 6.647754460351072e-06,
+      "loss": 0.5212,
+      "step": 4760
+    },
+    {
+      "epoch": 1.3202995008319467,
+      "grad_norm": 0.20835274457931519,
+      "learning_rate": 6.642875925901213e-06,
+      "loss": 0.5094,
+      "step": 4761
+    },
+    {
+      "epoch": 1.3205768164170826,
+      "grad_norm": 0.18857994675636292,
+      "learning_rate": 6.63799853433145e-06,
+      "loss": 0.4601,
+      "step": 4762
+    },
+    {
+      "epoch": 1.3208541320022185,
+      "grad_norm": 0.20583271980285645,
+      "learning_rate": 6.633122286593481e-06,
+      "loss": 0.5022,
+      "step": 4763
+    },
+    {
+      "epoch": 1.3211314475873543,
+      "grad_norm": 0.21066723763942719,
+      "learning_rate": 6.628247183638789e-06,
+      "loss": 0.4997,
+      "step": 4764
+    },
+    {
+      "epoch": 1.3214087631724902,
+      "grad_norm": 0.19231611490249634,
+      "learning_rate": 6.623373226418642e-06,
+      "loss": 0.4751,
+      "step": 4765
+    },
+    {
+      "epoch": 1.3216860787576261,
+      "grad_norm": 0.1879042237997055,
+      "learning_rate": 6.618500415884083e-06,
+      "loss": 0.4813,
+      "step": 4766
+    },
+    {
+      "epoch": 1.321963394342762,
+      "grad_norm": 0.20243264734745026,
+      "learning_rate": 6.613628752985912e-06,
+      "loss": 0.4951,
+      "step": 4767
+    },
+    {
+      "epoch": 1.3222407099278979,
+      "grad_norm": 0.20071490108966827,
+      "learning_rate": 6.608758238674733e-06,
+      "loss": 0.4892,
+      "step": 4768
+    },
+    {
+      "epoch": 1.3225180255130338,
+      "grad_norm": 0.213031604886055,
+      "learning_rate": 6.603888873900905e-06,
+      "loss": 0.5118,
+      "step": 4769
+    },
+    {
+      "epoch": 1.3227953410981697,
+      "grad_norm": 0.2158973515033722,
+      "learning_rate": 6.599020659614572e-06,
+      "loss": 0.4931,
+      "step": 4770
+    },
+    {
+      "epoch": 1.3230726566833055,
+      "grad_norm": 0.20497101545333862,
+      "learning_rate": 6.594153596765655e-06,
+      "loss": 0.5167,
+      "step": 4771
+    },
+    {
+      "epoch": 1.3233499722684414,
+      "grad_norm": 0.19923308491706848,
+      "learning_rate": 6.5892876863038385e-06,
+      "loss": 0.477,
+      "step": 4772
+    },
+    {
+      "epoch": 1.3236272878535773,
+      "grad_norm": 0.21297168731689453,
+      "learning_rate": 6.584422929178602e-06,
+      "loss": 0.5255,
+      "step": 4773
+    },
+    {
+      "epoch": 1.3239046034387132,
+      "grad_norm": 0.20591717958450317,
+      "learning_rate": 6.579559326339177e-06,
+      "loss": 0.5326,
+      "step": 4774
+    },
+    {
+      "epoch": 1.324181919023849,
+      "grad_norm": 0.20993812382221222,
+      "learning_rate": 6.574696878734592e-06,
+      "loss": 0.4997,
+      "step": 4775
+    },
+    {
+      "epoch": 1.324459234608985,
+      "grad_norm": 0.20869500935077667,
+      "learning_rate": 6.569835587313627e-06,
+      "loss": 0.5205,
+      "step": 4776
+    },
+    {
+      "epoch": 1.3247365501941208,
+      "grad_norm": 0.19962731003761292,
+      "learning_rate": 6.5649754530248575e-06,
+      "loss": 0.4994,
+      "step": 4777
+    },
+    {
+      "epoch": 1.3250138657792567,
+      "grad_norm": 0.20245380699634552,
+      "learning_rate": 6.560116476816627e-06,
+      "loss": 0.5062,
+      "step": 4778
+    },
+    {
+      "epoch": 1.3252911813643926,
+      "grad_norm": 0.20509305596351624,
+      "learning_rate": 6.5552586596370465e-06,
+      "loss": 0.501,
+      "step": 4779
+    },
+    {
+      "epoch": 1.3255684969495285,
+      "grad_norm": 0.19942280650138855,
+      "learning_rate": 6.5504020024340005e-06,
+      "loss": 0.4664,
+      "step": 4780
+    },
+    {
+      "epoch": 1.3258458125346644,
+      "grad_norm": 0.2001029998064041,
+      "learning_rate": 6.545546506155154e-06,
+      "loss": 0.4858,
+      "step": 4781
+    },
+    {
+      "epoch": 1.3261231281198003,
+      "grad_norm": 0.19130992889404297,
+      "learning_rate": 6.5406921717479474e-06,
+      "loss": 0.5081,
+      "step": 4782
+    },
+    {
+      "epoch": 1.3264004437049361,
+      "grad_norm": 0.20136502385139465,
+      "learning_rate": 6.53583900015959e-06,
+      "loss": 0.4889,
+      "step": 4783
+    },
+    {
+      "epoch": 1.326677759290072,
+      "grad_norm": 0.20278695225715637,
+      "learning_rate": 6.53098699233705e-06,
+      "loss": 0.5011,
+      "step": 4784
+    },
+    {
+      "epoch": 1.326955074875208,
+      "grad_norm": 0.1966562420129776,
+      "learning_rate": 6.5261361492271054e-06,
+      "loss": 0.4887,
+      "step": 4785
+    },
+    {
+      "epoch": 1.3272323904603438,
+      "grad_norm": 0.20651812851428986,
+      "learning_rate": 6.5212864717762696e-06,
+      "loss": 0.5006,
+      "step": 4786
+    },
+    {
+      "epoch": 1.3275097060454797,
+      "grad_norm": 0.210739865899086,
+      "learning_rate": 6.516437960930843e-06,
+      "loss": 0.5149,
+      "step": 4787
+    },
+    {
+      "epoch": 1.3277870216306156,
+      "grad_norm": 0.21981686353683472,
+      "learning_rate": 6.5115906176369025e-06,
+      "loss": 0.5234,
+      "step": 4788
+    },
+    {
+      "epoch": 1.3280643372157515,
+      "grad_norm": 0.19809675216674805,
+      "learning_rate": 6.506744442840296e-06,
+      "loss": 0.5058,
+      "step": 4789
+    },
+    {
+      "epoch": 1.3283416528008873,
+      "grad_norm": 0.19718238711357117,
+      "learning_rate": 6.501899437486637e-06,
+      "loss": 0.5016,
+      "step": 4790
+    },
+    {
+      "epoch": 1.3286189683860232,
+      "grad_norm": 0.19665461778640747,
+      "learning_rate": 6.4970556025213095e-06,
+      "loss": 0.4995,
+      "step": 4791
+    },
+    {
+      "epoch": 1.328896283971159,
+      "grad_norm": 0.1910347193479538,
+      "learning_rate": 6.492212938889481e-06,
+      "loss": 0.4835,
+      "step": 4792
+    },
+    {
+      "epoch": 1.329173599556295,
+      "grad_norm": 0.1989067643880844,
+      "learning_rate": 6.487371447536084e-06,
+      "loss": 0.4774,
+      "step": 4793
+    },
+    {
+      "epoch": 1.3294509151414309,
+      "grad_norm": 0.1986207813024521,
+      "learning_rate": 6.482531129405819e-06,
+      "loss": 0.5111,
+      "step": 4794
+    },
+    {
+      "epoch": 1.3297282307265668,
+      "grad_norm": 0.20336773991584778,
+      "learning_rate": 6.477691985443157e-06,
+      "loss": 0.5037,
+      "step": 4795
+    },
+    {
+      "epoch": 1.3300055463117026,
+      "grad_norm": 0.19637836515903473,
+      "learning_rate": 6.472854016592346e-06,
+      "loss": 0.5016,
+      "step": 4796
+    },
+    {
+      "epoch": 1.3302828618968385,
+      "grad_norm": 0.2004927545785904,
+      "learning_rate": 6.468017223797407e-06,
+      "loss": 0.509,
+      "step": 4797
+    },
+    {
+      "epoch": 1.3305601774819744,
+      "grad_norm": 0.2119477391242981,
+      "learning_rate": 6.463181608002118e-06,
+      "loss": 0.5365,
+      "step": 4798
+    },
+    {
+      "epoch": 1.3308374930671103,
+      "grad_norm": 0.20132261514663696,
+      "learning_rate": 6.4583471701500395e-06,
+      "loss": 0.5112,
+      "step": 4799
+    },
+    {
+      "epoch": 1.3311148086522462,
+      "grad_norm": 0.2103944718837738,
+      "learning_rate": 6.453513911184503e-06,
+      "loss": 0.4846,
+      "step": 4800
+    },
+    {
+      "epoch": 1.331392124237382,
+      "grad_norm": 0.18914499878883362,
+      "learning_rate": 6.448681832048603e-06,
+      "loss": 0.4821,
+      "step": 4801
+    },
+    {
+      "epoch": 1.331669439822518,
+      "grad_norm": 0.19525542855262756,
+      "learning_rate": 6.443850933685197e-06,
+      "loss": 0.5018,
+      "step": 4802
+    },
+    {
+      "epoch": 1.3319467554076538,
+      "grad_norm": 0.19284358620643616,
+      "learning_rate": 6.4390212170369305e-06,
+      "loss": 0.478,
+      "step": 4803
+    },
+    {
+      "epoch": 1.3322240709927897,
+      "grad_norm": 0.20030631124973297,
+      "learning_rate": 6.43419268304621e-06,
+      "loss": 0.5084,
+      "step": 4804
+    },
+    {
+      "epoch": 1.3325013865779256,
+      "grad_norm": 0.19547629356384277,
+      "learning_rate": 6.429365332655204e-06,
+      "loss": 0.4884,
+      "step": 4805
+    },
+    {
+      "epoch": 1.3327787021630615,
+      "grad_norm": 0.20050625503063202,
+      "learning_rate": 6.4245391668058655e-06,
+      "loss": 0.526,
+      "step": 4806
+    },
+    {
+      "epoch": 1.3330560177481974,
+      "grad_norm": 0.24080616235733032,
+      "learning_rate": 6.419714186439896e-06,
+      "loss": 0.4912,
+      "step": 4807
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.2031663954257965,
+      "learning_rate": 6.414890392498787e-06,
+      "loss": 0.5048,
+      "step": 4808
+    },
+    {
+      "epoch": 1.3336106489184691,
+      "grad_norm": 0.20676189661026,
+      "learning_rate": 6.410067785923779e-06,
+      "loss": 0.4905,
+      "step": 4809
+    },
+    {
+      "epoch": 1.333887964503605,
+      "grad_norm": 0.19812701642513275,
+      "learning_rate": 6.405246367655903e-06,
+      "loss": 0.5076,
+      "step": 4810
+    },
+    {
+      "epoch": 1.334165280088741,
+      "grad_norm": 0.20559370517730713,
+      "learning_rate": 6.4004261386359315e-06,
+      "loss": 0.4868,
+      "step": 4811
+    },
+    {
+      "epoch": 1.3344425956738768,
+      "grad_norm": 0.21239563822746277,
+      "learning_rate": 6.395607099804426e-06,
+      "loss": 0.4941,
+      "step": 4812
+    },
+    {
+      "epoch": 1.3347199112590127,
+      "grad_norm": 0.20791207253932953,
+      "learning_rate": 6.390789252101713e-06,
+      "loss": 0.5058,
+      "step": 4813
+    },
+    {
+      "epoch": 1.3349972268441486,
+      "grad_norm": 0.20702631771564484,
+      "learning_rate": 6.3859725964678735e-06,
+      "loss": 0.5146,
+      "step": 4814
+    },
+    {
+      "epoch": 1.3352745424292845,
+      "grad_norm": 0.1996176540851593,
+      "learning_rate": 6.381157133842772e-06,
+      "loss": 0.4976,
+      "step": 4815
+    },
+    {
+      "epoch": 1.3355518580144203,
+      "grad_norm": 0.20184668898582458,
+      "learning_rate": 6.376342865166024e-06,
+      "loss": 0.4912,
+      "step": 4816
+    },
+    {
+      "epoch": 1.3358291735995562,
+      "grad_norm": 0.20000986754894257,
+      "learning_rate": 6.371529791377031e-06,
+      "loss": 0.5215,
+      "step": 4817
+    },
+    {
+      "epoch": 1.336106489184692,
+      "grad_norm": 0.1989196389913559,
+      "learning_rate": 6.366717913414943e-06,
+      "loss": 0.4972,
+      "step": 4818
+    },
+    {
+      "epoch": 1.336383804769828,
+      "grad_norm": 0.19104993343353271,
+      "learning_rate": 6.361907232218689e-06,
+      "loss": 0.464,
+      "step": 4819
+    },
+    {
+      "epoch": 1.3366611203549639,
+      "grad_norm": 0.20721067488193512,
+      "learning_rate": 6.357097748726965e-06,
+      "loss": 0.5046,
+      "step": 4820
+    },
+    {
+      "epoch": 1.3369384359400998,
+      "grad_norm": 0.2302693873643875,
+      "learning_rate": 6.3522894638782204e-06,
+      "loss": 0.4778,
+      "step": 4821
+    },
+    {
+      "epoch": 1.3372157515252356,
+      "grad_norm": 0.20015782117843628,
+      "learning_rate": 6.347482378610678e-06,
+      "loss": 0.5031,
+      "step": 4822
+    },
+    {
+      "epoch": 1.3374930671103715,
+      "grad_norm": 0.19521984457969666,
+      "learning_rate": 6.342676493862332e-06,
+      "loss": 0.4701,
+      "step": 4823
+    },
+    {
+      "epoch": 1.3377703826955074,
+      "grad_norm": 0.20052647590637207,
+      "learning_rate": 6.337871810570943e-06,
+      "loss": 0.498,
+      "step": 4824
+    },
+    {
+      "epoch": 1.3380476982806433,
+      "grad_norm": 0.20175038278102875,
+      "learning_rate": 6.333068329674021e-06,
+      "loss": 0.5112,
+      "step": 4825
+    },
+    {
+      "epoch": 1.3383250138657792,
+      "grad_norm": 0.21088631451129913,
+      "learning_rate": 6.328266052108856e-06,
+      "loss": 0.5264,
+      "step": 4826
+    },
+    {
+      "epoch": 1.338602329450915,
+      "grad_norm": 0.20293080806732178,
+      "learning_rate": 6.323464978812507e-06,
+      "loss": 0.47,
+      "step": 4827
+    },
+    {
+      "epoch": 1.338879645036051,
+      "grad_norm": 0.22624416649341583,
+      "learning_rate": 6.318665110721786e-06,
+      "loss": 0.5097,
+      "step": 4828
+    },
+    {
+      "epoch": 1.3391569606211868,
+      "grad_norm": 0.20791591703891754,
+      "learning_rate": 6.3138664487732675e-06,
+      "loss": 0.4856,
+      "step": 4829
+    },
+    {
+      "epoch": 1.3394342762063227,
+      "grad_norm": 0.20461194217205048,
+      "learning_rate": 6.309068993903303e-06,
+      "loss": 0.4912,
+      "step": 4830
+    },
+    {
+      "epoch": 1.3397115917914586,
+      "grad_norm": 0.2024698108434677,
+      "learning_rate": 6.304272747048009e-06,
+      "loss": 0.4838,
+      "step": 4831
+    },
+    {
+      "epoch": 1.3399889073765945,
+      "grad_norm": 0.20816753804683685,
+      "learning_rate": 6.2994777091432535e-06,
+      "loss": 0.5021,
+      "step": 4832
+    },
+    {
+      "epoch": 1.3402662229617304,
+      "grad_norm": 0.2043784111738205,
+      "learning_rate": 6.2946838811246734e-06,
+      "loss": 0.5101,
+      "step": 4833
+    },
+    {
+      "epoch": 1.3405435385468663,
+      "grad_norm": 0.19572801887989044,
+      "learning_rate": 6.289891263927675e-06,
+      "loss": 0.4688,
+      "step": 4834
+    },
+    {
+      "epoch": 1.3408208541320021,
+      "grad_norm": 0.22099921107292175,
+      "learning_rate": 6.285099858487428e-06,
+      "loss": 0.4935,
+      "step": 4835
+    },
+    {
+      "epoch": 1.341098169717138,
+      "grad_norm": 0.19714292883872986,
+      "learning_rate": 6.280309665738854e-06,
+      "loss": 0.496,
+      "step": 4836
+    },
+    {
+      "epoch": 1.341375485302274,
+      "grad_norm": 0.21856801211833954,
+      "learning_rate": 6.275520686616654e-06,
+      "loss": 0.4743,
+      "step": 4837
+    },
+    {
+      "epoch": 1.3416528008874098,
+      "grad_norm": 0.2152533382177353,
+      "learning_rate": 6.270732922055286e-06,
+      "loss": 0.499,
+      "step": 4838
+    },
+    {
+      "epoch": 1.3419301164725457,
+      "grad_norm": 0.201574444770813,
+      "learning_rate": 6.2659463729889665e-06,
+      "loss": 0.4826,
+      "step": 4839
+    },
+    {
+      "epoch": 1.3422074320576816,
+      "grad_norm": 0.18966776132583618,
+      "learning_rate": 6.261161040351673e-06,
+      "loss": 0.4745,
+      "step": 4840
+    },
+    {
+      "epoch": 1.3424847476428174,
+      "grad_norm": 0.20077745616436005,
+      "learning_rate": 6.256376925077155e-06,
+      "loss": 0.4785,
+      "step": 4841
+    },
+    {
+      "epoch": 1.3427620632279533,
+      "grad_norm": 0.19951693713665009,
+      "learning_rate": 6.251594028098925e-06,
+      "loss": 0.4906,
+      "step": 4842
+    },
+    {
+      "epoch": 1.3430393788130892,
+      "grad_norm": 0.20605219900608063,
+      "learning_rate": 6.246812350350245e-06,
+      "loss": 0.5253,
+      "step": 4843
+    },
+    {
+      "epoch": 1.343316694398225,
+      "grad_norm": 0.1993149369955063,
+      "learning_rate": 6.242031892764156e-06,
+      "loss": 0.5034,
+      "step": 4844
+    },
+    {
+      "epoch": 1.343594009983361,
+      "grad_norm": 0.20574994385242462,
+      "learning_rate": 6.237252656273439e-06,
+      "loss": 0.5055,
+      "step": 4845
+    },
+    {
+      "epoch": 1.343871325568497,
+      "grad_norm": 0.20431512594223022,
+      "learning_rate": 6.232474641810664e-06,
+      "loss": 0.4775,
+      "step": 4846
+    },
+    {
+      "epoch": 1.344148641153633,
+      "grad_norm": 0.1970827877521515,
+      "learning_rate": 6.2276978503081355e-06,
+      "loss": 0.4963,
+      "step": 4847
+    },
+    {
+      "epoch": 1.3444259567387689,
+      "grad_norm": 0.20088131725788116,
+      "learning_rate": 6.222922282697944e-06,
+      "loss": 0.5,
+      "step": 4848
+    },
+    {
+      "epoch": 1.3447032723239047,
+      "grad_norm": 0.2057836651802063,
+      "learning_rate": 6.218147939911917e-06,
+      "loss": 0.4928,
+      "step": 4849
+    },
+    {
+      "epoch": 1.3449805879090406,
+      "grad_norm": 0.20138069987297058,
+      "learning_rate": 6.213374822881661e-06,
+      "loss": 0.4929,
+      "step": 4850
+    },
+    {
+      "epoch": 1.3452579034941765,
+      "grad_norm": 0.20331014692783356,
+      "learning_rate": 6.208602932538545e-06,
+      "loss": 0.492,
+      "step": 4851
+    },
+    {
+      "epoch": 1.3455352190793124,
+      "grad_norm": 0.20663729310035706,
+      "learning_rate": 6.203832269813678e-06,
+      "loss": 0.5118,
+      "step": 4852
+    },
+    {
+      "epoch": 1.3458125346644483,
+      "grad_norm": 0.20441415905952454,
+      "learning_rate": 6.1990628356379535e-06,
+      "loss": 0.5004,
+      "step": 4853
+    },
+    {
+      "epoch": 1.3460898502495842,
+      "grad_norm": 0.2063564509153366,
+      "learning_rate": 6.194294630942006e-06,
+      "loss": 0.5033,
+      "step": 4854
+    },
+    {
+      "epoch": 1.34636716583472,
+      "grad_norm": 0.2034982442855835,
+      "learning_rate": 6.1895276566562465e-06,
+      "loss": 0.4578,
+      "step": 4855
+    },
+    {
+      "epoch": 1.346644481419856,
+      "grad_norm": 0.19431711733341217,
+      "learning_rate": 6.184761913710829e-06,
+      "loss": 0.4876,
+      "step": 4856
+    },
+    {
+      "epoch": 1.3469217970049918,
+      "grad_norm": 0.2182120680809021,
+      "learning_rate": 6.179997403035682e-06,
+      "loss": 0.4799,
+      "step": 4857
+    },
+    {
+      "epoch": 1.3471991125901277,
+      "grad_norm": 0.19659878313541412,
+      "learning_rate": 6.175234125560492e-06,
+      "loss": 0.4867,
+      "step": 4858
+    },
+    {
+      "epoch": 1.3474764281752636,
+      "grad_norm": 0.21084415912628174,
+      "learning_rate": 6.1704720822146955e-06,
+      "loss": 0.4893,
+      "step": 4859
+    },
+    {
+      "epoch": 1.3477537437603995,
+      "grad_norm": 0.20155581831932068,
+      "learning_rate": 6.165711273927488e-06,
+      "loss": 0.5106,
+      "step": 4860
+    },
+    {
+      "epoch": 1.3480310593455354,
+      "grad_norm": 0.20054039359092712,
+      "learning_rate": 6.160951701627836e-06,
+      "loss": 0.4754,
+      "step": 4861
+    },
+    {
+      "epoch": 1.3483083749306712,
+      "grad_norm": 0.2048387974500656,
+      "learning_rate": 6.15619336624446e-06,
+      "loss": 0.4928,
+      "step": 4862
+    },
+    {
+      "epoch": 1.3485856905158071,
+      "grad_norm": 0.19565939903259277,
+      "learning_rate": 6.151436268705831e-06,
+      "loss": 0.4936,
+      "step": 4863
+    },
+    {
+      "epoch": 1.348863006100943,
+      "grad_norm": 0.21946604549884796,
+      "learning_rate": 6.1466804099401874e-06,
+      "loss": 0.4924,
+      "step": 4864
+    },
+    {
+      "epoch": 1.349140321686079,
+      "grad_norm": 0.20226682722568512,
+      "learning_rate": 6.141925790875529e-06,
+      "loss": 0.4983,
+      "step": 4865
+    },
+    {
+      "epoch": 1.3494176372712148,
+      "grad_norm": 0.191081702709198,
+      "learning_rate": 6.137172412439601e-06,
+      "loss": 0.5019,
+      "step": 4866
+    },
+    {
+      "epoch": 1.3496949528563507,
+      "grad_norm": 0.1935385912656784,
+      "learning_rate": 6.132420275559912e-06,
+      "loss": 0.4948,
+      "step": 4867
+    },
+    {
+      "epoch": 1.3499722684414865,
+      "grad_norm": 0.20531508326530457,
+      "learning_rate": 6.127669381163734e-06,
+      "loss": 0.5112,
+      "step": 4868
+    },
+    {
+      "epoch": 1.3502495840266224,
+      "grad_norm": 0.22489938139915466,
+      "learning_rate": 6.122919730178095e-06,
+      "loss": 0.5146,
+      "step": 4869
+    },
+    {
+      "epoch": 1.3505268996117583,
+      "grad_norm": 0.20186354219913483,
+      "learning_rate": 6.118171323529774e-06,
+      "loss": 0.4934,
+      "step": 4870
+    },
+    {
+      "epoch": 1.3508042151968942,
+      "grad_norm": 0.20553670823574066,
+      "learning_rate": 6.113424162145307e-06,
+      "loss": 0.5131,
+      "step": 4871
+    },
+    {
+      "epoch": 1.35108153078203,
+      "grad_norm": 0.20304057002067566,
+      "learning_rate": 6.108678246950994e-06,
+      "loss": 0.5157,
+      "step": 4872
+    },
+    {
+      "epoch": 1.351358846367166,
+      "grad_norm": 0.2151353359222412,
+      "learning_rate": 6.103933578872896e-06,
+      "loss": 0.4806,
+      "step": 4873
+    },
+    {
+      "epoch": 1.3516361619523019,
+      "grad_norm": 0.1945381909608841,
+      "learning_rate": 6.099190158836816e-06,
+      "loss": 0.5012,
+      "step": 4874
+    },
+    {
+      "epoch": 1.3519134775374377,
+      "grad_norm": 0.20316410064697266,
+      "learning_rate": 6.094447987768315e-06,
+      "loss": 0.4882,
+      "step": 4875
+    },
+    {
+      "epoch": 1.3521907931225736,
+      "grad_norm": 0.2029489129781723,
+      "learning_rate": 6.08970706659273e-06,
+      "loss": 0.5021,
+      "step": 4876
+    },
+    {
+      "epoch": 1.3524681087077095,
+      "grad_norm": 0.21302883327007294,
+      "learning_rate": 6.084967396235136e-06,
+      "loss": 0.527,
+      "step": 4877
+    },
+    {
+      "epoch": 1.3527454242928454,
+      "grad_norm": 0.204268679022789,
+      "learning_rate": 6.08022897762036e-06,
+      "loss": 0.521,
+      "step": 4878
+    },
+    {
+      "epoch": 1.3530227398779813,
+      "grad_norm": 0.20695167779922485,
+      "learning_rate": 6.0754918116730004e-06,
+      "loss": 0.4753,
+      "step": 4879
+    },
+    {
+      "epoch": 1.3533000554631172,
+      "grad_norm": 0.2412513643503189,
+      "learning_rate": 6.070755899317407e-06,
+      "loss": 0.4799,
+      "step": 4880
+    },
+    {
+      "epoch": 1.353577371048253,
+      "grad_norm": 0.1893726885318756,
+      "learning_rate": 6.066021241477676e-06,
+      "loss": 0.4823,
+      "step": 4881
+    },
+    {
+      "epoch": 1.353854686633389,
+      "grad_norm": 0.20982672274112701,
+      "learning_rate": 6.061287839077661e-06,
+      "loss": 0.5285,
+      "step": 4882
+    },
+    {
+      "epoch": 1.3541320022185248,
+      "grad_norm": 0.19612151384353638,
+      "learning_rate": 6.056555693040981e-06,
+      "loss": 0.5243,
+      "step": 4883
+    },
+    {
+      "epoch": 1.3544093178036607,
+      "grad_norm": 0.20173022150993347,
+      "learning_rate": 6.051824804291005e-06,
+      "loss": 0.5089,
+      "step": 4884
+    },
+    {
+      "epoch": 1.3546866333887966,
+      "grad_norm": 0.20177814364433289,
+      "learning_rate": 6.047095173750846e-06,
+      "loss": 0.508,
+      "step": 4885
+    },
+    {
+      "epoch": 1.3549639489739325,
+      "grad_norm": 0.21217969059944153,
+      "learning_rate": 6.042366802343389e-06,
+      "loss": 0.4961,
+      "step": 4886
+    },
+    {
+      "epoch": 1.3552412645590683,
+      "grad_norm": 0.21045343577861786,
+      "learning_rate": 6.0376396909912575e-06,
+      "loss": 0.471,
+      "step": 4887
+    },
+    {
+      "epoch": 1.3555185801442042,
+      "grad_norm": 0.20939841866493225,
+      "learning_rate": 6.032913840616843e-06,
+      "loss": 0.5066,
+      "step": 4888
+    },
+    {
+      "epoch": 1.3557958957293401,
+      "grad_norm": 0.20125700533390045,
+      "learning_rate": 6.028189252142276e-06,
+      "loss": 0.4869,
+      "step": 4889
+    },
+    {
+      "epoch": 1.356073211314476,
+      "grad_norm": 0.19696246087551117,
+      "learning_rate": 6.023465926489453e-06,
+      "loss": 0.5013,
+      "step": 4890
+    },
+    {
+      "epoch": 1.3563505268996119,
+      "grad_norm": 0.19802086055278778,
+      "learning_rate": 6.018743864580025e-06,
+      "loss": 0.4694,
+      "step": 4891
+    },
+    {
+      "epoch": 1.3566278424847478,
+      "grad_norm": 0.19850695133209229,
+      "learning_rate": 6.014023067335382e-06,
+      "loss": 0.5089,
+      "step": 4892
+    },
+    {
+      "epoch": 1.3569051580698837,
+      "grad_norm": 0.20037485659122467,
+      "learning_rate": 6.009303535676686e-06,
+      "loss": 0.4823,
+      "step": 4893
+    },
+    {
+      "epoch": 1.3571824736550195,
+      "grad_norm": 0.19831885397434235,
+      "learning_rate": 6.004585270524833e-06,
+      "loss": 0.4821,
+      "step": 4894
+    },
+    {
+      "epoch": 1.3574597892401554,
+      "grad_norm": 0.20965071022510529,
+      "learning_rate": 5.999868272800492e-06,
+      "loss": 0.4821,
+      "step": 4895
+    },
+    {
+      "epoch": 1.3577371048252913,
+      "grad_norm": 0.19735604524612427,
+      "learning_rate": 5.995152543424064e-06,
+      "loss": 0.472,
+      "step": 4896
+    },
+    {
+      "epoch": 1.3580144204104272,
+      "grad_norm": 0.20253437757492065,
+      "learning_rate": 5.990438083315721e-06,
+      "loss": 0.5013,
+      "step": 4897
+    },
+    {
+      "epoch": 1.358291735995563,
+      "grad_norm": 0.19695337116718292,
+      "learning_rate": 5.985724893395371e-06,
+      "loss": 0.4982,
+      "step": 4898
+    },
+    {
+      "epoch": 1.358569051580699,
+      "grad_norm": 0.2138115018606186,
+      "learning_rate": 5.981012974582688e-06,
+      "loss": 0.511,
+      "step": 4899
+    },
+    {
+      "epoch": 1.3588463671658348,
+      "grad_norm": 0.20029625296592712,
+      "learning_rate": 5.976302327797096e-06,
+      "loss": 0.5215,
+      "step": 4900
+    },
+    {
+      "epoch": 1.3591236827509707,
+      "grad_norm": 0.20940783619880676,
+      "learning_rate": 5.9715929539577595e-06,
+      "loss": 0.5094,
+      "step": 4901
+    },
+    {
+      "epoch": 1.3594009983361066,
+      "grad_norm": 0.2135714590549469,
+      "learning_rate": 5.966884853983597e-06,
+      "loss": 0.5108,
+      "step": 4902
+    },
+    {
+      "epoch": 1.3596783139212425,
+      "grad_norm": 0.19404084980487823,
+      "learning_rate": 5.9621780287932995e-06,
+      "loss": 0.4732,
+      "step": 4903
+    },
+    {
+      "epoch": 1.3599556295063784,
+      "grad_norm": 0.20225512981414795,
+      "learning_rate": 5.957472479305286e-06,
+      "loss": 0.4923,
+      "step": 4904
+    },
+    {
+      "epoch": 1.3602329450915143,
+      "grad_norm": 0.20398668944835663,
+      "learning_rate": 5.952768206437727e-06,
+      "loss": 0.478,
+      "step": 4905
+    },
+    {
+      "epoch": 1.3605102606766502,
+      "grad_norm": 0.2068692445755005,
+      "learning_rate": 5.9480652111085566e-06,
+      "loss": 0.5464,
+      "step": 4906
+    },
+    {
+      "epoch": 1.360787576261786,
+      "grad_norm": 0.20321063697338104,
+      "learning_rate": 5.9433634942354595e-06,
+      "loss": 0.4954,
+      "step": 4907
+    },
+    {
+      "epoch": 1.361064891846922,
+      "grad_norm": 0.21151407063007355,
+      "learning_rate": 5.938663056735859e-06,
+      "loss": 0.5176,
+      "step": 4908
+    },
+    {
+      "epoch": 1.3613422074320578,
+      "grad_norm": 0.20232000946998596,
+      "learning_rate": 5.93396389952693e-06,
+      "loss": 0.524,
+      "step": 4909
+    },
+    {
+      "epoch": 1.3616195230171937,
+      "grad_norm": 0.21282248198986053,
+      "learning_rate": 5.92926602352561e-06,
+      "loss": 0.5043,
+      "step": 4910
+    },
+    {
+      "epoch": 1.3618968386023296,
+      "grad_norm": 0.20285063982009888,
+      "learning_rate": 5.92456942964858e-06,
+      "loss": 0.4926,
+      "step": 4911
+    },
+    {
+      "epoch": 1.3621741541874655,
+      "grad_norm": 0.26317086815834045,
+      "learning_rate": 5.9198741188122675e-06,
+      "loss": 0.4847,
+      "step": 4912
+    },
+    {
+      "epoch": 1.3624514697726013,
+      "grad_norm": 0.20316633582115173,
+      "learning_rate": 5.915180091932843e-06,
+      "loss": 0.521,
+      "step": 4913
+    },
+    {
+      "epoch": 1.3627287853577372,
+      "grad_norm": 0.2053111493587494,
+      "learning_rate": 5.910487349926251e-06,
+      "loss": 0.4979,
+      "step": 4914
+    },
+    {
+      "epoch": 1.3630061009428731,
+      "grad_norm": 1.0747162103652954,
+      "learning_rate": 5.905795893708166e-06,
+      "loss": 0.4933,
+      "step": 4915
+    },
+    {
+      "epoch": 1.363283416528009,
+      "grad_norm": 0.19813428819179535,
+      "learning_rate": 5.901105724194006e-06,
+      "loss": 0.5115,
+      "step": 4916
+    },
+    {
+      "epoch": 1.3635607321131449,
+      "grad_norm": 0.2020360231399536,
+      "learning_rate": 5.896416842298953e-06,
+      "loss": 0.4945,
+      "step": 4917
+    },
+    {
+      "epoch": 1.3638380476982808,
+      "grad_norm": 0.20236392319202423,
+      "learning_rate": 5.891729248937938e-06,
+      "loss": 0.4844,
+      "step": 4918
+    },
+    {
+      "epoch": 1.3641153632834166,
+      "grad_norm": 0.20511947572231293,
+      "learning_rate": 5.8870429450256295e-06,
+      "loss": 0.4872,
+      "step": 4919
+    },
+    {
+      "epoch": 1.3643926788685525,
+      "grad_norm": 0.2103184312582016,
+      "learning_rate": 5.882357931476446e-06,
+      "loss": 0.519,
+      "step": 4920
+    },
+    {
+      "epoch": 1.3646699944536884,
+      "grad_norm": 0.20535695552825928,
+      "learning_rate": 5.877674209204559e-06,
+      "loss": 0.5072,
+      "step": 4921
+    },
+    {
+      "epoch": 1.3649473100388243,
+      "grad_norm": 0.20085355639457703,
+      "learning_rate": 5.872991779123894e-06,
+      "loss": 0.4931,
+      "step": 4922
+    },
+    {
+      "epoch": 1.3652246256239602,
+      "grad_norm": 0.19907495379447937,
+      "learning_rate": 5.8683106421481084e-06,
+      "loss": 0.5263,
+      "step": 4923
+    },
+    {
+      "epoch": 1.365501941209096,
+      "grad_norm": 0.2058345228433609,
+      "learning_rate": 5.863630799190624e-06,
+      "loss": 0.5022,
+      "step": 4924
+    },
+    {
+      "epoch": 1.365779256794232,
+      "grad_norm": 0.19591206312179565,
+      "learning_rate": 5.8589522511645944e-06,
+      "loss": 0.4979,
+      "step": 4925
+    },
+    {
+      "epoch": 1.3660565723793678,
+      "grad_norm": 0.212891086935997,
+      "learning_rate": 5.854274998982935e-06,
+      "loss": 0.5211,
+      "step": 4926
+    },
+    {
+      "epoch": 1.3663338879645037,
+      "grad_norm": 0.20310524106025696,
+      "learning_rate": 5.8495990435582945e-06,
+      "loss": 0.4923,
+      "step": 4927
+    },
+    {
+      "epoch": 1.3666112035496396,
+      "grad_norm": 0.2070825695991516,
+      "learning_rate": 5.844924385803078e-06,
+      "loss": 0.5088,
+      "step": 4928
+    },
+    {
+      "epoch": 1.3668885191347755,
+      "grad_norm": 0.1995309591293335,
+      "learning_rate": 5.8402510266294435e-06,
+      "loss": 0.4637,
+      "step": 4929
+    },
+    {
+      "epoch": 1.3671658347199114,
+      "grad_norm": 0.19737666845321655,
+      "learning_rate": 5.835578966949276e-06,
+      "loss": 0.5073,
+      "step": 4930
+    },
+    {
+      "epoch": 1.3674431503050473,
+      "grad_norm": 0.2089546024799347,
+      "learning_rate": 5.830908207674225e-06,
+      "loss": 0.4981,
+      "step": 4931
+    },
+    {
+      "epoch": 1.3677204658901831,
+      "grad_norm": 0.20078176259994507,
+      "learning_rate": 5.826238749715675e-06,
+      "loss": 0.4811,
+      "step": 4932
+    },
+    {
+      "epoch": 1.367997781475319,
+      "grad_norm": 0.20612461864948273,
+      "learning_rate": 5.821570593984765e-06,
+      "loss": 0.488,
+      "step": 4933
+    },
+    {
+      "epoch": 1.368275097060455,
+      "grad_norm": 0.19565999507904053,
+      "learning_rate": 5.816903741392371e-06,
+      "loss": 0.4613,
+      "step": 4934
+    },
+    {
+      "epoch": 1.3685524126455908,
+      "grad_norm": 0.20337355136871338,
+      "learning_rate": 5.812238192849126e-06,
+      "loss": 0.5004,
+      "step": 4935
+    },
+    {
+      "epoch": 1.3688297282307267,
+      "grad_norm": 0.1981084942817688,
+      "learning_rate": 5.8075739492653936e-06,
+      "loss": 0.4963,
+      "step": 4936
+    },
+    {
+      "epoch": 1.3691070438158626,
+      "grad_norm": 0.21550583839416504,
+      "learning_rate": 5.8029110115512975e-06,
+      "loss": 0.5181,
+      "step": 4937
+    },
+    {
+      "epoch": 1.3693843594009985,
+      "grad_norm": 0.2106407731771469,
+      "learning_rate": 5.7982493806167025e-06,
+      "loss": 0.5001,
+      "step": 4938
+    },
+    {
+      "epoch": 1.3696616749861343,
+      "grad_norm": 0.20755064487457275,
+      "learning_rate": 5.793589057371214e-06,
+      "loss": 0.4996,
+      "step": 4939
+    },
+    {
+      "epoch": 1.3699389905712702,
+      "grad_norm": 0.20589278638362885,
+      "learning_rate": 5.788930042724178e-06,
+      "loss": 0.4731,
+      "step": 4940
+    },
+    {
+      "epoch": 1.370216306156406,
+      "grad_norm": 0.19447284936904907,
+      "learning_rate": 5.7842723375846964e-06,
+      "loss": 0.4805,
+      "step": 4941
+    },
+    {
+      "epoch": 1.370493621741542,
+      "grad_norm": 0.18845054507255554,
+      "learning_rate": 5.779615942861617e-06,
+      "loss": 0.5183,
+      "step": 4942
+    },
+    {
+      "epoch": 1.3707709373266779,
+      "grad_norm": 0.2057630866765976,
+      "learning_rate": 5.774960859463516e-06,
+      "loss": 0.4762,
+      "step": 4943
+    },
+    {
+      "epoch": 1.3710482529118138,
+      "grad_norm": 0.20433704555034637,
+      "learning_rate": 5.770307088298728e-06,
+      "loss": 0.5085,
+      "step": 4944
+    },
+    {
+      "epoch": 1.3713255684969496,
+      "grad_norm": 0.20006833970546722,
+      "learning_rate": 5.76565463027533e-06,
+      "loss": 0.5217,
+      "step": 4945
+    },
+    {
+      "epoch": 1.3716028840820855,
+      "grad_norm": 0.1979144960641861,
+      "learning_rate": 5.761003486301138e-06,
+      "loss": 0.4903,
+      "step": 4946
+    },
+    {
+      "epoch": 1.3718801996672214,
+      "grad_norm": 0.20525884628295898,
+      "learning_rate": 5.756353657283707e-06,
+      "loss": 0.5108,
+      "step": 4947
+    },
+    {
+      "epoch": 1.3721575152523573,
+      "grad_norm": 0.20061984658241272,
+      "learning_rate": 5.7517051441303486e-06,
+      "loss": 0.5068,
+      "step": 4948
+    },
+    {
+      "epoch": 1.3724348308374932,
+      "grad_norm": 0.19711720943450928,
+      "learning_rate": 5.747057947748112e-06,
+      "loss": 0.5206,
+      "step": 4949
+    },
+    {
+      "epoch": 1.372712146422629,
+      "grad_norm": 0.20400893688201904,
+      "learning_rate": 5.742412069043786e-06,
+      "loss": 0.4879,
+      "step": 4950
+    },
+    {
+      "epoch": 1.372989462007765,
+      "grad_norm": 0.20518803596496582,
+      "learning_rate": 5.737767508923896e-06,
+      "loss": 0.5172,
+      "step": 4951
+    },
+    {
+      "epoch": 1.3732667775929008,
+      "grad_norm": 0.20601531863212585,
+      "learning_rate": 5.733124268294734e-06,
+      "loss": 0.535,
+      "step": 4952
+    },
+    {
+      "epoch": 1.3735440931780367,
+      "grad_norm": 0.21886709332466125,
+      "learning_rate": 5.728482348062314e-06,
+      "loss": 0.4945,
+      "step": 4953
+    },
+    {
+      "epoch": 1.3738214087631726,
+      "grad_norm": 0.2085854411125183,
+      "learning_rate": 5.723841749132395e-06,
+      "loss": 0.5162,
+      "step": 4954
+    },
+    {
+      "epoch": 1.3740987243483085,
+      "grad_norm": 0.2051091194152832,
+      "learning_rate": 5.719202472410475e-06,
+      "loss": 0.4917,
+      "step": 4955
+    },
+    {
+      "epoch": 1.3743760399334444,
+      "grad_norm": 0.20153647661209106,
+      "learning_rate": 5.714564518801813e-06,
+      "loss": 0.4679,
+      "step": 4956
+    },
+    {
+      "epoch": 1.3746533555185803,
+      "grad_norm": 0.20284593105316162,
+      "learning_rate": 5.709927889211391e-06,
+      "loss": 0.4813,
+      "step": 4957
+    },
+    {
+      "epoch": 1.3749306711037161,
+      "grad_norm": 0.20113231241703033,
+      "learning_rate": 5.705292584543932e-06,
+      "loss": 0.4958,
+      "step": 4958
+    },
+    {
+      "epoch": 1.375207986688852,
+      "grad_norm": 0.21188965439796448,
+      "learning_rate": 5.700658605703912e-06,
+      "loss": 0.5014,
+      "step": 4959
+    },
+    {
+      "epoch": 1.375485302273988,
+      "grad_norm": 0.20827704668045044,
+      "learning_rate": 5.696025953595549e-06,
+      "loss": 0.505,
+      "step": 4960
+    },
+    {
+      "epoch": 1.3757626178591238,
+      "grad_norm": 0.2128673493862152,
+      "learning_rate": 5.691394629122786e-06,
+      "loss": 0.5019,
+      "step": 4961
+    },
+    {
+      "epoch": 1.3760399334442597,
+      "grad_norm": 0.20192976295948029,
+      "learning_rate": 5.686764633189325e-06,
+      "loss": 0.4714,
+      "step": 4962
+    },
+    {
+      "epoch": 1.3763172490293956,
+      "grad_norm": 0.20420150458812714,
+      "learning_rate": 5.6821359666985925e-06,
+      "loss": 0.5052,
+      "step": 4963
+    },
+    {
+      "epoch": 1.3765945646145314,
+      "grad_norm": 0.19757725298404694,
+      "learning_rate": 5.677508630553774e-06,
+      "loss": 0.5136,
+      "step": 4964
+    },
+    {
+      "epoch": 1.3768718801996673,
+      "grad_norm": 0.19914944469928741,
+      "learning_rate": 5.672882625657776e-06,
+      "loss": 0.4925,
+      "step": 4965
+    },
+    {
+      "epoch": 1.3771491957848032,
+      "grad_norm": 0.20814815163612366,
+      "learning_rate": 5.668257952913259e-06,
+      "loss": 0.4955,
+      "step": 4966
+    },
+    {
+      "epoch": 1.377426511369939,
+      "grad_norm": 0.19726327061653137,
+      "learning_rate": 5.663634613222623e-06,
+      "loss": 0.4675,
+      "step": 4967
+    },
+    {
+      "epoch": 1.377703826955075,
+      "grad_norm": 0.20900078117847443,
+      "learning_rate": 5.659012607487994e-06,
+      "loss": 0.4912,
+      "step": 4968
+    },
+    {
+      "epoch": 1.3779811425402109,
+      "grad_norm": 0.21374890208244324,
+      "learning_rate": 5.65439193661126e-06,
+      "loss": 0.5166,
+      "step": 4969
+    },
+    {
+      "epoch": 1.3782584581253468,
+      "grad_norm": 0.21223044395446777,
+      "learning_rate": 5.649772601494026e-06,
+      "loss": 0.494,
+      "step": 4970
+    },
+    {
+      "epoch": 1.3785357737104826,
+      "grad_norm": 0.20285436511039734,
+      "learning_rate": 5.645154603037654e-06,
+      "loss": 0.5161,
+      "step": 4971
+    },
+    {
+      "epoch": 1.3788130892956185,
+      "grad_norm": 0.20432451367378235,
+      "learning_rate": 5.64053794214323e-06,
+      "loss": 0.4738,
+      "step": 4972
+    },
+    {
+      "epoch": 1.3790904048807544,
+      "grad_norm": 0.20761069655418396,
+      "learning_rate": 5.635922619711598e-06,
+      "loss": 0.5038,
+      "step": 4973
+    },
+    {
+      "epoch": 1.3793677204658903,
+      "grad_norm": 0.21306075155735016,
+      "learning_rate": 5.631308636643317e-06,
+      "loss": 0.4948,
+      "step": 4974
+    },
+    {
+      "epoch": 1.3796450360510262,
+      "grad_norm": 0.20453637838363647,
+      "learning_rate": 5.626695993838704e-06,
+      "loss": 0.5061,
+      "step": 4975
+    },
+    {
+      "epoch": 1.379922351636162,
+      "grad_norm": 0.20153799653053284,
+      "learning_rate": 5.622084692197811e-06,
+      "loss": 0.4923,
+      "step": 4976
+    },
+    {
+      "epoch": 1.380199667221298,
+      "grad_norm": 0.2005978673696518,
+      "learning_rate": 5.617474732620423e-06,
+      "loss": 0.4906,
+      "step": 4977
+    },
+    {
+      "epoch": 1.3804769828064338,
+      "grad_norm": 0.20294132828712463,
+      "learning_rate": 5.612866116006059e-06,
+      "loss": 0.4977,
+      "step": 4978
+    },
+    {
+      "epoch": 1.3807542983915697,
+      "grad_norm": 0.19424274563789368,
+      "learning_rate": 5.608258843253985e-06,
+      "loss": 0.4766,
+      "step": 4979
+    },
+    {
+      "epoch": 1.3810316139767056,
+      "grad_norm": 0.20564329624176025,
+      "learning_rate": 5.60365291526321e-06,
+      "loss": 0.5037,
+      "step": 4980
+    },
+    {
+      "epoch": 1.3813089295618415,
+      "grad_norm": 0.19803635776042938,
+      "learning_rate": 5.599048332932462e-06,
+      "loss": 0.4728,
+      "step": 4981
+    },
+    {
+      "epoch": 1.3815862451469774,
+      "grad_norm": 0.20995113253593445,
+      "learning_rate": 5.594445097160221e-06,
+      "loss": 0.5107,
+      "step": 4982
+    },
+    {
+      "epoch": 1.3818635607321132,
+      "grad_norm": 0.20244790613651276,
+      "learning_rate": 5.589843208844707e-06,
+      "loss": 0.4971,
+      "step": 4983
+    },
+    {
+      "epoch": 1.3821408763172491,
+      "grad_norm": 0.2114204615354538,
+      "learning_rate": 5.5852426688838625e-06,
+      "loss": 0.5059,
+      "step": 4984
+    },
+    {
+      "epoch": 1.382418191902385,
+      "grad_norm": 0.2145579308271408,
+      "learning_rate": 5.580643478175372e-06,
+      "loss": 0.4949,
+      "step": 4985
+    },
+    {
+      "epoch": 1.382695507487521,
+      "grad_norm": 0.1978285163640976,
+      "learning_rate": 5.576045637616663e-06,
+      "loss": 0.4941,
+      "step": 4986
+    },
+    {
+      "epoch": 1.3829728230726568,
+      "grad_norm": 0.1951082944869995,
+      "learning_rate": 5.571449148104903e-06,
+      "loss": 0.4821,
+      "step": 4987
+    },
+    {
+      "epoch": 1.3832501386577927,
+      "grad_norm": 0.4343617260456085,
+      "learning_rate": 5.5668540105369815e-06,
+      "loss": 0.5103,
+      "step": 4988
+    },
+    {
+      "epoch": 1.3835274542429286,
+      "grad_norm": 0.19941391050815582,
+      "learning_rate": 5.562260225809524e-06,
+      "loss": 0.4847,
+      "step": 4989
+    },
+    {
+      "epoch": 1.3838047698280644,
+      "grad_norm": 0.2054002434015274,
+      "learning_rate": 5.557667794818917e-06,
+      "loss": 0.5059,
+      "step": 4990
+    },
+    {
+      "epoch": 1.3840820854132003,
+      "grad_norm": 0.19655835628509521,
+      "learning_rate": 5.5530767184612584e-06,
+      "loss": 0.5045,
+      "step": 4991
+    },
+    {
+      "epoch": 1.3843594009983362,
+      "grad_norm": 0.2060847282409668,
+      "learning_rate": 5.548486997632386e-06,
+      "loss": 0.5122,
+      "step": 4992
+    },
+    {
+      "epoch": 1.384636716583472,
+      "grad_norm": 0.20054815709590912,
+      "learning_rate": 5.543898633227869e-06,
+      "loss": 0.4945,
+      "step": 4993
+    },
+    {
+      "epoch": 1.384914032168608,
+      "grad_norm": 0.20712971687316895,
+      "learning_rate": 5.539311626143034e-06,
+      "loss": 0.4962,
+      "step": 4994
+    },
+    {
+      "epoch": 1.3851913477537439,
+      "grad_norm": 0.20592884719371796,
+      "learning_rate": 5.534725977272923e-06,
+      "loss": 0.5046,
+      "step": 4995
+    },
+    {
+      "epoch": 1.3854686633388797,
+      "grad_norm": 0.19674254953861237,
+      "learning_rate": 5.530141687512311e-06,
+      "loss": 0.5111,
+      "step": 4996
+    },
+    {
+      "epoch": 1.3857459789240156,
+      "grad_norm": 0.20650383830070496,
+      "learning_rate": 5.525558757755716e-06,
+      "loss": 0.4801,
+      "step": 4997
+    },
+    {
+      "epoch": 1.3860232945091515,
+      "grad_norm": 0.20828871428966522,
+      "learning_rate": 5.520977188897398e-06,
+      "loss": 0.4849,
+      "step": 4998
+    },
+    {
+      "epoch": 1.3863006100942874,
+      "grad_norm": 0.20563003420829773,
+      "learning_rate": 5.516396981831337e-06,
+      "loss": 0.528,
+      "step": 4999
+    },
+    {
+      "epoch": 1.3865779256794233,
+      "grad_norm": 0.19634945690631866,
+      "learning_rate": 5.511818137451247e-06,
+      "loss": 0.482,
+      "step": 5000
+    },
+    {
+      "epoch": 1.3868552412645592,
+      "grad_norm": 0.19331537187099457,
+      "learning_rate": 5.507240656650586e-06,
+      "loss": 0.5174,
+      "step": 5001
+    },
+    {
+      "epoch": 1.387132556849695,
+      "grad_norm": 0.19204163551330566,
+      "learning_rate": 5.502664540322547e-06,
+      "loss": 0.4646,
+      "step": 5002
+    },
+    {
+      "epoch": 1.387409872434831,
+      "grad_norm": 0.19761690497398376,
+      "learning_rate": 5.498089789360043e-06,
+      "loss": 0.4901,
+      "step": 5003
+    },
+    {
+      "epoch": 1.3876871880199668,
+      "grad_norm": 0.22188632190227509,
+      "learning_rate": 5.493516404655733e-06,
+      "loss": 0.5084,
+      "step": 5004
+    },
+    {
+      "epoch": 1.3879645036051027,
+      "grad_norm": 0.20775288343429565,
+      "learning_rate": 5.48894438710201e-06,
+      "loss": 0.4741,
+      "step": 5005
+    },
+    {
+      "epoch": 1.3882418191902386,
+      "grad_norm": 0.2092396765947342,
+      "learning_rate": 5.484373737590992e-06,
+      "loss": 0.5163,
+      "step": 5006
+    },
+    {
+      "epoch": 1.3885191347753745,
+      "grad_norm": 0.20691785216331482,
+      "learning_rate": 5.479804457014528e-06,
+      "loss": 0.4954,
+      "step": 5007
+    },
+    {
+      "epoch": 1.3887964503605104,
+      "grad_norm": 0.20370697975158691,
+      "learning_rate": 5.4752365462642115e-06,
+      "loss": 0.492,
+      "step": 5008
+    },
+    {
+      "epoch": 1.3890737659456462,
+      "grad_norm": 0.19262473285198212,
+      "learning_rate": 5.4706700062313686e-06,
+      "loss": 0.5066,
+      "step": 5009
+    },
+    {
+      "epoch": 1.3893510815307821,
+      "grad_norm": 0.20191557705402374,
+      "learning_rate": 5.466104837807038e-06,
+      "loss": 0.4935,
+      "step": 5010
+    },
+    {
+      "epoch": 1.389628397115918,
+      "grad_norm": 0.2161003202199936,
+      "learning_rate": 5.461541041882021e-06,
+      "loss": 0.5021,
+      "step": 5011
+    },
+    {
+      "epoch": 1.389905712701054,
+      "grad_norm": 0.2024545818567276,
+      "learning_rate": 5.456978619346821e-06,
+      "loss": 0.4968,
+      "step": 5012
+    },
+    {
+      "epoch": 1.3901830282861898,
+      "grad_norm": 0.19628916680812836,
+      "learning_rate": 5.452417571091699e-06,
+      "loss": 0.5084,
+      "step": 5013
+    },
+    {
+      "epoch": 1.3904603438713257,
+      "grad_norm": 0.21482454240322113,
+      "learning_rate": 5.447857898006625e-06,
+      "loss": 0.5142,
+      "step": 5014
+    },
+    {
+      "epoch": 1.3907376594564616,
+      "grad_norm": 0.19978967308998108,
+      "learning_rate": 5.4432996009813235e-06,
+      "loss": 0.448,
+      "step": 5015
+    },
+    {
+      "epoch": 1.3910149750415974,
+      "grad_norm": 0.21920114755630493,
+      "learning_rate": 5.43874268090523e-06,
+      "loss": 0.5065,
+      "step": 5016
+    },
+    {
+      "epoch": 1.3912922906267333,
+      "grad_norm": 0.1980915367603302,
+      "learning_rate": 5.434187138667522e-06,
+      "loss": 0.5073,
+      "step": 5017
+    },
+    {
+      "epoch": 1.3915696062118692,
+      "grad_norm": 0.19898997247219086,
+      "learning_rate": 5.429632975157115e-06,
+      "loss": 0.5052,
+      "step": 5018
+    },
+    {
+      "epoch": 1.391846921797005,
+      "grad_norm": 0.20365798473358154,
+      "learning_rate": 5.425080191262634e-06,
+      "loss": 0.4903,
+      "step": 5019
+    },
+    {
+      "epoch": 1.392124237382141,
+      "grad_norm": 0.21267344057559967,
+      "learning_rate": 5.42052878787246e-06,
+      "loss": 0.5094,
+      "step": 5020
+    },
+    {
+      "epoch": 1.3924015529672769,
+      "grad_norm": 0.1986275613307953,
+      "learning_rate": 5.415978765874681e-06,
+      "loss": 0.5039,
+      "step": 5021
+    },
+    {
+      "epoch": 1.3926788685524127,
+      "grad_norm": 0.20506128668785095,
+      "learning_rate": 5.411430126157138e-06,
+      "loss": 0.5119,
+      "step": 5022
+    },
+    {
+      "epoch": 1.3929561841375486,
+      "grad_norm": 0.20554526150226593,
+      "learning_rate": 5.406882869607381e-06,
+      "loss": 0.4979,
+      "step": 5023
+    },
+    {
+      "epoch": 1.3932334997226845,
+      "grad_norm": 0.19483192265033722,
+      "learning_rate": 5.402336997112703e-06,
+      "loss": 0.4768,
+      "step": 5024
+    },
+    {
+      "epoch": 1.3935108153078204,
+      "grad_norm": 0.18851692974567413,
+      "learning_rate": 5.397792509560132e-06,
+      "loss": 0.4797,
+      "step": 5025
+    },
+    {
+      "epoch": 1.3937881308929563,
+      "grad_norm": 0.19844412803649902,
+      "learning_rate": 5.3932494078364125e-06,
+      "loss": 0.501,
+      "step": 5026
+    },
+    {
+      "epoch": 1.3940654464780922,
+      "grad_norm": 0.20119161903858185,
+      "learning_rate": 5.388707692828013e-06,
+      "loss": 0.4795,
+      "step": 5027
+    },
+    {
+      "epoch": 1.394342762063228,
+      "grad_norm": 0.1947944462299347,
+      "learning_rate": 5.384167365421161e-06,
+      "loss": 0.4779,
+      "step": 5028
+    },
+    {
+      "epoch": 1.394620077648364,
+      "grad_norm": 0.20363706350326538,
+      "learning_rate": 5.379628426501789e-06,
+      "loss": 0.4925,
+      "step": 5029
+    },
+    {
+      "epoch": 1.3948973932334998,
+      "grad_norm": 0.21294178068637848,
+      "learning_rate": 5.375090876955559e-06,
+      "loss": 0.4978,
+      "step": 5030
+    },
+    {
+      "epoch": 1.3951747088186357,
+      "grad_norm": 0.19777782261371613,
+      "learning_rate": 5.370554717667861e-06,
+      "loss": 0.4888,
+      "step": 5031
+    },
+    {
+      "epoch": 1.3954520244037716,
+      "grad_norm": 0.201828733086586,
+      "learning_rate": 5.36601994952384e-06,
+      "loss": 0.5061,
+      "step": 5032
+    },
+    {
+      "epoch": 1.3957293399889075,
+      "grad_norm": 0.20095033943653107,
+      "learning_rate": 5.3614865734083365e-06,
+      "loss": 0.4768,
+      "step": 5033
+    },
+    {
+      "epoch": 1.3960066555740434,
+      "grad_norm": 0.20809856057167053,
+      "learning_rate": 5.3569545902059285e-06,
+      "loss": 0.5032,
+      "step": 5034
+    },
+    {
+      "epoch": 1.3962839711591792,
+      "grad_norm": 0.20099328458309174,
+      "learning_rate": 5.352424000800934e-06,
+      "loss": 0.4971,
+      "step": 5035
+    },
+    {
+      "epoch": 1.3965612867443151,
+      "grad_norm": 0.20865300297737122,
+      "learning_rate": 5.34789480607739e-06,
+      "loss": 0.49,
+      "step": 5036
+    },
+    {
+      "epoch": 1.396838602329451,
+      "grad_norm": 0.28519347310066223,
+      "learning_rate": 5.3433670069190616e-06,
+      "loss": 0.4885,
+      "step": 5037
+    },
+    {
+      "epoch": 1.397115917914587,
+      "grad_norm": 0.19800283014774323,
+      "learning_rate": 5.338840604209438e-06,
+      "loss": 0.5033,
+      "step": 5038
+    },
+    {
+      "epoch": 1.3973932334997228,
+      "grad_norm": 0.2006731927394867,
+      "learning_rate": 5.334315598831743e-06,
+      "loss": 0.5022,
+      "step": 5039
+    },
+    {
+      "epoch": 1.3976705490848587,
+      "grad_norm": 0.200321763753891,
+      "learning_rate": 5.329791991668931e-06,
+      "loss": 0.4846,
+      "step": 5040
+    },
+    {
+      "epoch": 1.3979478646699945,
+      "grad_norm": 0.21275892853736877,
+      "learning_rate": 5.3252697836036675e-06,
+      "loss": 0.503,
+      "step": 5041
+    },
+    {
+      "epoch": 1.3982251802551304,
+      "grad_norm": 0.19826599955558777,
+      "learning_rate": 5.320748975518361e-06,
+      "loss": 0.4912,
+      "step": 5042
+    },
+    {
+      "epoch": 1.3985024958402663,
+      "grad_norm": 0.20074287056922913,
+      "learning_rate": 5.316229568295143e-06,
+      "loss": 0.5036,
+      "step": 5043
+    },
+    {
+      "epoch": 1.3987798114254022,
+      "grad_norm": 0.19842980802059174,
+      "learning_rate": 5.311711562815869e-06,
+      "loss": 0.5017,
+      "step": 5044
+    },
+    {
+      "epoch": 1.399057127010538,
+      "grad_norm": 0.19995532929897308,
+      "learning_rate": 5.307194959962112e-06,
+      "loss": 0.4696,
+      "step": 5045
+    },
+    {
+      "epoch": 1.399334442595674,
+      "grad_norm": 0.20109152793884277,
+      "learning_rate": 5.302679760615189e-06,
+      "loss": 0.4875,
+      "step": 5046
+    },
+    {
+      "epoch": 1.3996117581808099,
+      "grad_norm": 0.19944603741168976,
+      "learning_rate": 5.298165965656139e-06,
+      "loss": 0.485,
+      "step": 5047
+    },
+    {
+      "epoch": 1.3998890737659457,
+      "grad_norm": 0.19794021546840668,
+      "learning_rate": 5.293653575965714e-06,
+      "loss": 0.508,
+      "step": 5048
+    },
+    {
+      "epoch": 1.4001663893510816,
+      "grad_norm": 0.20664082467556,
+      "learning_rate": 5.2891425924244095e-06,
+      "loss": 0.498,
+      "step": 5049
+    },
+    {
+      "epoch": 1.4004437049362175,
+      "grad_norm": 0.2035246342420578,
+      "learning_rate": 5.284633015912428e-06,
+      "loss": 0.5023,
+      "step": 5050
+    },
+    {
+      "epoch": 1.4007210205213534,
+      "grad_norm": 0.2174990326166153,
+      "learning_rate": 5.280124847309717e-06,
+      "loss": 0.483,
+      "step": 5051
+    },
+    {
+      "epoch": 1.4009983361064893,
+      "grad_norm": 0.1980149745941162,
+      "learning_rate": 5.275618087495932e-06,
+      "loss": 0.4739,
+      "step": 5052
+    },
+    {
+      "epoch": 1.4012756516916252,
+      "grad_norm": 0.19647562503814697,
+      "learning_rate": 5.271112737350467e-06,
+      "loss": 0.4788,
+      "step": 5053
+    },
+    {
+      "epoch": 1.401552967276761,
+      "grad_norm": 0.1995796263217926,
+      "learning_rate": 5.266608797752429e-06,
+      "loss": 0.4577,
+      "step": 5054
+    },
+    {
+      "epoch": 1.401830282861897,
+      "grad_norm": 0.2159973829984665,
+      "learning_rate": 5.26210626958066e-06,
+      "loss": 0.5022,
+      "step": 5055
+    },
+    {
+      "epoch": 1.4021075984470328,
+      "grad_norm": 0.20078851282596588,
+      "learning_rate": 5.257605153713727e-06,
+      "loss": 0.4864,
+      "step": 5056
+    },
+    {
+      "epoch": 1.4023849140321687,
+      "grad_norm": 0.20937801897525787,
+      "learning_rate": 5.253105451029908e-06,
+      "loss": 0.4945,
+      "step": 5057
+    },
+    {
+      "epoch": 1.4026622296173046,
+      "grad_norm": 0.2047879546880722,
+      "learning_rate": 5.248607162407221e-06,
+      "loss": 0.5019,
+      "step": 5058
+    },
+    {
+      "epoch": 1.4029395452024405,
+      "grad_norm": 0.19715817272663116,
+      "learning_rate": 5.244110288723396e-06,
+      "loss": 0.492,
+      "step": 5059
+    },
+    {
+      "epoch": 1.4032168607875763,
+      "grad_norm": 0.20772488415241241,
+      "learning_rate": 5.2396148308558976e-06,
+      "loss": 0.5004,
+      "step": 5060
+    },
+    {
+      "epoch": 1.4034941763727122,
+      "grad_norm": 0.19815978407859802,
+      "learning_rate": 5.235120789681902e-06,
+      "loss": 0.452,
+      "step": 5061
+    },
+    {
+      "epoch": 1.4037714919578481,
+      "grad_norm": 0.2069665491580963,
+      "learning_rate": 5.23062816607832e-06,
+      "loss": 0.4955,
+      "step": 5062
+    },
+    {
+      "epoch": 1.404048807542984,
+      "grad_norm": 0.20308996737003326,
+      "learning_rate": 5.226136960921786e-06,
+      "loss": 0.5137,
+      "step": 5063
+    },
+    {
+      "epoch": 1.4043261231281199,
+      "grad_norm": 0.2034929245710373,
+      "learning_rate": 5.221647175088648e-06,
+      "loss": 0.4617,
+      "step": 5064
+    },
+    {
+      "epoch": 1.4046034387132558,
+      "grad_norm": 0.21139517426490784,
+      "learning_rate": 5.217158809454979e-06,
+      "loss": 0.5068,
+      "step": 5065
+    },
+    {
+      "epoch": 1.4048807542983917,
+      "grad_norm": 0.21125884354114532,
+      "learning_rate": 5.212671864896581e-06,
+      "loss": 0.537,
+      "step": 5066
+    },
+    {
+      "epoch": 1.4051580698835275,
+      "grad_norm": 0.21030350029468536,
+      "learning_rate": 5.208186342288979e-06,
+      "loss": 0.5214,
+      "step": 5067
+    },
+    {
+      "epoch": 1.4054353854686634,
+      "grad_norm": 0.19781124591827393,
+      "learning_rate": 5.203702242507416e-06,
+      "loss": 0.4858,
+      "step": 5068
+    },
+    {
+      "epoch": 1.4057127010537993,
+      "grad_norm": 0.20567484200000763,
+      "learning_rate": 5.199219566426848e-06,
+      "loss": 0.4904,
+      "step": 5069
+    },
+    {
+      "epoch": 1.4059900166389352,
+      "grad_norm": 0.267501562833786,
+      "learning_rate": 5.194738314921982e-06,
+      "loss": 0.4828,
+      "step": 5070
+    },
+    {
+      "epoch": 1.406267332224071,
+      "grad_norm": 0.2026844173669815,
+      "learning_rate": 5.1902584888672206e-06,
+      "loss": 0.4974,
+      "step": 5071
+    },
+    {
+      "epoch": 1.406544647809207,
+      "grad_norm": 0.20914383232593536,
+      "learning_rate": 5.185780089136691e-06,
+      "loss": 0.4762,
+      "step": 5072
+    },
+    {
+      "epoch": 1.4068219633943428,
+      "grad_norm": 0.20183809101581573,
+      "learning_rate": 5.181303116604253e-06,
+      "loss": 0.4782,
+      "step": 5073
+    },
+    {
+      "epoch": 1.4070992789794787,
+      "grad_norm": 0.20599482953548431,
+      "learning_rate": 5.176827572143486e-06,
+      "loss": 0.4911,
+      "step": 5074
+    },
+    {
+      "epoch": 1.4073765945646146,
+      "grad_norm": 0.1961178332567215,
+      "learning_rate": 5.172353456627683e-06,
+      "loss": 0.4784,
+      "step": 5075
+    },
+    {
+      "epoch": 1.4076539101497505,
+      "grad_norm": 0.2111572027206421,
+      "learning_rate": 5.1678807709298605e-06,
+      "loss": 0.4856,
+      "step": 5076
+    },
+    {
+      "epoch": 1.4079312257348864,
+      "grad_norm": 0.19795557856559753,
+      "learning_rate": 5.163409515922758e-06,
+      "loss": 0.4855,
+      "step": 5077
+    },
+    {
+      "epoch": 1.4082085413200223,
+      "grad_norm": 0.2135738879442215,
+      "learning_rate": 5.158939692478845e-06,
+      "loss": 0.5018,
+      "step": 5078
+    },
+    {
+      "epoch": 1.4084858569051582,
+      "grad_norm": 0.2012612372636795,
+      "learning_rate": 5.154471301470294e-06,
+      "loss": 0.4999,
+      "step": 5079
+    },
+    {
+      "epoch": 1.408763172490294,
+      "grad_norm": 0.20429253578186035,
+      "learning_rate": 5.150004343769001e-06,
+      "loss": 0.4955,
+      "step": 5080
+    },
+    {
+      "epoch": 1.40904048807543,
+      "grad_norm": 0.2064649909734726,
+      "learning_rate": 5.1455388202466025e-06,
+      "loss": 0.5041,
+      "step": 5081
+    },
+    {
+      "epoch": 1.4093178036605658,
+      "grad_norm": 0.19653840363025665,
+      "learning_rate": 5.141074731774433e-06,
+      "loss": 0.5023,
+      "step": 5082
+    },
+    {
+      "epoch": 1.4095951192457017,
+      "grad_norm": 0.20743335783481598,
+      "learning_rate": 5.13661207922355e-06,
+      "loss": 0.5033,
+      "step": 5083
+    },
+    {
+      "epoch": 1.4098724348308376,
+      "grad_norm": 0.21355509757995605,
+      "learning_rate": 5.13215086346474e-06,
+      "loss": 0.5117,
+      "step": 5084
+    },
+    {
+      "epoch": 1.4101497504159735,
+      "grad_norm": 0.21359796822071075,
+      "learning_rate": 5.127691085368508e-06,
+      "loss": 0.4959,
+      "step": 5085
+    },
+    {
+      "epoch": 1.4104270660011093,
+      "grad_norm": 0.21122997999191284,
+      "learning_rate": 5.123232745805067e-06,
+      "loss": 0.5095,
+      "step": 5086
+    },
+    {
+      "epoch": 1.4107043815862452,
+      "grad_norm": 0.20164433121681213,
+      "learning_rate": 5.118775845644365e-06,
+      "loss": 0.5069,
+      "step": 5087
+    },
+    {
+      "epoch": 1.410981697171381,
+      "grad_norm": 0.205289825797081,
+      "learning_rate": 5.114320385756052e-06,
+      "loss": 0.5191,
+      "step": 5088
+    },
+    {
+      "epoch": 1.411259012756517,
+      "grad_norm": 0.20195280015468597,
+      "learning_rate": 5.109866367009518e-06,
+      "loss": 0.4861,
+      "step": 5089
+    },
+    {
+      "epoch": 1.4115363283416529,
+      "grad_norm": 0.2103511095046997,
+      "learning_rate": 5.105413790273848e-06,
+      "loss": 0.5086,
+      "step": 5090
+    },
+    {
+      "epoch": 1.4118136439267888,
+      "grad_norm": 0.2056579887866974,
+      "learning_rate": 5.1009626564178685e-06,
+      "loss": 0.5043,
+      "step": 5091
+    },
+    {
+      "epoch": 1.4120909595119246,
+      "grad_norm": 0.22724954783916473,
+      "learning_rate": 5.096512966310103e-06,
+      "loss": 0.5088,
+      "step": 5092
+    },
+    {
+      "epoch": 1.4123682750970605,
+      "grad_norm": 0.20662304759025574,
+      "learning_rate": 5.0920647208188105e-06,
+      "loss": 0.4832,
+      "step": 5093
+    },
+    {
+      "epoch": 1.4126455906821964,
+      "grad_norm": 0.20837345719337463,
+      "learning_rate": 5.087617920811966e-06,
+      "loss": 0.5003,
+      "step": 5094
+    },
+    {
+      "epoch": 1.4129229062673323,
+      "grad_norm": 0.20730285346508026,
+      "learning_rate": 5.083172567157246e-06,
+      "loss": 0.4922,
+      "step": 5095
+    },
+    {
+      "epoch": 1.4132002218524682,
+      "grad_norm": 0.20552845299243927,
+      "learning_rate": 5.078728660722068e-06,
+      "loss": 0.5003,
+      "step": 5096
+    },
+    {
+      "epoch": 1.413477537437604,
+      "grad_norm": 0.21342763304710388,
+      "learning_rate": 5.074286202373547e-06,
+      "loss": 0.5156,
+      "step": 5097
+    },
+    {
+      "epoch": 1.41375485302274,
+      "grad_norm": 0.20104481279850006,
+      "learning_rate": 5.069845192978534e-06,
+      "loss": 0.5034,
+      "step": 5098
+    },
+    {
+      "epoch": 1.4140321686078758,
+      "grad_norm": 0.1914806067943573,
+      "learning_rate": 5.065405633403576e-06,
+      "loss": 0.474,
+      "step": 5099
+    },
+    {
+      "epoch": 1.4143094841930117,
+      "grad_norm": 0.19734273850917816,
+      "learning_rate": 5.060967524514956e-06,
+      "loss": 0.4821,
+      "step": 5100
+    },
+    {
+      "epoch": 1.4145867997781476,
+      "grad_norm": 0.2332550585269928,
+      "learning_rate": 5.05653086717867e-06,
+      "loss": 0.4929,
+      "step": 5101
+    },
+    {
+      "epoch": 1.4148641153632835,
+      "grad_norm": 0.20123951137065887,
+      "learning_rate": 5.052095662260421e-06,
+      "loss": 0.4948,
+      "step": 5102
+    },
+    {
+      "epoch": 1.4151414309484194,
+      "grad_norm": 0.20322628319263458,
+      "learning_rate": 5.047661910625634e-06,
+      "loss": 0.5161,
+      "step": 5103
+    },
+    {
+      "epoch": 1.4154187465335553,
+      "grad_norm": 0.2106594741344452,
+      "learning_rate": 5.043229613139454e-06,
+      "loss": 0.4567,
+      "step": 5104
+    },
+    {
+      "epoch": 1.4156960621186911,
+      "grad_norm": 0.21148160099983215,
+      "learning_rate": 5.038798770666744e-06,
+      "loss": 0.5119,
+      "step": 5105
+    },
+    {
+      "epoch": 1.415973377703827,
+      "grad_norm": 0.20894502103328705,
+      "learning_rate": 5.034369384072075e-06,
+      "loss": 0.4947,
+      "step": 5106
+    },
+    {
+      "epoch": 1.416250693288963,
+      "grad_norm": 0.20449240505695343,
+      "learning_rate": 5.029941454219728e-06,
+      "loss": 0.5044,
+      "step": 5107
+    },
+    {
+      "epoch": 1.4165280088740988,
+      "grad_norm": 0.21614056825637817,
+      "learning_rate": 5.025514981973728e-06,
+      "loss": 0.4952,
+      "step": 5108
+    },
+    {
+      "epoch": 1.4168053244592347,
+      "grad_norm": 0.19919288158416748,
+      "learning_rate": 5.0210899681977865e-06,
+      "loss": 0.5098,
+      "step": 5109
+    },
+    {
+      "epoch": 1.4170826400443706,
+      "grad_norm": 0.20138102769851685,
+      "learning_rate": 5.01666641375534e-06,
+      "loss": 0.4817,
+      "step": 5110
+    },
+    {
+      "epoch": 1.4173599556295065,
+      "grad_norm": 0.20684263110160828,
+      "learning_rate": 5.0122443195095416e-06,
+      "loss": 0.4934,
+      "step": 5111
+    },
+    {
+      "epoch": 1.4176372712146423,
+      "grad_norm": 0.20800793170928955,
+      "learning_rate": 5.007823686323267e-06,
+      "loss": 0.4999,
+      "step": 5112
+    },
+    {
+      "epoch": 1.4179145867997782,
+      "grad_norm": 0.20006629824638367,
+      "learning_rate": 5.0034045150590905e-06,
+      "loss": 0.4677,
+      "step": 5113
+    },
+    {
+      "epoch": 1.418191902384914,
+      "grad_norm": 0.20559728145599365,
+      "learning_rate": 4.998986806579309e-06,
+      "loss": 0.473,
+      "step": 5114
+    },
+    {
+      "epoch": 1.41846921797005,
+      "grad_norm": 0.2054857760667801,
+      "learning_rate": 4.994570561745936e-06,
+      "loss": 0.5104,
+      "step": 5115
+    },
+    {
+      "epoch": 1.4187465335551859,
+      "grad_norm": 0.21354462206363678,
+      "learning_rate": 4.990155781420704e-06,
+      "loss": 0.5001,
+      "step": 5116
+    },
+    {
+      "epoch": 1.4190238491403218,
+      "grad_norm": 0.1943047046661377,
+      "learning_rate": 4.985742466465047e-06,
+      "loss": 0.4839,
+      "step": 5117
+    },
+    {
+      "epoch": 1.4193011647254576,
+      "grad_norm": 0.2034555971622467,
+      "learning_rate": 4.981330617740118e-06,
+      "loss": 0.5107,
+      "step": 5118
+    },
+    {
+      "epoch": 1.4195784803105935,
+      "grad_norm": 0.2111237347126007,
+      "learning_rate": 4.9769202361067895e-06,
+      "loss": 0.5202,
+      "step": 5119
+    },
+    {
+      "epoch": 1.4198557958957294,
+      "grad_norm": 0.19280794262886047,
+      "learning_rate": 4.972511322425648e-06,
+      "loss": 0.5044,
+      "step": 5120
+    },
+    {
+      "epoch": 1.4201331114808653,
+      "grad_norm": 0.198442280292511,
+      "learning_rate": 4.968103877556979e-06,
+      "loss": 0.4788,
+      "step": 5121
+    },
+    {
+      "epoch": 1.4204104270660012,
+      "grad_norm": 0.20548267662525177,
+      "learning_rate": 4.963697902360798e-06,
+      "loss": 0.4731,
+      "step": 5122
+    },
+    {
+      "epoch": 1.420687742651137,
+      "grad_norm": 0.22825707495212555,
+      "learning_rate": 4.959293397696831e-06,
+      "loss": 0.4998,
+      "step": 5123
+    },
+    {
+      "epoch": 1.420965058236273,
+      "grad_norm": 0.20349940657615662,
+      "learning_rate": 4.954890364424508e-06,
+      "loss": 0.5058,
+      "step": 5124
+    },
+    {
+      "epoch": 1.4212423738214088,
+      "grad_norm": 0.20583848655223846,
+      "learning_rate": 4.950488803402975e-06,
+      "loss": 0.484,
+      "step": 5125
+    },
+    {
+      "epoch": 1.4215196894065447,
+      "grad_norm": 0.20237183570861816,
+      "learning_rate": 4.9460887154910985e-06,
+      "loss": 0.5071,
+      "step": 5126
+    },
+    {
+      "epoch": 1.4217970049916806,
+      "grad_norm": 0.20790375769138336,
+      "learning_rate": 4.941690101547454e-06,
+      "loss": 0.4937,
+      "step": 5127
+    },
+    {
+      "epoch": 1.4220743205768165,
+      "grad_norm": 0.22605054080486298,
+      "learning_rate": 4.9372929624303205e-06,
+      "loss": 0.4915,
+      "step": 5128
+    },
+    {
+      "epoch": 1.4223516361619524,
+      "grad_norm": 0.20419445633888245,
+      "learning_rate": 4.932897298997703e-06,
+      "loss": 0.4946,
+      "step": 5129
+    },
+    {
+      "epoch": 1.4226289517470883,
+      "grad_norm": 0.21240122616291046,
+      "learning_rate": 4.928503112107306e-06,
+      "loss": 0.5002,
+      "step": 5130
+    },
+    {
+      "epoch": 1.4229062673322241,
+      "grad_norm": 0.2051856964826584,
+      "learning_rate": 4.92411040261656e-06,
+      "loss": 0.4913,
+      "step": 5131
+    },
+    {
+      "epoch": 1.42318358291736,
+      "grad_norm": 0.20038112998008728,
+      "learning_rate": 4.919719171382588e-06,
+      "loss": 0.5047,
+      "step": 5132
+    },
+    {
+      "epoch": 1.423460898502496,
+      "grad_norm": 0.20497171580791473,
+      "learning_rate": 4.915329419262242e-06,
+      "loss": 0.5028,
+      "step": 5133
+    },
+    {
+      "epoch": 1.4237382140876318,
+      "grad_norm": 0.1987651139497757,
+      "learning_rate": 4.910941147112083e-06,
+      "loss": 0.4977,
+      "step": 5134
+    },
+    {
+      "epoch": 1.4240155296727677,
+      "grad_norm": 0.28679215908050537,
+      "learning_rate": 4.906554355788369e-06,
+      "loss": 0.4989,
+      "step": 5135
+    },
+    {
+      "epoch": 1.4242928452579036,
+      "grad_norm": 0.202947735786438,
+      "learning_rate": 4.90216904614709e-06,
+      "loss": 0.4802,
+      "step": 5136
+    },
+    {
+      "epoch": 1.4245701608430394,
+      "grad_norm": 0.2093580663204193,
+      "learning_rate": 4.897785219043927e-06,
+      "loss": 0.4787,
+      "step": 5137
+    },
+    {
+      "epoch": 1.4248474764281753,
+      "grad_norm": 0.23357978463172913,
+      "learning_rate": 4.893402875334288e-06,
+      "loss": 0.4944,
+      "step": 5138
+    },
+    {
+      "epoch": 1.4251247920133112,
+      "grad_norm": 0.20304684340953827,
+      "learning_rate": 4.889022015873277e-06,
+      "loss": 0.4846,
+      "step": 5139
+    },
+    {
+      "epoch": 1.425402107598447,
+      "grad_norm": 0.20766004920005798,
+      "learning_rate": 4.884642641515724e-06,
+      "loss": 0.4871,
+      "step": 5140
+    },
+    {
+      "epoch": 1.425679423183583,
+      "grad_norm": 0.20664039254188538,
+      "learning_rate": 4.880264753116153e-06,
+      "loss": 0.481,
+      "step": 5141
+    },
+    {
+      "epoch": 1.4259567387687189,
+      "grad_norm": 0.20463590323925018,
+      "learning_rate": 4.875888351528808e-06,
+      "loss": 0.5147,
+      "step": 5142
+    },
+    {
+      "epoch": 1.4262340543538548,
+      "grad_norm": 0.20236745476722717,
+      "learning_rate": 4.871513437607648e-06,
+      "loss": 0.4744,
+      "step": 5143
+    },
+    {
+      "epoch": 1.4265113699389906,
+      "grad_norm": 0.19818803668022156,
+      "learning_rate": 4.867140012206331e-06,
+      "loss": 0.5014,
+      "step": 5144
+    },
+    {
+      "epoch": 1.4267886855241265,
+      "grad_norm": 0.22346088290214539,
+      "learning_rate": 4.86276807617822e-06,
+      "loss": 0.6068,
+      "step": 5145
+    },
+    {
+      "epoch": 1.4270660011092624,
+      "grad_norm": 0.21238963305950165,
+      "learning_rate": 4.858397630376402e-06,
+      "loss": 0.4987,
+      "step": 5146
+    },
+    {
+      "epoch": 1.4273433166943983,
+      "grad_norm": 0.20392395555973053,
+      "learning_rate": 4.854028675653673e-06,
+      "loss": 0.4896,
+      "step": 5147
+    },
+    {
+      "epoch": 1.4276206322795342,
+      "grad_norm": 0.20558500289916992,
+      "learning_rate": 4.849661212862519e-06,
+      "loss": 0.485,
+      "step": 5148
+    },
+    {
+      "epoch": 1.42789794786467,
+      "grad_norm": 0.20195691287517548,
+      "learning_rate": 4.8452952428551555e-06,
+      "loss": 0.4898,
+      "step": 5149
+    },
+    {
+      "epoch": 1.428175263449806,
+      "grad_norm": 0.20537883043289185,
+      "learning_rate": 4.8409307664835005e-06,
+      "loss": 0.51,
+      "step": 5150
+    },
+    {
+      "epoch": 1.4284525790349418,
+      "grad_norm": 0.1957499235868454,
+      "learning_rate": 4.836567784599177e-06,
+      "loss": 0.4805,
+      "step": 5151
+    },
+    {
+      "epoch": 1.4287298946200777,
+      "grad_norm": 0.20477133989334106,
+      "learning_rate": 4.832206298053514e-06,
+      "loss": 0.5113,
+      "step": 5152
+    },
+    {
+      "epoch": 1.4290072102052136,
+      "grad_norm": 0.20341917872428894,
+      "learning_rate": 4.827846307697556e-06,
+      "loss": 0.4972,
+      "step": 5153
+    },
+    {
+      "epoch": 1.4292845257903495,
+      "grad_norm": 0.20808421075344086,
+      "learning_rate": 4.823487814382058e-06,
+      "loss": 0.5044,
+      "step": 5154
+    },
+    {
+      "epoch": 1.4295618413754854,
+      "grad_norm": 0.19239382445812225,
+      "learning_rate": 4.819130818957472e-06,
+      "loss": 0.47,
+      "step": 5155
+    },
+    {
+      "epoch": 1.4298391569606212,
+      "grad_norm": 0.20638294517993927,
+      "learning_rate": 4.814775322273961e-06,
+      "loss": 0.5008,
+      "step": 5156
+    },
+    {
+      "epoch": 1.4301164725457571,
+      "grad_norm": 0.20612825453281403,
+      "learning_rate": 4.8104213251814e-06,
+      "loss": 0.4738,
+      "step": 5157
+    },
+    {
+      "epoch": 1.430393788130893,
+      "grad_norm": 0.20103727281093597,
+      "learning_rate": 4.806068828529374e-06,
+      "loss": 0.4854,
+      "step": 5158
+    },
+    {
+      "epoch": 1.430671103716029,
+      "grad_norm": 0.20140987634658813,
+      "learning_rate": 4.801717833167162e-06,
+      "loss": 0.498,
+      "step": 5159
+    },
+    {
+      "epoch": 1.4309484193011648,
+      "grad_norm": 0.20108608901500702,
+      "learning_rate": 4.797368339943763e-06,
+      "loss": 0.4587,
+      "step": 5160
+    },
+    {
+      "epoch": 1.4312257348863007,
+      "grad_norm": 0.20430995523929596,
+      "learning_rate": 4.793020349707883e-06,
+      "loss": 0.502,
+      "step": 5161
+    },
+    {
+      "epoch": 1.4315030504714366,
+      "grad_norm": 0.1962215155363083,
+      "learning_rate": 4.7886738633079254e-06,
+      "loss": 0.4873,
+      "step": 5162
+    },
+    {
+      "epoch": 1.4317803660565724,
+      "grad_norm": 0.2041066735982895,
+      "learning_rate": 4.784328881592e-06,
+      "loss": 0.4807,
+      "step": 5163
+    },
+    {
+      "epoch": 1.4320576816417083,
+      "grad_norm": 0.2137051373720169,
+      "learning_rate": 4.779985405407933e-06,
+      "loss": 0.4924,
+      "step": 5164
+    },
+    {
+      "epoch": 1.4323349972268442,
+      "grad_norm": 0.20929081737995148,
+      "learning_rate": 4.775643435603255e-06,
+      "loss": 0.5052,
+      "step": 5165
+    },
+    {
+      "epoch": 1.43261231281198,
+      "grad_norm": 0.2020164579153061,
+      "learning_rate": 4.7713029730251925e-06,
+      "loss": 0.5179,
+      "step": 5166
+    },
+    {
+      "epoch": 1.432889628397116,
+      "grad_norm": 0.21634337306022644,
+      "learning_rate": 4.766964018520691e-06,
+      "loss": 0.5043,
+      "step": 5167
+    },
+    {
+      "epoch": 1.4331669439822519,
+      "grad_norm": 0.20789554715156555,
+      "learning_rate": 4.762626572936389e-06,
+      "loss": 0.4949,
+      "step": 5168
+    },
+    {
+      "epoch": 1.4334442595673877,
+      "grad_norm": 0.20186547935009003,
+      "learning_rate": 4.7582906371186435e-06,
+      "loss": 0.4865,
+      "step": 5169
+    },
+    {
+      "epoch": 1.4337215751525236,
+      "grad_norm": 0.20182502269744873,
+      "learning_rate": 4.753956211913504e-06,
+      "loss": 0.4684,
+      "step": 5170
+    },
+    {
+      "epoch": 1.4339988907376595,
+      "grad_norm": 0.20346789062023163,
+      "learning_rate": 4.749623298166736e-06,
+      "loss": 0.498,
+      "step": 5171
+    },
+    {
+      "epoch": 1.4342762063227954,
+      "grad_norm": 0.21296192705631256,
+      "learning_rate": 4.745291896723808e-06,
+      "loss": 0.4982,
+      "step": 5172
+    },
+    {
+      "epoch": 1.4345535219079313,
+      "grad_norm": 0.21046607196331024,
+      "learning_rate": 4.740962008429885e-06,
+      "loss": 0.5046,
+      "step": 5173
+    },
+    {
+      "epoch": 1.4348308374930672,
+      "grad_norm": 0.20590196549892426,
+      "learning_rate": 4.736633634129849e-06,
+      "loss": 0.4919,
+      "step": 5174
+    },
+    {
+      "epoch": 1.435108153078203,
+      "grad_norm": 0.18597187101840973,
+      "learning_rate": 4.732306774668274e-06,
+      "loss": 0.492,
+      "step": 5175
+    },
+    {
+      "epoch": 1.435385468663339,
+      "grad_norm": 0.20251090824604034,
+      "learning_rate": 4.727981430889452e-06,
+      "loss": 0.534,
+      "step": 5176
+    },
+    {
+      "epoch": 1.4356627842484748,
+      "grad_norm": 0.19426211714744568,
+      "learning_rate": 4.723657603637364e-06,
+      "loss": 0.4626,
+      "step": 5177
+    },
+    {
+      "epoch": 1.4359400998336107,
+      "grad_norm": 0.20881786942481995,
+      "learning_rate": 4.7193352937557125e-06,
+      "loss": 0.4843,
+      "step": 5178
+    },
+    {
+      "epoch": 1.4362174154187466,
+      "grad_norm": 0.19851280748844147,
+      "learning_rate": 4.7150145020878865e-06,
+      "loss": 0.5019,
+      "step": 5179
+    },
+    {
+      "epoch": 1.4364947310038825,
+      "grad_norm": 0.2078125774860382,
+      "learning_rate": 4.71069522947699e-06,
+      "loss": 0.4671,
+      "step": 5180
+    },
+    {
+      "epoch": 1.4367720465890184,
+      "grad_norm": 0.20463737845420837,
+      "learning_rate": 4.706377476765832e-06,
+      "loss": 0.5038,
+      "step": 5181
+    },
+    {
+      "epoch": 1.4370493621741542,
+      "grad_norm": 0.20316970348358154,
+      "learning_rate": 4.702061244796916e-06,
+      "loss": 0.4726,
+      "step": 5182
+    },
+    {
+      "epoch": 1.4373266777592901,
+      "grad_norm": 0.1965150386095047,
+      "learning_rate": 4.69774653441245e-06,
+      "loss": 0.4753,
+      "step": 5183
+    },
+    {
+      "epoch": 1.437603993344426,
+      "grad_norm": 0.19861292839050293,
+      "learning_rate": 4.693433346454352e-06,
+      "loss": 0.491,
+      "step": 5184
+    },
+    {
+      "epoch": 1.437881308929562,
+      "grad_norm": 0.2055879682302475,
+      "learning_rate": 4.689121681764243e-06,
+      "loss": 0.496,
+      "step": 5185
+    },
+    {
+      "epoch": 1.4381586245146978,
+      "grad_norm": 0.20754766464233398,
+      "learning_rate": 4.684811541183436e-06,
+      "loss": 0.5039,
+      "step": 5186
+    },
+    {
+      "epoch": 1.4384359400998337,
+      "grad_norm": 0.2045608013868332,
+      "learning_rate": 4.680502925552956e-06,
+      "loss": 0.4964,
+      "step": 5187
+    },
+    {
+      "epoch": 1.4387132556849695,
+      "grad_norm": 0.20918679237365723,
+      "learning_rate": 4.676195835713533e-06,
+      "loss": 0.505,
+      "step": 5188
+    },
+    {
+      "epoch": 1.4389905712701054,
+      "grad_norm": 0.21245136857032776,
+      "learning_rate": 4.67189027250559e-06,
+      "loss": 0.4968,
+      "step": 5189
+    },
+    {
+      "epoch": 1.4392678868552413,
+      "grad_norm": 0.30057552456855774,
+      "learning_rate": 4.667586236769253e-06,
+      "loss": 0.4925,
+      "step": 5190
+    },
+    {
+      "epoch": 1.4395452024403772,
+      "grad_norm": 0.2126583307981491,
+      "learning_rate": 4.6632837293443575e-06,
+      "loss": 0.4906,
+      "step": 5191
+    },
+    {
+      "epoch": 1.439822518025513,
+      "grad_norm": 0.20761823654174805,
+      "learning_rate": 4.65898275107044e-06,
+      "loss": 0.507,
+      "step": 5192
+    },
+    {
+      "epoch": 1.440099833610649,
+      "grad_norm": 0.19843849539756775,
+      "learning_rate": 4.65468330278673e-06,
+      "loss": 0.4824,
+      "step": 5193
+    },
+    {
+      "epoch": 1.4403771491957849,
+      "grad_norm": 0.2050592452287674,
+      "learning_rate": 4.650385385332163e-06,
+      "loss": 0.4696,
+      "step": 5194
+    },
+    {
+      "epoch": 1.4406544647809207,
+      "grad_norm": 0.20630647242069244,
+      "learning_rate": 4.646088999545378e-06,
+      "loss": 0.514,
+      "step": 5195
+    },
+    {
+      "epoch": 1.4409317803660566,
+      "grad_norm": 0.20964011549949646,
+      "learning_rate": 4.6417941462647206e-06,
+      "loss": 0.5106,
+      "step": 5196
+    },
+    {
+      "epoch": 1.4412090959511925,
+      "grad_norm": 0.20389756560325623,
+      "learning_rate": 4.637500826328223e-06,
+      "loss": 0.5036,
+      "step": 5197
+    },
+    {
+      "epoch": 1.4414864115363284,
+      "grad_norm": 0.21767324209213257,
+      "learning_rate": 4.633209040573619e-06,
+      "loss": 0.5197,
+      "step": 5198
+    },
+    {
+      "epoch": 1.4417637271214643,
+      "grad_norm": 0.22654587030410767,
+      "learning_rate": 4.628918789838367e-06,
+      "loss": 0.5067,
+      "step": 5199
+    },
+    {
+      "epoch": 1.4420410427066002,
+      "grad_norm": 0.19964653253555298,
+      "learning_rate": 4.624630074959599e-06,
+      "loss": 0.4778,
+      "step": 5200
+    },
+    {
+      "epoch": 1.442318358291736,
+      "grad_norm": 0.20028294622898102,
+      "learning_rate": 4.620342896774152e-06,
+      "loss": 0.4896,
+      "step": 5201
+    },
+    {
+      "epoch": 1.442595673876872,
+      "grad_norm": 0.20315535366535187,
+      "learning_rate": 4.616057256118575e-06,
+      "loss": 0.495,
+      "step": 5202
+    },
+    {
+      "epoch": 1.4428729894620078,
+      "grad_norm": 0.23156633973121643,
+      "learning_rate": 4.611773153829111e-06,
+      "loss": 0.4772,
+      "step": 5203
+    },
+    {
+      "epoch": 1.4431503050471437,
+      "grad_norm": 0.22991783916950226,
+      "learning_rate": 4.607490590741702e-06,
+      "loss": 0.5134,
+      "step": 5204
+    },
+    {
+      "epoch": 1.4434276206322796,
+      "grad_norm": 0.20309005677700043,
+      "learning_rate": 4.603209567691979e-06,
+      "loss": 0.4889,
+      "step": 5205
+    },
+    {
+      "epoch": 1.4437049362174155,
+      "grad_norm": 0.20435002446174622,
+      "learning_rate": 4.598930085515293e-06,
+      "loss": 0.4943,
+      "step": 5206
+    },
+    {
+      "epoch": 1.4439822518025514,
+      "grad_norm": 0.19982177019119263,
+      "learning_rate": 4.594652145046688e-06,
+      "loss": 0.4995,
+      "step": 5207
+    },
+    {
+      "epoch": 1.4442595673876872,
+      "grad_norm": 0.22092601656913757,
+      "learning_rate": 4.5903757471208914e-06,
+      "loss": 0.5108,
+      "step": 5208
+    },
+    {
+      "epoch": 1.4445368829728231,
+      "grad_norm": 0.2120787650346756,
+      "learning_rate": 4.586100892572352e-06,
+      "loss": 0.48,
+      "step": 5209
+    },
+    {
+      "epoch": 1.444814198557959,
+      "grad_norm": 0.209842249751091,
+      "learning_rate": 4.5818275822352e-06,
+      "loss": 0.4888,
+      "step": 5210
+    },
+    {
+      "epoch": 1.445091514143095,
+      "grad_norm": 0.1947280764579773,
+      "learning_rate": 4.577555816943279e-06,
+      "loss": 0.5057,
+      "step": 5211
+    },
+    {
+      "epoch": 1.4453688297282308,
+      "grad_norm": 0.20627637207508087,
+      "learning_rate": 4.573285597530114e-06,
+      "loss": 0.5058,
+      "step": 5212
+    },
+    {
+      "epoch": 1.4456461453133667,
+      "grad_norm": 0.2139754295349121,
+      "learning_rate": 4.569016924828945e-06,
+      "loss": 0.4867,
+      "step": 5213
+    },
+    {
+      "epoch": 1.4459234608985025,
+      "grad_norm": 0.21095266938209534,
+      "learning_rate": 4.564749799672705e-06,
+      "loss": 0.5132,
+      "step": 5214
+    },
+    {
+      "epoch": 1.4462007764836384,
+      "grad_norm": 0.21572549641132355,
+      "learning_rate": 4.560484222894014e-06,
+      "loss": 0.4752,
+      "step": 5215
+    },
+    {
+      "epoch": 1.4464780920687743,
+      "grad_norm": 0.21467210352420807,
+      "learning_rate": 4.55622019532521e-06,
+      "loss": 0.5311,
+      "step": 5216
+    },
+    {
+      "epoch": 1.4467554076539102,
+      "grad_norm": 0.21415258944034576,
+      "learning_rate": 4.551957717798308e-06,
+      "loss": 0.4949,
+      "step": 5217
+    },
+    {
+      "epoch": 1.447032723239046,
+      "grad_norm": 0.2082109898328781,
+      "learning_rate": 4.5476967911450345e-06,
+      "loss": 0.501,
+      "step": 5218
+    },
+    {
+      "epoch": 1.447310038824182,
+      "grad_norm": 0.22765327990055084,
+      "learning_rate": 4.543437416196814e-06,
+      "loss": 0.4966,
+      "step": 5219
+    },
+    {
+      "epoch": 1.4475873544093179,
+      "grad_norm": 0.20551997423171997,
+      "learning_rate": 4.539179593784758e-06,
+      "loss": 0.4974,
+      "step": 5220
+    },
+    {
+      "epoch": 1.4478646699944537,
+      "grad_norm": 0.20596256852149963,
+      "learning_rate": 4.534923324739677e-06,
+      "loss": 0.4995,
+      "step": 5221
+    },
+    {
+      "epoch": 1.4481419855795896,
+      "grad_norm": 0.1983633041381836,
+      "learning_rate": 4.530668609892087e-06,
+      "loss": 0.4959,
+      "step": 5222
+    },
+    {
+      "epoch": 1.4484193011647255,
+      "grad_norm": 0.19731079041957855,
+      "learning_rate": 4.526415450072198e-06,
+      "loss": 0.4974,
+      "step": 5223
+    },
+    {
+      "epoch": 1.4486966167498614,
+      "grad_norm": 0.21283119916915894,
+      "learning_rate": 4.522163846109908e-06,
+      "loss": 0.521,
+      "step": 5224
+    },
+    {
+      "epoch": 1.4489739323349973,
+      "grad_norm": 0.21046127378940582,
+      "learning_rate": 4.517913798834818e-06,
+      "loss": 0.483,
+      "step": 5225
+    },
+    {
+      "epoch": 1.4492512479201332,
+      "grad_norm": 0.22251440584659576,
+      "learning_rate": 4.513665309076233e-06,
+      "loss": 0.4978,
+      "step": 5226
+    },
+    {
+      "epoch": 1.449528563505269,
+      "grad_norm": 0.20175869762897491,
+      "learning_rate": 4.50941837766314e-06,
+      "loss": 0.5356,
+      "step": 5227
+    },
+    {
+      "epoch": 1.449805879090405,
+      "grad_norm": 0.21276865899562836,
+      "learning_rate": 4.505173005424224e-06,
+      "loss": 0.4813,
+      "step": 5228
+    },
+    {
+      "epoch": 1.4500831946755408,
+      "grad_norm": 0.20545652508735657,
+      "learning_rate": 4.500929193187872e-06,
+      "loss": 0.4697,
+      "step": 5229
+    },
+    {
+      "epoch": 1.4503605102606767,
+      "grad_norm": 0.21193096041679382,
+      "learning_rate": 4.496686941782172e-06,
+      "loss": 0.5156,
+      "step": 5230
+    },
+    {
+      "epoch": 1.4506378258458126,
+      "grad_norm": 0.2051508128643036,
+      "learning_rate": 4.492446252034893e-06,
+      "loss": 0.5015,
+      "step": 5231
+    },
+    {
+      "epoch": 1.4509151414309485,
+      "grad_norm": 0.2122107744216919,
+      "learning_rate": 4.488207124773501e-06,
+      "loss": 0.4912,
+      "step": 5232
+    },
+    {
+      "epoch": 1.4511924570160843,
+      "grad_norm": 0.19480577111244202,
+      "learning_rate": 4.483969560825169e-06,
+      "loss": 0.4876,
+      "step": 5233
+    },
+    {
+      "epoch": 1.4514697726012202,
+      "grad_norm": 0.20153845846652985,
+      "learning_rate": 4.479733561016759e-06,
+      "loss": 0.5092,
+      "step": 5234
+    },
+    {
+      "epoch": 1.4517470881863561,
+      "grad_norm": 0.21945208311080933,
+      "learning_rate": 4.475499126174826e-06,
+      "loss": 0.5147,
+      "step": 5235
+    },
+    {
+      "epoch": 1.452024403771492,
+      "grad_norm": 0.215095654129982,
+      "learning_rate": 4.471266257125609e-06,
+      "loss": 0.5222,
+      "step": 5236
+    },
+    {
+      "epoch": 1.4523017193566279,
+      "grad_norm": 0.20627257227897644,
+      "learning_rate": 4.467034954695071e-06,
+      "loss": 0.4668,
+      "step": 5237
+    },
+    {
+      "epoch": 1.4525790349417638,
+      "grad_norm": 0.2098441869020462,
+      "learning_rate": 4.462805219708843e-06,
+      "loss": 0.4668,
+      "step": 5238
+    },
+    {
+      "epoch": 1.4528563505268997,
+      "grad_norm": 0.22378282248973846,
+      "learning_rate": 4.4585770529922535e-06,
+      "loss": 0.4728,
+      "step": 5239
+    },
+    {
+      "epoch": 1.4531336661120355,
+      "grad_norm": 0.21106509864330292,
+      "learning_rate": 4.454350455370334e-06,
+      "loss": 0.5006,
+      "step": 5240
+    },
+    {
+      "epoch": 1.4534109816971714,
+      "grad_norm": 0.21677586436271667,
+      "learning_rate": 4.450125427667812e-06,
+      "loss": 0.4977,
+      "step": 5241
+    },
+    {
+      "epoch": 1.4536882972823073,
+      "grad_norm": 0.2078382968902588,
+      "learning_rate": 4.4459019707090956e-06,
+      "loss": 0.5159,
+      "step": 5242
+    },
+    {
+      "epoch": 1.4539656128674432,
+      "grad_norm": 0.19706477224826813,
+      "learning_rate": 4.441680085318289e-06,
+      "loss": 0.4935,
+      "step": 5243
+    },
+    {
+      "epoch": 1.454242928452579,
+      "grad_norm": 0.19929809868335724,
+      "learning_rate": 4.437459772319201e-06,
+      "loss": 0.4725,
+      "step": 5244
+    },
+    {
+      "epoch": 1.454520244037715,
+      "grad_norm": 0.19897820055484772,
+      "learning_rate": 4.4332410325353265e-06,
+      "loss": 0.4693,
+      "step": 5245
+    },
+    {
+      "epoch": 1.4547975596228508,
+      "grad_norm": 0.19526724517345428,
+      "learning_rate": 4.429023866789848e-06,
+      "loss": 0.4856,
+      "step": 5246
+    },
+    {
+      "epoch": 1.4550748752079867,
+      "grad_norm": 0.2073434442281723,
+      "learning_rate": 4.424808275905654e-06,
+      "loss": 0.4909,
+      "step": 5247
+    },
+    {
+      "epoch": 1.4553521907931226,
+      "grad_norm": 0.2015712559223175,
+      "learning_rate": 4.420594260705309e-06,
+      "loss": 0.4805,
+      "step": 5248
+    },
+    {
+      "epoch": 1.4556295063782585,
+      "grad_norm": 0.20347121357917786,
+      "learning_rate": 4.416381822011087e-06,
+      "loss": 0.5058,
+      "step": 5249
+    },
+    {
+      "epoch": 1.4559068219633944,
+      "grad_norm": 0.20566286146640778,
+      "learning_rate": 4.412170960644939e-06,
+      "loss": 0.4688,
+      "step": 5250
+    },
+    {
+      "epoch": 1.4561841375485303,
+      "grad_norm": 0.2034814953804016,
+      "learning_rate": 4.407961677428521e-06,
+      "loss": 0.5307,
+      "step": 5251
+    },
+    {
+      "epoch": 1.4564614531336662,
+      "grad_norm": 0.21038807928562164,
+      "learning_rate": 4.403753973183177e-06,
+      "loss": 0.5287,
+      "step": 5252
+    },
+    {
+      "epoch": 1.456738768718802,
+      "grad_norm": 0.2552982270717621,
+      "learning_rate": 4.399547848729935e-06,
+      "loss": 0.4838,
+      "step": 5253
+    },
+    {
+      "epoch": 1.457016084303938,
+      "grad_norm": 0.194267138838768,
+      "learning_rate": 4.395343304889529e-06,
+      "loss": 0.4981,
+      "step": 5254
+    },
+    {
+      "epoch": 1.4572933998890738,
+      "grad_norm": 0.20130032300949097,
+      "learning_rate": 4.391140342482369e-06,
+      "loss": 0.4863,
+      "step": 5255
+    },
+    {
+      "epoch": 1.4575707154742097,
+      "grad_norm": 0.21114356815814972,
+      "learning_rate": 4.3869389623285725e-06,
+      "loss": 0.511,
+      "step": 5256
+    },
+    {
+      "epoch": 1.4578480310593456,
+      "grad_norm": 0.19720643758773804,
+      "learning_rate": 4.382739165247933e-06,
+      "loss": 0.4973,
+      "step": 5257
+    },
+    {
+      "epoch": 1.4581253466444815,
+      "grad_norm": 0.1979367434978485,
+      "learning_rate": 4.378540952059948e-06,
+      "loss": 0.5119,
+      "step": 5258
+    },
+    {
+      "epoch": 1.4584026622296173,
+      "grad_norm": 0.1983027458190918,
+      "learning_rate": 4.374344323583793e-06,
+      "loss": 0.5073,
+      "step": 5259
+    },
+    {
+      "epoch": 1.4586799778147532,
+      "grad_norm": 0.20961186289787292,
+      "learning_rate": 4.370149280638347e-06,
+      "loss": 0.5181,
+      "step": 5260
+    },
+    {
+      "epoch": 1.458957293399889,
+      "grad_norm": 0.1928398460149765,
+      "learning_rate": 4.3659558240421755e-06,
+      "loss": 0.504,
+      "step": 5261
+    },
+    {
+      "epoch": 1.459234608985025,
+      "grad_norm": 0.20917226374149323,
+      "learning_rate": 4.361763954613526e-06,
+      "loss": 0.5018,
+      "step": 5262
+    },
+    {
+      "epoch": 1.4595119245701609,
+      "grad_norm": 0.19321578741073608,
+      "learning_rate": 4.357573673170352e-06,
+      "loss": 0.4699,
+      "step": 5263
+    },
+    {
+      "epoch": 1.4597892401552968,
+      "grad_norm": 0.19718189537525177,
+      "learning_rate": 4.35338498053028e-06,
+      "loss": 0.4754,
+      "step": 5264
+    },
+    {
+      "epoch": 1.4600665557404326,
+      "grad_norm": 0.20244868099689484,
+      "learning_rate": 4.349197877510643e-06,
+      "loss": 0.4869,
+      "step": 5265
+    },
+    {
+      "epoch": 1.4603438713255685,
+      "grad_norm": 0.20539158582687378,
+      "learning_rate": 4.345012364928447e-06,
+      "loss": 0.5209,
+      "step": 5266
+    },
+    {
+      "epoch": 1.4606211869107044,
+      "grad_norm": 0.19396114349365234,
+      "learning_rate": 4.340828443600401e-06,
+      "loss": 0.4901,
+      "step": 5267
+    },
+    {
+      "epoch": 1.4608985024958403,
+      "grad_norm": 0.21171221137046814,
+      "learning_rate": 4.336646114342903e-06,
+      "loss": 0.49,
+      "step": 5268
+    },
+    {
+      "epoch": 1.4611758180809762,
+      "grad_norm": 0.20075739920139313,
+      "learning_rate": 4.332465377972031e-06,
+      "loss": 0.5127,
+      "step": 5269
+    },
+    {
+      "epoch": 1.461453133666112,
+      "grad_norm": 0.2100810706615448,
+      "learning_rate": 4.328286235303555e-06,
+      "loss": 0.506,
+      "step": 5270
+    },
+    {
+      "epoch": 1.461730449251248,
+      "grad_norm": 0.224534273147583,
+      "learning_rate": 4.324108687152941e-06,
+      "loss": 0.5033,
+      "step": 5271
+    },
+    {
+      "epoch": 1.4620077648363838,
+      "grad_norm": 0.20113049447536469,
+      "learning_rate": 4.3199327343353415e-06,
+      "loss": 0.4867,
+      "step": 5272
+    },
+    {
+      "epoch": 1.4622850804215197,
+      "grad_norm": 0.20711010694503784,
+      "learning_rate": 4.315758377665592e-06,
+      "loss": 0.4818,
+      "step": 5273
+    },
+    {
+      "epoch": 1.4625623960066556,
+      "grad_norm": 0.20650602877140045,
+      "learning_rate": 4.311585617958214e-06,
+      "loss": 0.4794,
+      "step": 5274
+    },
+    {
+      "epoch": 1.4628397115917915,
+      "grad_norm": 0.19701525568962097,
+      "learning_rate": 4.307414456027437e-06,
+      "loss": 0.5037,
+      "step": 5275
+    },
+    {
+      "epoch": 1.4631170271769274,
+      "grad_norm": 0.2053639143705368,
+      "learning_rate": 4.303244892687157e-06,
+      "loss": 0.4924,
+      "step": 5276
+    },
+    {
+      "epoch": 1.4633943427620633,
+      "grad_norm": 0.20747579634189606,
+      "learning_rate": 4.299076928750964e-06,
+      "loss": 0.4847,
+      "step": 5277
+    },
+    {
+      "epoch": 1.4636716583471991,
+      "grad_norm": 0.19542749226093292,
+      "learning_rate": 4.294910565032143e-06,
+      "loss": 0.4799,
+      "step": 5278
+    },
+    {
+      "epoch": 1.463948973932335,
+      "grad_norm": 0.20481598377227783,
+      "learning_rate": 4.290745802343663e-06,
+      "loss": 0.4998,
+      "step": 5279
+    },
+    {
+      "epoch": 1.464226289517471,
+      "grad_norm": 0.20202511548995972,
+      "learning_rate": 4.286582641498177e-06,
+      "loss": 0.5196,
+      "step": 5280
+    },
+    {
+      "epoch": 1.4645036051026068,
+      "grad_norm": 0.20724746584892273,
+      "learning_rate": 4.282421083308024e-06,
+      "loss": 0.4854,
+      "step": 5281
+    },
+    {
+      "epoch": 1.4647809206877427,
+      "grad_norm": 0.22080543637275696,
+      "learning_rate": 4.2782611285852386e-06,
+      "loss": 0.5355,
+      "step": 5282
+    },
+    {
+      "epoch": 1.4650582362728786,
+      "grad_norm": 0.20695441961288452,
+      "learning_rate": 4.274102778141542e-06,
+      "loss": 0.4853,
+      "step": 5283
+    },
+    {
+      "epoch": 1.4653355518580145,
+      "grad_norm": 0.20421764254570007,
+      "learning_rate": 4.26994603278833e-06,
+      "loss": 0.4865,
+      "step": 5284
+    },
+    {
+      "epoch": 1.4656128674431503,
+      "grad_norm": 0.19895236194133759,
+      "learning_rate": 4.265790893336702e-06,
+      "loss": 0.5041,
+      "step": 5285
+    },
+    {
+      "epoch": 1.4658901830282862,
+      "grad_norm": 0.20470494031906128,
+      "learning_rate": 4.261637360597428e-06,
+      "loss": 0.4982,
+      "step": 5286
+    },
+    {
+      "epoch": 1.466167498613422,
+      "grad_norm": 0.20590339601039886,
+      "learning_rate": 4.257485435380981e-06,
+      "loss": 0.4576,
+      "step": 5287
+    },
+    {
+      "epoch": 1.466444814198558,
+      "grad_norm": 0.20538951456546783,
+      "learning_rate": 4.253335118497503e-06,
+      "loss": 0.513,
+      "step": 5288
+    },
+    {
+      "epoch": 1.4667221297836939,
+      "grad_norm": 0.19989731907844543,
+      "learning_rate": 4.249186410756834e-06,
+      "loss": 0.4728,
+      "step": 5289
+    },
+    {
+      "epoch": 1.4669994453688298,
+      "grad_norm": 0.19829894602298737,
+      "learning_rate": 4.245039312968502e-06,
+      "loss": 0.5188,
+      "step": 5290
+    },
+    {
+      "epoch": 1.4672767609539656,
+      "grad_norm": 0.2066100686788559,
+      "learning_rate": 4.240893825941707e-06,
+      "loss": 0.5106,
+      "step": 5291
+    },
+    {
+      "epoch": 1.4675540765391015,
+      "grad_norm": 0.2136521190404892,
+      "learning_rate": 4.236749950485351e-06,
+      "loss": 0.4995,
+      "step": 5292
+    },
+    {
+      "epoch": 1.4678313921242374,
+      "grad_norm": 0.2011151909828186,
+      "learning_rate": 4.232607687408007e-06,
+      "loss": 0.4716,
+      "step": 5293
+    },
+    {
+      "epoch": 1.4681087077093733,
+      "grad_norm": 0.20019252598285675,
+      "learning_rate": 4.228467037517945e-06,
+      "loss": 0.5025,
+      "step": 5294
+    },
+    {
+      "epoch": 1.4683860232945092,
+      "grad_norm": 0.21552008390426636,
+      "learning_rate": 4.224328001623114e-06,
+      "loss": 0.4994,
+      "step": 5295
+    },
+    {
+      "epoch": 1.468663338879645,
+      "grad_norm": 0.2044316530227661,
+      "learning_rate": 4.2201905805311515e-06,
+      "loss": 0.4901,
+      "step": 5296
+    },
+    {
+      "epoch": 1.468940654464781,
+      "grad_norm": 0.22182868421077728,
+      "learning_rate": 4.216054775049372e-06,
+      "loss": 0.533,
+      "step": 5297
+    },
+    {
+      "epoch": 1.4692179700499168,
+      "grad_norm": 0.20859482884407043,
+      "learning_rate": 4.211920585984786e-06,
+      "loss": 0.5134,
+      "step": 5298
+    },
+    {
+      "epoch": 1.4694952856350527,
+      "grad_norm": 0.2023824155330658,
+      "learning_rate": 4.207788014144084e-06,
+      "loss": 0.5222,
+      "step": 5299
+    },
+    {
+      "epoch": 1.4697726012201886,
+      "grad_norm": 0.23869916796684265,
+      "learning_rate": 4.203657060333641e-06,
+      "loss": 0.5025,
+      "step": 5300
+    },
+    {
+      "epoch": 1.4700499168053245,
+      "grad_norm": 0.23270224034786224,
+      "learning_rate": 4.199527725359508e-06,
+      "loss": 0.5166,
+      "step": 5301
+    },
+    {
+      "epoch": 1.4703272323904604,
+      "grad_norm": 0.22280660271644592,
+      "learning_rate": 4.195400010027432e-06,
+      "loss": 0.4891,
+      "step": 5302
+    },
+    {
+      "epoch": 1.4706045479755963,
+      "grad_norm": 0.21119210124015808,
+      "learning_rate": 4.191273915142846e-06,
+      "loss": 0.5021,
+      "step": 5303
+    },
+    {
+      "epoch": 1.4708818635607321,
+      "grad_norm": 0.21453891694545746,
+      "learning_rate": 4.187149441510848e-06,
+      "loss": 0.52,
+      "step": 5304
+    },
+    {
+      "epoch": 1.471159179145868,
+      "grad_norm": 0.20360562205314636,
+      "learning_rate": 4.183026589936241e-06,
+      "loss": 0.5032,
+      "step": 5305
+    },
+    {
+      "epoch": 1.471436494731004,
+      "grad_norm": 0.20575298368930817,
+      "learning_rate": 4.178905361223505e-06,
+      "loss": 0.4982,
+      "step": 5306
+    },
+    {
+      "epoch": 1.4717138103161398,
+      "grad_norm": 0.20205079019069672,
+      "learning_rate": 4.174785756176794e-06,
+      "loss": 0.5021,
+      "step": 5307
+    },
+    {
+      "epoch": 1.4719911259012757,
+      "grad_norm": 0.22563353180885315,
+      "learning_rate": 4.170667775599951e-06,
+      "loss": 0.4909,
+      "step": 5308
+    },
+    {
+      "epoch": 1.4722684414864116,
+      "grad_norm": 0.19961145520210266,
+      "learning_rate": 4.166551420296508e-06,
+      "loss": 0.5033,
+      "step": 5309
+    },
+    {
+      "epoch": 1.4725457570715474,
+      "grad_norm": 0.20603464543819427,
+      "learning_rate": 4.162436691069676e-06,
+      "loss": 0.4849,
+      "step": 5310
+    },
+    {
+      "epoch": 1.4728230726566833,
+      "grad_norm": 0.20677094161510468,
+      "learning_rate": 4.1583235887223455e-06,
+      "loss": 0.481,
+      "step": 5311
+    },
+    {
+      "epoch": 1.4731003882418192,
+      "grad_norm": 0.20227313041687012,
+      "learning_rate": 4.154212114057082e-06,
+      "loss": 0.5013,
+      "step": 5312
+    },
+    {
+      "epoch": 1.473377703826955,
+      "grad_norm": 0.2114361971616745,
+      "learning_rate": 4.150102267876163e-06,
+      "loss": 0.4746,
+      "step": 5313
+    },
+    {
+      "epoch": 1.473655019412091,
+      "grad_norm": 0.20628516376018524,
+      "learning_rate": 4.145994050981516e-06,
+      "loss": 0.5033,
+      "step": 5314
+    },
+    {
+      "epoch": 1.4739323349972269,
+      "grad_norm": 0.1993936002254486,
+      "learning_rate": 4.141887464174763e-06,
+      "loss": 0.4915,
+      "step": 5315
+    },
+    {
+      "epoch": 1.4742096505823628,
+      "grad_norm": 0.200067400932312,
+      "learning_rate": 4.137782508257207e-06,
+      "loss": 0.4849,
+      "step": 5316
+    },
+    {
+      "epoch": 1.4744869661674986,
+      "grad_norm": 0.20835205912590027,
+      "learning_rate": 4.1336791840298425e-06,
+      "loss": 0.5053,
+      "step": 5317
+    },
+    {
+      "epoch": 1.4747642817526345,
+      "grad_norm": 0.207503080368042,
+      "learning_rate": 4.1295774922933295e-06,
+      "loss": 0.4963,
+      "step": 5318
+    },
+    {
+      "epoch": 1.4750415973377704,
+      "grad_norm": 0.20725563168525696,
+      "learning_rate": 4.125477433848014e-06,
+      "loss": 0.5063,
+      "step": 5319
+    },
+    {
+      "epoch": 1.4753189129229063,
+      "grad_norm": 0.20372618734836578,
+      "learning_rate": 4.121379009493931e-06,
+      "loss": 0.4971,
+      "step": 5320
+    },
+    {
+      "epoch": 1.4755962285080422,
+      "grad_norm": 0.20692381262779236,
+      "learning_rate": 4.117282220030794e-06,
+      "loss": 0.5218,
+      "step": 5321
+    },
+    {
+      "epoch": 1.475873544093178,
+      "grad_norm": 0.19848695397377014,
+      "learning_rate": 4.113187066257991e-06,
+      "loss": 0.5201,
+      "step": 5322
+    },
+    {
+      "epoch": 1.476150859678314,
+      "grad_norm": 0.20803263783454895,
+      "learning_rate": 4.109093548974592e-06,
+      "loss": 0.5094,
+      "step": 5323
+    },
+    {
+      "epoch": 1.4764281752634498,
+      "grad_norm": 0.20934459567070007,
+      "learning_rate": 4.105001668979355e-06,
+      "loss": 0.4812,
+      "step": 5324
+    },
+    {
+      "epoch": 1.4767054908485857,
+      "grad_norm": 0.22413517534732819,
+      "learning_rate": 4.100911427070718e-06,
+      "loss": 0.5026,
+      "step": 5325
+    },
+    {
+      "epoch": 1.4769828064337216,
+      "grad_norm": 0.19864612817764282,
+      "learning_rate": 4.096822824046787e-06,
+      "loss": 0.5218,
+      "step": 5326
+    },
+    {
+      "epoch": 1.4772601220188575,
+      "grad_norm": 0.20662501454353333,
+      "learning_rate": 4.09273586070536e-06,
+      "loss": 0.5018,
+      "step": 5327
+    },
+    {
+      "epoch": 1.4775374376039934,
+      "grad_norm": 0.20286211371421814,
+      "learning_rate": 4.088650537843919e-06,
+      "loss": 0.4921,
+      "step": 5328
+    },
+    {
+      "epoch": 1.4778147531891292,
+      "grad_norm": 0.21929144859313965,
+      "learning_rate": 4.084566856259611e-06,
+      "loss": 0.483,
+      "step": 5329
+    },
+    {
+      "epoch": 1.4780920687742651,
+      "grad_norm": 0.20432603359222412,
+      "learning_rate": 4.080484816749268e-06,
+      "loss": 0.4813,
+      "step": 5330
+    },
+    {
+      "epoch": 1.478369384359401,
+      "grad_norm": 0.19982703030109406,
+      "learning_rate": 4.076404420109409e-06,
+      "loss": 0.4792,
+      "step": 5331
+    },
+    {
+      "epoch": 1.478646699944537,
+      "grad_norm": 0.20570825040340424,
+      "learning_rate": 4.072325667136228e-06,
+      "loss": 0.4854,
+      "step": 5332
+    },
+    {
+      "epoch": 1.4789240155296728,
+      "grad_norm": 0.2098425328731537,
+      "learning_rate": 4.068248558625595e-06,
+      "loss": 0.484,
+      "step": 5333
+    },
+    {
+      "epoch": 1.4792013311148087,
+      "grad_norm": 0.20449112355709076,
+      "learning_rate": 4.064173095373067e-06,
+      "loss": 0.5152,
+      "step": 5334
+    },
+    {
+      "epoch": 1.4794786466999446,
+      "grad_norm": 0.2116931974887848,
+      "learning_rate": 4.060099278173867e-06,
+      "loss": 0.4783,
+      "step": 5335
+    },
+    {
+      "epoch": 1.4797559622850804,
+      "grad_norm": 0.1945703774690628,
+      "learning_rate": 4.056027107822911e-06,
+      "loss": 0.4861,
+      "step": 5336
+    },
+    {
+      "epoch": 1.4800332778702163,
+      "grad_norm": 0.20104999840259552,
+      "learning_rate": 4.051956585114783e-06,
+      "loss": 0.4833,
+      "step": 5337
+    },
+    {
+      "epoch": 1.4803105934553522,
+      "grad_norm": 0.2020789533853531,
+      "learning_rate": 4.047887710843756e-06,
+      "loss": 0.4986,
+      "step": 5338
+    },
+    {
+      "epoch": 1.480587909040488,
+      "grad_norm": 0.2103864550590515,
+      "learning_rate": 4.043820485803766e-06,
+      "loss": 0.4997,
+      "step": 5339
+    },
+    {
+      "epoch": 1.480865224625624,
+      "grad_norm": 0.21654918789863586,
+      "learning_rate": 4.039754910788442e-06,
+      "loss": 0.4647,
+      "step": 5340
+    },
+    {
+      "epoch": 1.4811425402107599,
+      "grad_norm": 0.2058803290128708,
+      "learning_rate": 4.0356909865910895e-06,
+      "loss": 0.5266,
+      "step": 5341
+    },
+    {
+      "epoch": 1.4814198557958957,
+      "grad_norm": 0.20043110847473145,
+      "learning_rate": 4.031628714004678e-06,
+      "loss": 0.514,
+      "step": 5342
+    },
+    {
+      "epoch": 1.4816971713810316,
+      "grad_norm": 0.2056841254234314,
+      "learning_rate": 4.0275680938218705e-06,
+      "loss": 0.5026,
+      "step": 5343
+    },
+    {
+      "epoch": 1.4819744869661675,
+      "grad_norm": 0.2606765627861023,
+      "learning_rate": 4.023509126835004e-06,
+      "loss": 0.4976,
+      "step": 5344
+    },
+    {
+      "epoch": 1.4822518025513034,
+      "grad_norm": 0.20417386293411255,
+      "learning_rate": 4.019451813836088e-06,
+      "loss": 0.5086,
+      "step": 5345
+    },
+    {
+      "epoch": 1.4825291181364393,
+      "grad_norm": 0.21264471113681793,
+      "learning_rate": 4.015396155616806e-06,
+      "loss": 0.4835,
+      "step": 5346
+    },
+    {
+      "epoch": 1.4828064337215752,
+      "grad_norm": 0.19947127997875214,
+      "learning_rate": 4.011342152968531e-06,
+      "loss": 0.4885,
+      "step": 5347
+    },
+    {
+      "epoch": 1.483083749306711,
+      "grad_norm": 0.20617523789405823,
+      "learning_rate": 4.007289806682307e-06,
+      "loss": 0.5123,
+      "step": 5348
+    },
+    {
+      "epoch": 1.483361064891847,
+      "grad_norm": 0.21049796044826508,
+      "learning_rate": 4.003239117548853e-06,
+      "loss": 0.5122,
+      "step": 5349
+    },
+    {
+      "epoch": 1.4836383804769828,
+      "grad_norm": 0.22191189229488373,
+      "learning_rate": 3.999190086358556e-06,
+      "loss": 0.5074,
+      "step": 5350
+    },
+    {
+      "epoch": 1.4839156960621187,
+      "grad_norm": 0.20709733664989471,
+      "learning_rate": 3.995142713901506e-06,
+      "loss": 0.507,
+      "step": 5351
+    },
+    {
+      "epoch": 1.4841930116472546,
+      "grad_norm": 0.21513330936431885,
+      "learning_rate": 3.9910970009674445e-06,
+      "loss": 0.5193,
+      "step": 5352
+    },
+    {
+      "epoch": 1.4844703272323905,
+      "grad_norm": 0.2153635323047638,
+      "learning_rate": 3.987052948345797e-06,
+      "loss": 0.4891,
+      "step": 5353
+    },
+    {
+      "epoch": 1.4847476428175264,
+      "grad_norm": 0.21485386788845062,
+      "learning_rate": 3.983010556825658e-06,
+      "loss": 0.5012,
+      "step": 5354
+    },
+    {
+      "epoch": 1.4850249584026622,
+      "grad_norm": 0.20851318538188934,
+      "learning_rate": 3.978969827195821e-06,
+      "loss": 0.4958,
+      "step": 5355
+    },
+    {
+      "epoch": 1.4853022739877981,
+      "grad_norm": 0.2146236151456833,
+      "learning_rate": 3.97493076024473e-06,
+      "loss": 0.4802,
+      "step": 5356
+    },
+    {
+      "epoch": 1.485579589572934,
+      "grad_norm": 0.22128230333328247,
+      "learning_rate": 3.970893356760512e-06,
+      "loss": 0.52,
+      "step": 5357
+    },
+    {
+      "epoch": 1.48585690515807,
+      "grad_norm": 0.20890596508979797,
+      "learning_rate": 3.966857617530974e-06,
+      "loss": 0.4722,
+      "step": 5358
+    },
+    {
+      "epoch": 1.4861342207432058,
+      "grad_norm": 0.20855508744716644,
+      "learning_rate": 3.9628235433436e-06,
+      "loss": 0.4897,
+      "step": 5359
+    },
+    {
+      "epoch": 1.4864115363283417,
+      "grad_norm": 0.21735462546348572,
+      "learning_rate": 3.958791134985541e-06,
+      "loss": 0.5027,
+      "step": 5360
+    },
+    {
+      "epoch": 1.4866888519134775,
+      "grad_norm": 0.20674702525138855,
+      "learning_rate": 3.954760393243623e-06,
+      "loss": 0.4892,
+      "step": 5361
+    },
+    {
+      "epoch": 1.4869661674986134,
+      "grad_norm": 0.20703266561031342,
+      "learning_rate": 3.950731318904355e-06,
+      "loss": 0.5145,
+      "step": 5362
+    },
+    {
+      "epoch": 1.4872434830837493,
+      "grad_norm": 0.20337079465389252,
+      "learning_rate": 3.946703912753917e-06,
+      "loss": 0.5017,
+      "step": 5363
+    },
+    {
+      "epoch": 1.4875207986688852,
+      "grad_norm": 0.2117006629705429,
+      "learning_rate": 3.942678175578159e-06,
+      "loss": 0.5023,
+      "step": 5364
+    },
+    {
+      "epoch": 1.487798114254021,
+      "grad_norm": 0.20295751094818115,
+      "learning_rate": 3.938654108162611e-06,
+      "loss": 0.4966,
+      "step": 5365
+    },
+    {
+      "epoch": 1.488075429839157,
+      "grad_norm": 0.20612047612667084,
+      "learning_rate": 3.934631711292477e-06,
+      "loss": 0.4974,
+      "step": 5366
+    },
+    {
+      "epoch": 1.4883527454242929,
+      "grad_norm": 0.1999398022890091,
+      "learning_rate": 3.930610985752633e-06,
+      "loss": 0.5003,
+      "step": 5367
+    },
+    {
+      "epoch": 1.4886300610094287,
+      "grad_norm": 0.20393617451190948,
+      "learning_rate": 3.926591932327622e-06,
+      "loss": 0.499,
+      "step": 5368
+    },
+    {
+      "epoch": 1.4889073765945646,
+      "grad_norm": 0.20448175072669983,
+      "learning_rate": 3.922574551801675e-06,
+      "loss": 0.4882,
+      "step": 5369
+    },
+    {
+      "epoch": 1.4891846921797005,
+      "grad_norm": 0.20072409510612488,
+      "learning_rate": 3.918558844958691e-06,
+      "loss": 0.4902,
+      "step": 5370
+    },
+    {
+      "epoch": 1.4894620077648364,
+      "grad_norm": 0.2017168551683426,
+      "learning_rate": 3.9145448125822325e-06,
+      "loss": 0.4926,
+      "step": 5371
+    },
+    {
+      "epoch": 1.4897393233499723,
+      "grad_norm": 0.20040208101272583,
+      "learning_rate": 3.910532455455553e-06,
+      "loss": 0.5075,
+      "step": 5372
+    },
+    {
+      "epoch": 1.4900166389351082,
+      "grad_norm": 0.19632667303085327,
+      "learning_rate": 3.9065217743615605e-06,
+      "loss": 0.4889,
+      "step": 5373
+    },
+    {
+      "epoch": 1.490293954520244,
+      "grad_norm": 0.21401472389698029,
+      "learning_rate": 3.902512770082854e-06,
+      "loss": 0.4837,
+      "step": 5374
+    },
+    {
+      "epoch": 1.49057127010538,
+      "grad_norm": 0.21348023414611816,
+      "learning_rate": 3.898505443401689e-06,
+      "loss": 0.5165,
+      "step": 5375
+    },
+    {
+      "epoch": 1.4908485856905158,
+      "grad_norm": 0.20388244092464447,
+      "learning_rate": 3.8944997951000066e-06,
+      "loss": 0.5116,
+      "step": 5376
+    },
+    {
+      "epoch": 1.4911259012756517,
+      "grad_norm": 0.20777450501918793,
+      "learning_rate": 3.89049582595941e-06,
+      "loss": 0.5097,
+      "step": 5377
+    },
+    {
+      "epoch": 1.4914032168607876,
+      "grad_norm": 0.20261326432228088,
+      "learning_rate": 3.886493536761182e-06,
+      "loss": 0.5027,
+      "step": 5378
+    },
+    {
+      "epoch": 1.4916805324459235,
+      "grad_norm": 0.20818550884723663,
+      "learning_rate": 3.882492928286279e-06,
+      "loss": 0.4847,
+      "step": 5379
+    },
+    {
+      "epoch": 1.4919578480310594,
+      "grad_norm": 0.1958109438419342,
+      "learning_rate": 3.878494001315319e-06,
+      "loss": 0.4782,
+      "step": 5380
+    },
+    {
+      "epoch": 1.4922351636161952,
+      "grad_norm": 0.20695748925209045,
+      "learning_rate": 3.874496756628606e-06,
+      "loss": 0.4889,
+      "step": 5381
+    },
+    {
+      "epoch": 1.4925124792013311,
+      "grad_norm": 0.20474323630332947,
+      "learning_rate": 3.8705011950060985e-06,
+      "loss": 0.4796,
+      "step": 5382
+    },
+    {
+      "epoch": 1.492789794786467,
+      "grad_norm": 0.20276211202144623,
+      "learning_rate": 3.8665073172274465e-06,
+      "loss": 0.4914,
+      "step": 5383
+    },
+    {
+      "epoch": 1.493067110371603,
+      "grad_norm": 0.20544062554836273,
+      "learning_rate": 3.862515124071954e-06,
+      "loss": 0.5119,
+      "step": 5384
+    },
+    {
+      "epoch": 1.4933444259567388,
+      "grad_norm": 0.19975237548351288,
+      "learning_rate": 3.858524616318607e-06,
+      "loss": 0.4804,
+      "step": 5385
+    },
+    {
+      "epoch": 1.4936217415418747,
+      "grad_norm": 0.20508365333080292,
+      "learning_rate": 3.8545357947460625e-06,
+      "loss": 0.4953,
+      "step": 5386
+    },
+    {
+      "epoch": 1.4938990571270105,
+      "grad_norm": 0.20914685726165771,
+      "learning_rate": 3.850548660132641e-06,
+      "loss": 0.4684,
+      "step": 5387
+    },
+    {
+      "epoch": 1.4941763727121464,
+      "grad_norm": 0.23906511068344116,
+      "learning_rate": 3.846563213256335e-06,
+      "loss": 0.5059,
+      "step": 5388
+    },
+    {
+      "epoch": 1.4944536882972823,
+      "grad_norm": 0.240016907453537,
+      "learning_rate": 3.842579454894815e-06,
+      "loss": 0.5089,
+      "step": 5389
+    },
+    {
+      "epoch": 1.4947310038824182,
+      "grad_norm": 0.2028864324092865,
+      "learning_rate": 3.838597385825421e-06,
+      "loss": 0.4835,
+      "step": 5390
+    },
+    {
+      "epoch": 1.495008319467554,
+      "grad_norm": 0.21252113580703735,
+      "learning_rate": 3.834617006825157e-06,
+      "loss": 0.4773,
+      "step": 5391
+    },
+    {
+      "epoch": 1.49528563505269,
+      "grad_norm": 0.20465324819087982,
+      "learning_rate": 3.830638318670691e-06,
+      "loss": 0.4668,
+      "step": 5392
+    },
+    {
+      "epoch": 1.4955629506378258,
+      "grad_norm": 0.20127691328525543,
+      "learning_rate": 3.826661322138389e-06,
+      "loss": 0.5119,
+      "step": 5393
+    },
+    {
+      "epoch": 1.4958402662229617,
+      "grad_norm": 0.2034253180027008,
+      "learning_rate": 3.822686018004258e-06,
+      "loss": 0.5025,
+      "step": 5394
+    },
+    {
+      "epoch": 1.4961175818080976,
+      "grad_norm": 0.2095792442560196,
+      "learning_rate": 3.8187124070439815e-06,
+      "loss": 0.4719,
+      "step": 5395
+    },
+    {
+      "epoch": 1.4963948973932335,
+      "grad_norm": 0.21357692778110504,
+      "learning_rate": 3.814740490032921e-06,
+      "loss": 0.5078,
+      "step": 5396
+    },
+    {
+      "epoch": 1.4966722129783694,
+      "grad_norm": 0.2135164737701416,
+      "learning_rate": 3.8107702677461067e-06,
+      "loss": 0.4999,
+      "step": 5397
+    },
+    {
+      "epoch": 1.4969495285635053,
+      "grad_norm": 0.21254348754882812,
+      "learning_rate": 3.8068017409582294e-06,
+      "loss": 0.4971,
+      "step": 5398
+    },
+    {
+      "epoch": 1.4972268441486412,
+      "grad_norm": 0.2134813517332077,
+      "learning_rate": 3.802834910443652e-06,
+      "loss": 0.5112,
+      "step": 5399
+    },
+    {
+      "epoch": 1.497504159733777,
+      "grad_norm": 0.20528295636177063,
+      "learning_rate": 3.798869776976409e-06,
+      "loss": 0.4773,
+      "step": 5400
+    },
+    {
+      "epoch": 1.497781475318913,
+      "grad_norm": 0.20741385221481323,
+      "learning_rate": 3.7949063413302093e-06,
+      "loss": 0.5007,
+      "step": 5401
+    },
+    {
+      "epoch": 1.4980587909040488,
+      "grad_norm": 0.2020622044801712,
+      "learning_rate": 3.7909446042784193e-06,
+      "loss": 0.4935,
+      "step": 5402
+    },
+    {
+      "epoch": 1.4983361064891847,
+      "grad_norm": 0.20889051258563995,
+      "learning_rate": 3.786984566594071e-06,
+      "loss": 0.471,
+      "step": 5403
+    },
+    {
+      "epoch": 1.4986134220743206,
+      "grad_norm": 0.19613035023212433,
+      "learning_rate": 3.7830262290498896e-06,
+      "loss": 0.5077,
+      "step": 5404
+    },
+    {
+      "epoch": 1.4988907376594565,
+      "grad_norm": 0.2048519104719162,
+      "learning_rate": 3.7790695924182413e-06,
+      "loss": 0.4703,
+      "step": 5405
+    },
+    {
+      "epoch": 1.4991680532445923,
+      "grad_norm": 0.22040173411369324,
+      "learning_rate": 3.775114657471168e-06,
+      "loss": 0.4972,
+      "step": 5406
+    },
+    {
+      "epoch": 1.4994453688297282,
+      "grad_norm": 0.2038019597530365,
+      "learning_rate": 3.771161424980388e-06,
+      "loss": 0.5195,
+      "step": 5407
+    },
+    {
+      "epoch": 1.4997226844148641,
+      "grad_norm": 0.20490224659442902,
+      "learning_rate": 3.7672098957172846e-06,
+      "loss": 0.4884,
+      "step": 5408
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.20087428390979767,
+      "learning_rate": 3.763260070452895e-06,
+      "loss": 0.4781,
+      "step": 5409
+    },
+    {
+      "epoch": 1.5002773155851359,
+      "grad_norm": 0.21157024800777435,
+      "learning_rate": 3.759311949957947e-06,
+      "loss": 0.5003,
+      "step": 5410
+    },
+    {
+      "epoch": 1.5005546311702718,
+      "grad_norm": 0.23590406775474548,
+      "learning_rate": 3.755365535002814e-06,
+      "loss": 0.4798,
+      "step": 5411
+    },
+    {
+      "epoch": 1.5008319467554077,
+      "grad_norm": 0.21228978037834167,
+      "learning_rate": 3.751420826357553e-06,
+      "loss": 0.5178,
+      "step": 5412
+    },
+    {
+      "epoch": 1.5011092623405435,
+      "grad_norm": 0.21098542213439941,
+      "learning_rate": 3.7474778247918743e-06,
+      "loss": 0.4764,
+      "step": 5413
+    },
+    {
+      "epoch": 1.5013865779256794,
+      "grad_norm": 0.20355211198329926,
+      "learning_rate": 3.743536531075169e-06,
+      "loss": 0.4938,
+      "step": 5414
+    },
+    {
+      "epoch": 1.5016638935108153,
+      "grad_norm": 0.20924827456474304,
+      "learning_rate": 3.739596945976481e-06,
+      "loss": 0.5048,
+      "step": 5415
+    },
+    {
+      "epoch": 1.5019412090959512,
+      "grad_norm": 0.20819483697414398,
+      "learning_rate": 3.7356590702645327e-06,
+      "loss": 0.4889,
+      "step": 5416
+    },
+    {
+      "epoch": 1.502218524681087,
+      "grad_norm": 0.20322836935520172,
+      "learning_rate": 3.7317229047077086e-06,
+      "loss": 0.4886,
+      "step": 5417
+    },
+    {
+      "epoch": 1.502495840266223,
+      "grad_norm": 0.20074166357517242,
+      "learning_rate": 3.7277884500740543e-06,
+      "loss": 0.5019,
+      "step": 5418
+    },
+    {
+      "epoch": 1.5027731558513588,
+      "grad_norm": 0.21029093861579895,
+      "learning_rate": 3.723855707131292e-06,
+      "loss": 0.4973,
+      "step": 5419
+    },
+    {
+      "epoch": 1.5030504714364947,
+      "grad_norm": 0.20609499514102936,
+      "learning_rate": 3.7199246766467964e-06,
+      "loss": 0.5115,
+      "step": 5420
+    },
+    {
+      "epoch": 1.5033277870216306,
+      "grad_norm": 0.21178114414215088,
+      "learning_rate": 3.715995359387625e-06,
+      "loss": 0.5378,
+      "step": 5421
+    },
+    {
+      "epoch": 1.5036051026067665,
+      "grad_norm": 0.2018701434135437,
+      "learning_rate": 3.7120677561204823e-06,
+      "loss": 0.4945,
+      "step": 5422
+    },
+    {
+      "epoch": 1.5038824181919024,
+      "grad_norm": 0.2033861130475998,
+      "learning_rate": 3.708141867611753e-06,
+      "loss": 0.4795,
+      "step": 5423
+    },
+    {
+      "epoch": 1.5041597337770383,
+      "grad_norm": 0.20554760098457336,
+      "learning_rate": 3.7042176946274846e-06,
+      "loss": 0.4864,
+      "step": 5424
+    },
+    {
+      "epoch": 1.5044370493621742,
+      "grad_norm": 0.20681151747703552,
+      "learning_rate": 3.7002952379333837e-06,
+      "loss": 0.4868,
+      "step": 5425
+    },
+    {
+      "epoch": 1.50471436494731,
+      "grad_norm": 0.1939842700958252,
+      "learning_rate": 3.6963744982948216e-06,
+      "loss": 0.4813,
+      "step": 5426
+    },
+    {
+      "epoch": 1.504991680532446,
+      "grad_norm": 0.20737822353839874,
+      "learning_rate": 3.6924554764768428e-06,
+      "loss": 0.5158,
+      "step": 5427
+    },
+    {
+      "epoch": 1.5052689961175818,
+      "grad_norm": 0.20026959478855133,
+      "learning_rate": 3.688538173244156e-06,
+      "loss": 0.493,
+      "step": 5428
+    },
+    {
+      "epoch": 1.5055463117027177,
+      "grad_norm": 0.20325352251529694,
+      "learning_rate": 3.6846225893611265e-06,
+      "loss": 0.5011,
+      "step": 5429
+    },
+    {
+      "epoch": 1.5058236272878536,
+      "grad_norm": 0.20845289528369904,
+      "learning_rate": 3.680708725591782e-06,
+      "loss": 0.4736,
+      "step": 5430
+    },
+    {
+      "epoch": 1.5061009428729895,
+      "grad_norm": 0.20399589836597443,
+      "learning_rate": 3.6767965826998345e-06,
+      "loss": 0.5033,
+      "step": 5431
+    },
+    {
+      "epoch": 1.5063782584581253,
+      "grad_norm": 0.19747604429721832,
+      "learning_rate": 3.672886161448641e-06,
+      "loss": 0.4793,
+      "step": 5432
+    },
+    {
+      "epoch": 1.5066555740432612,
+      "grad_norm": 0.2051791548728943,
+      "learning_rate": 3.6689774626012224e-06,
+      "loss": 0.5145,
+      "step": 5433
+    },
+    {
+      "epoch": 1.506932889628397,
+      "grad_norm": 0.1973484754562378,
+      "learning_rate": 3.665070486920276e-06,
+      "loss": 0.4764,
+      "step": 5434
+    },
+    {
+      "epoch": 1.507210205213533,
+      "grad_norm": 0.195805624127388,
+      "learning_rate": 3.6611652351681568e-06,
+      "loss": 0.4818,
+      "step": 5435
+    },
+    {
+      "epoch": 1.5074875207986689,
+      "grad_norm": 0.20444104075431824,
+      "learning_rate": 3.657261708106882e-06,
+      "loss": 0.4878,
+      "step": 5436
+    },
+    {
+      "epoch": 1.5077648363838048,
+      "grad_norm": 0.20944221317768097,
+      "learning_rate": 3.653359906498127e-06,
+      "loss": 0.5183,
+      "step": 5437
+    },
+    {
+      "epoch": 1.5080421519689406,
+      "grad_norm": 0.2081797569990158,
+      "learning_rate": 3.6494598311032415e-06,
+      "loss": 0.4885,
+      "step": 5438
+    },
+    {
+      "epoch": 1.5083194675540765,
+      "grad_norm": 0.20127315819263458,
+      "learning_rate": 3.645561482683238e-06,
+      "loss": 0.4935,
+      "step": 5439
+    },
+    {
+      "epoch": 1.5085967831392124,
+      "grad_norm": 0.20761513710021973,
+      "learning_rate": 3.6416648619987837e-06,
+      "loss": 0.4706,
+      "step": 5440
+    },
+    {
+      "epoch": 1.5088740987243483,
+      "grad_norm": 0.2054562270641327,
+      "learning_rate": 3.6377699698102035e-06,
+      "loss": 0.5033,
+      "step": 5441
+    },
+    {
+      "epoch": 1.5091514143094842,
+      "grad_norm": 0.21789734065532684,
+      "learning_rate": 3.6338768068775104e-06,
+      "loss": 0.4885,
+      "step": 5442
+    },
+    {
+      "epoch": 1.50942872989462,
+      "grad_norm": 0.21268032491207123,
+      "learning_rate": 3.6299853739603555e-06,
+      "loss": 0.4779,
+      "step": 5443
+    },
+    {
+      "epoch": 1.509706045479756,
+      "grad_norm": 0.19880391657352448,
+      "learning_rate": 3.6260956718180568e-06,
+      "loss": 0.4655,
+      "step": 5444
+    },
+    {
+      "epoch": 1.5099833610648918,
+      "grad_norm": 0.21325719356536865,
+      "learning_rate": 3.6222077012096026e-06,
+      "loss": 0.5004,
+      "step": 5445
+    },
+    {
+      "epoch": 1.5102606766500277,
+      "grad_norm": 0.20476926863193512,
+      "learning_rate": 3.6183214628936425e-06,
+      "loss": 0.4965,
+      "step": 5446
+    },
+    {
+      "epoch": 1.5105379922351636,
+      "grad_norm": 0.20836953818798065,
+      "learning_rate": 3.614436957628481e-06,
+      "loss": 0.4832,
+      "step": 5447
+    },
+    {
+      "epoch": 1.5108153078202995,
+      "grad_norm": 0.2125098705291748,
+      "learning_rate": 3.6105541861720826e-06,
+      "loss": 0.5014,
+      "step": 5448
+    },
+    {
+      "epoch": 1.5110926234054354,
+      "grad_norm": 0.2126840204000473,
+      "learning_rate": 3.6066731492820844e-06,
+      "loss": 0.5311,
+      "step": 5449
+    },
+    {
+      "epoch": 1.5113699389905713,
+      "grad_norm": 0.2019418478012085,
+      "learning_rate": 3.6027938477157838e-06,
+      "loss": 0.5007,
+      "step": 5450
+    },
+    {
+      "epoch": 1.5116472545757071,
+      "grad_norm": 0.25397753715515137,
+      "learning_rate": 3.598916282230126e-06,
+      "loss": 0.5272,
+      "step": 5451
+    },
+    {
+      "epoch": 1.511924570160843,
+      "grad_norm": 0.20074354112148285,
+      "learning_rate": 3.595040453581734e-06,
+      "loss": 0.4877,
+      "step": 5452
+    },
+    {
+      "epoch": 1.512201885745979,
+      "grad_norm": 0.2004215121269226,
+      "learning_rate": 3.5911663625268792e-06,
+      "loss": 0.4733,
+      "step": 5453
+    },
+    {
+      "epoch": 1.5124792013311148,
+      "grad_norm": 0.21349315345287323,
+      "learning_rate": 3.587294009821507e-06,
+      "loss": 0.5166,
+      "step": 5454
+    },
+    {
+      "epoch": 1.5127565169162507,
+      "grad_norm": 0.20270389318466187,
+      "learning_rate": 3.5834233962212056e-06,
+      "loss": 0.5115,
+      "step": 5455
+    },
+    {
+      "epoch": 1.5130338325013866,
+      "grad_norm": 0.20797984302043915,
+      "learning_rate": 3.5795545224812405e-06,
+      "loss": 0.4746,
+      "step": 5456
+    },
+    {
+      "epoch": 1.5133111480865225,
+      "grad_norm": 0.2141650915145874,
+      "learning_rate": 3.575687389356534e-06,
+      "loss": 0.5169,
+      "step": 5457
+    },
+    {
+      "epoch": 1.5135884636716583,
+      "grad_norm": 0.20348010957241058,
+      "learning_rate": 3.5718219976016614e-06,
+      "loss": 0.5052,
+      "step": 5458
+    },
+    {
+      "epoch": 1.5138657792567942,
+      "grad_norm": 0.2087784856557846,
+      "learning_rate": 3.5679583479708664e-06,
+      "loss": 0.5135,
+      "step": 5459
+    },
+    {
+      "epoch": 1.51414309484193,
+      "grad_norm": 0.2124669849872589,
+      "learning_rate": 3.564096441218044e-06,
+      "loss": 0.513,
+      "step": 5460
+    },
+    {
+      "epoch": 1.514420410427066,
+      "grad_norm": 0.20236825942993164,
+      "learning_rate": 3.5602362780967624e-06,
+      "loss": 0.4963,
+      "step": 5461
+    },
+    {
+      "epoch": 1.5146977260122019,
+      "grad_norm": 0.1994817554950714,
+      "learning_rate": 3.556377859360234e-06,
+      "loss": 0.4795,
+      "step": 5462
+    },
+    {
+      "epoch": 1.5149750415973378,
+      "grad_norm": 0.20891886949539185,
+      "learning_rate": 3.5525211857613454e-06,
+      "loss": 0.5117,
+      "step": 5463
+    },
+    {
+      "epoch": 1.5152523571824736,
+      "grad_norm": 0.19349178671836853,
+      "learning_rate": 3.5486662580526285e-06,
+      "loss": 0.4596,
+      "step": 5464
+    },
+    {
+      "epoch": 1.5155296727676095,
+      "grad_norm": 0.21563927829265594,
+      "learning_rate": 3.5448130769862864e-06,
+      "loss": 0.4859,
+      "step": 5465
+    },
+    {
+      "epoch": 1.5158069883527454,
+      "grad_norm": 0.2011108696460724,
+      "learning_rate": 3.5409616433141793e-06,
+      "loss": 0.4963,
+      "step": 5466
+    },
+    {
+      "epoch": 1.5160843039378813,
+      "grad_norm": 0.19179309904575348,
+      "learning_rate": 3.537111957787821e-06,
+      "loss": 0.4804,
+      "step": 5467
+    },
+    {
+      "epoch": 1.5163616195230172,
+      "grad_norm": 0.194498673081398,
+      "learning_rate": 3.5332640211583807e-06,
+      "loss": 0.4814,
+      "step": 5468
+    },
+    {
+      "epoch": 1.516638935108153,
+      "grad_norm": 0.2188744843006134,
+      "learning_rate": 3.5294178341767043e-06,
+      "loss": 0.5059,
+      "step": 5469
+    },
+    {
+      "epoch": 1.516916250693289,
+      "grad_norm": 0.2129710465669632,
+      "learning_rate": 3.525573397593282e-06,
+      "loss": 0.4924,
+      "step": 5470
+    },
+    {
+      "epoch": 1.5171935662784248,
+      "grad_norm": 0.21179741621017456,
+      "learning_rate": 3.521730712158257e-06,
+      "loss": 0.4987,
+      "step": 5471
+    },
+    {
+      "epoch": 1.5174708818635607,
+      "grad_norm": 0.20567160844802856,
+      "learning_rate": 3.517889778621446e-06,
+      "loss": 0.5164,
+      "step": 5472
+    },
+    {
+      "epoch": 1.5177481974486966,
+      "grad_norm": 0.2281576544046402,
+      "learning_rate": 3.5140505977323186e-06,
+      "loss": 0.47,
+      "step": 5473
+    },
+    {
+      "epoch": 1.5180255130338325,
+      "grad_norm": 0.20740900933742523,
+      "learning_rate": 3.5102131702399983e-06,
+      "loss": 0.4853,
+      "step": 5474
+    },
+    {
+      "epoch": 1.5183028286189684,
+      "grad_norm": 0.20481230318546295,
+      "learning_rate": 3.506377496893265e-06,
+      "loss": 0.5248,
+      "step": 5475
+    },
+    {
+      "epoch": 1.5185801442041043,
+      "grad_norm": 0.20815472304821014,
+      "learning_rate": 3.502543578440562e-06,
+      "loss": 0.4855,
+      "step": 5476
+    },
+    {
+      "epoch": 1.5188574597892401,
+      "grad_norm": 0.20264998078346252,
+      "learning_rate": 3.498711415629996e-06,
+      "loss": 0.4765,
+      "step": 5477
+    },
+    {
+      "epoch": 1.519134775374376,
+      "grad_norm": 0.2170276939868927,
+      "learning_rate": 3.494881009209315e-06,
+      "loss": 0.5036,
+      "step": 5478
+    },
+    {
+      "epoch": 1.519412090959512,
+      "grad_norm": 0.2021361142396927,
+      "learning_rate": 3.4910523599259294e-06,
+      "loss": 0.4986,
+      "step": 5479
+    },
+    {
+      "epoch": 1.5196894065446478,
+      "grad_norm": 0.195895254611969,
+      "learning_rate": 3.487225468526924e-06,
+      "loss": 0.4801,
+      "step": 5480
+    },
+    {
+      "epoch": 1.5199667221297837,
+      "grad_norm": 0.21378639340400696,
+      "learning_rate": 3.483400335759017e-06,
+      "loss": 0.5152,
+      "step": 5481
+    },
+    {
+      "epoch": 1.5202440377149196,
+      "grad_norm": 0.213482066988945,
+      "learning_rate": 3.4795769623685912e-06,
+      "loss": 0.5034,
+      "step": 5482
+    },
+    {
+      "epoch": 1.5205213533000554,
+      "grad_norm": 0.2165437638759613,
+      "learning_rate": 3.475755349101692e-06,
+      "loss": 0.4841,
+      "step": 5483
+    },
+    {
+      "epoch": 1.5207986688851913,
+      "grad_norm": 0.20139548182487488,
+      "learning_rate": 3.471935496704022e-06,
+      "loss": 0.4711,
+      "step": 5484
+    },
+    {
+      "epoch": 1.5210759844703272,
+      "grad_norm": 0.2036607712507248,
+      "learning_rate": 3.4681174059209288e-06,
+      "loss": 0.4757,
+      "step": 5485
+    },
+    {
+      "epoch": 1.521353300055463,
+      "grad_norm": 0.19882218539714813,
+      "learning_rate": 3.4643010774974226e-06,
+      "loss": 0.5082,
+      "step": 5486
+    },
+    {
+      "epoch": 1.521630615640599,
+      "grad_norm": 0.21727705001831055,
+      "learning_rate": 3.460486512178171e-06,
+      "loss": 0.474,
+      "step": 5487
+    },
+    {
+      "epoch": 1.5219079312257349,
+      "grad_norm": 0.21604587137699127,
+      "learning_rate": 3.456673710707503e-06,
+      "loss": 0.4752,
+      "step": 5488
+    },
+    {
+      "epoch": 1.5221852468108708,
+      "grad_norm": 0.20462778210639954,
+      "learning_rate": 3.4528626738293893e-06,
+      "loss": 0.5011,
+      "step": 5489
+    },
+    {
+      "epoch": 1.5224625623960066,
+      "grad_norm": 0.20358699560165405,
+      "learning_rate": 3.4490534022874706e-06,
+      "loss": 0.4666,
+      "step": 5490
+    },
+    {
+      "epoch": 1.5227398779811425,
+      "grad_norm": 0.2057787925004959,
+      "learning_rate": 3.4452458968250293e-06,
+      "loss": 0.5309,
+      "step": 5491
+    },
+    {
+      "epoch": 1.5230171935662784,
+      "grad_norm": 0.1948128193616867,
+      "learning_rate": 3.441440158185018e-06,
+      "loss": 0.486,
+      "step": 5492
+    },
+    {
+      "epoch": 1.5232945091514143,
+      "grad_norm": 0.22251975536346436,
+      "learning_rate": 3.4376361871100313e-06,
+      "loss": 0.505,
+      "step": 5493
+    },
+    {
+      "epoch": 1.5235718247365502,
+      "grad_norm": 0.20903460681438446,
+      "learning_rate": 3.433833984342326e-06,
+      "loss": 0.5135,
+      "step": 5494
+    },
+    {
+      "epoch": 1.523849140321686,
+      "grad_norm": 0.20052765309810638,
+      "learning_rate": 3.430033550623818e-06,
+      "loss": 0.5181,
+      "step": 5495
+    },
+    {
+      "epoch": 1.524126455906822,
+      "grad_norm": 0.20799602568149567,
+      "learning_rate": 3.4262348866960655e-06,
+      "loss": 0.5056,
+      "step": 5496
+    },
+    {
+      "epoch": 1.5244037714919578,
+      "grad_norm": 0.21278195083141327,
+      "learning_rate": 3.4224379933002953e-06,
+      "loss": 0.4942,
+      "step": 5497
+    },
+    {
+      "epoch": 1.5246810870770937,
+      "grad_norm": 0.30083736777305603,
+      "learning_rate": 3.418642871177373e-06,
+      "loss": 0.4991,
+      "step": 5498
+    },
+    {
+      "epoch": 1.5249584026622296,
+      "grad_norm": 0.20429125428199768,
+      "learning_rate": 3.4148495210678366e-06,
+      "loss": 0.4891,
+      "step": 5499
+    },
+    {
+      "epoch": 1.5252357182473655,
+      "grad_norm": 0.2550550401210785,
+      "learning_rate": 3.4110579437118624e-06,
+      "loss": 0.4804,
+      "step": 5500
+    },
+    {
+      "epoch": 1.5255130338325014,
+      "grad_norm": 0.20311662554740906,
+      "learning_rate": 3.4072681398492942e-06,
+      "loss": 0.4673,
+      "step": 5501
+    },
+    {
+      "epoch": 1.5257903494176372,
+      "grad_norm": 0.1889600306749344,
+      "learning_rate": 3.4034801102196146e-06,
+      "loss": 0.4716,
+      "step": 5502
+    },
+    {
+      "epoch": 1.5260676650027731,
+      "grad_norm": 0.21437427401542664,
+      "learning_rate": 3.399693855561974e-06,
+      "loss": 0.5053,
+      "step": 5503
+    },
+    {
+      "epoch": 1.526344980587909,
+      "grad_norm": 0.21157221496105194,
+      "learning_rate": 3.395909376615172e-06,
+      "loss": 0.4717,
+      "step": 5504
+    },
+    {
+      "epoch": 1.526622296173045,
+      "grad_norm": 0.2132854014635086,
+      "learning_rate": 3.3921266741176614e-06,
+      "loss": 0.4751,
+      "step": 5505
+    },
+    {
+      "epoch": 1.5268996117581808,
+      "grad_norm": 0.21224236488342285,
+      "learning_rate": 3.3883457488075406e-06,
+      "loss": 0.4927,
+      "step": 5506
+    },
+    {
+      "epoch": 1.5271769273433167,
+      "grad_norm": 0.20786736905574799,
+      "learning_rate": 3.384566601422573e-06,
+      "loss": 0.4666,
+      "step": 5507
+    },
+    {
+      "epoch": 1.5274542429284526,
+      "grad_norm": 0.21545729041099548,
+      "learning_rate": 3.3807892327001737e-06,
+      "loss": 0.4783,
+      "step": 5508
+    },
+    {
+      "epoch": 1.5277315585135884,
+      "grad_norm": 0.20959782600402832,
+      "learning_rate": 3.377013643377401e-06,
+      "loss": 0.5053,
+      "step": 5509
+    },
+    {
+      "epoch": 1.5280088740987243,
+      "grad_norm": 0.20486906170845032,
+      "learning_rate": 3.373239834190975e-06,
+      "loss": 0.4873,
+      "step": 5510
+    },
+    {
+      "epoch": 1.5282861896838602,
+      "grad_norm": 0.2007468342781067,
+      "learning_rate": 3.3694678058772704e-06,
+      "loss": 0.4899,
+      "step": 5511
+    },
+    {
+      "epoch": 1.528563505268996,
+      "grad_norm": 0.19690293073654175,
+      "learning_rate": 3.365697559172304e-06,
+      "loss": 0.4941,
+      "step": 5512
+    },
+    {
+      "epoch": 1.528840820854132,
+      "grad_norm": 0.2142125815153122,
+      "learning_rate": 3.3619290948117513e-06,
+      "loss": 0.4746,
+      "step": 5513
+    },
+    {
+      "epoch": 1.5291181364392679,
+      "grad_norm": 0.20775122940540314,
+      "learning_rate": 3.3581624135309395e-06,
+      "loss": 0.5111,
+      "step": 5514
+    },
+    {
+      "epoch": 1.5293954520244037,
+      "grad_norm": 0.21404637396335602,
+      "learning_rate": 3.3543975160648526e-06,
+      "loss": 0.4836,
+      "step": 5515
+    },
+    {
+      "epoch": 1.5296727676095396,
+      "grad_norm": 0.20010797679424286,
+      "learning_rate": 3.3506344031481187e-06,
+      "loss": 0.5035,
+      "step": 5516
+    },
+    {
+      "epoch": 1.5299500831946755,
+      "grad_norm": 0.20686432719230652,
+      "learning_rate": 3.346873075515014e-06,
+      "loss": 0.4676,
+      "step": 5517
+    },
+    {
+      "epoch": 1.5302273987798114,
+      "grad_norm": 0.1945461481809616,
+      "learning_rate": 3.3431135338994864e-06,
+      "loss": 0.491,
+      "step": 5518
+    },
+    {
+      "epoch": 1.5305047143649473,
+      "grad_norm": 0.20243096351623535,
+      "learning_rate": 3.3393557790351167e-06,
+      "loss": 0.5112,
+      "step": 5519
+    },
+    {
+      "epoch": 1.5307820299500832,
+      "grad_norm": 0.19688890874385834,
+      "learning_rate": 3.3355998116551395e-06,
+      "loss": 0.4847,
+      "step": 5520
+    },
+    {
+      "epoch": 1.531059345535219,
+      "grad_norm": 0.20034383237361908,
+      "learning_rate": 3.331845632492439e-06,
+      "loss": 0.4914,
+      "step": 5521
+    },
+    {
+      "epoch": 1.531336661120355,
+      "grad_norm": 0.21226158738136292,
+      "learning_rate": 3.328093242279569e-06,
+      "loss": 0.5064,
+      "step": 5522
+    },
+    {
+      "epoch": 1.5316139767054908,
+      "grad_norm": 0.20493634045124054,
+      "learning_rate": 3.3243426417487107e-06,
+      "loss": 0.5008,
+      "step": 5523
+    },
+    {
+      "epoch": 1.5318912922906267,
+      "grad_norm": 0.2078971415758133,
+      "learning_rate": 3.3205938316317047e-06,
+      "loss": 0.4862,
+      "step": 5524
+    },
+    {
+      "epoch": 1.5321686078757626,
+      "grad_norm": 0.19189336895942688,
+      "learning_rate": 3.316846812660046e-06,
+      "loss": 0.4914,
+      "step": 5525
+    },
+    {
+      "epoch": 1.5324459234608985,
+      "grad_norm": 0.18428561091423035,
+      "learning_rate": 3.313101585564882e-06,
+      "loss": 0.4801,
+      "step": 5526
+    },
+    {
+      "epoch": 1.5327232390460344,
+      "grad_norm": 0.19429829716682434,
+      "learning_rate": 3.3093581510769995e-06,
+      "loss": 0.4672,
+      "step": 5527
+    },
+    {
+      "epoch": 1.5330005546311702,
+      "grad_norm": 0.20795489847660065,
+      "learning_rate": 3.3056165099268398e-06,
+      "loss": 0.5051,
+      "step": 5528
+    },
+    {
+      "epoch": 1.5332778702163061,
+      "grad_norm": 0.20628464221954346,
+      "learning_rate": 3.3018766628445e-06,
+      "loss": 0.5073,
+      "step": 5529
+    },
+    {
+      "epoch": 1.533555185801442,
+      "grad_norm": 0.1983129382133484,
+      "learning_rate": 3.2981386105597256e-06,
+      "loss": 0.4834,
+      "step": 5530
+    },
+    {
+      "epoch": 1.533832501386578,
+      "grad_norm": 0.20056569576263428,
+      "learning_rate": 3.294402353801905e-06,
+      "loss": 0.4883,
+      "step": 5531
+    },
+    {
+      "epoch": 1.5341098169717138,
+      "grad_norm": 0.20827257633209229,
+      "learning_rate": 3.2906678933000813e-06,
+      "loss": 0.4962,
+      "step": 5532
+    },
+    {
+      "epoch": 1.5343871325568497,
+      "grad_norm": 0.20605547726154327,
+      "learning_rate": 3.2869352297829532e-06,
+      "loss": 0.4792,
+      "step": 5533
+    },
+    {
+      "epoch": 1.5346644481419855,
+      "grad_norm": 0.20513512194156647,
+      "learning_rate": 3.283204363978852e-06,
+      "loss": 0.507,
+      "step": 5534
+    },
+    {
+      "epoch": 1.5349417637271214,
+      "grad_norm": 0.2062227725982666,
+      "learning_rate": 3.2794752966157793e-06,
+      "loss": 0.4994,
+      "step": 5535
+    },
+    {
+      "epoch": 1.5352190793122573,
+      "grad_norm": 0.2113962173461914,
+      "learning_rate": 3.2757480284213646e-06,
+      "loss": 0.4754,
+      "step": 5536
+    },
+    {
+      "epoch": 1.5354963948973932,
+      "grad_norm": 0.21231204271316528,
+      "learning_rate": 3.272022560122906e-06,
+      "loss": 0.488,
+      "step": 5537
+    },
+    {
+      "epoch": 1.535773710482529,
+      "grad_norm": 0.19757109880447388,
+      "learning_rate": 3.2682988924473316e-06,
+      "loss": 0.4951,
+      "step": 5538
+    },
+    {
+      "epoch": 1.536051026067665,
+      "grad_norm": 0.20043742656707764,
+      "learning_rate": 3.264577026121238e-06,
+      "loss": 0.5039,
+      "step": 5539
+    },
+    {
+      "epoch": 1.5363283416528009,
+      "grad_norm": 0.1949152946472168,
+      "learning_rate": 3.2608569618708494e-06,
+      "loss": 0.5021,
+      "step": 5540
+    },
+    {
+      "epoch": 1.5366056572379367,
+      "grad_norm": 0.20842109620571136,
+      "learning_rate": 3.257138700422055e-06,
+      "loss": 0.5008,
+      "step": 5541
+    },
+    {
+      "epoch": 1.5368829728230726,
+      "grad_norm": 0.21400795876979828,
+      "learning_rate": 3.253422242500388e-06,
+      "loss": 0.5048,
+      "step": 5542
+    },
+    {
+      "epoch": 1.5371602884082085,
+      "grad_norm": 0.21231138706207275,
+      "learning_rate": 3.249707588831025e-06,
+      "loss": 0.4919,
+      "step": 5543
+    },
+    {
+      "epoch": 1.5374376039933444,
+      "grad_norm": 0.2112995684146881,
+      "learning_rate": 3.24599474013879e-06,
+      "loss": 0.5135,
+      "step": 5544
+    },
+    {
+      "epoch": 1.5377149195784803,
+      "grad_norm": 0.20431232452392578,
+      "learning_rate": 3.2422836971481616e-06,
+      "loss": 0.5028,
+      "step": 5545
+    },
+    {
+      "epoch": 1.5379922351636162,
+      "grad_norm": 0.20688453316688538,
+      "learning_rate": 3.238574460583266e-06,
+      "loss": 0.4814,
+      "step": 5546
+    },
+    {
+      "epoch": 1.538269550748752,
+      "grad_norm": 0.21401961147785187,
+      "learning_rate": 3.234867031167865e-06,
+      "loss": 0.5026,
+      "step": 5547
+    },
+    {
+      "epoch": 1.538546866333888,
+      "grad_norm": 0.20731821656227112,
+      "learning_rate": 3.231161409625383e-06,
+      "loss": 0.5069,
+      "step": 5548
+    },
+    {
+      "epoch": 1.5388241819190238,
+      "grad_norm": 0.20831851661205292,
+      "learning_rate": 3.227457596678886e-06,
+      "loss": 0.4883,
+      "step": 5549
+    },
+    {
+      "epoch": 1.5391014975041597,
+      "grad_norm": 0.2025631219148636,
+      "learning_rate": 3.223755593051084e-06,
+      "loss": 0.4687,
+      "step": 5550
+    },
+    {
+      "epoch": 1.5393788130892956,
+      "grad_norm": 0.21386443078517914,
+      "learning_rate": 3.2200553994643307e-06,
+      "loss": 0.4807,
+      "step": 5551
+    },
+    {
+      "epoch": 1.5396561286744315,
+      "grad_norm": 0.22764462232589722,
+      "learning_rate": 3.2163570166406366e-06,
+      "loss": 0.4884,
+      "step": 5552
+    },
+    {
+      "epoch": 1.5399334442595674,
+      "grad_norm": 0.20263278484344482,
+      "learning_rate": 3.2126604453016574e-06,
+      "loss": 0.5059,
+      "step": 5553
+    },
+    {
+      "epoch": 1.5402107598447032,
+      "grad_norm": 0.22203831374645233,
+      "learning_rate": 3.20896568616869e-06,
+      "loss": 0.5071,
+      "step": 5554
+    },
+    {
+      "epoch": 1.5404880754298391,
+      "grad_norm": 0.21261858940124512,
+      "learning_rate": 3.205272739962674e-06,
+      "loss": 0.4967,
+      "step": 5555
+    },
+    {
+      "epoch": 1.540765391014975,
+      "grad_norm": 0.2063218355178833,
+      "learning_rate": 3.2015816074042052e-06,
+      "loss": 0.4983,
+      "step": 5556
+    },
+    {
+      "epoch": 1.541042706600111,
+      "grad_norm": 0.21330100297927856,
+      "learning_rate": 3.197892289213526e-06,
+      "loss": 0.4947,
+      "step": 5557
+    },
+    {
+      "epoch": 1.5413200221852468,
+      "grad_norm": 0.2076280117034912,
+      "learning_rate": 3.1942047861105163e-06,
+      "loss": 0.5006,
+      "step": 5558
+    },
+    {
+      "epoch": 1.5415973377703827,
+      "grad_norm": 0.2099052220582962,
+      "learning_rate": 3.190519098814697e-06,
+      "loss": 0.4869,
+      "step": 5559
+    },
+    {
+      "epoch": 1.5418746533555185,
+      "grad_norm": 0.20259883999824524,
+      "learning_rate": 3.1868352280452595e-06,
+      "loss": 0.4829,
+      "step": 5560
+    },
+    {
+      "epoch": 1.5421519689406544,
+      "grad_norm": 0.20924805104732513,
+      "learning_rate": 3.1831531745210168e-06,
+      "loss": 0.4715,
+      "step": 5561
+    },
+    {
+      "epoch": 1.5424292845257903,
+      "grad_norm": 0.2099631428718567,
+      "learning_rate": 3.1794729389604304e-06,
+      "loss": 0.491,
+      "step": 5562
+    },
+    {
+      "epoch": 1.5427066001109262,
+      "grad_norm": 0.2203865945339203,
+      "learning_rate": 3.1757945220816167e-06,
+      "loss": 0.5099,
+      "step": 5563
+    },
+    {
+      "epoch": 1.542983915696062,
+      "grad_norm": 0.20883382856845856,
+      "learning_rate": 3.1721179246023356e-06,
+      "loss": 0.4756,
+      "step": 5564
+    },
+    {
+      "epoch": 1.543261231281198,
+      "grad_norm": 0.24059654772281647,
+      "learning_rate": 3.1684431472399856e-06,
+      "loss": 0.4966,
+      "step": 5565
+    },
+    {
+      "epoch": 1.5435385468663338,
+      "grad_norm": 0.20498476922512054,
+      "learning_rate": 3.164770190711608e-06,
+      "loss": 0.5164,
+      "step": 5566
+    },
+    {
+      "epoch": 1.5438158624514697,
+      "grad_norm": 0.22245949506759644,
+      "learning_rate": 3.1610990557338987e-06,
+      "loss": 0.502,
+      "step": 5567
+    },
+    {
+      "epoch": 1.5440931780366056,
+      "grad_norm": 0.2156648188829422,
+      "learning_rate": 3.1574297430231977e-06,
+      "loss": 0.5193,
+      "step": 5568
+    },
+    {
+      "epoch": 1.5443704936217415,
+      "grad_norm": 0.22104239463806152,
+      "learning_rate": 3.153762253295475e-06,
+      "loss": 0.5048,
+      "step": 5569
+    },
+    {
+      "epoch": 1.5446478092068774,
+      "grad_norm": 0.20960746705532074,
+      "learning_rate": 3.1500965872663628e-06,
+      "loss": 0.5105,
+      "step": 5570
+    },
+    {
+      "epoch": 1.5449251247920133,
+      "grad_norm": 0.20040079951286316,
+      "learning_rate": 3.1464327456511288e-06,
+      "loss": 0.5048,
+      "step": 5571
+    },
+    {
+      "epoch": 1.5452024403771492,
+      "grad_norm": 0.20150801539421082,
+      "learning_rate": 3.142770729164686e-06,
+      "loss": 0.4976,
+      "step": 5572
+    },
+    {
+      "epoch": 1.545479755962285,
+      "grad_norm": 0.20172013342380524,
+      "learning_rate": 3.1391105385215847e-06,
+      "loss": 0.4958,
+      "step": 5573
+    },
+    {
+      "epoch": 1.545757071547421,
+      "grad_norm": 0.2224036306142807,
+      "learning_rate": 3.1354521744360295e-06,
+      "loss": 0.5085,
+      "step": 5574
+    },
+    {
+      "epoch": 1.5460343871325568,
+      "grad_norm": 0.21245881915092468,
+      "learning_rate": 3.131795637621868e-06,
+      "loss": 0.4828,
+      "step": 5575
+    },
+    {
+      "epoch": 1.5463117027176927,
+      "grad_norm": 0.21667277812957764,
+      "learning_rate": 3.1281409287925793e-06,
+      "loss": 0.492,
+      "step": 5576
+    },
+    {
+      "epoch": 1.5465890183028286,
+      "grad_norm": 0.2037910372018814,
+      "learning_rate": 3.1244880486613017e-06,
+      "loss": 0.5103,
+      "step": 5577
+    },
+    {
+      "epoch": 1.5468663338879645,
+      "grad_norm": 0.3222982585430145,
+      "learning_rate": 3.1208369979408013e-06,
+      "loss": 0.4977,
+      "step": 5578
+    },
+    {
+      "epoch": 1.5471436494731003,
+      "grad_norm": 0.2243606150150299,
+      "learning_rate": 3.117187777343504e-06,
+      "loss": 0.4867,
+      "step": 5579
+    },
+    {
+      "epoch": 1.5474209650582362,
+      "grad_norm": 0.2110065221786499,
+      "learning_rate": 3.1135403875814593e-06,
+      "loss": 0.5011,
+      "step": 5580
+    },
+    {
+      "epoch": 1.5476982806433721,
+      "grad_norm": 0.1963575929403305,
+      "learning_rate": 3.109894829366379e-06,
+      "loss": 0.4776,
+      "step": 5581
+    },
+    {
+      "epoch": 1.547975596228508,
+      "grad_norm": 0.20565971732139587,
+      "learning_rate": 3.1062511034095993e-06,
+      "loss": 0.4844,
+      "step": 5582
+    },
+    {
+      "epoch": 1.5482529118136439,
+      "grad_norm": 0.20696739852428436,
+      "learning_rate": 3.1026092104221124e-06,
+      "loss": 0.4766,
+      "step": 5583
+    },
+    {
+      "epoch": 1.5485302273987798,
+      "grad_norm": 0.27509939670562744,
+      "learning_rate": 3.098969151114552e-06,
+      "loss": 0.4783,
+      "step": 5584
+    },
+    {
+      "epoch": 1.5488075429839157,
+      "grad_norm": 0.20480765402317047,
+      "learning_rate": 3.0953309261971804e-06,
+      "loss": 0.4953,
+      "step": 5585
+    },
+    {
+      "epoch": 1.5490848585690515,
+      "grad_norm": 0.38247838616371155,
+      "learning_rate": 3.0916945363799214e-06,
+      "loss": 0.4516,
+      "step": 5586
+    },
+    {
+      "epoch": 1.5493621741541874,
+      "grad_norm": 0.21146711707115173,
+      "learning_rate": 3.088059982372324e-06,
+      "loss": 0.4855,
+      "step": 5587
+    },
+    {
+      "epoch": 1.5496394897393233,
+      "grad_norm": 0.2040729522705078,
+      "learning_rate": 3.0844272648835908e-06,
+      "loss": 0.5106,
+      "step": 5588
+    },
+    {
+      "epoch": 1.5499168053244592,
+      "grad_norm": 0.2061615288257599,
+      "learning_rate": 3.080796384622557e-06,
+      "loss": 0.4986,
+      "step": 5589
+    },
+    {
+      "epoch": 1.550194120909595,
+      "grad_norm": 0.19639424979686737,
+      "learning_rate": 3.0771673422977044e-06,
+      "loss": 0.5096,
+      "step": 5590
+    },
+    {
+      "epoch": 1.550471436494731,
+      "grad_norm": 0.22984932363033295,
+      "learning_rate": 3.0735401386171618e-06,
+      "loss": 0.4803,
+      "step": 5591
+    },
+    {
+      "epoch": 1.5507487520798668,
+      "grad_norm": 0.23276609182357788,
+      "learning_rate": 3.0699147742886862e-06,
+      "loss": 0.4857,
+      "step": 5592
+    },
+    {
+      "epoch": 1.5510260676650027,
+      "grad_norm": 0.22025956213474274,
+      "learning_rate": 3.0662912500196804e-06,
+      "loss": 0.5067,
+      "step": 5593
+    },
+    {
+      "epoch": 1.5513033832501386,
+      "grad_norm": 0.19536203145980835,
+      "learning_rate": 3.0626695665171927e-06,
+      "loss": 0.4849,
+      "step": 5594
+    },
+    {
+      "epoch": 1.5515806988352745,
+      "grad_norm": 0.2104620635509491,
+      "learning_rate": 3.059049724487914e-06,
+      "loss": 0.5033,
+      "step": 5595
+    },
+    {
+      "epoch": 1.5518580144204104,
+      "grad_norm": 0.2038435935974121,
+      "learning_rate": 3.0554317246381664e-06,
+      "loss": 0.5164,
+      "step": 5596
+    },
+    {
+      "epoch": 1.5521353300055463,
+      "grad_norm": 0.23535074293613434,
+      "learning_rate": 3.051815567673913e-06,
+      "loss": 0.4933,
+      "step": 5597
+    },
+    {
+      "epoch": 1.5524126455906821,
+      "grad_norm": 0.20647317171096802,
+      "learning_rate": 3.0482012543007725e-06,
+      "loss": 0.5152,
+      "step": 5598
+    },
+    {
+      "epoch": 1.552689961175818,
+      "grad_norm": 0.19642433524131775,
+      "learning_rate": 3.0445887852239897e-06,
+      "loss": 0.4896,
+      "step": 5599
+    },
+    {
+      "epoch": 1.552967276760954,
+      "grad_norm": 0.20343849062919617,
+      "learning_rate": 3.040978161148446e-06,
+      "loss": 0.4754,
+      "step": 5600
+    },
+    {
+      "epoch": 1.5532445923460898,
+      "grad_norm": 0.20736227929592133,
+      "learning_rate": 3.0373693827786766e-06,
+      "loss": 0.4728,
+      "step": 5601
+    },
+    {
+      "epoch": 1.5535219079312257,
+      "grad_norm": 0.2885718047618866,
+      "learning_rate": 3.03376245081885e-06,
+      "loss": 0.4844,
+      "step": 5602
+    },
+    {
+      "epoch": 1.5537992235163616,
+      "grad_norm": 0.21258218586444855,
+      "learning_rate": 3.0301573659727746e-06,
+      "loss": 0.4675,
+      "step": 5603
+    },
+    {
+      "epoch": 1.5540765391014975,
+      "grad_norm": 0.2054533213376999,
+      "learning_rate": 3.026554128943891e-06,
+      "loss": 0.4944,
+      "step": 5604
+    },
+    {
+      "epoch": 1.5543538546866333,
+      "grad_norm": 0.21457917988300323,
+      "learning_rate": 3.022952740435292e-06,
+      "loss": 0.4545,
+      "step": 5605
+    },
+    {
+      "epoch": 1.5546311702717692,
+      "grad_norm": 0.2189054936170578,
+      "learning_rate": 3.019353201149705e-06,
+      "loss": 0.4932,
+      "step": 5606
+    },
+    {
+      "epoch": 1.554908485856905,
+      "grad_norm": 0.19639572501182556,
+      "learning_rate": 3.0157555117894907e-06,
+      "loss": 0.4758,
+      "step": 5607
+    },
+    {
+      "epoch": 1.555185801442041,
+      "grad_norm": 0.21322979032993317,
+      "learning_rate": 3.012159673056661e-06,
+      "loss": 0.4928,
+      "step": 5608
+    },
+    {
+      "epoch": 1.5554631170271769,
+      "grad_norm": 0.2102145254611969,
+      "learning_rate": 3.008565685652849e-06,
+      "loss": 0.515,
+      "step": 5609
+    },
+    {
+      "epoch": 1.5557404326123128,
+      "grad_norm": 0.22351384162902832,
+      "learning_rate": 3.004973550279348e-06,
+      "loss": 0.5084,
+      "step": 5610
+    },
+    {
+      "epoch": 1.5560177481974486,
+      "grad_norm": 0.2032230645418167,
+      "learning_rate": 3.001383267637069e-06,
+      "loss": 0.4957,
+      "step": 5611
+    },
+    {
+      "epoch": 1.5562950637825845,
+      "grad_norm": 0.2006012350320816,
+      "learning_rate": 2.997794838426575e-06,
+      "loss": 0.481,
+      "step": 5612
+    },
+    {
+      "epoch": 1.5565723793677204,
+      "grad_norm": 0.20619052648544312,
+      "learning_rate": 2.9942082633480696e-06,
+      "loss": 0.4755,
+      "step": 5613
+    },
+    {
+      "epoch": 1.5568496949528563,
+      "grad_norm": 0.21772894263267517,
+      "learning_rate": 2.990623543101377e-06,
+      "loss": 0.4795,
+      "step": 5614
+    },
+    {
+      "epoch": 1.5571270105379922,
+      "grad_norm": 0.19506295025348663,
+      "learning_rate": 2.9870406783859827e-06,
+      "loss": 0.4655,
+      "step": 5615
+    },
+    {
+      "epoch": 1.557404326123128,
+      "grad_norm": 0.20197400450706482,
+      "learning_rate": 2.98345966990099e-06,
+      "loss": 0.5146,
+      "step": 5616
+    },
+    {
+      "epoch": 1.557681641708264,
+      "grad_norm": 0.20348283648490906,
+      "learning_rate": 2.979880518345153e-06,
+      "loss": 0.4377,
+      "step": 5617
+    },
+    {
+      "epoch": 1.5579589572933998,
+      "grad_norm": 0.20409370958805084,
+      "learning_rate": 2.976303224416856e-06,
+      "loss": 0.5216,
+      "step": 5618
+    },
+    {
+      "epoch": 1.5582362728785357,
+      "grad_norm": 0.21259050071239471,
+      "learning_rate": 2.972727788814128e-06,
+      "loss": 0.4988,
+      "step": 5619
+    },
+    {
+      "epoch": 1.5585135884636716,
+      "grad_norm": 0.19778795540332794,
+      "learning_rate": 2.9691542122346262e-06,
+      "loss": 0.5165,
+      "step": 5620
+    },
+    {
+      "epoch": 1.5587909040488075,
+      "grad_norm": 0.27131521701812744,
+      "learning_rate": 2.9655824953756517e-06,
+      "loss": 0.4913,
+      "step": 5621
+    },
+    {
+      "epoch": 1.5590682196339434,
+      "grad_norm": 0.2034967541694641,
+      "learning_rate": 2.962012638934146e-06,
+      "loss": 0.4725,
+      "step": 5622
+    },
+    {
+      "epoch": 1.5593455352190793,
+      "grad_norm": 0.19683615863323212,
+      "learning_rate": 2.9584446436066752e-06,
+      "loss": 0.4949,
+      "step": 5623
+    },
+    {
+      "epoch": 1.5596228508042151,
+      "grad_norm": 0.21673937141895294,
+      "learning_rate": 2.9548785100894565e-06,
+      "loss": 0.5125,
+      "step": 5624
+    },
+    {
+      "epoch": 1.559900166389351,
+      "grad_norm": 0.22336265444755554,
+      "learning_rate": 2.9513142390783303e-06,
+      "loss": 0.4803,
+      "step": 5625
+    },
+    {
+      "epoch": 1.560177481974487,
+      "grad_norm": 0.21071960031986237,
+      "learning_rate": 2.947751831268787e-06,
+      "loss": 0.5069,
+      "step": 5626
+    },
+    {
+      "epoch": 1.5604547975596228,
+      "grad_norm": 0.2002846896648407,
+      "learning_rate": 2.9441912873559406e-06,
+      "loss": 0.4857,
+      "step": 5627
+    },
+    {
+      "epoch": 1.5607321131447587,
+      "grad_norm": 0.20336318016052246,
+      "learning_rate": 2.940632608034549e-06,
+      "loss": 0.5015,
+      "step": 5628
+    },
+    {
+      "epoch": 1.5610094287298946,
+      "grad_norm": 0.199977844953537,
+      "learning_rate": 2.9370757939990114e-06,
+      "loss": 0.5078,
+      "step": 5629
+    },
+    {
+      "epoch": 1.5612867443150305,
+      "grad_norm": 0.20519019663333893,
+      "learning_rate": 2.933520845943351e-06,
+      "loss": 0.5052,
+      "step": 5630
+    },
+    {
+      "epoch": 1.5615640599001663,
+      "grad_norm": 0.21687155961990356,
+      "learning_rate": 2.9299677645612282e-06,
+      "loss": 0.4759,
+      "step": 5631
+    },
+    {
+      "epoch": 1.5618413754853022,
+      "grad_norm": 0.2080409973859787,
+      "learning_rate": 2.9264165505459476e-06,
+      "loss": 0.4918,
+      "step": 5632
+    },
+    {
+      "epoch": 1.562118691070438,
+      "grad_norm": 0.2009446918964386,
+      "learning_rate": 2.92286720459045e-06,
+      "loss": 0.4763,
+      "step": 5633
+    },
+    {
+      "epoch": 1.562396006655574,
+      "grad_norm": 0.22965089976787567,
+      "learning_rate": 2.9193197273873014e-06,
+      "loss": 0.483,
+      "step": 5634
+    },
+    {
+      "epoch": 1.5626733222407099,
+      "grad_norm": 0.20917722582817078,
+      "learning_rate": 2.9157741196287017e-06,
+      "loss": 0.496,
+      "step": 5635
+    },
+    {
+      "epoch": 1.5629506378258458,
+      "grad_norm": 0.23011514544487,
+      "learning_rate": 2.9122303820065087e-06,
+      "loss": 0.4844,
+      "step": 5636
+    },
+    {
+      "epoch": 1.5632279534109816,
+      "grad_norm": 0.2208072394132614,
+      "learning_rate": 2.908688515212191e-06,
+      "loss": 0.4847,
+      "step": 5637
+    },
+    {
+      "epoch": 1.5635052689961175,
+      "grad_norm": 0.2128543108701706,
+      "learning_rate": 2.905148519936857e-06,
+      "loss": 0.5051,
+      "step": 5638
+    },
+    {
+      "epoch": 1.5637825845812534,
+      "grad_norm": 0.2076890915632248,
+      "learning_rate": 2.9016103968712584e-06,
+      "loss": 0.5162,
+      "step": 5639
+    },
+    {
+      "epoch": 1.5640599001663893,
+      "grad_norm": 0.20999306440353394,
+      "learning_rate": 2.8980741467057786e-06,
+      "loss": 0.4999,
+      "step": 5640
+    },
+    {
+      "epoch": 1.5643372157515252,
+      "grad_norm": 0.22848869860172272,
+      "learning_rate": 2.8945397701304304e-06,
+      "loss": 0.4806,
+      "step": 5641
+    },
+    {
+      "epoch": 1.564614531336661,
+      "grad_norm": 0.23232677578926086,
+      "learning_rate": 2.8910072678348625e-06,
+      "loss": 0.4944,
+      "step": 5642
+    },
+    {
+      "epoch": 1.564891846921797,
+      "grad_norm": 0.2025647908449173,
+      "learning_rate": 2.887476640508363e-06,
+      "loss": 0.4707,
+      "step": 5643
+    },
+    {
+      "epoch": 1.5651691625069328,
+      "grad_norm": 0.20394767820835114,
+      "learning_rate": 2.8839478888398523e-06,
+      "loss": 0.5031,
+      "step": 5644
+    },
+    {
+      "epoch": 1.5654464780920687,
+      "grad_norm": 0.2066267430782318,
+      "learning_rate": 2.880421013517881e-06,
+      "loss": 0.4955,
+      "step": 5645
+    },
+    {
+      "epoch": 1.5657237936772046,
+      "grad_norm": 0.20270580053329468,
+      "learning_rate": 2.876896015230632e-06,
+      "loss": 0.474,
+      "step": 5646
+    },
+    {
+      "epoch": 1.5660011092623405,
+      "grad_norm": 0.2070900797843933,
+      "learning_rate": 2.8733728946659293e-06,
+      "loss": 0.4942,
+      "step": 5647
+    },
+    {
+      "epoch": 1.5662784248474764,
+      "grad_norm": 0.20855194330215454,
+      "learning_rate": 2.8698516525112324e-06,
+      "loss": 0.4801,
+      "step": 5648
+    },
+    {
+      "epoch": 1.5665557404326123,
+      "grad_norm": 0.2164842188358307,
+      "learning_rate": 2.86633228945362e-06,
+      "loss": 0.4946,
+      "step": 5649
+    },
+    {
+      "epoch": 1.5668330560177481,
+      "grad_norm": 0.21440735459327698,
+      "learning_rate": 2.862814806179817e-06,
+      "loss": 0.5068,
+      "step": 5650
+    },
+    {
+      "epoch": 1.567110371602884,
+      "grad_norm": 0.2091311663389206,
+      "learning_rate": 2.8592992033761814e-06,
+      "loss": 0.4693,
+      "step": 5651
+    },
+    {
+      "epoch": 1.56738768718802,
+      "grad_norm": 0.20943525433540344,
+      "learning_rate": 2.855785481728697e-06,
+      "loss": 0.5165,
+      "step": 5652
+    },
+    {
+      "epoch": 1.5676650027731558,
+      "grad_norm": 0.20559410750865936,
+      "learning_rate": 2.85227364192298e-06,
+      "loss": 0.5084,
+      "step": 5653
+    },
+    {
+      "epoch": 1.5679423183582917,
+      "grad_norm": 0.20715293288230896,
+      "learning_rate": 2.8487636846442873e-06,
+      "loss": 0.5199,
+      "step": 5654
+    },
+    {
+      "epoch": 1.5682196339434276,
+      "grad_norm": 0.21139737963676453,
+      "learning_rate": 2.8452556105775073e-06,
+      "loss": 0.5422,
+      "step": 5655
+    },
+    {
+      "epoch": 1.5684969495285634,
+      "grad_norm": 0.20338012278079987,
+      "learning_rate": 2.8417494204071526e-06,
+      "loss": 0.5006,
+      "step": 5656
+    },
+    {
+      "epoch": 1.5687742651136993,
+      "grad_norm": 0.21125024557113647,
+      "learning_rate": 2.8382451148173798e-06,
+      "loss": 0.5144,
+      "step": 5657
+    },
+    {
+      "epoch": 1.5690515806988352,
+      "grad_norm": 0.2064410001039505,
+      "learning_rate": 2.8347426944919637e-06,
+      "loss": 0.4919,
+      "step": 5658
+    },
+    {
+      "epoch": 1.569328896283971,
+      "grad_norm": 0.19719727337360382,
+      "learning_rate": 2.8312421601143267e-06,
+      "loss": 0.4966,
+      "step": 5659
+    },
+    {
+      "epoch": 1.569606211869107,
+      "grad_norm": 0.2056899517774582,
+      "learning_rate": 2.8277435123675144e-06,
+      "loss": 0.5109,
+      "step": 5660
+    },
+    {
+      "epoch": 1.5698835274542429,
+      "grad_norm": 0.20407631993293762,
+      "learning_rate": 2.8242467519342022e-06,
+      "loss": 0.476,
+      "step": 5661
+    },
+    {
+      "epoch": 1.5701608430393788,
+      "grad_norm": 0.2034579962491989,
+      "learning_rate": 2.8207518794967054e-06,
+      "loss": 0.5173,
+      "step": 5662
+    },
+    {
+      "epoch": 1.5704381586245146,
+      "grad_norm": 0.21342724561691284,
+      "learning_rate": 2.81725889573696e-06,
+      "loss": 0.4815,
+      "step": 5663
+    },
+    {
+      "epoch": 1.5707154742096505,
+      "grad_norm": 0.2078382819890976,
+      "learning_rate": 2.813767801336549e-06,
+      "loss": 0.5132,
+      "step": 5664
+    },
+    {
+      "epoch": 1.5709927897947864,
+      "grad_norm": 0.20430020987987518,
+      "learning_rate": 2.810278596976666e-06,
+      "loss": 0.5055,
+      "step": 5665
+    },
+    {
+      "epoch": 1.5712701053799223,
+      "grad_norm": 0.2025614082813263,
+      "learning_rate": 2.806791283338155e-06,
+      "loss": 0.4966,
+      "step": 5666
+    },
+    {
+      "epoch": 1.5715474209650582,
+      "grad_norm": 0.20510238409042358,
+      "learning_rate": 2.8033058611014838e-06,
+      "loss": 0.489,
+      "step": 5667
+    },
+    {
+      "epoch": 1.571824736550194,
+      "grad_norm": 0.2076101154088974,
+      "learning_rate": 2.7998223309467484e-06,
+      "loss": 0.5016,
+      "step": 5668
+    },
+    {
+      "epoch": 1.57210205213533,
+      "grad_norm": 0.19899241626262665,
+      "learning_rate": 2.7963406935536728e-06,
+      "loss": 0.5041,
+      "step": 5669
+    },
+    {
+      "epoch": 1.5723793677204658,
+      "grad_norm": 0.22239477932453156,
+      "learning_rate": 2.7928609496016235e-06,
+      "loss": 0.5068,
+      "step": 5670
+    },
+    {
+      "epoch": 1.5726566833056017,
+      "grad_norm": 0.20445238053798676,
+      "learning_rate": 2.789383099769591e-06,
+      "loss": 0.4886,
+      "step": 5671
+    },
+    {
+      "epoch": 1.5729339988907376,
+      "grad_norm": 0.20267120003700256,
+      "learning_rate": 2.785907144736194e-06,
+      "loss": 0.4763,
+      "step": 5672
+    },
+    {
+      "epoch": 1.5732113144758735,
+      "grad_norm": 0.20123063027858734,
+      "learning_rate": 2.7824330851796755e-06,
+      "loss": 0.4932,
+      "step": 5673
+    },
+    {
+      "epoch": 1.5734886300610094,
+      "grad_norm": 0.20822611451148987,
+      "learning_rate": 2.7789609217779316e-06,
+      "loss": 0.47,
+      "step": 5674
+    },
+    {
+      "epoch": 1.5737659456461452,
+      "grad_norm": 0.21520821750164032,
+      "learning_rate": 2.7754906552084667e-06,
+      "loss": 0.4951,
+      "step": 5675
+    },
+    {
+      "epoch": 1.5740432612312811,
+      "grad_norm": 0.20591013133525848,
+      "learning_rate": 2.7720222861484167e-06,
+      "loss": 0.4938,
+      "step": 5676
+    },
+    {
+      "epoch": 1.574320576816417,
+      "grad_norm": 0.22870129346847534,
+      "learning_rate": 2.768555815274557e-06,
+      "loss": 0.4814,
+      "step": 5677
+    },
+    {
+      "epoch": 1.574597892401553,
+      "grad_norm": 0.1999206393957138,
+      "learning_rate": 2.76509124326329e-06,
+      "loss": 0.5062,
+      "step": 5678
+    },
+    {
+      "epoch": 1.5748752079866888,
+      "grad_norm": 0.21404382586479187,
+      "learning_rate": 2.7616285707906447e-06,
+      "loss": 0.5205,
+      "step": 5679
+    },
+    {
+      "epoch": 1.5751525235718247,
+      "grad_norm": 0.22264225780963898,
+      "learning_rate": 2.7581677985322742e-06,
+      "loss": 0.4951,
+      "step": 5680
+    },
+    {
+      "epoch": 1.5754298391569606,
+      "grad_norm": 0.21465402841567993,
+      "learning_rate": 2.75470892716347e-06,
+      "loss": 0.4647,
+      "step": 5681
+    },
+    {
+      "epoch": 1.5757071547420964,
+      "grad_norm": 0.21319536864757538,
+      "learning_rate": 2.751251957359155e-06,
+      "loss": 0.5177,
+      "step": 5682
+    },
+    {
+      "epoch": 1.5759844703272323,
+      "grad_norm": 0.21107381582260132,
+      "learning_rate": 2.7477968897938717e-06,
+      "loss": 0.4781,
+      "step": 5683
+    },
+    {
+      "epoch": 1.5762617859123682,
+      "grad_norm": 0.22035937011241913,
+      "learning_rate": 2.744343725141792e-06,
+      "loss": 0.4566,
+      "step": 5684
+    },
+    {
+      "epoch": 1.576539101497504,
+      "grad_norm": 0.21590951085090637,
+      "learning_rate": 2.7408924640767218e-06,
+      "loss": 0.5025,
+      "step": 5685
+    },
+    {
+      "epoch": 1.57681641708264,
+      "grad_norm": 0.20270653069019318,
+      "learning_rate": 2.7374431072720975e-06,
+      "loss": 0.529,
+      "step": 5686
+    },
+    {
+      "epoch": 1.5770937326677759,
+      "grad_norm": 0.2021293044090271,
+      "learning_rate": 2.733995655400974e-06,
+      "loss": 0.4969,
+      "step": 5687
+    },
+    {
+      "epoch": 1.5773710482529117,
+      "grad_norm": 0.19878076016902924,
+      "learning_rate": 2.7305501091360444e-06,
+      "loss": 0.476,
+      "step": 5688
+    },
+    {
+      "epoch": 1.5776483638380476,
+      "grad_norm": 0.2047017365694046,
+      "learning_rate": 2.7271064691496277e-06,
+      "loss": 0.4912,
+      "step": 5689
+    },
+    {
+      "epoch": 1.5779256794231835,
+      "grad_norm": 0.20984232425689697,
+      "learning_rate": 2.723664736113668e-06,
+      "loss": 0.4956,
+      "step": 5690
+    },
+    {
+      "epoch": 1.5782029950083194,
+      "grad_norm": 0.2142646759748459,
+      "learning_rate": 2.720224910699733e-06,
+      "loss": 0.5026,
+      "step": 5691
+    },
+    {
+      "epoch": 1.5784803105934553,
+      "grad_norm": 0.2094864547252655,
+      "learning_rate": 2.7167869935790276e-06,
+      "loss": 0.4846,
+      "step": 5692
+    },
+    {
+      "epoch": 1.5787576261785912,
+      "grad_norm": 0.21176081895828247,
+      "learning_rate": 2.713350985422386e-06,
+      "loss": 0.5049,
+      "step": 5693
+    },
+    {
+      "epoch": 1.579034941763727,
+      "grad_norm": 0.1998070776462555,
+      "learning_rate": 2.7099168869002543e-06,
+      "loss": 0.4521,
+      "step": 5694
+    },
+    {
+      "epoch": 1.579312257348863,
+      "grad_norm": 0.20673444867134094,
+      "learning_rate": 2.7064846986827264e-06,
+      "loss": 0.511,
+      "step": 5695
+    },
+    {
+      "epoch": 1.5795895729339988,
+      "grad_norm": 0.21131543815135956,
+      "learning_rate": 2.7030544214395035e-06,
+      "loss": 0.4869,
+      "step": 5696
+    },
+    {
+      "epoch": 1.5798668885191347,
+      "grad_norm": 0.21434102952480316,
+      "learning_rate": 2.6996260558399324e-06,
+      "loss": 0.4945,
+      "step": 5697
+    },
+    {
+      "epoch": 1.5801442041042706,
+      "grad_norm": 0.2118789553642273,
+      "learning_rate": 2.696199602552971e-06,
+      "loss": 0.4696,
+      "step": 5698
+    },
+    {
+      "epoch": 1.5804215196894065,
+      "grad_norm": 0.21816542744636536,
+      "learning_rate": 2.6927750622472157e-06,
+      "loss": 0.4896,
+      "step": 5699
+    },
+    {
+      "epoch": 1.5806988352745424,
+      "grad_norm": 0.2012248933315277,
+      "learning_rate": 2.6893524355908804e-06,
+      "loss": 0.4781,
+      "step": 5700
+    },
+    {
+      "epoch": 1.5809761508596782,
+      "grad_norm": 0.1999814510345459,
+      "learning_rate": 2.685931723251814e-06,
+      "loss": 0.507,
+      "step": 5701
+    },
+    {
+      "epoch": 1.5812534664448141,
+      "grad_norm": 0.21244245767593384,
+      "learning_rate": 2.682512925897489e-06,
+      "loss": 0.4788,
+      "step": 5702
+    },
+    {
+      "epoch": 1.58153078202995,
+      "grad_norm": 0.19930778443813324,
+      "learning_rate": 2.6790960441949996e-06,
+      "loss": 0.5082,
+      "step": 5703
+    },
+    {
+      "epoch": 1.581808097615086,
+      "grad_norm": 0.19802865386009216,
+      "learning_rate": 2.675681078811075e-06,
+      "loss": 0.4891,
+      "step": 5704
+    },
+    {
+      "epoch": 1.5820854132002218,
+      "grad_norm": 0.2145322859287262,
+      "learning_rate": 2.6722680304120575e-06,
+      "loss": 0.5113,
+      "step": 5705
+    },
+    {
+      "epoch": 1.5823627287853577,
+      "grad_norm": 0.20376405119895935,
+      "learning_rate": 2.6688568996639337e-06,
+      "loss": 0.4751,
+      "step": 5706
+    },
+    {
+      "epoch": 1.5826400443704935,
+      "grad_norm": 0.21193109452724457,
+      "learning_rate": 2.6654476872322948e-06,
+      "loss": 0.4846,
+      "step": 5707
+    },
+    {
+      "epoch": 1.5829173599556294,
+      "grad_norm": 0.21111689507961273,
+      "learning_rate": 2.662040393782375e-06,
+      "loss": 0.4803,
+      "step": 5708
+    },
+    {
+      "epoch": 1.5831946755407653,
+      "grad_norm": 0.204023540019989,
+      "learning_rate": 2.658635019979029e-06,
+      "loss": 0.5126,
+      "step": 5709
+    },
+    {
+      "epoch": 1.5834719911259012,
+      "grad_norm": 0.20971745252609253,
+      "learning_rate": 2.655231566486732e-06,
+      "loss": 0.5089,
+      "step": 5710
+    },
+    {
+      "epoch": 1.583749306711037,
+      "grad_norm": 0.19988863170146942,
+      "learning_rate": 2.6518300339695865e-06,
+      "loss": 0.5121,
+      "step": 5711
+    },
+    {
+      "epoch": 1.584026622296173,
+      "grad_norm": 0.21533909440040588,
+      "learning_rate": 2.6484304230913236e-06,
+      "loss": 0.4912,
+      "step": 5712
+    },
+    {
+      "epoch": 1.5843039378813089,
+      "grad_norm": 0.20329493284225464,
+      "learning_rate": 2.6450327345153e-06,
+      "loss": 0.4747,
+      "step": 5713
+    },
+    {
+      "epoch": 1.5845812534664447,
+      "grad_norm": 0.2205643355846405,
+      "learning_rate": 2.6416369689044903e-06,
+      "loss": 0.5216,
+      "step": 5714
+    },
+    {
+      "epoch": 1.5848585690515806,
+      "grad_norm": 0.20560097694396973,
+      "learning_rate": 2.638243126921498e-06,
+      "loss": 0.4795,
+      "step": 5715
+    },
+    {
+      "epoch": 1.5851358846367165,
+      "grad_norm": 0.19872471690177917,
+      "learning_rate": 2.6348512092285603e-06,
+      "loss": 0.516,
+      "step": 5716
+    },
+    {
+      "epoch": 1.5854132002218524,
+      "grad_norm": 0.2076081484556198,
+      "learning_rate": 2.6314612164875213e-06,
+      "loss": 0.4939,
+      "step": 5717
+    },
+    {
+      "epoch": 1.5856905158069883,
+      "grad_norm": 0.19979238510131836,
+      "learning_rate": 2.6280731493598596e-06,
+      "loss": 0.4665,
+      "step": 5718
+    },
+    {
+      "epoch": 1.5859678313921242,
+      "grad_norm": 0.23124083876609802,
+      "learning_rate": 2.6246870085066764e-06,
+      "loss": 0.5132,
+      "step": 5719
+    },
+    {
+      "epoch": 1.58624514697726,
+      "grad_norm": 0.22173020243644714,
+      "learning_rate": 2.6213027945887035e-06,
+      "loss": 0.5085,
+      "step": 5720
+    },
+    {
+      "epoch": 1.586522462562396,
+      "grad_norm": 0.19916951656341553,
+      "learning_rate": 2.6179205082662862e-06,
+      "loss": 0.4924,
+      "step": 5721
+    },
+    {
+      "epoch": 1.5867997781475318,
+      "grad_norm": 0.21253123879432678,
+      "learning_rate": 2.614540150199396e-06,
+      "loss": 0.5012,
+      "step": 5722
+    },
+    {
+      "epoch": 1.5870770937326677,
+      "grad_norm": 0.21266770362854004,
+      "learning_rate": 2.611161721047632e-06,
+      "loss": 0.51,
+      "step": 5723
+    },
+    {
+      "epoch": 1.5873544093178036,
+      "grad_norm": 0.20849168300628662,
+      "learning_rate": 2.60778522147022e-06,
+      "loss": 0.5161,
+      "step": 5724
+    },
+    {
+      "epoch": 1.5876317249029395,
+      "grad_norm": 0.20374304056167603,
+      "learning_rate": 2.6044106521259963e-06,
+      "loss": 0.4899,
+      "step": 5725
+    },
+    {
+      "epoch": 1.5879090404880754,
+      "grad_norm": 0.2087586522102356,
+      "learning_rate": 2.6010380136734347e-06,
+      "loss": 0.5035,
+      "step": 5726
+    },
+    {
+      "epoch": 1.5881863560732112,
+      "grad_norm": 0.20148667693138123,
+      "learning_rate": 2.5976673067706262e-06,
+      "loss": 0.504,
+      "step": 5727
+    },
+    {
+      "epoch": 1.5884636716583471,
+      "grad_norm": 0.2141498327255249,
+      "learning_rate": 2.5942985320752843e-06,
+      "loss": 0.5007,
+      "step": 5728
+    },
+    {
+      "epoch": 1.588740987243483,
+      "grad_norm": 0.20762969553470612,
+      "learning_rate": 2.5909316902447426e-06,
+      "loss": 0.4928,
+      "step": 5729
+    },
+    {
+      "epoch": 1.589018302828619,
+      "grad_norm": 0.20631739497184753,
+      "learning_rate": 2.5875667819359626e-06,
+      "loss": 0.4968,
+      "step": 5730
+    },
+    {
+      "epoch": 1.5892956184137548,
+      "grad_norm": 0.2064218521118164,
+      "learning_rate": 2.5842038078055327e-06,
+      "loss": 0.4878,
+      "step": 5731
+    },
+    {
+      "epoch": 1.5895729339988907,
+      "grad_norm": 0.20431697368621826,
+      "learning_rate": 2.5808427685096505e-06,
+      "loss": 0.4973,
+      "step": 5732
+    },
+    {
+      "epoch": 1.5898502495840265,
+      "grad_norm": 0.2072911262512207,
+      "learning_rate": 2.5774836647041515e-06,
+      "loss": 0.5057,
+      "step": 5733
+    },
+    {
+      "epoch": 1.5901275651691624,
+      "grad_norm": 0.22611966729164124,
+      "learning_rate": 2.5741264970444768e-06,
+      "loss": 0.4918,
+      "step": 5734
+    },
+    {
+      "epoch": 1.5904048807542983,
+      "grad_norm": 0.20940294861793518,
+      "learning_rate": 2.570771266185708e-06,
+      "loss": 0.4741,
+      "step": 5735
+    },
+    {
+      "epoch": 1.5906821963394342,
+      "grad_norm": 0.21070222556591034,
+      "learning_rate": 2.5674179727825307e-06,
+      "loss": 0.4841,
+      "step": 5736
+    },
+    {
+      "epoch": 1.59095951192457,
+      "grad_norm": 0.21693362295627594,
+      "learning_rate": 2.564066617489269e-06,
+      "loss": 0.4725,
+      "step": 5737
+    },
+    {
+      "epoch": 1.591236827509706,
+      "grad_norm": 0.20319637656211853,
+      "learning_rate": 2.5607172009598556e-06,
+      "loss": 0.5041,
+      "step": 5738
+    },
+    {
+      "epoch": 1.5915141430948418,
+      "grad_norm": 0.2054455578327179,
+      "learning_rate": 2.5573697238478523e-06,
+      "loss": 0.4975,
+      "step": 5739
+    },
+    {
+      "epoch": 1.5917914586799777,
+      "grad_norm": 0.20702321827411652,
+      "learning_rate": 2.5540241868064434e-06,
+      "loss": 0.4761,
+      "step": 5740
+    },
+    {
+      "epoch": 1.5920687742651136,
+      "grad_norm": 0.23018468916416168,
+      "learning_rate": 2.5506805904884272e-06,
+      "loss": 0.5258,
+      "step": 5741
+    },
+    {
+      "epoch": 1.5923460898502495,
+      "grad_norm": 0.2070692479610443,
+      "learning_rate": 2.5473389355462325e-06,
+      "loss": 0.5011,
+      "step": 5742
+    },
+    {
+      "epoch": 1.5926234054353854,
+      "grad_norm": 0.21105840802192688,
+      "learning_rate": 2.543999222631899e-06,
+      "loss": 0.491,
+      "step": 5743
+    },
+    {
+      "epoch": 1.5929007210205213,
+      "grad_norm": 0.21551688015460968,
+      "learning_rate": 2.5406614523971e-06,
+      "loss": 0.4878,
+      "step": 5744
+    },
+    {
+      "epoch": 1.5931780366056572,
+      "grad_norm": 0.21947219967842102,
+      "learning_rate": 2.537325625493116e-06,
+      "loss": 0.4982,
+      "step": 5745
+    },
+    {
+      "epoch": 1.593455352190793,
+      "grad_norm": 0.21670997142791748,
+      "learning_rate": 2.5339917425708584e-06,
+      "loss": 0.5216,
+      "step": 5746
+    },
+    {
+      "epoch": 1.593732667775929,
+      "grad_norm": 0.22349973022937775,
+      "learning_rate": 2.5306598042808592e-06,
+      "loss": 0.5216,
+      "step": 5747
+    },
+    {
+      "epoch": 1.5940099833610648,
+      "grad_norm": 0.20010913908481598,
+      "learning_rate": 2.5273298112732657e-06,
+      "loss": 0.467,
+      "step": 5748
+    },
+    {
+      "epoch": 1.5942872989462007,
+      "grad_norm": 0.2064429521560669,
+      "learning_rate": 2.5240017641978435e-06,
+      "loss": 0.4931,
+      "step": 5749
+    },
+    {
+      "epoch": 1.5945646145313366,
+      "grad_norm": 0.19768312573432922,
+      "learning_rate": 2.520675663703985e-06,
+      "loss": 0.4852,
+      "step": 5750
+    },
+    {
+      "epoch": 1.5948419301164725,
+      "grad_norm": 0.21115142107009888,
+      "learning_rate": 2.517351510440706e-06,
+      "loss": 0.5125,
+      "step": 5751
+    },
+    {
+      "epoch": 1.5951192457016083,
+      "grad_norm": 0.19108986854553223,
+      "learning_rate": 2.5140293050566295e-06,
+      "loss": 0.4878,
+      "step": 5752
+    },
+    {
+      "epoch": 1.5953965612867442,
+      "grad_norm": 0.20708754658699036,
+      "learning_rate": 2.510709048200009e-06,
+      "loss": 0.4871,
+      "step": 5753
+    },
+    {
+      "epoch": 1.5956738768718801,
+      "grad_norm": 0.20507743954658508,
+      "learning_rate": 2.507390740518717e-06,
+      "loss": 0.5031,
+      "step": 5754
+    },
+    {
+      "epoch": 1.595951192457016,
+      "grad_norm": 0.2134825587272644,
+      "learning_rate": 2.5040743826602405e-06,
+      "loss": 0.4753,
+      "step": 5755
+    },
+    {
+      "epoch": 1.5962285080421519,
+      "grad_norm": 0.20486515760421753,
+      "learning_rate": 2.5007599752716867e-06,
+      "loss": 0.5022,
+      "step": 5756
+    },
+    {
+      "epoch": 1.5965058236272878,
+      "grad_norm": 0.20787964761257172,
+      "learning_rate": 2.4974475189997862e-06,
+      "loss": 0.498,
+      "step": 5757
+    },
+    {
+      "epoch": 1.5967831392124237,
+      "grad_norm": 0.20221605896949768,
+      "learning_rate": 2.494137014490891e-06,
+      "loss": 0.518,
+      "step": 5758
+    },
+    {
+      "epoch": 1.5970604547975595,
+      "grad_norm": 0.2075101137161255,
+      "learning_rate": 2.4908284623909638e-06,
+      "loss": 0.5026,
+      "step": 5759
+    },
+    {
+      "epoch": 1.5973377703826954,
+      "grad_norm": 0.21558356285095215,
+      "learning_rate": 2.487521863345589e-06,
+      "loss": 0.488,
+      "step": 5760
+    },
+    {
+      "epoch": 1.5976150859678313,
+      "grad_norm": 0.20336772501468658,
+      "learning_rate": 2.4842172179999736e-06,
+      "loss": 0.4776,
+      "step": 5761
+    },
+    {
+      "epoch": 1.5978924015529672,
+      "grad_norm": 0.2211672067642212,
+      "learning_rate": 2.480914526998945e-06,
+      "loss": 0.4935,
+      "step": 5762
+    },
+    {
+      "epoch": 1.598169717138103,
+      "grad_norm": 0.2278861552476883,
+      "learning_rate": 2.4776137909869434e-06,
+      "loss": 0.4833,
+      "step": 5763
+    },
+    {
+      "epoch": 1.598447032723239,
+      "grad_norm": 0.20449037849903107,
+      "learning_rate": 2.4743150106080233e-06,
+      "loss": 0.4874,
+      "step": 5764
+    },
+    {
+      "epoch": 1.5987243483083748,
+      "grad_norm": 0.20173722505569458,
+      "learning_rate": 2.471018186505876e-06,
+      "loss": 0.4963,
+      "step": 5765
+    },
+    {
+      "epoch": 1.5990016638935107,
+      "grad_norm": 0.2080666571855545,
+      "learning_rate": 2.4677233193237945e-06,
+      "loss": 0.4766,
+      "step": 5766
+    },
+    {
+      "epoch": 1.5992789794786466,
+      "grad_norm": 0.2013276368379593,
+      "learning_rate": 2.4644304097046892e-06,
+      "loss": 0.4742,
+      "step": 5767
+    },
+    {
+      "epoch": 1.5995562950637825,
+      "grad_norm": 0.21188431978225708,
+      "learning_rate": 2.461139458291098e-06,
+      "loss": 0.4795,
+      "step": 5768
+    },
+    {
+      "epoch": 1.5998336106489184,
+      "grad_norm": 0.20740027725696564,
+      "learning_rate": 2.457850465725177e-06,
+      "loss": 0.4796,
+      "step": 5769
+    },
+    {
+      "epoch": 1.6001109262340543,
+      "grad_norm": 0.20362474024295807,
+      "learning_rate": 2.454563432648692e-06,
+      "loss": 0.5114,
+      "step": 5770
+    },
+    {
+      "epoch": 1.6003882418191901,
+      "grad_norm": 0.3291257619857788,
+      "learning_rate": 2.451278359703027e-06,
+      "loss": 0.4933,
+      "step": 5771
+    },
+    {
+      "epoch": 1.600665557404326,
+      "grad_norm": 0.20942147076129913,
+      "learning_rate": 2.447995247529189e-06,
+      "loss": 0.4955,
+      "step": 5772
+    },
+    {
+      "epoch": 1.600942872989462,
+      "grad_norm": 0.20903481543064117,
+      "learning_rate": 2.4447140967678057e-06,
+      "loss": 0.4991,
+      "step": 5773
+    },
+    {
+      "epoch": 1.6012201885745978,
+      "grad_norm": 0.20149755477905273,
+      "learning_rate": 2.4414349080591064e-06,
+      "loss": 0.4864,
+      "step": 5774
+    },
+    {
+      "epoch": 1.6014975041597337,
+      "grad_norm": 0.20722270011901855,
+      "learning_rate": 2.4381576820429588e-06,
+      "loss": 0.4994,
+      "step": 5775
+    },
+    {
+      "epoch": 1.6017748197448696,
+      "grad_norm": 0.21165424585342407,
+      "learning_rate": 2.434882419358826e-06,
+      "loss": 0.4968,
+      "step": 5776
+    },
+    {
+      "epoch": 1.6020521353300055,
+      "grad_norm": 0.2054947316646576,
+      "learning_rate": 2.4316091206458073e-06,
+      "loss": 0.48,
+      "step": 5777
+    },
+    {
+      "epoch": 1.6023294509151413,
+      "grad_norm": 0.2067498415708542,
+      "learning_rate": 2.428337786542603e-06,
+      "loss": 0.4804,
+      "step": 5778
+    },
+    {
+      "epoch": 1.6026067665002772,
+      "grad_norm": 0.19922077655792236,
+      "learning_rate": 2.42506841768754e-06,
+      "loss": 0.4946,
+      "step": 5779
+    },
+    {
+      "epoch": 1.602884082085413,
+      "grad_norm": 0.202239990234375,
+      "learning_rate": 2.4218010147185625e-06,
+      "loss": 0.4929,
+      "step": 5780
+    },
+    {
+      "epoch": 1.603161397670549,
+      "grad_norm": 0.2116711288690567,
+      "learning_rate": 2.4185355782732205e-06,
+      "loss": 0.4702,
+      "step": 5781
+    },
+    {
+      "epoch": 1.6034387132556849,
+      "grad_norm": 0.1949220895767212,
+      "learning_rate": 2.4152721089886933e-06,
+      "loss": 0.4611,
+      "step": 5782
+    },
+    {
+      "epoch": 1.6037160288408208,
+      "grad_norm": 0.20025965571403503,
+      "learning_rate": 2.412010607501765e-06,
+      "loss": 0.5017,
+      "step": 5783
+    },
+    {
+      "epoch": 1.6039933444259566,
+      "grad_norm": 0.22020164132118225,
+      "learning_rate": 2.4087510744488465e-06,
+      "loss": 0.4937,
+      "step": 5784
+    },
+    {
+      "epoch": 1.6042706600110925,
+      "grad_norm": 0.21315798163414001,
+      "learning_rate": 2.4054935104659533e-06,
+      "loss": 0.4847,
+      "step": 5785
+    },
+    {
+      "epoch": 1.6045479755962284,
+      "grad_norm": 0.22190795838832855,
+      "learning_rate": 2.4022379161887265e-06,
+      "loss": 0.4754,
+      "step": 5786
+    },
+    {
+      "epoch": 1.6048252911813643,
+      "grad_norm": 0.21109223365783691,
+      "learning_rate": 2.3989842922524154e-06,
+      "loss": 0.5015,
+      "step": 5787
+    },
+    {
+      "epoch": 1.6051026067665002,
+      "grad_norm": 0.21155421435832977,
+      "learning_rate": 2.3957326392918906e-06,
+      "loss": 0.4992,
+      "step": 5788
+    },
+    {
+      "epoch": 1.605379922351636,
+      "grad_norm": 0.2093130350112915,
+      "learning_rate": 2.392482957941637e-06,
+      "loss": 0.5007,
+      "step": 5789
+    },
+    {
+      "epoch": 1.605657237936772,
+      "grad_norm": 0.20884621143341064,
+      "learning_rate": 2.389235248835754e-06,
+      "loss": 0.5155,
+      "step": 5790
+    },
+    {
+      "epoch": 1.6059345535219078,
+      "grad_norm": 0.20897847414016724,
+      "learning_rate": 2.385989512607946e-06,
+      "loss": 0.5158,
+      "step": 5791
+    },
+    {
+      "epoch": 1.6062118691070437,
+      "grad_norm": 0.2077287882566452,
+      "learning_rate": 2.382745749891556e-06,
+      "loss": 0.4918,
+      "step": 5792
+    },
+    {
+      "epoch": 1.6064891846921796,
+      "grad_norm": 0.2046043872833252,
+      "learning_rate": 2.379503961319522e-06,
+      "loss": 0.4631,
+      "step": 5793
+    },
+    {
+      "epoch": 1.6067665002773155,
+      "grad_norm": 0.22556503117084503,
+      "learning_rate": 2.3762641475244e-06,
+      "loss": 0.4901,
+      "step": 5794
+    },
+    {
+      "epoch": 1.6070438158624514,
+      "grad_norm": 0.20669564604759216,
+      "learning_rate": 2.3730263091383654e-06,
+      "loss": 0.5145,
+      "step": 5795
+    },
+    {
+      "epoch": 1.6073211314475873,
+      "grad_norm": 0.20489123463630676,
+      "learning_rate": 2.3697904467932104e-06,
+      "loss": 0.49,
+      "step": 5796
+    },
+    {
+      "epoch": 1.6075984470327231,
+      "grad_norm": 0.2056889533996582,
+      "learning_rate": 2.366556561120334e-06,
+      "loss": 0.5178,
+      "step": 5797
+    },
+    {
+      "epoch": 1.607875762617859,
+      "grad_norm": 0.22281832993030548,
+      "learning_rate": 2.3633246527507507e-06,
+      "loss": 0.4962,
+      "step": 5798
+    },
+    {
+      "epoch": 1.608153078202995,
+      "grad_norm": 0.20517009496688843,
+      "learning_rate": 2.3600947223150926e-06,
+      "loss": 0.517,
+      "step": 5799
+    },
+    {
+      "epoch": 1.6084303937881308,
+      "grad_norm": 0.2156504988670349,
+      "learning_rate": 2.3568667704436096e-06,
+      "loss": 0.4901,
+      "step": 5800
+    },
+    {
+      "epoch": 1.6087077093732667,
+      "grad_norm": 0.20395740866661072,
+      "learning_rate": 2.3536407977661573e-06,
+      "loss": 0.4848,
+      "step": 5801
+    },
+    {
+      "epoch": 1.6089850249584026,
+      "grad_norm": 0.20527851581573486,
+      "learning_rate": 2.3504168049122006e-06,
+      "loss": 0.5139,
+      "step": 5802
+    },
+    {
+      "epoch": 1.6092623405435384,
+      "grad_norm": 0.20943453907966614,
+      "learning_rate": 2.34719479251084e-06,
+      "loss": 0.5047,
+      "step": 5803
+    },
+    {
+      "epoch": 1.6095396561286743,
+      "grad_norm": 0.19371294975280762,
+      "learning_rate": 2.34397476119077e-06,
+      "loss": 0.4913,
+      "step": 5804
+    },
+    {
+      "epoch": 1.6098169717138102,
+      "grad_norm": 0.20842455327510834,
+      "learning_rate": 2.3407567115802983e-06,
+      "loss": 0.5115,
+      "step": 5805
+    },
+    {
+      "epoch": 1.610094287298946,
+      "grad_norm": 0.21230155229568481,
+      "learning_rate": 2.337540644307358e-06,
+      "loss": 0.4878,
+      "step": 5806
+    },
+    {
+      "epoch": 1.610371602884082,
+      "grad_norm": 0.20435665547847748,
+      "learning_rate": 2.334326559999489e-06,
+      "loss": 0.5108,
+      "step": 5807
+    },
+    {
+      "epoch": 1.6106489184692179,
+      "grad_norm": 0.20020738244056702,
+      "learning_rate": 2.3311144592838425e-06,
+      "loss": 0.4754,
+      "step": 5808
+    },
+    {
+      "epoch": 1.6109262340543538,
+      "grad_norm": 0.20572920143604279,
+      "learning_rate": 2.327904342787182e-06,
+      "loss": 0.5035,
+      "step": 5809
+    },
+    {
+      "epoch": 1.6112035496394896,
+      "grad_norm": 0.2309185266494751,
+      "learning_rate": 2.324696211135889e-06,
+      "loss": 0.5065,
+      "step": 5810
+    },
+    {
+      "epoch": 1.6114808652246255,
+      "grad_norm": 0.22252123057842255,
+      "learning_rate": 2.3214900649559572e-06,
+      "loss": 0.5174,
+      "step": 5811
+    },
+    {
+      "epoch": 1.6117581808097614,
+      "grad_norm": 0.1993841528892517,
+      "learning_rate": 2.3182859048729856e-06,
+      "loss": 0.4873,
+      "step": 5812
+    },
+    {
+      "epoch": 1.6120354963948973,
+      "grad_norm": 0.2186397761106491,
+      "learning_rate": 2.3150837315121966e-06,
+      "loss": 0.5177,
+      "step": 5813
+    },
+    {
+      "epoch": 1.6123128119800332,
+      "grad_norm": 0.21579131484031677,
+      "learning_rate": 2.3118835454984126e-06,
+      "loss": 0.4963,
+      "step": 5814
+    },
+    {
+      "epoch": 1.612590127565169,
+      "grad_norm": 0.23498544096946716,
+      "learning_rate": 2.3086853474560814e-06,
+      "loss": 0.4963,
+      "step": 5815
+    },
+    {
+      "epoch": 1.612867443150305,
+      "grad_norm": 0.20521143078804016,
+      "learning_rate": 2.305489138009248e-06,
+      "loss": 0.4848,
+      "step": 5816
+    },
+    {
+      "epoch": 1.6131447587354408,
+      "grad_norm": 0.1934400200843811,
+      "learning_rate": 2.302294917781583e-06,
+      "loss": 0.4906,
+      "step": 5817
+    },
+    {
+      "epoch": 1.6134220743205767,
+      "grad_norm": 0.20979085564613342,
+      "learning_rate": 2.2991026873963676e-06,
+      "loss": 0.4736,
+      "step": 5818
+    },
+    {
+      "epoch": 1.6136993899057126,
+      "grad_norm": 0.2139722853899002,
+      "learning_rate": 2.295912447476481e-06,
+      "loss": 0.4949,
+      "step": 5819
+    },
+    {
+      "epoch": 1.6139767054908485,
+      "grad_norm": 0.21068985760211945,
+      "learning_rate": 2.2927241986444303e-06,
+      "loss": 0.4978,
+      "step": 5820
+    },
+    {
+      "epoch": 1.6142540210759844,
+      "grad_norm": 0.20163820683956146,
+      "learning_rate": 2.289537941522324e-06,
+      "loss": 0.5178,
+      "step": 5821
+    },
+    {
+      "epoch": 1.6145313366611203,
+      "grad_norm": 0.20904479920864105,
+      "learning_rate": 2.286353676731888e-06,
+      "loss": 0.5101,
+      "step": 5822
+    },
+    {
+      "epoch": 1.6148086522462561,
+      "grad_norm": 0.2054595649242401,
+      "learning_rate": 2.283171404894452e-06,
+      "loss": 0.4829,
+      "step": 5823
+    },
+    {
+      "epoch": 1.615085967831392,
+      "grad_norm": 0.21114782989025116,
+      "learning_rate": 2.279991126630969e-06,
+      "loss": 0.4995,
+      "step": 5824
+    },
+    {
+      "epoch": 1.615363283416528,
+      "grad_norm": 0.20004698634147644,
+      "learning_rate": 2.2768128425619858e-06,
+      "loss": 0.4823,
+      "step": 5825
+    },
+    {
+      "epoch": 1.6156405990016638,
+      "grad_norm": 0.21248266100883484,
+      "learning_rate": 2.273636553307677e-06,
+      "loss": 0.5148,
+      "step": 5826
+    },
+    {
+      "epoch": 1.6159179145867997,
+      "grad_norm": 0.21467168629169464,
+      "learning_rate": 2.2704622594878225e-06,
+      "loss": 0.4923,
+      "step": 5827
+    },
+    {
+      "epoch": 1.6161952301719356,
+      "grad_norm": 0.21232888102531433,
+      "learning_rate": 2.2672899617218065e-06,
+      "loss": 0.5014,
+      "step": 5828
+    },
+    {
+      "epoch": 1.6164725457570714,
+      "grad_norm": 0.2272992879152298,
+      "learning_rate": 2.2641196606286274e-06,
+      "loss": 0.5026,
+      "step": 5829
+    },
+    {
+      "epoch": 1.6167498613422073,
+      "grad_norm": 0.2556884288787842,
+      "learning_rate": 2.2609513568268958e-06,
+      "loss": 0.5019,
+      "step": 5830
+    },
+    {
+      "epoch": 1.6170271769273432,
+      "grad_norm": 0.2090773582458496,
+      "learning_rate": 2.257785050934838e-06,
+      "loss": 0.5294,
+      "step": 5831
+    },
+    {
+      "epoch": 1.617304492512479,
+      "grad_norm": 0.2042931616306305,
+      "learning_rate": 2.2546207435702738e-06,
+      "loss": 0.507,
+      "step": 5832
+    },
+    {
+      "epoch": 1.617581808097615,
+      "grad_norm": 0.21114638447761536,
+      "learning_rate": 2.2514584353506505e-06,
+      "loss": 0.4889,
+      "step": 5833
+    },
+    {
+      "epoch": 1.6178591236827509,
+      "grad_norm": 0.21107198297977448,
+      "learning_rate": 2.2482981268930183e-06,
+      "loss": 0.4965,
+      "step": 5834
+    },
+    {
+      "epoch": 1.6181364392678868,
+      "grad_norm": 0.22077830135822296,
+      "learning_rate": 2.2451398188140365e-06,
+      "loss": 0.4955,
+      "step": 5835
+    },
+    {
+      "epoch": 1.6184137548530226,
+      "grad_norm": 0.20404209196567535,
+      "learning_rate": 2.2419835117299682e-06,
+      "loss": 0.4862,
+      "step": 5836
+    },
+    {
+      "epoch": 1.6186910704381585,
+      "grad_norm": 0.2278210073709488,
+      "learning_rate": 2.2388292062567e-06,
+      "loss": 0.4899,
+      "step": 5837
+    },
+    {
+      "epoch": 1.6189683860232944,
+      "grad_norm": 0.20038627088069916,
+      "learning_rate": 2.2356769030097203e-06,
+      "loss": 0.4979,
+      "step": 5838
+    },
+    {
+      "epoch": 1.6192457016084303,
+      "grad_norm": 0.2179410755634308,
+      "learning_rate": 2.232526602604125e-06,
+      "loss": 0.4971,
+      "step": 5839
+    },
+    {
+      "epoch": 1.6195230171935662,
+      "grad_norm": 0.2113763988018036,
+      "learning_rate": 2.2293783056546156e-06,
+      "loss": 0.5079,
+      "step": 5840
+    },
+    {
+      "epoch": 1.619800332778702,
+      "grad_norm": 0.2023729383945465,
+      "learning_rate": 2.2262320127755184e-06,
+      "loss": 0.5081,
+      "step": 5841
+    },
+    {
+      "epoch": 1.620077648363838,
+      "grad_norm": 0.2042120099067688,
+      "learning_rate": 2.2230877245807553e-06,
+      "loss": 0.4965,
+      "step": 5842
+    },
+    {
+      "epoch": 1.6203549639489738,
+      "grad_norm": 0.21033427119255066,
+      "learning_rate": 2.2199454416838585e-06,
+      "loss": 0.4882,
+      "step": 5843
+    },
+    {
+      "epoch": 1.6206322795341097,
+      "grad_norm": 0.2538556456565857,
+      "learning_rate": 2.2168051646979647e-06,
+      "loss": 0.4803,
+      "step": 5844
+    },
+    {
+      "epoch": 1.6209095951192456,
+      "grad_norm": 0.20632927119731903,
+      "learning_rate": 2.2136668942358373e-06,
+      "loss": 0.4971,
+      "step": 5845
+    },
+    {
+      "epoch": 1.6211869107043815,
+      "grad_norm": 0.20020337402820587,
+      "learning_rate": 2.210530630909831e-06,
+      "loss": 0.4935,
+      "step": 5846
+    },
+    {
+      "epoch": 1.6214642262895174,
+      "grad_norm": 0.21856513619422913,
+      "learning_rate": 2.2073963753319076e-06,
+      "loss": 0.4761,
+      "step": 5847
+    },
+    {
+      "epoch": 1.6217415418746532,
+      "grad_norm": 0.2104158103466034,
+      "learning_rate": 2.2042641281136508e-06,
+      "loss": 0.4971,
+      "step": 5848
+    },
+    {
+      "epoch": 1.6220188574597891,
+      "grad_norm": 0.20630767941474915,
+      "learning_rate": 2.2011338898662458e-06,
+      "loss": 0.4878,
+      "step": 5849
+    },
+    {
+      "epoch": 1.622296173044925,
+      "grad_norm": 0.20518629252910614,
+      "learning_rate": 2.198005661200478e-06,
+      "loss": 0.5087,
+      "step": 5850
+    },
+    {
+      "epoch": 1.622573488630061,
+      "grad_norm": 0.25231507420539856,
+      "learning_rate": 2.1948794427267565e-06,
+      "loss": 0.514,
+      "step": 5851
+    },
+    {
+      "epoch": 1.6228508042151968,
+      "grad_norm": 0.21174240112304688,
+      "learning_rate": 2.1917552350550803e-06,
+      "loss": 0.4977,
+      "step": 5852
+    },
+    {
+      "epoch": 1.6231281198003327,
+      "grad_norm": 0.2215232104063034,
+      "learning_rate": 2.1886330387950737e-06,
+      "loss": 0.4799,
+      "step": 5853
+    },
+    {
+      "epoch": 1.6234054353854686,
+      "grad_norm": 0.2032269835472107,
+      "learning_rate": 2.1855128545559517e-06,
+      "loss": 0.4942,
+      "step": 5854
+    },
+    {
+      "epoch": 1.6236827509706044,
+      "grad_norm": 0.2067827433347702,
+      "learning_rate": 2.1823946829465496e-06,
+      "loss": 0.5164,
+      "step": 5855
+    },
+    {
+      "epoch": 1.6239600665557403,
+      "grad_norm": 0.19948525726795197,
+      "learning_rate": 2.1792785245753082e-06,
+      "loss": 0.4838,
+      "step": 5856
+    },
+    {
+      "epoch": 1.6242373821408762,
+      "grad_norm": 0.21956472098827362,
+      "learning_rate": 2.176164380050265e-06,
+      "loss": 0.4999,
+      "step": 5857
+    },
+    {
+      "epoch": 1.624514697726012,
+      "grad_norm": 0.20982100069522858,
+      "learning_rate": 2.1730522499790793e-06,
+      "loss": 0.4661,
+      "step": 5858
+    },
+    {
+      "epoch": 1.624792013311148,
+      "grad_norm": 0.20440787076950073,
+      "learning_rate": 2.169942134969004e-06,
+      "loss": 0.4907,
+      "step": 5859
+    },
+    {
+      "epoch": 1.6250693288962839,
+      "grad_norm": 0.21445617079734802,
+      "learning_rate": 2.1668340356269107e-06,
+      "loss": 0.5016,
+      "step": 5860
+    },
+    {
+      "epoch": 1.6253466444814197,
+      "grad_norm": 0.25135523080825806,
+      "learning_rate": 2.163727952559266e-06,
+      "loss": 0.4958,
+      "step": 5861
+    },
+    {
+      "epoch": 1.6256239600665556,
+      "grad_norm": 0.20553235709667206,
+      "learning_rate": 2.1606238863721568e-06,
+      "loss": 0.4758,
+      "step": 5862
+    },
+    {
+      "epoch": 1.6259012756516915,
+      "grad_norm": 0.1968127191066742,
+      "learning_rate": 2.157521837671259e-06,
+      "loss": 0.4919,
+      "step": 5863
+    },
+    {
+      "epoch": 1.6261785912368274,
+      "grad_norm": 0.28879934549331665,
+      "learning_rate": 2.1544218070618695e-06,
+      "loss": 0.4916,
+      "step": 5864
+    },
+    {
+      "epoch": 1.6264559068219633,
+      "grad_norm": 0.2082158625125885,
+      "learning_rate": 2.1513237951488907e-06,
+      "loss": 0.4969,
+      "step": 5865
+    },
+    {
+      "epoch": 1.6267332224070992,
+      "grad_norm": 0.20625808835029602,
+      "learning_rate": 2.1482278025368214e-06,
+      "loss": 0.4714,
+      "step": 5866
+    },
+    {
+      "epoch": 1.627010537992235,
+      "grad_norm": 0.2138909250497818,
+      "learning_rate": 2.1451338298297706e-06,
+      "loss": 0.4911,
+      "step": 5867
+    },
+    {
+      "epoch": 1.627287853577371,
+      "grad_norm": 0.20718157291412354,
+      "learning_rate": 2.1420418776314565e-06,
+      "loss": 0.4918,
+      "step": 5868
+    },
+    {
+      "epoch": 1.6275651691625068,
+      "grad_norm": 0.2110764980316162,
+      "learning_rate": 2.1389519465452035e-06,
+      "loss": 0.4946,
+      "step": 5869
+    },
+    {
+      "epoch": 1.6278424847476427,
+      "grad_norm": 0.21094083786010742,
+      "learning_rate": 2.135864037173933e-06,
+      "loss": 0.4799,
+      "step": 5870
+    },
+    {
+      "epoch": 1.6281198003327786,
+      "grad_norm": 0.20451469719409943,
+      "learning_rate": 2.132778150120182e-06,
+      "loss": 0.4805,
+      "step": 5871
+    },
+    {
+      "epoch": 1.6283971159179145,
+      "grad_norm": 0.20834794640541077,
+      "learning_rate": 2.129694285986092e-06,
+      "loss": 0.5009,
+      "step": 5872
+    },
+    {
+      "epoch": 1.6286744315030504,
+      "grad_norm": 0.20821170508861542,
+      "learning_rate": 2.126612445373402e-06,
+      "loss": 0.5267,
+      "step": 5873
+    },
+    {
+      "epoch": 1.6289517470881862,
+      "grad_norm": 0.2115286886692047,
+      "learning_rate": 2.1235326288834595e-06,
+      "loss": 0.5084,
+      "step": 5874
+    },
+    {
+      "epoch": 1.6292290626733221,
+      "grad_norm": 0.2055329829454422,
+      "learning_rate": 2.1204548371172194e-06,
+      "loss": 0.4826,
+      "step": 5875
+    },
+    {
+      "epoch": 1.629506378258458,
+      "grad_norm": 0.20218388736248016,
+      "learning_rate": 2.117379070675245e-06,
+      "loss": 0.5188,
+      "step": 5876
+    },
+    {
+      "epoch": 1.629783693843594,
+      "grad_norm": 0.22304154932498932,
+      "learning_rate": 2.1143053301576954e-06,
+      "loss": 0.5008,
+      "step": 5877
+    },
+    {
+      "epoch": 1.6300610094287298,
+      "grad_norm": 0.1911652535200119,
+      "learning_rate": 2.1112336161643347e-06,
+      "loss": 0.4933,
+      "step": 5878
+    },
+    {
+      "epoch": 1.6303383250138657,
+      "grad_norm": 0.20941410958766937,
+      "learning_rate": 2.108163929294546e-06,
+      "loss": 0.5065,
+      "step": 5879
+    },
+    {
+      "epoch": 1.6306156405990015,
+      "grad_norm": 0.20953522622585297,
+      "learning_rate": 2.1050962701473014e-06,
+      "loss": 0.4801,
+      "step": 5880
+    },
+    {
+      "epoch": 1.6308929561841374,
+      "grad_norm": 0.2114289551973343,
+      "learning_rate": 2.102030639321183e-06,
+      "loss": 0.4981,
+      "step": 5881
+    },
+    {
+      "epoch": 1.6311702717692733,
+      "grad_norm": 0.21842321753501892,
+      "learning_rate": 2.0989670374143693e-06,
+      "loss": 0.5037,
+      "step": 5882
+    },
+    {
+      "epoch": 1.6314475873544092,
+      "grad_norm": 0.2003999501466751,
+      "learning_rate": 2.0959054650246626e-06,
+      "loss": 0.4975,
+      "step": 5883
+    },
+    {
+      "epoch": 1.631724902939545,
+      "grad_norm": 0.21246321499347687,
+      "learning_rate": 2.0928459227494505e-06,
+      "loss": 0.4951,
+      "step": 5884
+    },
+    {
+      "epoch": 1.632002218524681,
+      "grad_norm": 0.1984589397907257,
+      "learning_rate": 2.0897884111857292e-06,
+      "loss": 0.4977,
+      "step": 5885
+    },
+    {
+      "epoch": 1.6322795341098169,
+      "grad_norm": 0.20242911577224731,
+      "learning_rate": 2.086732930930102e-06,
+      "loss": 0.4695,
+      "step": 5886
+    },
+    {
+      "epoch": 1.6325568496949527,
+      "grad_norm": 0.2037610560655594,
+      "learning_rate": 2.0836794825787763e-06,
+      "loss": 0.4774,
+      "step": 5887
+    },
+    {
+      "epoch": 1.6328341652800886,
+      "grad_norm": 0.23012907803058624,
+      "learning_rate": 2.0806280667275594e-06,
+      "loss": 0.4965,
+      "step": 5888
+    },
+    {
+      "epoch": 1.6331114808652245,
+      "grad_norm": 0.21143199503421783,
+      "learning_rate": 2.0775786839718607e-06,
+      "loss": 0.4985,
+      "step": 5889
+    },
+    {
+      "epoch": 1.6333887964503604,
+      "grad_norm": 0.20616495609283447,
+      "learning_rate": 2.074531334906696e-06,
+      "loss": 0.5273,
+      "step": 5890
+    },
+    {
+      "epoch": 1.6336661120354963,
+      "grad_norm": 0.2136596441268921,
+      "learning_rate": 2.0714860201266895e-06,
+      "loss": 0.4846,
+      "step": 5891
+    },
+    {
+      "epoch": 1.6339434276206322,
+      "grad_norm": 0.22074300050735474,
+      "learning_rate": 2.068442740226055e-06,
+      "loss": 0.5065,
+      "step": 5892
+    },
+    {
+      "epoch": 1.634220743205768,
+      "grad_norm": 0.20592841506004333,
+      "learning_rate": 2.065401495798622e-06,
+      "loss": 0.4774,
+      "step": 5893
+    },
+    {
+      "epoch": 1.634498058790904,
+      "grad_norm": 0.20553089678287506,
+      "learning_rate": 2.062362287437818e-06,
+      "loss": 0.4521,
+      "step": 5894
+    },
+    {
+      "epoch": 1.6347753743760398,
+      "grad_norm": 0.21127671003341675,
+      "learning_rate": 2.0593251157366727e-06,
+      "loss": 0.5082,
+      "step": 5895
+    },
+    {
+      "epoch": 1.6350526899611757,
+      "grad_norm": 0.20608097314834595,
+      "learning_rate": 2.056289981287815e-06,
+      "loss": 0.4999,
+      "step": 5896
+    },
+    {
+      "epoch": 1.6353300055463116,
+      "grad_norm": 0.20263780653476715,
+      "learning_rate": 2.0532568846834825e-06,
+      "loss": 0.469,
+      "step": 5897
+    },
+    {
+      "epoch": 1.6356073211314475,
+      "grad_norm": 0.2072770744562149,
+      "learning_rate": 2.050225826515516e-06,
+      "loss": 0.476,
+      "step": 5898
+    },
+    {
+      "epoch": 1.6358846367165834,
+      "grad_norm": 0.20764364302158356,
+      "learning_rate": 2.0471968073753483e-06,
+      "loss": 0.4831,
+      "step": 5899
+    },
+    {
+      "epoch": 1.6361619523017192,
+      "grad_norm": 0.20213009417057037,
+      "learning_rate": 2.0441698278540296e-06,
+      "loss": 0.4584,
+      "step": 5900
+    },
+    {
+      "epoch": 1.6364392678868551,
+      "grad_norm": 0.19651676714420319,
+      "learning_rate": 2.041144888542196e-06,
+      "loss": 0.4947,
+      "step": 5901
+    },
+    {
+      "epoch": 1.636716583471991,
+      "grad_norm": 0.20257742702960968,
+      "learning_rate": 2.0381219900300993e-06,
+      "loss": 0.5024,
+      "step": 5902
+    },
+    {
+      "epoch": 1.6369938990571269,
+      "grad_norm": 0.20740802586078644,
+      "learning_rate": 2.0351011329075823e-06,
+      "loss": 0.5088,
+      "step": 5903
+    },
+    {
+      "epoch": 1.6372712146422628,
+      "grad_norm": 0.21403908729553223,
+      "learning_rate": 2.0320823177641e-06,
+      "loss": 0.4876,
+      "step": 5904
+    },
+    {
+      "epoch": 1.6375485302273987,
+      "grad_norm": 0.2182864248752594,
+      "learning_rate": 2.0290655451886965e-06,
+      "loss": 0.4772,
+      "step": 5905
+    },
+    {
+      "epoch": 1.6378258458125345,
+      "grad_norm": 0.21262075006961823,
+      "learning_rate": 2.0260508157700266e-06,
+      "loss": 0.4797,
+      "step": 5906
+    },
+    {
+      "epoch": 1.6381031613976704,
+      "grad_norm": 0.21089348196983337,
+      "learning_rate": 2.023038130096347e-06,
+      "loss": 0.4934,
+      "step": 5907
+    },
+    {
+      "epoch": 1.6383804769828063,
+      "grad_norm": 0.2085971236228943,
+      "learning_rate": 2.020027488755509e-06,
+      "loss": 0.4815,
+      "step": 5908
+    },
+    {
+      "epoch": 1.6386577925679422,
+      "grad_norm": 0.2214841991662979,
+      "learning_rate": 2.017018892334971e-06,
+      "loss": 0.5057,
+      "step": 5909
+    },
+    {
+      "epoch": 1.638935108153078,
+      "grad_norm": 0.2010078728199005,
+      "learning_rate": 2.0140123414217867e-06,
+      "loss": 0.4997,
+      "step": 5910
+    },
+    {
+      "epoch": 1.639212423738214,
+      "grad_norm": 0.21052546799182892,
+      "learning_rate": 2.0110078366026173e-06,
+      "loss": 0.4891,
+      "step": 5911
+    },
+    {
+      "epoch": 1.6394897393233498,
+      "grad_norm": 0.2032831907272339,
+      "learning_rate": 2.008005378463716e-06,
+      "loss": 0.5115,
+      "step": 5912
+    },
+    {
+      "epoch": 1.6397670549084857,
+      "grad_norm": 0.20649638772010803,
+      "learning_rate": 2.0050049675909467e-06,
+      "loss": 0.4735,
+      "step": 5913
+    },
+    {
+      "epoch": 1.6400443704936216,
+      "grad_norm": 0.20801763236522675,
+      "learning_rate": 2.0020066045697714e-06,
+      "loss": 0.4761,
+      "step": 5914
+    },
+    {
+      "epoch": 1.6403216860787575,
+      "grad_norm": 0.19369389116764069,
+      "learning_rate": 1.999010289985247e-06,
+      "loss": 0.5107,
+      "step": 5915
+    },
+    {
+      "epoch": 1.6405990016638934,
+      "grad_norm": 0.20587508380413055,
+      "learning_rate": 1.9960160244220263e-06,
+      "loss": 0.4925,
+      "step": 5916
+    },
+    {
+      "epoch": 1.6408763172490293,
+      "grad_norm": 0.20227836072444916,
+      "learning_rate": 1.993023808464382e-06,
+      "loss": 0.5093,
+      "step": 5917
+    },
+    {
+      "epoch": 1.6411536328341652,
+      "grad_norm": 0.18990160524845123,
+      "learning_rate": 1.990033642696172e-06,
+      "loss": 0.4542,
+      "step": 5918
+    },
+    {
+      "epoch": 1.641430948419301,
+      "grad_norm": 0.20094534754753113,
+      "learning_rate": 1.9870455277008536e-06,
+      "loss": 0.4991,
+      "step": 5919
+    },
+    {
+      "epoch": 1.641708264004437,
+      "grad_norm": 0.22286152839660645,
+      "learning_rate": 1.9840594640614816e-06,
+      "loss": 0.5301,
+      "step": 5920
+    },
+    {
+      "epoch": 1.6419855795895728,
+      "grad_norm": 0.19467279314994812,
+      "learning_rate": 1.9810754523607296e-06,
+      "loss": 0.4759,
+      "step": 5921
+    },
+    {
+      "epoch": 1.6422628951747087,
+      "grad_norm": 0.21738329529762268,
+      "learning_rate": 1.9780934931808506e-06,
+      "loss": 0.4728,
+      "step": 5922
+    },
+    {
+      "epoch": 1.6425402107598446,
+      "grad_norm": 0.21225155889987946,
+      "learning_rate": 1.9751135871036995e-06,
+      "loss": 0.4957,
+      "step": 5923
+    },
+    {
+      "epoch": 1.6428175263449805,
+      "grad_norm": 0.20787325501441956,
+      "learning_rate": 1.9721357347107406e-06,
+      "loss": 0.4905,
+      "step": 5924
+    },
+    {
+      "epoch": 1.6430948419301163,
+      "grad_norm": 0.21133233606815338,
+      "learning_rate": 1.9691599365830328e-06,
+      "loss": 0.4878,
+      "step": 5925
+    },
+    {
+      "epoch": 1.6433721575152522,
+      "grad_norm": 0.19063884019851685,
+      "learning_rate": 1.966186193301231e-06,
+      "loss": 0.4926,
+      "step": 5926
+    },
+    {
+      "epoch": 1.6436494731003881,
+      "grad_norm": 0.2176242172718048,
+      "learning_rate": 1.9632145054455873e-06,
+      "loss": 0.4622,
+      "step": 5927
+    },
+    {
+      "epoch": 1.643926788685524,
+      "grad_norm": 0.2013779729604721,
+      "learning_rate": 1.960244873595961e-06,
+      "loss": 0.4814,
+      "step": 5928
+    },
+    {
+      "epoch": 1.6442041042706599,
+      "grad_norm": 0.21530881524085999,
+      "learning_rate": 1.957277298331808e-06,
+      "loss": 0.5058,
+      "step": 5929
+    },
+    {
+      "epoch": 1.6444814198557958,
+      "grad_norm": 0.20872762799263,
+      "learning_rate": 1.9543117802321764e-06,
+      "loss": 0.4875,
+      "step": 5930
+    },
+    {
+      "epoch": 1.6447587354409317,
+      "grad_norm": 0.21378107368946075,
+      "learning_rate": 1.9513483198757176e-06,
+      "loss": 0.4703,
+      "step": 5931
+    },
+    {
+      "epoch": 1.6450360510260675,
+      "grad_norm": 0.22160130739212036,
+      "learning_rate": 1.9483869178406875e-06,
+      "loss": 0.4991,
+      "step": 5932
+    },
+    {
+      "epoch": 1.6453133666112034,
+      "grad_norm": 0.21556591987609863,
+      "learning_rate": 1.945427574704928e-06,
+      "loss": 0.5146,
+      "step": 5933
+    },
+    {
+      "epoch": 1.6455906821963393,
+      "grad_norm": 0.2028489112854004,
+      "learning_rate": 1.9424702910458837e-06,
+      "loss": 0.5054,
+      "step": 5934
+    },
+    {
+      "epoch": 1.6458679977814752,
+      "grad_norm": 0.2313559353351593,
+      "learning_rate": 1.939515067440603e-06,
+      "loss": 0.5166,
+      "step": 5935
+    },
+    {
+      "epoch": 1.646145313366611,
+      "grad_norm": 0.20201341807842255,
+      "learning_rate": 1.9365619044657306e-06,
+      "loss": 0.478,
+      "step": 5936
+    },
+    {
+      "epoch": 1.646422628951747,
+      "grad_norm": 0.20596764981746674,
+      "learning_rate": 1.9336108026975e-06,
+      "loss": 0.4794,
+      "step": 5937
+    },
+    {
+      "epoch": 1.6466999445368828,
+      "grad_norm": 0.20622918009757996,
+      "learning_rate": 1.9306617627117567e-06,
+      "loss": 0.4991,
+      "step": 5938
+    },
+    {
+      "epoch": 1.6469772601220187,
+      "grad_norm": 0.20188166201114655,
+      "learning_rate": 1.927714785083928e-06,
+      "loss": 0.5167,
+      "step": 5939
+    },
+    {
+      "epoch": 1.6472545757071546,
+      "grad_norm": 0.2264070063829422,
+      "learning_rate": 1.9247698703890566e-06,
+      "loss": 0.5025,
+      "step": 5940
+    },
+    {
+      "epoch": 1.6475318912922905,
+      "grad_norm": 0.20895101130008698,
+      "learning_rate": 1.921827019201766e-06,
+      "loss": 0.5183,
+      "step": 5941
+    },
+    {
+      "epoch": 1.6478092068774264,
+      "grad_norm": 0.2241641879081726,
+      "learning_rate": 1.91888623209629e-06,
+      "loss": 0.4804,
+      "step": 5942
+    },
+    {
+      "epoch": 1.6480865224625623,
+      "grad_norm": 0.1995384395122528,
+      "learning_rate": 1.9159475096464484e-06,
+      "loss": 0.4742,
+      "step": 5943
+    },
+    {
+      "epoch": 1.6483638380476981,
+      "grad_norm": 0.213314950466156,
+      "learning_rate": 1.9130108524256672e-06,
+      "loss": 0.4683,
+      "step": 5944
+    },
+    {
+      "epoch": 1.648641153632834,
+      "grad_norm": 0.220163494348526,
+      "learning_rate": 1.9100762610069684e-06,
+      "loss": 0.5043,
+      "step": 5945
+    },
+    {
+      "epoch": 1.64891846921797,
+      "grad_norm": 0.2084074467420578,
+      "learning_rate": 1.907143735962963e-06,
+      "loss": 0.4717,
+      "step": 5946
+    },
+    {
+      "epoch": 1.6491957848031058,
+      "grad_norm": 0.21064861118793488,
+      "learning_rate": 1.9042132778658698e-06,
+      "loss": 0.4935,
+      "step": 5947
+    },
+    {
+      "epoch": 1.6494731003882417,
+      "grad_norm": 0.20477135479450226,
+      "learning_rate": 1.9012848872874938e-06,
+      "loss": 0.4861,
+      "step": 5948
+    },
+    {
+      "epoch": 1.6497504159733776,
+      "grad_norm": 0.21649906039237976,
+      "learning_rate": 1.8983585647992463e-06,
+      "loss": 0.4822,
+      "step": 5949
+    },
+    {
+      "epoch": 1.6500277315585135,
+      "grad_norm": 0.21001774072647095,
+      "learning_rate": 1.8954343109721245e-06,
+      "loss": 0.4892,
+      "step": 5950
+    },
+    {
+      "epoch": 1.6503050471436493,
+      "grad_norm": 0.19450335204601288,
+      "learning_rate": 1.8925121263767317e-06,
+      "loss": 0.4973,
+      "step": 5951
+    },
+    {
+      "epoch": 1.6505823627287852,
+      "grad_norm": 0.21450437605381012,
+      "learning_rate": 1.8895920115832675e-06,
+      "loss": 0.4925,
+      "step": 5952
+    },
+    {
+      "epoch": 1.650859678313921,
+      "grad_norm": 0.21557024121284485,
+      "learning_rate": 1.8866739671615175e-06,
+      "loss": 0.4697,
+      "step": 5953
+    },
+    {
+      "epoch": 1.651136993899057,
+      "grad_norm": 0.21621811389923096,
+      "learning_rate": 1.8837579936808695e-06,
+      "loss": 0.4718,
+      "step": 5954
+    },
+    {
+      "epoch": 1.6514143094841929,
+      "grad_norm": 0.20242778956890106,
+      "learning_rate": 1.8808440917103085e-06,
+      "loss": 0.4827,
+      "step": 5955
+    },
+    {
+      "epoch": 1.6516916250693288,
+      "grad_norm": 0.21023069322109222,
+      "learning_rate": 1.8779322618184173e-06,
+      "loss": 0.514,
+      "step": 5956
+    },
+    {
+      "epoch": 1.6519689406544646,
+      "grad_norm": 0.20898282527923584,
+      "learning_rate": 1.8750225045733678e-06,
+      "loss": 0.5226,
+      "step": 5957
+    },
+    {
+      "epoch": 1.6522462562396005,
+      "grad_norm": 0.21227650344371796,
+      "learning_rate": 1.872114820542925e-06,
+      "loss": 0.5026,
+      "step": 5958
+    },
+    {
+      "epoch": 1.6525235718247364,
+      "grad_norm": 0.21009580790996552,
+      "learning_rate": 1.8692092102944674e-06,
+      "loss": 0.4846,
+      "step": 5959
+    },
+    {
+      "epoch": 1.6528008874098723,
+      "grad_norm": 0.2119254618883133,
+      "learning_rate": 1.8663056743949512e-06,
+      "loss": 0.4779,
+      "step": 5960
+    },
+    {
+      "epoch": 1.6530782029950082,
+      "grad_norm": 0.21257485449314117,
+      "learning_rate": 1.8634042134109285e-06,
+      "loss": 0.5017,
+      "step": 5961
+    },
+    {
+      "epoch": 1.653355518580144,
+      "grad_norm": 0.216257706284523,
+      "learning_rate": 1.860504827908556e-06,
+      "loss": 0.5239,
+      "step": 5962
+    },
+    {
+      "epoch": 1.65363283416528,
+      "grad_norm": 0.20020192861557007,
+      "learning_rate": 1.8576075184535815e-06,
+      "loss": 0.5074,
+      "step": 5963
+    },
+    {
+      "epoch": 1.6539101497504158,
+      "grad_norm": 0.21256963908672333,
+      "learning_rate": 1.8547122856113458e-06,
+      "loss": 0.5388,
+      "step": 5964
+    },
+    {
+      "epoch": 1.6541874653355517,
+      "grad_norm": 0.21079207956790924,
+      "learning_rate": 1.8518191299467815e-06,
+      "loss": 0.5012,
+      "step": 5965
+    },
+    {
+      "epoch": 1.6544647809206876,
+      "grad_norm": 0.21106182038784027,
+      "learning_rate": 1.8489280520244235e-06,
+      "loss": 0.4973,
+      "step": 5966
+    },
+    {
+      "epoch": 1.6547420965058235,
+      "grad_norm": 0.2153145670890808,
+      "learning_rate": 1.8460390524083992e-06,
+      "loss": 0.5018,
+      "step": 5967
+    },
+    {
+      "epoch": 1.6550194120909594,
+      "grad_norm": 0.2027408331632614,
+      "learning_rate": 1.843152131662429e-06,
+      "loss": 0.4777,
+      "step": 5968
+    },
+    {
+      "epoch": 1.6552967276760953,
+      "grad_norm": 0.2015599012374878,
+      "learning_rate": 1.8402672903498188e-06,
+      "loss": 0.4992,
+      "step": 5969
+    },
+    {
+      "epoch": 1.6555740432612311,
+      "grad_norm": 0.24570131301879883,
+      "learning_rate": 1.8373845290334896e-06,
+      "loss": 0.5061,
+      "step": 5970
+    },
+    {
+      "epoch": 1.655851358846367,
+      "grad_norm": 0.21255463361740112,
+      "learning_rate": 1.834503848275941e-06,
+      "loss": 0.4945,
+      "step": 5971
+    },
+    {
+      "epoch": 1.656128674431503,
+      "grad_norm": 0.21138375997543335,
+      "learning_rate": 1.8316252486392654e-06,
+      "loss": 0.4811,
+      "step": 5972
+    },
+    {
+      "epoch": 1.6564059900166388,
+      "grad_norm": 0.2128453552722931,
+      "learning_rate": 1.8287487306851564e-06,
+      "loss": 0.4978,
+      "step": 5973
+    },
+    {
+      "epoch": 1.6566833056017747,
+      "grad_norm": 0.20760896801948547,
+      "learning_rate": 1.8258742949749024e-06,
+      "loss": 0.4547,
+      "step": 5974
+    },
+    {
+      "epoch": 1.6569606211869106,
+      "grad_norm": 0.2109159529209137,
+      "learning_rate": 1.8230019420693758e-06,
+      "loss": 0.4891,
+      "step": 5975
+    },
+    {
+      "epoch": 1.6572379367720464,
+      "grad_norm": 0.21184983849525452,
+      "learning_rate": 1.820131672529056e-06,
+      "loss": 0.4846,
+      "step": 5976
+    },
+    {
+      "epoch": 1.6575152523571823,
+      "grad_norm": 0.20627717673778534,
+      "learning_rate": 1.8172634869140018e-06,
+      "loss": 0.4988,
+      "step": 5977
+    },
+    {
+      "epoch": 1.6577925679423182,
+      "grad_norm": 0.1983606368303299,
+      "learning_rate": 1.8143973857838768e-06,
+      "loss": 0.4799,
+      "step": 5978
+    },
+    {
+      "epoch": 1.658069883527454,
+      "grad_norm": 0.2059786021709442,
+      "learning_rate": 1.8115333696979293e-06,
+      "loss": 0.4843,
+      "step": 5979
+    },
+    {
+      "epoch": 1.65834719911259,
+      "grad_norm": 0.20751167833805084,
+      "learning_rate": 1.8086714392150096e-06,
+      "loss": 0.4832,
+      "step": 5980
+    },
+    {
+      "epoch": 1.6586245146977259,
+      "grad_norm": 0.20422019064426422,
+      "learning_rate": 1.8058115948935514e-06,
+      "loss": 0.488,
+      "step": 5981
+    },
+    {
+      "epoch": 1.6589018302828618,
+      "grad_norm": 0.2069856822490692,
+      "learning_rate": 1.8029538372915878e-06,
+      "loss": 0.5008,
+      "step": 5982
+    },
+    {
+      "epoch": 1.6591791458679976,
+      "grad_norm": 0.21730239689350128,
+      "learning_rate": 1.800098166966746e-06,
+      "loss": 0.4896,
+      "step": 5983
+    },
+    {
+      "epoch": 1.6594564614531335,
+      "grad_norm": 0.20635366439819336,
+      "learning_rate": 1.7972445844762376e-06,
+      "loss": 0.4924,
+      "step": 5984
+    },
+    {
+      "epoch": 1.6597337770382694,
+      "grad_norm": 0.21054959297180176,
+      "learning_rate": 1.794393090376878e-06,
+      "loss": 0.4737,
+      "step": 5985
+    },
+    {
+      "epoch": 1.6600110926234053,
+      "grad_norm": 0.22848398983478546,
+      "learning_rate": 1.7915436852250625e-06,
+      "loss": 0.5273,
+      "step": 5986
+    },
+    {
+      "epoch": 1.6602884082085412,
+      "grad_norm": 0.20154784619808197,
+      "learning_rate": 1.7886963695767921e-06,
+      "loss": 0.487,
+      "step": 5987
+    },
+    {
+      "epoch": 1.660565723793677,
+      "grad_norm": 0.213270902633667,
+      "learning_rate": 1.7858511439876491e-06,
+      "loss": 0.4962,
+      "step": 5988
+    },
+    {
+      "epoch": 1.660843039378813,
+      "grad_norm": 0.21345144510269165,
+      "learning_rate": 1.7830080090128127e-06,
+      "loss": 0.5156,
+      "step": 5989
+    },
+    {
+      "epoch": 1.6611203549639488,
+      "grad_norm": 0.1976012885570526,
+      "learning_rate": 1.7801669652070585e-06,
+      "loss": 0.4916,
+      "step": 5990
+    },
+    {
+      "epoch": 1.6613976705490847,
+      "grad_norm": 0.21380779147148132,
+      "learning_rate": 1.7773280131247461e-06,
+      "loss": 0.5073,
+      "step": 5991
+    },
+    {
+      "epoch": 1.6616749861342206,
+      "grad_norm": 0.20816531777381897,
+      "learning_rate": 1.7744911533198266e-06,
+      "loss": 0.4932,
+      "step": 5992
+    },
+    {
+      "epoch": 1.6619523017193565,
+      "grad_norm": 0.21241043508052826,
+      "learning_rate": 1.771656386345852e-06,
+      "loss": 0.4989,
+      "step": 5993
+    },
+    {
+      "epoch": 1.6622296173044924,
+      "grad_norm": 0.2513413727283478,
+      "learning_rate": 1.76882371275596e-06,
+      "loss": 0.494,
+      "step": 5994
+    },
+    {
+      "epoch": 1.6625069328896283,
+      "grad_norm": 0.20039229094982147,
+      "learning_rate": 1.7659931331028777e-06,
+      "loss": 0.5074,
+      "step": 5995
+    },
+    {
+      "epoch": 1.6627842484747641,
+      "grad_norm": 0.1997574418783188,
+      "learning_rate": 1.7631646479389224e-06,
+      "loss": 0.5249,
+      "step": 5996
+    },
+    {
+      "epoch": 1.6630615640599,
+      "grad_norm": 0.20580746233463287,
+      "learning_rate": 1.7603382578160174e-06,
+      "loss": 0.5056,
+      "step": 5997
+    },
+    {
+      "epoch": 1.663338879645036,
+      "grad_norm": 0.21222086250782013,
+      "learning_rate": 1.7575139632856604e-06,
+      "loss": 0.4943,
+      "step": 5998
+    },
+    {
+      "epoch": 1.6636161952301718,
+      "grad_norm": 0.21751612424850464,
+      "learning_rate": 1.7546917648989428e-06,
+      "loss": 0.4876,
+      "step": 5999
+    },
+    {
+      "epoch": 1.6638935108153077,
+      "grad_norm": 0.20124460756778717,
+      "learning_rate": 1.7518716632065544e-06,
+      "loss": 0.4757,
+      "step": 6000
+    },
+    {
+      "epoch": 1.6641708264004436,
+      "grad_norm": 0.20955534279346466,
+      "learning_rate": 1.7490536587587716e-06,
+      "loss": 0.5009,
+      "step": 6001
+    },
+    {
+      "epoch": 1.6644481419855794,
+      "grad_norm": 0.21060660481452942,
+      "learning_rate": 1.7462377521054633e-06,
+      "loss": 0.5176,
+      "step": 6002
+    },
+    {
+      "epoch": 1.6647254575707153,
+      "grad_norm": 0.2049666792154312,
+      "learning_rate": 1.7434239437960797e-06,
+      "loss": 0.5014,
+      "step": 6003
+    },
+    {
+      "epoch": 1.6650027731558512,
+      "grad_norm": 0.2187904566526413,
+      "learning_rate": 1.7406122343796766e-06,
+      "loss": 0.4969,
+      "step": 6004
+    },
+    {
+      "epoch": 1.665280088740987,
+      "grad_norm": 0.20527362823486328,
+      "learning_rate": 1.737802624404894e-06,
+      "loss": 0.469,
+      "step": 6005
+    },
+    {
+      "epoch": 1.665557404326123,
+      "grad_norm": 0.21435390412807465,
+      "learning_rate": 1.7349951144199572e-06,
+      "loss": 0.5112,
+      "step": 6006
+    },
+    {
+      "epoch": 1.6658347199112589,
+      "grad_norm": 0.2521720826625824,
+      "learning_rate": 1.732189704972685e-06,
+      "loss": 0.5026,
+      "step": 6007
+    },
+    {
+      "epoch": 1.6661120354963947,
+      "grad_norm": 0.21735185384750366,
+      "learning_rate": 1.7293863966104898e-06,
+      "loss": 0.5284,
+      "step": 6008
+    },
+    {
+      "epoch": 1.6663893510815306,
+      "grad_norm": 0.20416045188903809,
+      "learning_rate": 1.7265851898803725e-06,
+      "loss": 0.4943,
+      "step": 6009
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.19750310480594635,
+      "learning_rate": 1.7237860853289183e-06,
+      "loss": 0.4829,
+      "step": 6010
+    },
+    {
+      "epoch": 1.6669439822518024,
+      "grad_norm": 0.21233929693698883,
+      "learning_rate": 1.7209890835023086e-06,
+      "loss": 0.4907,
+      "step": 6011
+    },
+    {
+      "epoch": 1.6672212978369383,
+      "grad_norm": 0.20838162302970886,
+      "learning_rate": 1.7181941849463173e-06,
+      "loss": 0.4875,
+      "step": 6012
+    },
+    {
+      "epoch": 1.6674986134220742,
+      "grad_norm": 0.21194593608379364,
+      "learning_rate": 1.7154013902062977e-06,
+      "loss": 0.4902,
+      "step": 6013
+    },
+    {
+      "epoch": 1.66777592900721,
+      "grad_norm": 0.208678737282753,
+      "learning_rate": 1.7126106998271968e-06,
+      "loss": 0.4778,
+      "step": 6014
+    },
+    {
+      "epoch": 1.668053244592346,
+      "grad_norm": 0.2077416628599167,
+      "learning_rate": 1.7098221143535557e-06,
+      "loss": 0.4905,
+      "step": 6015
+    },
+    {
+      "epoch": 1.6683305601774818,
+      "grad_norm": 0.21134360134601593,
+      "learning_rate": 1.7070356343295026e-06,
+      "loss": 0.4539,
+      "step": 6016
+    },
+    {
+      "epoch": 1.6686078757626177,
+      "grad_norm": 0.21801994740962982,
+      "learning_rate": 1.7042512602987489e-06,
+      "loss": 0.4891,
+      "step": 6017
+    },
+    {
+      "epoch": 1.6688851913477536,
+      "grad_norm": 0.21269913017749786,
+      "learning_rate": 1.7014689928046056e-06,
+      "loss": 0.4759,
+      "step": 6018
+    },
+    {
+      "epoch": 1.6691625069328895,
+      "grad_norm": 0.21309658885002136,
+      "learning_rate": 1.6986888323899594e-06,
+      "loss": 0.477,
+      "step": 6019
+    },
+    {
+      "epoch": 1.6694398225180254,
+      "grad_norm": 0.21096886694431305,
+      "learning_rate": 1.6959107795973011e-06,
+      "loss": 0.5187,
+      "step": 6020
+    },
+    {
+      "epoch": 1.6697171381031612,
+      "grad_norm": 0.20901210606098175,
+      "learning_rate": 1.693134834968696e-06,
+      "loss": 0.4737,
+      "step": 6021
+    },
+    {
+      "epoch": 1.6699944536882971,
+      "grad_norm": 0.21308885514736176,
+      "learning_rate": 1.6903609990458063e-06,
+      "loss": 0.4976,
+      "step": 6022
+    },
+    {
+      "epoch": 1.670271769273433,
+      "grad_norm": 0.21009613573551178,
+      "learning_rate": 1.6875892723698855e-06,
+      "loss": 0.5045,
+      "step": 6023
+    },
+    {
+      "epoch": 1.670549084858569,
+      "grad_norm": 0.2145100235939026,
+      "learning_rate": 1.6848196554817633e-06,
+      "loss": 0.4772,
+      "step": 6024
+    },
+    {
+      "epoch": 1.6708264004437048,
+      "grad_norm": 0.21058829128742218,
+      "learning_rate": 1.6820521489218728e-06,
+      "loss": 0.5044,
+      "step": 6025
+    },
+    {
+      "epoch": 1.6711037160288407,
+      "grad_norm": 0.21063855290412903,
+      "learning_rate": 1.6792867532302207e-06,
+      "loss": 0.5194,
+      "step": 6026
+    },
+    {
+      "epoch": 1.6713810316139766,
+      "grad_norm": 0.21397702395915985,
+      "learning_rate": 1.6765234689464157e-06,
+      "loss": 0.4903,
+      "step": 6027
+    },
+    {
+      "epoch": 1.6716583471991124,
+      "grad_norm": 0.20812270045280457,
+      "learning_rate": 1.6737622966096405e-06,
+      "loss": 0.4869,
+      "step": 6028
+    },
+    {
+      "epoch": 1.6719356627842483,
+      "grad_norm": 0.19779033958911896,
+      "learning_rate": 1.671003236758681e-06,
+      "loss": 0.4884,
+      "step": 6029
+    },
+    {
+      "epoch": 1.6722129783693842,
+      "grad_norm": 0.2069951891899109,
+      "learning_rate": 1.6682462899318962e-06,
+      "loss": 0.5154,
+      "step": 6030
+    },
+    {
+      "epoch": 1.67249029395452,
+      "grad_norm": 0.20514649152755737,
+      "learning_rate": 1.665491456667241e-06,
+      "loss": 0.509,
+      "step": 6031
+    },
+    {
+      "epoch": 1.672767609539656,
+      "grad_norm": 0.20265156030654907,
+      "learning_rate": 1.6627387375022605e-06,
+      "loss": 0.5008,
+      "step": 6032
+    },
+    {
+      "epoch": 1.6730449251247919,
+      "grad_norm": 0.20257985591888428,
+      "learning_rate": 1.6599881329740817e-06,
+      "loss": 0.4914,
+      "step": 6033
+    },
+    {
+      "epoch": 1.6733222407099277,
+      "grad_norm": 0.19979064166545868,
+      "learning_rate": 1.657239643619414e-06,
+      "loss": 0.477,
+      "step": 6034
+    },
+    {
+      "epoch": 1.6735995562950636,
+      "grad_norm": 0.20505757629871368,
+      "learning_rate": 1.6544932699745663e-06,
+      "loss": 0.4945,
+      "step": 6035
+    },
+    {
+      "epoch": 1.6738768718801995,
+      "grad_norm": 0.21343204379081726,
+      "learning_rate": 1.6517490125754307e-06,
+      "loss": 0.4913,
+      "step": 6036
+    },
+    {
+      "epoch": 1.6741541874653354,
+      "grad_norm": 0.20039796829223633,
+      "learning_rate": 1.6490068719574787e-06,
+      "loss": 0.4874,
+      "step": 6037
+    },
+    {
+      "epoch": 1.6744315030504713,
+      "grad_norm": 0.3097711503505707,
+      "learning_rate": 1.646266848655778e-06,
+      "loss": 0.464,
+      "step": 6038
+    },
+    {
+      "epoch": 1.6747088186356072,
+      "grad_norm": 0.21286475658416748,
+      "learning_rate": 1.6435289432049818e-06,
+      "loss": 0.4889,
+      "step": 6039
+    },
+    {
+      "epoch": 1.674986134220743,
+      "grad_norm": 0.2054137885570526,
+      "learning_rate": 1.6407931561393253e-06,
+      "loss": 0.5036,
+      "step": 6040
+    },
+    {
+      "epoch": 1.675263449805879,
+      "grad_norm": 0.20179954171180725,
+      "learning_rate": 1.638059487992631e-06,
+      "loss": 0.5136,
+      "step": 6041
+    },
+    {
+      "epoch": 1.6755407653910148,
+      "grad_norm": 0.21015551686286926,
+      "learning_rate": 1.6353279392983117e-06,
+      "loss": 0.4785,
+      "step": 6042
+    },
+    {
+      "epoch": 1.6758180809761507,
+      "grad_norm": 0.2488449513912201,
+      "learning_rate": 1.63259851058937e-06,
+      "loss": 0.4889,
+      "step": 6043
+    },
+    {
+      "epoch": 1.6760953965612866,
+      "grad_norm": 0.20739984512329102,
+      "learning_rate": 1.6298712023983837e-06,
+      "loss": 0.4849,
+      "step": 6044
+    },
+    {
+      "epoch": 1.6763727121464225,
+      "grad_norm": 0.21058540046215057,
+      "learning_rate": 1.627146015257522e-06,
+      "loss": 0.522,
+      "step": 6045
+    },
+    {
+      "epoch": 1.6766500277315584,
+      "grad_norm": 0.1949351280927658,
+      "learning_rate": 1.6244229496985426e-06,
+      "loss": 0.4794,
+      "step": 6046
+    },
+    {
+      "epoch": 1.6769273433166942,
+      "grad_norm": 0.21308818459510803,
+      "learning_rate": 1.6217020062527927e-06,
+      "loss": 0.4913,
+      "step": 6047
+    },
+    {
+      "epoch": 1.6772046589018301,
+      "grad_norm": 0.20299099385738373,
+      "learning_rate": 1.6189831854511937e-06,
+      "loss": 0.4901,
+      "step": 6048
+    },
+    {
+      "epoch": 1.677481974486966,
+      "grad_norm": 0.21231001615524292,
+      "learning_rate": 1.616266487824261e-06,
+      "loss": 0.4885,
+      "step": 6049
+    },
+    {
+      "epoch": 1.677759290072102,
+      "grad_norm": 0.20558898150920868,
+      "learning_rate": 1.6135519139021005e-06,
+      "loss": 0.5032,
+      "step": 6050
+    },
+    {
+      "epoch": 1.6780366056572378,
+      "grad_norm": 0.2060549259185791,
+      "learning_rate": 1.6108394642143907e-06,
+      "loss": 0.5098,
+      "step": 6051
+    },
+    {
+      "epoch": 1.6783139212423737,
+      "grad_norm": 0.20870618522167206,
+      "learning_rate": 1.6081291392904027e-06,
+      "loss": 0.4791,
+      "step": 6052
+    },
+    {
+      "epoch": 1.6785912368275095,
+      "grad_norm": 0.20835687220096588,
+      "learning_rate": 1.6054209396589929e-06,
+      "loss": 0.4865,
+      "step": 6053
+    },
+    {
+      "epoch": 1.6788685524126454,
+      "grad_norm": 0.2111913561820984,
+      "learning_rate": 1.6027148658486077e-06,
+      "loss": 0.4962,
+      "step": 6054
+    },
+    {
+      "epoch": 1.6791458679977813,
+      "grad_norm": 0.21386343240737915,
+      "learning_rate": 1.600010918387268e-06,
+      "loss": 0.4999,
+      "step": 6055
+    },
+    {
+      "epoch": 1.6794231835829172,
+      "grad_norm": 0.20479026436805725,
+      "learning_rate": 1.5973090978025906e-06,
+      "loss": 0.4723,
+      "step": 6056
+    },
+    {
+      "epoch": 1.679700499168053,
+      "grad_norm": 0.20431126654148102,
+      "learning_rate": 1.5946094046217664e-06,
+      "loss": 0.4792,
+      "step": 6057
+    },
+    {
+      "epoch": 1.679977814753189,
+      "grad_norm": 0.21851181983947754,
+      "learning_rate": 1.5919118393715834e-06,
+      "loss": 0.4986,
+      "step": 6058
+    },
+    {
+      "epoch": 1.6802551303383249,
+      "grad_norm": 0.3242267668247223,
+      "learning_rate": 1.5892164025784015e-06,
+      "loss": 0.5032,
+      "step": 6059
+    },
+    {
+      "epoch": 1.6805324459234607,
+      "grad_norm": 0.2177019715309143,
+      "learning_rate": 1.5865230947681762e-06,
+      "loss": 0.4851,
+      "step": 6060
+    },
+    {
+      "epoch": 1.6808097615085966,
+      "grad_norm": 0.20672672986984253,
+      "learning_rate": 1.5838319164664438e-06,
+      "loss": 0.4711,
+      "step": 6061
+    },
+    {
+      "epoch": 1.6810870770937325,
+      "grad_norm": 0.1982792168855667,
+      "learning_rate": 1.581142868198321e-06,
+      "loss": 0.466,
+      "step": 6062
+    },
+    {
+      "epoch": 1.6813643926788684,
+      "grad_norm": 0.20962117612361908,
+      "learning_rate": 1.5784559504885166e-06,
+      "loss": 0.5069,
+      "step": 6063
+    },
+    {
+      "epoch": 1.6816417082640043,
+      "grad_norm": 0.2054154872894287,
+      "learning_rate": 1.5757711638613143e-06,
+      "loss": 0.5134,
+      "step": 6064
+    },
+    {
+      "epoch": 1.6819190238491402,
+      "grad_norm": 0.19651010632514954,
+      "learning_rate": 1.5730885088405922e-06,
+      "loss": 0.4967,
+      "step": 6065
+    },
+    {
+      "epoch": 1.682196339434276,
+      "grad_norm": 0.20710954070091248,
+      "learning_rate": 1.570407985949804e-06,
+      "loss": 0.4772,
+      "step": 6066
+    },
+    {
+      "epoch": 1.682473655019412,
+      "grad_norm": 0.2049458771944046,
+      "learning_rate": 1.5677295957119934e-06,
+      "loss": 0.4758,
+      "step": 6067
+    },
+    {
+      "epoch": 1.6827509706045478,
+      "grad_norm": 0.21262648701667786,
+      "learning_rate": 1.5650533386497801e-06,
+      "loss": 0.5058,
+      "step": 6068
+    },
+    {
+      "epoch": 1.6830282861896837,
+      "grad_norm": 0.20999044179916382,
+      "learning_rate": 1.5623792152853783e-06,
+      "loss": 0.5048,
+      "step": 6069
+    },
+    {
+      "epoch": 1.6833056017748196,
+      "grad_norm": 0.21131351590156555,
+      "learning_rate": 1.559707226140579e-06,
+      "loss": 0.4729,
+      "step": 6070
+    },
+    {
+      "epoch": 1.6835829173599555,
+      "grad_norm": 0.22507424652576447,
+      "learning_rate": 1.5570373717367594e-06,
+      "loss": 0.4815,
+      "step": 6071
+    },
+    {
+      "epoch": 1.6838602329450914,
+      "grad_norm": 0.2068501114845276,
+      "learning_rate": 1.5543696525948726e-06,
+      "loss": 0.5036,
+      "step": 6072
+    },
+    {
+      "epoch": 1.6841375485302272,
+      "grad_norm": 0.20177114009857178,
+      "learning_rate": 1.551704069235467e-06,
+      "loss": 0.4747,
+      "step": 6073
+    },
+    {
+      "epoch": 1.6844148641153631,
+      "grad_norm": 0.20704385638237,
+      "learning_rate": 1.5490406221786686e-06,
+      "loss": 0.4812,
+      "step": 6074
+    },
+    {
+      "epoch": 1.684692179700499,
+      "grad_norm": 0.21657253801822662,
+      "learning_rate": 1.5463793119441835e-06,
+      "loss": 0.4854,
+      "step": 6075
+    },
+    {
+      "epoch": 1.6849694952856349,
+      "grad_norm": 0.2125793993473053,
+      "learning_rate": 1.543720139051305e-06,
+      "loss": 0.4749,
+      "step": 6076
+    },
+    {
+      "epoch": 1.6852468108707708,
+      "grad_norm": 0.20819686353206635,
+      "learning_rate": 1.541063104018911e-06,
+      "loss": 0.5036,
+      "step": 6077
+    },
+    {
+      "epoch": 1.6855241264559067,
+      "grad_norm": 0.21309512853622437,
+      "learning_rate": 1.5384082073654564e-06,
+      "loss": 0.5012,
+      "step": 6078
+    },
+    {
+      "epoch": 1.6858014420410425,
+      "grad_norm": 0.20586422085762024,
+      "learning_rate": 1.5357554496089805e-06,
+      "loss": 0.4803,
+      "step": 6079
+    },
+    {
+      "epoch": 1.6860787576261784,
+      "grad_norm": 0.2012133002281189,
+      "learning_rate": 1.5331048312671085e-06,
+      "loss": 0.4771,
+      "step": 6080
+    },
+    {
+      "epoch": 1.6863560732113143,
+      "grad_norm": 0.2082866132259369,
+      "learning_rate": 1.5304563528570488e-06,
+      "loss": 0.4793,
+      "step": 6081
+    },
+    {
+      "epoch": 1.6866333887964502,
+      "grad_norm": 0.21011273562908173,
+      "learning_rate": 1.5278100148955865e-06,
+      "loss": 0.4703,
+      "step": 6082
+    },
+    {
+      "epoch": 1.686910704381586,
+      "grad_norm": 0.20084689557552338,
+      "learning_rate": 1.5251658178990908e-06,
+      "loss": 0.4969,
+      "step": 6083
+    },
+    {
+      "epoch": 1.687188019966722,
+      "grad_norm": 0.22017468512058258,
+      "learning_rate": 1.5225237623835167e-06,
+      "loss": 0.4982,
+      "step": 6084
+    },
+    {
+      "epoch": 1.6874653355518578,
+      "grad_norm": 0.21381786465644836,
+      "learning_rate": 1.5198838488644036e-06,
+      "loss": 0.4626,
+      "step": 6085
+    },
+    {
+      "epoch": 1.687742651136994,
+      "grad_norm": 0.2041536420583725,
+      "learning_rate": 1.5172460778568626e-06,
+      "loss": 0.4923,
+      "step": 6086
+    },
+    {
+      "epoch": 1.6880199667221298,
+      "grad_norm": 0.2171899676322937,
+      "learning_rate": 1.5146104498755891e-06,
+      "loss": 0.5064,
+      "step": 6087
+    },
+    {
+      "epoch": 1.6882972823072657,
+      "grad_norm": 0.217244952917099,
+      "learning_rate": 1.5119769654348748e-06,
+      "loss": 0.5021,
+      "step": 6088
+    },
+    {
+      "epoch": 1.6885745978924016,
+      "grad_norm": 0.20813634991645813,
+      "learning_rate": 1.5093456250485764e-06,
+      "loss": 0.4623,
+      "step": 6089
+    },
+    {
+      "epoch": 1.6888519134775375,
+      "grad_norm": 0.2243288904428482,
+      "learning_rate": 1.5067164292301358e-06,
+      "loss": 0.5251,
+      "step": 6090
+    },
+    {
+      "epoch": 1.6891292290626734,
+      "grad_norm": 0.21137182414531708,
+      "learning_rate": 1.504089378492582e-06,
+      "loss": 0.5257,
+      "step": 6091
+    },
+    {
+      "epoch": 1.6894065446478093,
+      "grad_norm": 0.20429569482803345,
+      "learning_rate": 1.501464473348524e-06,
+      "loss": 0.4699,
+      "step": 6092
+    },
+    {
+      "epoch": 1.6896838602329451,
+      "grad_norm": 0.20688970386981964,
+      "learning_rate": 1.498841714310148e-06,
+      "loss": 0.5066,
+      "step": 6093
+    },
+    {
+      "epoch": 1.689961175818081,
+      "grad_norm": 0.20542113482952118,
+      "learning_rate": 1.496221101889221e-06,
+      "loss": 0.4753,
+      "step": 6094
+    },
+    {
+      "epoch": 1.690238491403217,
+      "grad_norm": 0.20209279656410217,
+      "learning_rate": 1.4936026365970968e-06,
+      "loss": 0.5048,
+      "step": 6095
+    },
+    {
+      "epoch": 1.6905158069883528,
+      "grad_norm": 0.2007710188627243,
+      "learning_rate": 1.4909863189447093e-06,
+      "loss": 0.4646,
+      "step": 6096
+    },
+    {
+      "epoch": 1.6907931225734887,
+      "grad_norm": 0.19921134412288666,
+      "learning_rate": 1.488372149442567e-06,
+      "loss": 0.5055,
+      "step": 6097
+    },
+    {
+      "epoch": 1.6910704381586246,
+      "grad_norm": 0.2226364016532898,
+      "learning_rate": 1.485760128600769e-06,
+      "loss": 0.4918,
+      "step": 6098
+    },
+    {
+      "epoch": 1.6913477537437605,
+      "grad_norm": 0.21037879586219788,
+      "learning_rate": 1.4831502569289834e-06,
+      "loss": 0.4959,
+      "step": 6099
+    },
+    {
+      "epoch": 1.6916250693288963,
+      "grad_norm": 0.20453481376171112,
+      "learning_rate": 1.4805425349364716e-06,
+      "loss": 0.5035,
+      "step": 6100
+    },
+    {
+      "epoch": 1.6919023849140322,
+      "grad_norm": 0.2033025324344635,
+      "learning_rate": 1.4779369631320637e-06,
+      "loss": 0.4661,
+      "step": 6101
+    },
+    {
+      "epoch": 1.692179700499168,
+      "grad_norm": 0.20237600803375244,
+      "learning_rate": 1.475333542024178e-06,
+      "loss": 0.4769,
+      "step": 6102
+    },
+    {
+      "epoch": 1.692457016084304,
+      "grad_norm": 0.20004519820213318,
+      "learning_rate": 1.4727322721208136e-06,
+      "loss": 0.4957,
+      "step": 6103
+    },
+    {
+      "epoch": 1.6927343316694399,
+      "grad_norm": 0.2055417001247406,
+      "learning_rate": 1.4701331539295426e-06,
+      "loss": 0.4975,
+      "step": 6104
+    },
+    {
+      "epoch": 1.6930116472545758,
+      "grad_norm": 0.21751341223716736,
+      "learning_rate": 1.4675361879575271e-06,
+      "loss": 0.4675,
+      "step": 6105
+    },
+    {
+      "epoch": 1.6932889628397116,
+      "grad_norm": 0.21179702877998352,
+      "learning_rate": 1.4649413747114982e-06,
+      "loss": 0.5041,
+      "step": 6106
+    },
+    {
+      "epoch": 1.6935662784248475,
+      "grad_norm": 0.20675987005233765,
+      "learning_rate": 1.4623487146977754e-06,
+      "loss": 0.5094,
+      "step": 6107
+    },
+    {
+      "epoch": 1.6938435940099834,
+      "grad_norm": 0.20587822794914246,
+      "learning_rate": 1.4597582084222571e-06,
+      "loss": 0.5026,
+      "step": 6108
+    },
+    {
+      "epoch": 1.6941209095951193,
+      "grad_norm": 0.19946245849132538,
+      "learning_rate": 1.4571698563904196e-06,
+      "loss": 0.4706,
+      "step": 6109
+    },
+    {
+      "epoch": 1.6943982251802552,
+      "grad_norm": 0.2001856118440628,
+      "learning_rate": 1.4545836591073129e-06,
+      "loss": 0.5065,
+      "step": 6110
+    },
+    {
+      "epoch": 1.694675540765391,
+      "grad_norm": 0.21339738368988037,
+      "learning_rate": 1.4519996170775791e-06,
+      "loss": 0.5259,
+      "step": 6111
+    },
+    {
+      "epoch": 1.694952856350527,
+      "grad_norm": 0.2282184213399887,
+      "learning_rate": 1.4494177308054315e-06,
+      "loss": 0.4954,
+      "step": 6112
+    },
+    {
+      "epoch": 1.6952301719356628,
+      "grad_norm": 0.2113092988729477,
+      "learning_rate": 1.4468380007946633e-06,
+      "loss": 0.4853,
+      "step": 6113
+    },
+    {
+      "epoch": 1.6955074875207987,
+      "grad_norm": 0.20249804854393005,
+      "learning_rate": 1.4442604275486493e-06,
+      "loss": 0.4781,
+      "step": 6114
+    },
+    {
+      "epoch": 1.6957848031059346,
+      "grad_norm": 0.21009276807308197,
+      "learning_rate": 1.4416850115703442e-06,
+      "loss": 0.5062,
+      "step": 6115
+    },
+    {
+      "epoch": 1.6960621186910705,
+      "grad_norm": 0.21459093689918518,
+      "learning_rate": 1.439111753362278e-06,
+      "loss": 0.4612,
+      "step": 6116
+    },
+    {
+      "epoch": 1.6963394342762064,
+      "grad_norm": 0.20765627920627594,
+      "learning_rate": 1.4365406534265587e-06,
+      "loss": 0.4704,
+      "step": 6117
+    },
+    {
+      "epoch": 1.6966167498613423,
+      "grad_norm": 0.20445479452610016,
+      "learning_rate": 1.4339717122648797e-06,
+      "loss": 0.4637,
+      "step": 6118
+    },
+    {
+      "epoch": 1.6968940654464781,
+      "grad_norm": 0.19960498809814453,
+      "learning_rate": 1.431404930378509e-06,
+      "loss": 0.4579,
+      "step": 6119
+    },
+    {
+      "epoch": 1.697171381031614,
+      "grad_norm": 0.22552312910556793,
+      "learning_rate": 1.428840308268295e-06,
+      "loss": 0.4842,
+      "step": 6120
+    },
+    {
+      "epoch": 1.69744869661675,
+      "grad_norm": 0.2081850916147232,
+      "learning_rate": 1.4262778464346593e-06,
+      "loss": 0.4912,
+      "step": 6121
+    },
+    {
+      "epoch": 1.6977260122018858,
+      "grad_norm": 0.22280938923358917,
+      "learning_rate": 1.4237175453776077e-06,
+      "loss": 0.492,
+      "step": 6122
+    },
+    {
+      "epoch": 1.6980033277870217,
+      "grad_norm": 0.20667922496795654,
+      "learning_rate": 1.4211594055967252e-06,
+      "loss": 0.5025,
+      "step": 6123
+    },
+    {
+      "epoch": 1.6982806433721576,
+      "grad_norm": 0.20905707776546478,
+      "learning_rate": 1.4186034275911726e-06,
+      "loss": 0.4897,
+      "step": 6124
+    },
+    {
+      "epoch": 1.6985579589572934,
+      "grad_norm": 0.22128935158252716,
+      "learning_rate": 1.4160496118596823e-06,
+      "loss": 0.5004,
+      "step": 6125
+    },
+    {
+      "epoch": 1.6988352745424293,
+      "grad_norm": 0.225121408700943,
+      "learning_rate": 1.413497958900581e-06,
+      "loss": 0.4908,
+      "step": 6126
+    },
+    {
+      "epoch": 1.6991125901275652,
+      "grad_norm": 0.21258188784122467,
+      "learning_rate": 1.4109484692117592e-06,
+      "loss": 0.5058,
+      "step": 6127
+    },
+    {
+      "epoch": 1.699389905712701,
+      "grad_norm": 0.21206620335578918,
+      "learning_rate": 1.408401143290687e-06,
+      "loss": 0.5058,
+      "step": 6128
+    },
+    {
+      "epoch": 1.699667221297837,
+      "grad_norm": 0.21642054617404938,
+      "learning_rate": 1.4058559816344186e-06,
+      "loss": 0.47,
+      "step": 6129
+    },
+    {
+      "epoch": 1.6999445368829729,
+      "grad_norm": 0.20300422608852386,
+      "learning_rate": 1.403312984739584e-06,
+      "loss": 0.492,
+      "step": 6130
+    },
+    {
+      "epoch": 1.7002218524681088,
+      "grad_norm": 0.20064608752727509,
+      "learning_rate": 1.400772153102388e-06,
+      "loss": 0.4729,
+      "step": 6131
+    },
+    {
+      "epoch": 1.7004991680532446,
+      "grad_norm": 0.22698557376861572,
+      "learning_rate": 1.3982334872186101e-06,
+      "loss": 0.4651,
+      "step": 6132
+    },
+    {
+      "epoch": 1.7007764836383805,
+      "grad_norm": 0.21828146278858185,
+      "learning_rate": 1.3956969875836155e-06,
+      "loss": 0.4998,
+      "step": 6133
+    },
+    {
+      "epoch": 1.7010537992235164,
+      "grad_norm": 0.20743338763713837,
+      "learning_rate": 1.3931626546923426e-06,
+      "loss": 0.5038,
+      "step": 6134
+    },
+    {
+      "epoch": 1.7013311148086523,
+      "grad_norm": 0.19973209500312805,
+      "learning_rate": 1.3906304890393047e-06,
+      "loss": 0.475,
+      "step": 6135
+    },
+    {
+      "epoch": 1.7016084303937882,
+      "grad_norm": 0.20409800112247467,
+      "learning_rate": 1.3881004911185976e-06,
+      "loss": 0.4745,
+      "step": 6136
+    },
+    {
+      "epoch": 1.701885745978924,
+      "grad_norm": 0.21515069901943207,
+      "learning_rate": 1.3855726614238868e-06,
+      "loss": 0.5089,
+      "step": 6137
+    },
+    {
+      "epoch": 1.70216306156406,
+      "grad_norm": 0.20320101082324982,
+      "learning_rate": 1.383047000448423e-06,
+      "loss": 0.4882,
+      "step": 6138
+    },
+    {
+      "epoch": 1.7024403771491958,
+      "grad_norm": 0.20085285604000092,
+      "learning_rate": 1.3805235086850249e-06,
+      "loss": 0.4702,
+      "step": 6139
+    },
+    {
+      "epoch": 1.7027176927343317,
+      "grad_norm": 0.2055083066225052,
+      "learning_rate": 1.3780021866260955e-06,
+      "loss": 0.4807,
+      "step": 6140
+    },
+    {
+      "epoch": 1.7029950083194676,
+      "grad_norm": 0.21570569276809692,
+      "learning_rate": 1.3754830347636138e-06,
+      "loss": 0.4849,
+      "step": 6141
+    },
+    {
+      "epoch": 1.7032723239046035,
+      "grad_norm": 0.20811456441879272,
+      "learning_rate": 1.3729660535891282e-06,
+      "loss": 0.4693,
+      "step": 6142
+    },
+    {
+      "epoch": 1.7035496394897394,
+      "grad_norm": 0.2096460610628128,
+      "learning_rate": 1.3704512435937734e-06,
+      "loss": 0.5394,
+      "step": 6143
+    },
+    {
+      "epoch": 1.7038269550748752,
+      "grad_norm": 0.20185941457748413,
+      "learning_rate": 1.3679386052682499e-06,
+      "loss": 0.5063,
+      "step": 6144
+    },
+    {
+      "epoch": 1.7041042706600111,
+      "grad_norm": 0.26697927713394165,
+      "learning_rate": 1.365428139102845e-06,
+      "loss": 0.5086,
+      "step": 6145
+    },
+    {
+      "epoch": 1.704381586245147,
+      "grad_norm": 0.21113193035125732,
+      "learning_rate": 1.362919845587414e-06,
+      "loss": 0.5187,
+      "step": 6146
+    },
+    {
+      "epoch": 1.704658901830283,
+      "grad_norm": 0.21000999212265015,
+      "learning_rate": 1.360413725211393e-06,
+      "loss": 0.4919,
+      "step": 6147
+    },
+    {
+      "epoch": 1.7049362174154188,
+      "grad_norm": 0.21595942974090576,
+      "learning_rate": 1.3579097784637908e-06,
+      "loss": 0.4821,
+      "step": 6148
+    },
+    {
+      "epoch": 1.7052135330005547,
+      "grad_norm": 0.20245753228664398,
+      "learning_rate": 1.3554080058331947e-06,
+      "loss": 0.4946,
+      "step": 6149
+    },
+    {
+      "epoch": 1.7054908485856906,
+      "grad_norm": 0.19732367992401123,
+      "learning_rate": 1.3529084078077695e-06,
+      "loss": 0.4822,
+      "step": 6150
+    },
+    {
+      "epoch": 1.7057681641708264,
+      "grad_norm": 0.2121272087097168,
+      "learning_rate": 1.3504109848752485e-06,
+      "loss": 0.5231,
+      "step": 6151
+    },
+    {
+      "epoch": 1.7060454797559623,
+      "grad_norm": 0.21215039491653442,
+      "learning_rate": 1.3479157375229493e-06,
+      "loss": 0.4893,
+      "step": 6152
+    },
+    {
+      "epoch": 1.7063227953410982,
+      "grad_norm": 0.20783814787864685,
+      "learning_rate": 1.3454226662377555e-06,
+      "loss": 0.4933,
+      "step": 6153
+    },
+    {
+      "epoch": 1.706600110926234,
+      "grad_norm": 0.20154152810573578,
+      "learning_rate": 1.3429317715061367e-06,
+      "loss": 0.4944,
+      "step": 6154
+    },
+    {
+      "epoch": 1.70687742651137,
+      "grad_norm": 0.20301756262779236,
+      "learning_rate": 1.340443053814129e-06,
+      "loss": 0.4592,
+      "step": 6155
+    },
+    {
+      "epoch": 1.7071547420965059,
+      "grad_norm": 0.20453648269176483,
+      "learning_rate": 1.3379565136473482e-06,
+      "loss": 0.4897,
+      "step": 6156
+    },
+    {
+      "epoch": 1.7074320576816417,
+      "grad_norm": 0.21424759924411774,
+      "learning_rate": 1.3354721514909865e-06,
+      "loss": 0.4816,
+      "step": 6157
+    },
+    {
+      "epoch": 1.7077093732667776,
+      "grad_norm": 0.20235998928546906,
+      "learning_rate": 1.3329899678298063e-06,
+      "loss": 0.4907,
+      "step": 6158
+    },
+    {
+      "epoch": 1.7079866888519135,
+      "grad_norm": 0.20267461240291595,
+      "learning_rate": 1.3305099631481453e-06,
+      "loss": 0.4633,
+      "step": 6159
+    },
+    {
+      "epoch": 1.7082640044370494,
+      "grad_norm": 0.21311825513839722,
+      "learning_rate": 1.3280321379299215e-06,
+      "loss": 0.4865,
+      "step": 6160
+    },
+    {
+      "epoch": 1.7085413200221853,
+      "grad_norm": 0.2097369134426117,
+      "learning_rate": 1.325556492658625e-06,
+      "loss": 0.5052,
+      "step": 6161
+    },
+    {
+      "epoch": 1.7088186356073212,
+      "grad_norm": 0.2084108293056488,
+      "learning_rate": 1.3230830278173178e-06,
+      "loss": 0.496,
+      "step": 6162
+    },
+    {
+      "epoch": 1.709095951192457,
+      "grad_norm": 0.20339664816856384,
+      "learning_rate": 1.3206117438886333e-06,
+      "loss": 0.5049,
+      "step": 6163
+    },
+    {
+      "epoch": 1.709373266777593,
+      "grad_norm": 0.2198762446641922,
+      "learning_rate": 1.3181426413547955e-06,
+      "loss": 0.4874,
+      "step": 6164
+    },
+    {
+      "epoch": 1.7096505823627288,
+      "grad_norm": 0.21083226799964905,
+      "learning_rate": 1.3156757206975873e-06,
+      "loss": 0.4878,
+      "step": 6165
+    },
+    {
+      "epoch": 1.7099278979478647,
+      "grad_norm": 0.23222039639949799,
+      "learning_rate": 1.313210982398365e-06,
+      "loss": 0.4773,
+      "step": 6166
+    },
+    {
+      "epoch": 1.7102052135330006,
+      "grad_norm": 0.20829123258590698,
+      "learning_rate": 1.3107484269380688e-06,
+      "loss": 0.4823,
+      "step": 6167
+    },
+    {
+      "epoch": 1.7104825291181365,
+      "grad_norm": 0.20589673519134521,
+      "learning_rate": 1.3082880547972104e-06,
+      "loss": 0.4466,
+      "step": 6168
+    },
+    {
+      "epoch": 1.7107598447032724,
+      "grad_norm": 0.21213486790657043,
+      "learning_rate": 1.3058298664558725e-06,
+      "loss": 0.4807,
+      "step": 6169
+    },
+    {
+      "epoch": 1.7110371602884082,
+      "grad_norm": 0.2168436497449875,
+      "learning_rate": 1.3033738623937072e-06,
+      "loss": 0.4891,
+      "step": 6170
+    },
+    {
+      "epoch": 1.7113144758735441,
+      "grad_norm": 0.1989026516675949,
+      "learning_rate": 1.300920043089951e-06,
+      "loss": 0.5083,
+      "step": 6171
+    },
+    {
+      "epoch": 1.71159179145868,
+      "grad_norm": 0.19995690882205963,
+      "learning_rate": 1.2984684090234122e-06,
+      "loss": 0.4747,
+      "step": 6172
+    },
+    {
+      "epoch": 1.711869107043816,
+      "grad_norm": 0.20262081921100616,
+      "learning_rate": 1.2960189606724613e-06,
+      "loss": 0.4834,
+      "step": 6173
+    },
+    {
+      "epoch": 1.7121464226289518,
+      "grad_norm": 0.22770211100578308,
+      "learning_rate": 1.2935716985150587e-06,
+      "loss": 0.5197,
+      "step": 6174
+    },
+    {
+      "epoch": 1.7124237382140877,
+      "grad_norm": 0.21298766136169434,
+      "learning_rate": 1.2911266230287239e-06,
+      "loss": 0.5022,
+      "step": 6175
+    },
+    {
+      "epoch": 1.7127010537992235,
+      "grad_norm": 0.21138055622577667,
+      "learning_rate": 1.2886837346905615e-06,
+      "loss": 0.472,
+      "step": 6176
+    },
+    {
+      "epoch": 1.7129783693843594,
+      "grad_norm": 0.20885500311851501,
+      "learning_rate": 1.2862430339772372e-06,
+      "loss": 0.4925,
+      "step": 6177
+    },
+    {
+      "epoch": 1.7132556849694953,
+      "grad_norm": 0.2066037952899933,
+      "learning_rate": 1.2838045213650008e-06,
+      "loss": 0.5073,
+      "step": 6178
+    },
+    {
+      "epoch": 1.7135330005546312,
+      "grad_norm": 0.2201787531375885,
+      "learning_rate": 1.2813681973296714e-06,
+      "loss": 0.4771,
+      "step": 6179
+    },
+    {
+      "epoch": 1.713810316139767,
+      "grad_norm": 0.20949430763721466,
+      "learning_rate": 1.2789340623466358e-06,
+      "loss": 0.4767,
+      "step": 6180
+    },
+    {
+      "epoch": 1.714087631724903,
+      "grad_norm": 0.20830868184566498,
+      "learning_rate": 1.276502116890864e-06,
+      "loss": 0.4684,
+      "step": 6181
+    },
+    {
+      "epoch": 1.7143649473100389,
+      "grad_norm": 0.197309672832489,
+      "learning_rate": 1.2740723614368886e-06,
+      "loss": 0.4686,
+      "step": 6182
+    },
+    {
+      "epoch": 1.7146422628951747,
+      "grad_norm": 0.22104878723621368,
+      "learning_rate": 1.2716447964588222e-06,
+      "loss": 0.4738,
+      "step": 6183
+    },
+    {
+      "epoch": 1.7149195784803106,
+      "grad_norm": 0.20894010365009308,
+      "learning_rate": 1.2692194224303442e-06,
+      "loss": 0.4781,
+      "step": 6184
+    },
+    {
+      "epoch": 1.7151968940654465,
+      "grad_norm": 0.2051091194152832,
+      "learning_rate": 1.266796239824712e-06,
+      "loss": 0.4777,
+      "step": 6185
+    },
+    {
+      "epoch": 1.7154742096505824,
+      "grad_norm": 0.26800915598869324,
+      "learning_rate": 1.2643752491147505e-06,
+      "loss": 0.5087,
+      "step": 6186
+    },
+    {
+      "epoch": 1.7157515252357183,
+      "grad_norm": 0.20493799448013306,
+      "learning_rate": 1.2619564507728595e-06,
+      "loss": 0.4678,
+      "step": 6187
+    },
+    {
+      "epoch": 1.7160288408208542,
+      "grad_norm": 0.22277140617370605,
+      "learning_rate": 1.2595398452710128e-06,
+      "loss": 0.5145,
+      "step": 6188
+    },
+    {
+      "epoch": 1.71630615640599,
+      "grad_norm": 0.2096967250108719,
+      "learning_rate": 1.2571254330807538e-06,
+      "loss": 0.4936,
+      "step": 6189
+    },
+    {
+      "epoch": 1.716583471991126,
+      "grad_norm": 0.2071332186460495,
+      "learning_rate": 1.254713214673195e-06,
+      "loss": 0.4964,
+      "step": 6190
+    },
+    {
+      "epoch": 1.7168607875762618,
+      "grad_norm": 0.20031912624835968,
+      "learning_rate": 1.252303190519026e-06,
+      "loss": 0.5048,
+      "step": 6191
+    },
+    {
+      "epoch": 1.7171381031613977,
+      "grad_norm": 0.2249787598848343,
+      "learning_rate": 1.2498953610885087e-06,
+      "loss": 0.5219,
+      "step": 6192
+    },
+    {
+      "epoch": 1.7174154187465336,
+      "grad_norm": 0.20989800989627838,
+      "learning_rate": 1.2474897268514696e-06,
+      "loss": 0.4814,
+      "step": 6193
+    },
+    {
+      "epoch": 1.7176927343316695,
+      "grad_norm": 0.21632020175457,
+      "learning_rate": 1.2450862882773154e-06,
+      "loss": 0.4855,
+      "step": 6194
+    },
+    {
+      "epoch": 1.7179700499168054,
+      "grad_norm": 0.21728786826133728,
+      "learning_rate": 1.2426850458350208e-06,
+      "loss": 0.504,
+      "step": 6195
+    },
+    {
+      "epoch": 1.7182473655019412,
+      "grad_norm": 0.20542924106121063,
+      "learning_rate": 1.240285999993132e-06,
+      "loss": 0.5027,
+      "step": 6196
+    },
+    {
+      "epoch": 1.7185246810870771,
+      "grad_norm": 0.2109268456697464,
+      "learning_rate": 1.237889151219762e-06,
+      "loss": 0.5254,
+      "step": 6197
+    },
+    {
+      "epoch": 1.718801996672213,
+      "grad_norm": 0.21600918471813202,
+      "learning_rate": 1.2354944999826022e-06,
+      "loss": 0.5359,
+      "step": 6198
+    },
+    {
+      "epoch": 1.719079312257349,
+      "grad_norm": 0.21099236607551575,
+      "learning_rate": 1.2331020467489157e-06,
+      "loss": 0.5269,
+      "step": 6199
+    },
+    {
+      "epoch": 1.7193566278424848,
+      "grad_norm": 0.21038185060024261,
+      "learning_rate": 1.230711791985531e-06,
+      "loss": 0.5292,
+      "step": 6200
+    },
+    {
+      "epoch": 1.7196339434276207,
+      "grad_norm": 0.21021394431591034,
+      "learning_rate": 1.2283237361588442e-06,
+      "loss": 0.5175,
+      "step": 6201
+    },
+    {
+      "epoch": 1.7199112590127565,
+      "grad_norm": 0.20856258273124695,
+      "learning_rate": 1.2259378797348397e-06,
+      "loss": 0.5112,
+      "step": 6202
+    },
+    {
+      "epoch": 1.7201885745978924,
+      "grad_norm": 0.22244808077812195,
+      "learning_rate": 1.2235542231790548e-06,
+      "loss": 0.5144,
+      "step": 6203
+    },
+    {
+      "epoch": 1.7204658901830283,
+      "grad_norm": 0.20829738676548004,
+      "learning_rate": 1.2211727669566034e-06,
+      "loss": 0.5132,
+      "step": 6204
+    },
+    {
+      "epoch": 1.7207432057681642,
+      "grad_norm": 0.20939025282859802,
+      "learning_rate": 1.2187935115321708e-06,
+      "loss": 0.4784,
+      "step": 6205
+    },
+    {
+      "epoch": 1.7210205213533,
+      "grad_norm": 0.20607073605060577,
+      "learning_rate": 1.2164164573700162e-06,
+      "loss": 0.4863,
+      "step": 6206
+    },
+    {
+      "epoch": 1.721297836938436,
+      "grad_norm": 0.20456166565418243,
+      "learning_rate": 1.2140416049339644e-06,
+      "loss": 0.5155,
+      "step": 6207
+    },
+    {
+      "epoch": 1.7215751525235718,
+      "grad_norm": 0.21939820051193237,
+      "learning_rate": 1.2116689546874088e-06,
+      "loss": 0.4917,
+      "step": 6208
+    },
+    {
+      "epoch": 1.7218524681087077,
+      "grad_norm": 0.2015402466058731,
+      "learning_rate": 1.209298507093319e-06,
+      "loss": 0.4664,
+      "step": 6209
+    },
+    {
+      "epoch": 1.7221297836938436,
+      "grad_norm": 0.20761947333812714,
+      "learning_rate": 1.2069302626142352e-06,
+      "loss": 0.4844,
+      "step": 6210
+    },
+    {
+      "epoch": 1.7224070992789795,
+      "grad_norm": 0.21084575355052948,
+      "learning_rate": 1.2045642217122594e-06,
+      "loss": 0.5133,
+      "step": 6211
+    },
+    {
+      "epoch": 1.7226844148641154,
+      "grad_norm": 0.20690487325191498,
+      "learning_rate": 1.2022003848490699e-06,
+      "loss": 0.5062,
+      "step": 6212
+    },
+    {
+      "epoch": 1.7229617304492513,
+      "grad_norm": 0.2094261348247528,
+      "learning_rate": 1.1998387524859141e-06,
+      "loss": 0.5167,
+      "step": 6213
+    },
+    {
+      "epoch": 1.7232390460343872,
+      "grad_norm": 0.20428799092769623,
+      "learning_rate": 1.1974793250836128e-06,
+      "loss": 0.4953,
+      "step": 6214
+    },
+    {
+      "epoch": 1.723516361619523,
+      "grad_norm": 0.200530007481575,
+      "learning_rate": 1.1951221031025473e-06,
+      "loss": 0.5019,
+      "step": 6215
+    },
+    {
+      "epoch": 1.723793677204659,
+      "grad_norm": 0.21505190432071686,
+      "learning_rate": 1.1927670870026762e-06,
+      "loss": 0.5069,
+      "step": 6216
+    },
+    {
+      "epoch": 1.7240709927897948,
+      "grad_norm": 0.20491503179073334,
+      "learning_rate": 1.190414277243529e-06,
+      "loss": 0.4872,
+      "step": 6217
+    },
+    {
+      "epoch": 1.7243483083749307,
+      "grad_norm": 0.20874449610710144,
+      "learning_rate": 1.188063674284197e-06,
+      "loss": 0.4886,
+      "step": 6218
+    },
+    {
+      "epoch": 1.7246256239600666,
+      "grad_norm": 0.21290519833564758,
+      "learning_rate": 1.1857152785833451e-06,
+      "loss": 0.478,
+      "step": 6219
+    },
+    {
+      "epoch": 1.7249029395452025,
+      "grad_norm": 0.20770753920078278,
+      "learning_rate": 1.1833690905992081e-06,
+      "loss": 0.4946,
+      "step": 6220
+    },
+    {
+      "epoch": 1.7251802551303383,
+      "grad_norm": 0.21284344792366028,
+      "learning_rate": 1.1810251107895923e-06,
+      "loss": 0.4757,
+      "step": 6221
+    },
+    {
+      "epoch": 1.7254575707154742,
+      "grad_norm": 0.20383362472057343,
+      "learning_rate": 1.1786833396118664e-06,
+      "loss": 0.4774,
+      "step": 6222
+    },
+    {
+      "epoch": 1.7257348863006101,
+      "grad_norm": 0.2078067809343338,
+      "learning_rate": 1.1763437775229744e-06,
+      "loss": 0.4934,
+      "step": 6223
+    },
+    {
+      "epoch": 1.726012201885746,
+      "grad_norm": 0.2041560709476471,
+      "learning_rate": 1.174006424979425e-06,
+      "loss": 0.5165,
+      "step": 6224
+    },
+    {
+      "epoch": 1.7262895174708819,
+      "grad_norm": 0.21720266342163086,
+      "learning_rate": 1.1716712824373006e-06,
+      "loss": 0.4729,
+      "step": 6225
+    },
+    {
+      "epoch": 1.7265668330560178,
+      "grad_norm": 0.19432541728019714,
+      "learning_rate": 1.1693383503522435e-06,
+      "loss": 0.4889,
+      "step": 6226
+    },
+    {
+      "epoch": 1.7268441486411537,
+      "grad_norm": 0.20884351432323456,
+      "learning_rate": 1.1670076291794785e-06,
+      "loss": 0.4874,
+      "step": 6227
+    },
+    {
+      "epoch": 1.7271214642262895,
+      "grad_norm": 0.20704929530620575,
+      "learning_rate": 1.1646791193737848e-06,
+      "loss": 0.4934,
+      "step": 6228
+    },
+    {
+      "epoch": 1.7273987798114254,
+      "grad_norm": 0.2176746129989624,
+      "learning_rate": 1.1623528213895174e-06,
+      "loss": 0.4709,
+      "step": 6229
+    },
+    {
+      "epoch": 1.7276760953965613,
+      "grad_norm": 0.20258481800556183,
+      "learning_rate": 1.160028735680603e-06,
+      "loss": 0.4843,
+      "step": 6230
+    },
+    {
+      "epoch": 1.7279534109816972,
+      "grad_norm": 0.20934736728668213,
+      "learning_rate": 1.1577068627005264e-06,
+      "loss": 0.4852,
+      "step": 6231
+    },
+    {
+      "epoch": 1.728230726566833,
+      "grad_norm": 0.20559830963611603,
+      "learning_rate": 1.1553872029023498e-06,
+      "loss": 0.4676,
+      "step": 6232
+    },
+    {
+      "epoch": 1.728508042151969,
+      "grad_norm": 0.20282293856143951,
+      "learning_rate": 1.1530697567387019e-06,
+      "loss": 0.474,
+      "step": 6233
+    },
+    {
+      "epoch": 1.7287853577371048,
+      "grad_norm": 0.20379206538200378,
+      "learning_rate": 1.1507545246617763e-06,
+      "loss": 0.4831,
+      "step": 6234
+    },
+    {
+      "epoch": 1.7290626733222407,
+      "grad_norm": 0.20645445585250854,
+      "learning_rate": 1.1484415071233322e-06,
+      "loss": 0.4729,
+      "step": 6235
+    },
+    {
+      "epoch": 1.7293399889073766,
+      "grad_norm": 0.20266611874103546,
+      "learning_rate": 1.1461307045747035e-06,
+      "loss": 0.4816,
+      "step": 6236
+    },
+    {
+      "epoch": 1.7296173044925125,
+      "grad_norm": 0.20587818324565887,
+      "learning_rate": 1.1438221174667931e-06,
+      "loss": 0.5129,
+      "step": 6237
+    },
+    {
+      "epoch": 1.7298946200776484,
+      "grad_norm": 0.21184997260570526,
+      "learning_rate": 1.1415157462500631e-06,
+      "loss": 0.5172,
+      "step": 6238
+    },
+    {
+      "epoch": 1.7301719356627843,
+      "grad_norm": 0.21965019404888153,
+      "learning_rate": 1.1392115913745436e-06,
+      "loss": 0.5091,
+      "step": 6239
+    },
+    {
+      "epoch": 1.7304492512479202,
+      "grad_norm": 0.2110959142446518,
+      "learning_rate": 1.1369096532898458e-06,
+      "loss": 0.4752,
+      "step": 6240
+    },
+    {
+      "epoch": 1.730726566833056,
+      "grad_norm": 0.20495730638504028,
+      "learning_rate": 1.134609932445134e-06,
+      "loss": 0.4793,
+      "step": 6241
+    },
+    {
+      "epoch": 1.731003882418192,
+      "grad_norm": 0.20782862603664398,
+      "learning_rate": 1.132312429289145e-06,
+      "loss": 0.5282,
+      "step": 6242
+    },
+    {
+      "epoch": 1.7312811980033278,
+      "grad_norm": 0.20263516902923584,
+      "learning_rate": 1.1300171442701776e-06,
+      "loss": 0.4969,
+      "step": 6243
+    },
+    {
+      "epoch": 1.7315585135884637,
+      "grad_norm": 0.2159367799758911,
+      "learning_rate": 1.127724077836112e-06,
+      "loss": 0.489,
+      "step": 6244
+    },
+    {
+      "epoch": 1.7318358291735996,
+      "grad_norm": 0.2180434763431549,
+      "learning_rate": 1.1254332304343806e-06,
+      "loss": 0.4872,
+      "step": 6245
+    },
+    {
+      "epoch": 1.7321131447587355,
+      "grad_norm": 0.20098507404327393,
+      "learning_rate": 1.1231446025119885e-06,
+      "loss": 0.5097,
+      "step": 6246
+    },
+    {
+      "epoch": 1.7323904603438713,
+      "grad_norm": 0.20740365982055664,
+      "learning_rate": 1.1208581945155075e-06,
+      "loss": 0.4777,
+      "step": 6247
+    },
+    {
+      "epoch": 1.7326677759290072,
+      "grad_norm": 0.2009047120809555,
+      "learning_rate": 1.1185740068910807e-06,
+      "loss": 0.4916,
+      "step": 6248
+    },
+    {
+      "epoch": 1.732945091514143,
+      "grad_norm": 0.2049127072095871,
+      "learning_rate": 1.1162920400844102e-06,
+      "loss": 0.493,
+      "step": 6249
+    },
+    {
+      "epoch": 1.733222407099279,
+      "grad_norm": 0.21437804400920868,
+      "learning_rate": 1.1140122945407644e-06,
+      "loss": 0.4864,
+      "step": 6250
+    },
+    {
+      "epoch": 1.7334997226844149,
+      "grad_norm": 0.20494471490383148,
+      "learning_rate": 1.1117347707049876e-06,
+      "loss": 0.4845,
+      "step": 6251
+    },
+    {
+      "epoch": 1.7337770382695508,
+      "grad_norm": 0.2116573303937912,
+      "learning_rate": 1.1094594690214858e-06,
+      "loss": 0.4745,
+      "step": 6252
+    },
+    {
+      "epoch": 1.7340543538546866,
+      "grad_norm": 0.2070380300283432,
+      "learning_rate": 1.1071863899342255e-06,
+      "loss": 0.4956,
+      "step": 6253
+    },
+    {
+      "epoch": 1.7343316694398225,
+      "grad_norm": 0.2108084261417389,
+      "learning_rate": 1.1049155338867466e-06,
+      "loss": 0.4917,
+      "step": 6254
+    },
+    {
+      "epoch": 1.7346089850249584,
+      "grad_norm": 0.20470231771469116,
+      "learning_rate": 1.1026469013221574e-06,
+      "loss": 0.4788,
+      "step": 6255
+    },
+    {
+      "epoch": 1.7348863006100943,
+      "grad_norm": 0.20482300221920013,
+      "learning_rate": 1.1003804926831243e-06,
+      "loss": 0.4933,
+      "step": 6256
+    },
+    {
+      "epoch": 1.7351636161952302,
+      "grad_norm": 0.23574180901050568,
+      "learning_rate": 1.0981163084118825e-06,
+      "loss": 0.4953,
+      "step": 6257
+    },
+    {
+      "epoch": 1.735440931780366,
+      "grad_norm": 0.19923263788223267,
+      "learning_rate": 1.095854348950237e-06,
+      "loss": 0.4594,
+      "step": 6258
+    },
+    {
+      "epoch": 1.735718247365502,
+      "grad_norm": 0.23000024259090424,
+      "learning_rate": 1.0935946147395556e-06,
+      "loss": 0.4972,
+      "step": 6259
+    },
+    {
+      "epoch": 1.7359955629506378,
+      "grad_norm": 0.2056475728750229,
+      "learning_rate": 1.0913371062207702e-06,
+      "loss": 0.4875,
+      "step": 6260
+    },
+    {
+      "epoch": 1.7362728785357737,
+      "grad_norm": 0.21127453446388245,
+      "learning_rate": 1.0890818238343856e-06,
+      "loss": 0.4894,
+      "step": 6261
+    },
+    {
+      "epoch": 1.7365501941209096,
+      "grad_norm": 0.21046780049800873,
+      "learning_rate": 1.0868287680204606e-06,
+      "loss": 0.499,
+      "step": 6262
+    },
+    {
+      "epoch": 1.7368275097060455,
+      "grad_norm": 0.2049138844013214,
+      "learning_rate": 1.0845779392186311e-06,
+      "loss": 0.5112,
+      "step": 6263
+    },
+    {
+      "epoch": 1.7371048252911814,
+      "grad_norm": 0.20450134575366974,
+      "learning_rate": 1.0823293378680904e-06,
+      "loss": 0.4916,
+      "step": 6264
+    },
+    {
+      "epoch": 1.7373821408763173,
+      "grad_norm": 0.20806051790714264,
+      "learning_rate": 1.0800829644076027e-06,
+      "loss": 0.4837,
+      "step": 6265
+    },
+    {
+      "epoch": 1.7376594564614531,
+      "grad_norm": 0.20882248878479004,
+      "learning_rate": 1.0778388192754909e-06,
+      "loss": 0.4949,
+      "step": 6266
+    },
+    {
+      "epoch": 1.737936772046589,
+      "grad_norm": 0.2119029462337494,
+      "learning_rate": 1.0755969029096508e-06,
+      "loss": 0.4759,
+      "step": 6267
+    },
+    {
+      "epoch": 1.738214087631725,
+      "grad_norm": 0.20091450214385986,
+      "learning_rate": 1.0733572157475405e-06,
+      "loss": 0.4907,
+      "step": 6268
+    },
+    {
+      "epoch": 1.7384914032168608,
+      "grad_norm": 0.295280784368515,
+      "learning_rate": 1.0711197582261786e-06,
+      "loss": 0.5013,
+      "step": 6269
+    },
+    {
+      "epoch": 1.7387687188019967,
+      "grad_norm": 0.21129965782165527,
+      "learning_rate": 1.0688845307821577e-06,
+      "loss": 0.5146,
+      "step": 6270
+    },
+    {
+      "epoch": 1.7390460343871326,
+      "grad_norm": 0.20647571980953217,
+      "learning_rate": 1.0666515338516232e-06,
+      "loss": 0.4848,
+      "step": 6271
+    },
+    {
+      "epoch": 1.7393233499722685,
+      "grad_norm": 0.20426985621452332,
+      "learning_rate": 1.064420767870297e-06,
+      "loss": 0.4568,
+      "step": 6272
+    },
+    {
+      "epoch": 1.7396006655574043,
+      "grad_norm": 0.20671120285987854,
+      "learning_rate": 1.0621922332734568e-06,
+      "loss": 0.477,
+      "step": 6273
+    },
+    {
+      "epoch": 1.7398779811425402,
+      "grad_norm": 0.20746487379074097,
+      "learning_rate": 1.0599659304959514e-06,
+      "loss": 0.5096,
+      "step": 6274
+    },
+    {
+      "epoch": 1.740155296727676,
+      "grad_norm": 0.21173517405986786,
+      "learning_rate": 1.057741859972193e-06,
+      "loss": 0.4759,
+      "step": 6275
+    },
+    {
+      "epoch": 1.740432612312812,
+      "grad_norm": 0.20655114948749542,
+      "learning_rate": 1.0555200221361556e-06,
+      "loss": 0.4914,
+      "step": 6276
+    },
+    {
+      "epoch": 1.7407099278979479,
+      "grad_norm": 0.2177557796239853,
+      "learning_rate": 1.053300417421374e-06,
+      "loss": 0.4762,
+      "step": 6277
+    },
+    {
+      "epoch": 1.7409872434830838,
+      "grad_norm": 0.1989573985338211,
+      "learning_rate": 1.0510830462609561e-06,
+      "loss": 0.4646,
+      "step": 6278
+    },
+    {
+      "epoch": 1.7412645590682196,
+      "grad_norm": 0.20482690632343292,
+      "learning_rate": 1.0488679090875711e-06,
+      "loss": 0.4705,
+      "step": 6279
+    },
+    {
+      "epoch": 1.7415418746533555,
+      "grad_norm": 0.19922995567321777,
+      "learning_rate": 1.0466550063334497e-06,
+      "loss": 0.4826,
+      "step": 6280
+    },
+    {
+      "epoch": 1.7418191902384914,
+      "grad_norm": 0.2140817493200302,
+      "learning_rate": 1.0444443384303823e-06,
+      "loss": 0.4991,
+      "step": 6281
+    },
+    {
+      "epoch": 1.7420965058236273,
+      "grad_norm": 0.2041434794664383,
+      "learning_rate": 1.0422359058097382e-06,
+      "loss": 0.4831,
+      "step": 6282
+    },
+    {
+      "epoch": 1.7423738214087632,
+      "grad_norm": 0.2063605785369873,
+      "learning_rate": 1.0400297089024373e-06,
+      "loss": 0.5162,
+      "step": 6283
+    },
+    {
+      "epoch": 1.742651136993899,
+      "grad_norm": 0.21836838126182556,
+      "learning_rate": 1.0378257481389639e-06,
+      "loss": 0.4971,
+      "step": 6284
+    },
+    {
+      "epoch": 1.742928452579035,
+      "grad_norm": 0.2118816375732422,
+      "learning_rate": 1.0356240239493705e-06,
+      "loss": 0.4815,
+      "step": 6285
+    },
+    {
+      "epoch": 1.7432057681641708,
+      "grad_norm": 0.20386700332164764,
+      "learning_rate": 1.0334245367632764e-06,
+      "loss": 0.4998,
+      "step": 6286
+    },
+    {
+      "epoch": 1.7434830837493067,
+      "grad_norm": 0.2046600878238678,
+      "learning_rate": 1.031227287009856e-06,
+      "loss": 0.4958,
+      "step": 6287
+    },
+    {
+      "epoch": 1.7437603993344426,
+      "grad_norm": 0.20780619978904724,
+      "learning_rate": 1.0290322751178486e-06,
+      "loss": 0.5133,
+      "step": 6288
+    },
+    {
+      "epoch": 1.7440377149195785,
+      "grad_norm": 0.22479791939258575,
+      "learning_rate": 1.0268395015155613e-06,
+      "loss": 0.4936,
+      "step": 6289
+    },
+    {
+      "epoch": 1.7443150305047144,
+      "grad_norm": 0.2186380922794342,
+      "learning_rate": 1.0246489666308672e-06,
+      "loss": 0.5165,
+      "step": 6290
+    },
+    {
+      "epoch": 1.7445923460898503,
+      "grad_norm": 0.2129991352558136,
+      "learning_rate": 1.0224606708911913e-06,
+      "loss": 0.5162,
+      "step": 6291
+    },
+    {
+      "epoch": 1.7448696616749861,
+      "grad_norm": 0.2101704329252243,
+      "learning_rate": 1.0202746147235265e-06,
+      "loss": 0.5055,
+      "step": 6292
+    },
+    {
+      "epoch": 1.745146977260122,
+      "grad_norm": 0.21138980984687805,
+      "learning_rate": 1.0180907985544387e-06,
+      "loss": 0.4946,
+      "step": 6293
+    },
+    {
+      "epoch": 1.745424292845258,
+      "grad_norm": 0.20843324065208435,
+      "learning_rate": 1.0159092228100437e-06,
+      "loss": 0.4729,
+      "step": 6294
+    },
+    {
+      "epoch": 1.7457016084303938,
+      "grad_norm": 0.20566268265247345,
+      "learning_rate": 1.0137298879160221e-06,
+      "loss": 0.4515,
+      "step": 6295
+    },
+    {
+      "epoch": 1.7459789240155297,
+      "grad_norm": 0.2100181132555008,
+      "learning_rate": 1.0115527942976224e-06,
+      "loss": 0.4772,
+      "step": 6296
+    },
+    {
+      "epoch": 1.7462562396006656,
+      "grad_norm": 0.2049798220396042,
+      "learning_rate": 1.009377942379655e-06,
+      "loss": 0.4771,
+      "step": 6297
+    },
+    {
+      "epoch": 1.7465335551858014,
+      "grad_norm": 0.2053791582584381,
+      "learning_rate": 1.0072053325864875e-06,
+      "loss": 0.4837,
+      "step": 6298
+    },
+    {
+      "epoch": 1.7468108707709373,
+      "grad_norm": 0.20660558342933655,
+      "learning_rate": 1.005034965342057e-06,
+      "loss": 0.5195,
+      "step": 6299
+    },
+    {
+      "epoch": 1.7470881863560732,
+      "grad_norm": 0.21594339609146118,
+      "learning_rate": 1.0028668410698564e-06,
+      "loss": 0.4937,
+      "step": 6300
+    },
+    {
+      "epoch": 1.747365501941209,
+      "grad_norm": 0.21394188702106476,
+      "learning_rate": 1.000700960192949e-06,
+      "loss": 0.5009,
+      "step": 6301
+    },
+    {
+      "epoch": 1.747642817526345,
+      "grad_norm": 0.2106638103723526,
+      "learning_rate": 9.985373231339496e-07,
+      "loss": 0.4723,
+      "step": 6302
+    },
+    {
+      "epoch": 1.7479201331114809,
+      "grad_norm": 0.20718468725681305,
+      "learning_rate": 9.963759303150453e-07,
+      "loss": 0.4694,
+      "step": 6303
+    },
+    {
+      "epoch": 1.7481974486966168,
+      "grad_norm": 0.2988225817680359,
+      "learning_rate": 9.942167821579778e-07,
+      "loss": 0.5089,
+      "step": 6304
+    },
+    {
+      "epoch": 1.7484747642817526,
+      "grad_norm": 0.21094036102294922,
+      "learning_rate": 9.920598790840562e-07,
+      "loss": 0.4992,
+      "step": 6305
+    },
+    {
+      "epoch": 1.7487520798668885,
+      "grad_norm": 0.20664121210575104,
+      "learning_rate": 9.899052215141522e-07,
+      "loss": 0.4853,
+      "step": 6306
+    },
+    {
+      "epoch": 1.7490293954520244,
+      "grad_norm": 0.21847885847091675,
+      "learning_rate": 9.877528098686917e-07,
+      "loss": 0.5011,
+      "step": 6307
+    },
+    {
+      "epoch": 1.7493067110371603,
+      "grad_norm": 0.22634607553482056,
+      "learning_rate": 9.856026445676709e-07,
+      "loss": 0.508,
+      "step": 6308
+    },
+    {
+      "epoch": 1.7495840266222962,
+      "grad_norm": 0.21793290972709656,
+      "learning_rate": 9.8345472603064e-07,
+      "loss": 0.5396,
+      "step": 6309
+    },
+    {
+      "epoch": 1.749861342207432,
+      "grad_norm": 0.2086942344903946,
+      "learning_rate": 9.813090546767184e-07,
+      "loss": 0.5171,
+      "step": 6310
+    },
+    {
+      "epoch": 1.750138657792568,
+      "grad_norm": 0.21406447887420654,
+      "learning_rate": 9.79165630924582e-07,
+      "loss": 0.5021,
+      "step": 6311
+    },
+    {
+      "epoch": 1.7504159733777038,
+      "grad_norm": 0.20880061388015747,
+      "learning_rate": 9.770244551924683e-07,
+      "loss": 0.5135,
+      "step": 6312
+    },
+    {
+      "epoch": 1.7506932889628397,
+      "grad_norm": 0.2138943374156952,
+      "learning_rate": 9.748855278981811e-07,
+      "loss": 0.5149,
+      "step": 6313
+    },
+    {
+      "epoch": 1.7509706045479756,
+      "grad_norm": 0.20010744035243988,
+      "learning_rate": 9.727488494590805e-07,
+      "loss": 0.5055,
+      "step": 6314
+    },
+    {
+      "epoch": 1.7512479201331115,
+      "grad_norm": 0.2039041370153427,
+      "learning_rate": 9.706144202920848e-07,
+      "loss": 0.4796,
+      "step": 6315
+    },
+    {
+      "epoch": 1.7515252357182474,
+      "grad_norm": 0.1957438588142395,
+      "learning_rate": 9.684822408136813e-07,
+      "loss": 0.4513,
+      "step": 6316
+    },
+    {
+      "epoch": 1.7518025513033832,
+      "grad_norm": 0.20559154450893402,
+      "learning_rate": 9.663523114399179e-07,
+      "loss": 0.5066,
+      "step": 6317
+    },
+    {
+      "epoch": 1.7520798668885191,
+      "grad_norm": 0.20115143060684204,
+      "learning_rate": 9.642246325863954e-07,
+      "loss": 0.53,
+      "step": 6318
+    },
+    {
+      "epoch": 1.752357182473655,
+      "grad_norm": 0.22017818689346313,
+      "learning_rate": 9.620992046682778e-07,
+      "loss": 0.4881,
+      "step": 6319
+    },
+    {
+      "epoch": 1.752634498058791,
+      "grad_norm": 0.20881177484989166,
+      "learning_rate": 9.599760281002997e-07,
+      "loss": 0.4777,
+      "step": 6320
+    },
+    {
+      "epoch": 1.7529118136439268,
+      "grad_norm": 0.24889224767684937,
+      "learning_rate": 9.578551032967466e-07,
+      "loss": 0.4785,
+      "step": 6321
+    },
+    {
+      "epoch": 1.7531891292290627,
+      "grad_norm": 0.21844631433486938,
+      "learning_rate": 9.557364306714638e-07,
+      "loss": 0.4774,
+      "step": 6322
+    },
+    {
+      "epoch": 1.7534664448141986,
+      "grad_norm": 0.2035273164510727,
+      "learning_rate": 9.536200106378637e-07,
+      "loss": 0.4977,
+      "step": 6323
+    },
+    {
+      "epoch": 1.7537437603993344,
+      "grad_norm": 0.19780419766902924,
+      "learning_rate": 9.515058436089158e-07,
+      "loss": 0.4964,
+      "step": 6324
+    },
+    {
+      "epoch": 1.7540210759844703,
+      "grad_norm": 0.19891899824142456,
+      "learning_rate": 9.493939299971499e-07,
+      "loss": 0.4999,
+      "step": 6325
+    },
+    {
+      "epoch": 1.7542983915696062,
+      "grad_norm": 0.20503740012645721,
+      "learning_rate": 9.472842702146545e-07,
+      "loss": 0.4901,
+      "step": 6326
+    },
+    {
+      "epoch": 1.754575707154742,
+      "grad_norm": 0.22530865669250488,
+      "learning_rate": 9.451768646730808e-07,
+      "loss": 0.4997,
+      "step": 6327
+    },
+    {
+      "epoch": 1.754853022739878,
+      "grad_norm": 0.21066060662269592,
+      "learning_rate": 9.430717137836414e-07,
+      "loss": 0.4874,
+      "step": 6328
+    },
+    {
+      "epoch": 1.7551303383250139,
+      "grad_norm": 0.2237623631954193,
+      "learning_rate": 9.409688179571066e-07,
+      "loss": 0.4899,
+      "step": 6329
+    },
+    {
+      "epoch": 1.7554076539101497,
+      "grad_norm": 0.2304733246564865,
+      "learning_rate": 9.388681776038022e-07,
+      "loss": 0.4858,
+      "step": 6330
+    },
+    {
+      "epoch": 1.7556849694952856,
+      "grad_norm": 0.21019956469535828,
+      "learning_rate": 9.367697931336266e-07,
+      "loss": 0.5145,
+      "step": 6331
+    },
+    {
+      "epoch": 1.7559622850804215,
+      "grad_norm": 0.2206750512123108,
+      "learning_rate": 9.346736649560262e-07,
+      "loss": 0.5251,
+      "step": 6332
+    },
+    {
+      "epoch": 1.7562396006655574,
+      "grad_norm": 0.21418681740760803,
+      "learning_rate": 9.325797934800082e-07,
+      "loss": 0.4837,
+      "step": 6333
+    },
+    {
+      "epoch": 1.7565169162506933,
+      "grad_norm": 0.20882757008075714,
+      "learning_rate": 9.304881791141474e-07,
+      "loss": 0.5064,
+      "step": 6334
+    },
+    {
+      "epoch": 1.7567942318358292,
+      "grad_norm": 0.22608445584774017,
+      "learning_rate": 9.283988222665726e-07,
+      "loss": 0.5077,
+      "step": 6335
+    },
+    {
+      "epoch": 1.757071547420965,
+      "grad_norm": 0.20677009224891663,
+      "learning_rate": 9.263117233449706e-07,
+      "loss": 0.5135,
+      "step": 6336
+    },
+    {
+      "epoch": 1.757348863006101,
+      "grad_norm": 0.2091902196407318,
+      "learning_rate": 9.24226882756589e-07,
+      "loss": 0.5079,
+      "step": 6337
+    },
+    {
+      "epoch": 1.7576261785912368,
+      "grad_norm": 0.2035513073205948,
+      "learning_rate": 9.22144300908237e-07,
+      "loss": 0.4837,
+      "step": 6338
+    },
+    {
+      "epoch": 1.7579034941763727,
+      "grad_norm": 0.21183468401432037,
+      "learning_rate": 9.200639782062842e-07,
+      "loss": 0.4968,
+      "step": 6339
+    },
+    {
+      "epoch": 1.7581808097615086,
+      "grad_norm": 0.20480120182037354,
+      "learning_rate": 9.179859150566503e-07,
+      "loss": 0.5,
+      "step": 6340
+    },
+    {
+      "epoch": 1.7584581253466445,
+      "grad_norm": 0.20064327120780945,
+      "learning_rate": 9.159101118648276e-07,
+      "loss": 0.4982,
+      "step": 6341
+    },
+    {
+      "epoch": 1.7587354409317804,
+      "grad_norm": 0.22214539349079132,
+      "learning_rate": 9.13836569035853e-07,
+      "loss": 0.5245,
+      "step": 6342
+    },
+    {
+      "epoch": 1.7590127565169162,
+      "grad_norm": 0.23857805132865906,
+      "learning_rate": 9.117652869743365e-07,
+      "loss": 0.5035,
+      "step": 6343
+    },
+    {
+      "epoch": 1.7592900721020521,
+      "grad_norm": 0.21097144484519958,
+      "learning_rate": 9.096962660844352e-07,
+      "loss": 0.4751,
+      "step": 6344
+    },
+    {
+      "epoch": 1.759567387687188,
+      "grad_norm": 0.20220641791820526,
+      "learning_rate": 9.076295067698707e-07,
+      "loss": 0.4813,
+      "step": 6345
+    },
+    {
+      "epoch": 1.759844703272324,
+      "grad_norm": 0.20024532079696655,
+      "learning_rate": 9.055650094339257e-07,
+      "loss": 0.4912,
+      "step": 6346
+    },
+    {
+      "epoch": 1.7601220188574598,
+      "grad_norm": 0.2002626657485962,
+      "learning_rate": 9.035027744794353e-07,
+      "loss": 0.4897,
+      "step": 6347
+    },
+    {
+      "epoch": 1.7603993344425957,
+      "grad_norm": 0.20744280517101288,
+      "learning_rate": 9.014428023087981e-07,
+      "loss": 0.5057,
+      "step": 6348
+    },
+    {
+      "epoch": 1.7606766500277315,
+      "grad_norm": 0.21660728752613068,
+      "learning_rate": 8.993850933239664e-07,
+      "loss": 0.5014,
+      "step": 6349
+    },
+    {
+      "epoch": 1.7609539656128674,
+      "grad_norm": 0.22033831477165222,
+      "learning_rate": 8.973296479264564e-07,
+      "loss": 0.4623,
+      "step": 6350
+    },
+    {
+      "epoch": 1.7612312811980033,
+      "grad_norm": 0.1961023062467575,
+      "learning_rate": 8.952764665173378e-07,
+      "loss": 0.458,
+      "step": 6351
+    },
+    {
+      "epoch": 1.7615085967831392,
+      "grad_norm": 0.20602113008499146,
+      "learning_rate": 8.932255494972452e-07,
+      "loss": 0.487,
+      "step": 6352
+    },
+    {
+      "epoch": 1.761785912368275,
+      "grad_norm": 0.21287092566490173,
+      "learning_rate": 8.911768972663603e-07,
+      "loss": 0.5232,
+      "step": 6353
+    },
+    {
+      "epoch": 1.762063227953411,
+      "grad_norm": 0.20058588683605194,
+      "learning_rate": 8.891305102244326e-07,
+      "loss": 0.4714,
+      "step": 6354
+    },
+    {
+      "epoch": 1.7623405435385469,
+      "grad_norm": 0.22055946290493011,
+      "learning_rate": 8.870863887707678e-07,
+      "loss": 0.5222,
+      "step": 6355
+    },
+    {
+      "epoch": 1.7626178591236827,
+      "grad_norm": 0.2032008171081543,
+      "learning_rate": 8.850445333042287e-07,
+      "loss": 0.5174,
+      "step": 6356
+    },
+    {
+      "epoch": 1.7628951747088186,
+      "grad_norm": 0.22134113311767578,
+      "learning_rate": 8.830049442232285e-07,
+      "loss": 0.5117,
+      "step": 6357
+    },
+    {
+      "epoch": 1.7631724902939545,
+      "grad_norm": 0.2134564369916916,
+      "learning_rate": 8.809676219257546e-07,
+      "loss": 0.5069,
+      "step": 6358
+    },
+    {
+      "epoch": 1.7634498058790904,
+      "grad_norm": 0.21325929462909698,
+      "learning_rate": 8.789325668093374e-07,
+      "loss": 0.5023,
+      "step": 6359
+    },
+    {
+      "epoch": 1.7637271214642263,
+      "grad_norm": 0.20757299661636353,
+      "learning_rate": 8.768997792710689e-07,
+      "loss": 0.5152,
+      "step": 6360
+    },
+    {
+      "epoch": 1.7640044370493622,
+      "grad_norm": 0.2015492469072342,
+      "learning_rate": 8.74869259707603e-07,
+      "loss": 0.4596,
+      "step": 6361
+    },
+    {
+      "epoch": 1.764281752634498,
+      "grad_norm": 0.23414163291454315,
+      "learning_rate": 8.728410085151476e-07,
+      "loss": 0.4936,
+      "step": 6362
+    },
+    {
+      "epoch": 1.764559068219634,
+      "grad_norm": 0.21143180131912231,
+      "learning_rate": 8.708150260894666e-07,
+      "loss": 0.5004,
+      "step": 6363
+    },
+    {
+      "epoch": 1.7648363838047698,
+      "grad_norm": 0.2277495414018631,
+      "learning_rate": 8.687913128258826e-07,
+      "loss": 0.4884,
+      "step": 6364
+    },
+    {
+      "epoch": 1.7651136993899057,
+      "grad_norm": 0.21437212824821472,
+      "learning_rate": 8.667698691192758e-07,
+      "loss": 0.504,
+      "step": 6365
+    },
+    {
+      "epoch": 1.7653910149750416,
+      "grad_norm": 0.20440204441547394,
+      "learning_rate": 8.64750695364086e-07,
+      "loss": 0.4866,
+      "step": 6366
+    },
+    {
+      "epoch": 1.7656683305601775,
+      "grad_norm": 0.2076803594827652,
+      "learning_rate": 8.627337919543066e-07,
+      "loss": 0.5107,
+      "step": 6367
+    },
+    {
+      "epoch": 1.7659456461453134,
+      "grad_norm": 0.21426992118358612,
+      "learning_rate": 8.607191592834824e-07,
+      "loss": 0.4913,
+      "step": 6368
+    },
+    {
+      "epoch": 1.7662229617304492,
+      "grad_norm": 0.24520544707775116,
+      "learning_rate": 8.587067977447321e-07,
+      "loss": 0.5185,
+      "step": 6369
+    },
+    {
+      "epoch": 1.7665002773155851,
+      "grad_norm": 0.20242126286029816,
+      "learning_rate": 8.566967077307156e-07,
+      "loss": 0.4838,
+      "step": 6370
+    },
+    {
+      "epoch": 1.766777592900721,
+      "grad_norm": 0.2145501673221588,
+      "learning_rate": 8.546888896336536e-07,
+      "loss": 0.4861,
+      "step": 6371
+    },
+    {
+      "epoch": 1.767054908485857,
+      "grad_norm": 0.21382984519004822,
+      "learning_rate": 8.526833438453258e-07,
+      "loss": 0.4972,
+      "step": 6372
+    },
+    {
+      "epoch": 1.7673322240709928,
+      "grad_norm": 0.20749729871749878,
+      "learning_rate": 8.506800707570691e-07,
+      "loss": 0.4799,
+      "step": 6373
+    },
+    {
+      "epoch": 1.7676095396561287,
+      "grad_norm": 0.20935972034931183,
+      "learning_rate": 8.486790707597725e-07,
+      "loss": 0.4861,
+      "step": 6374
+    },
+    {
+      "epoch": 1.7678868552412645,
+      "grad_norm": 0.21835561096668243,
+      "learning_rate": 8.466803442438844e-07,
+      "loss": 0.5149,
+      "step": 6375
+    },
+    {
+      "epoch": 1.7681641708264004,
+      "grad_norm": 0.2116965502500534,
+      "learning_rate": 8.446838915994099e-07,
+      "loss": 0.5302,
+      "step": 6376
+    },
+    {
+      "epoch": 1.7684414864115363,
+      "grad_norm": 0.20836491882801056,
+      "learning_rate": 8.42689713215912e-07,
+      "loss": 0.5131,
+      "step": 6377
+    },
+    {
+      "epoch": 1.7687188019966722,
+      "grad_norm": 0.19809886813163757,
+      "learning_rate": 8.406978094825033e-07,
+      "loss": 0.5147,
+      "step": 6378
+    },
+    {
+      "epoch": 1.768996117581808,
+      "grad_norm": 0.20016132295131683,
+      "learning_rate": 8.387081807878602e-07,
+      "loss": 0.4531,
+      "step": 6379
+    },
+    {
+      "epoch": 1.769273433166944,
+      "grad_norm": 0.20489495992660522,
+      "learning_rate": 8.367208275202112e-07,
+      "loss": 0.4955,
+      "step": 6380
+    },
+    {
+      "epoch": 1.7695507487520798,
+      "grad_norm": 0.22036077082157135,
+      "learning_rate": 8.34735750067342e-07,
+      "loss": 0.4954,
+      "step": 6381
+    },
+    {
+      "epoch": 1.7698280643372157,
+      "grad_norm": 0.19575130939483643,
+      "learning_rate": 8.327529488165925e-07,
+      "loss": 0.4695,
+      "step": 6382
+    },
+    {
+      "epoch": 1.7701053799223516,
+      "grad_norm": 0.20686204731464386,
+      "learning_rate": 8.30772424154859e-07,
+      "loss": 0.4713,
+      "step": 6383
+    },
+    {
+      "epoch": 1.7703826955074875,
+      "grad_norm": 0.2088318020105362,
+      "learning_rate": 8.287941764685989e-07,
+      "loss": 0.5319,
+      "step": 6384
+    },
+    {
+      "epoch": 1.7706600110926234,
+      "grad_norm": 0.20468567311763763,
+      "learning_rate": 8.26818206143816e-07,
+      "loss": 0.5164,
+      "step": 6385
+    },
+    {
+      "epoch": 1.7709373266777593,
+      "grad_norm": 0.22691620886325836,
+      "learning_rate": 8.248445135660782e-07,
+      "loss": 0.5026,
+      "step": 6386
+    },
+    {
+      "epoch": 1.7712146422628952,
+      "grad_norm": 0.2066589593887329,
+      "learning_rate": 8.22873099120501e-07,
+      "loss": 0.5169,
+      "step": 6387
+    },
+    {
+      "epoch": 1.771491957848031,
+      "grad_norm": 0.2105691283941269,
+      "learning_rate": 8.209039631917631e-07,
+      "loss": 0.5144,
+      "step": 6388
+    },
+    {
+      "epoch": 1.771769273433167,
+      "grad_norm": 0.23009151220321655,
+      "learning_rate": 8.18937106164093e-07,
+      "loss": 0.5291,
+      "step": 6389
+    },
+    {
+      "epoch": 1.7720465890183028,
+      "grad_norm": 0.20932504534721375,
+      "learning_rate": 8.169725284212781e-07,
+      "loss": 0.5105,
+      "step": 6390
+    },
+    {
+      "epoch": 1.7723239046034387,
+      "grad_norm": 0.2107200175523758,
+      "learning_rate": 8.150102303466578e-07,
+      "loss": 0.4886,
+      "step": 6391
+    },
+    {
+      "epoch": 1.7726012201885746,
+      "grad_norm": 0.20422813296318054,
+      "learning_rate": 8.130502123231285e-07,
+      "loss": 0.5055,
+      "step": 6392
+    },
+    {
+      "epoch": 1.7728785357737105,
+      "grad_norm": 0.20422576367855072,
+      "learning_rate": 8.110924747331467e-07,
+      "loss": 0.5034,
+      "step": 6393
+    },
+    {
+      "epoch": 1.7731558513588463,
+      "grad_norm": 0.21044956147670746,
+      "learning_rate": 8.091370179587124e-07,
+      "loss": 0.5029,
+      "step": 6394
+    },
+    {
+      "epoch": 1.7734331669439822,
+      "grad_norm": 0.20264363288879395,
+      "learning_rate": 8.071838423813885e-07,
+      "loss": 0.5073,
+      "step": 6395
+    },
+    {
+      "epoch": 1.7737104825291181,
+      "grad_norm": 0.2157750427722931,
+      "learning_rate": 8.052329483822924e-07,
+      "loss": 0.4861,
+      "step": 6396
+    },
+    {
+      "epoch": 1.773987798114254,
+      "grad_norm": 0.20734168589115143,
+      "learning_rate": 8.032843363420972e-07,
+      "loss": 0.5013,
+      "step": 6397
+    },
+    {
+      "epoch": 1.7742651136993899,
+      "grad_norm": 0.23018255829811096,
+      "learning_rate": 8.013380066410253e-07,
+      "loss": 0.5124,
+      "step": 6398
+    },
+    {
+      "epoch": 1.7745424292845258,
+      "grad_norm": 0.19958136975765228,
+      "learning_rate": 7.993939596588576e-07,
+      "loss": 0.4944,
+      "step": 6399
+    },
+    {
+      "epoch": 1.7748197448696617,
+      "grad_norm": 0.2005843073129654,
+      "learning_rate": 7.974521957749309e-07,
+      "loss": 0.4886,
+      "step": 6400
+    },
+    {
+      "epoch": 1.7750970604547975,
+      "grad_norm": 0.20511426031589508,
+      "learning_rate": 7.955127153681352e-07,
+      "loss": 0.4816,
+      "step": 6401
+    },
+    {
+      "epoch": 1.7753743760399334,
+      "grad_norm": 0.19508033990859985,
+      "learning_rate": 7.935755188169122e-07,
+      "loss": 0.4817,
+      "step": 6402
+    },
+    {
+      "epoch": 1.7756516916250693,
+      "grad_norm": 0.2073792815208435,
+      "learning_rate": 7.916406064992596e-07,
+      "loss": 0.4902,
+      "step": 6403
+    },
+    {
+      "epoch": 1.7759290072102052,
+      "grad_norm": 0.2069348245859146,
+      "learning_rate": 7.89707978792735e-07,
+      "loss": 0.5026,
+      "step": 6404
+    },
+    {
+      "epoch": 1.776206322795341,
+      "grad_norm": 0.21495787799358368,
+      "learning_rate": 7.87777636074441e-07,
+      "loss": 0.4647,
+      "step": 6405
+    },
+    {
+      "epoch": 1.776483638380477,
+      "grad_norm": 0.20726212859153748,
+      "learning_rate": 7.858495787210373e-07,
+      "loss": 0.4705,
+      "step": 6406
+    },
+    {
+      "epoch": 1.7767609539656128,
+      "grad_norm": 0.2127843052148819,
+      "learning_rate": 7.83923807108744e-07,
+      "loss": 0.5182,
+      "step": 6407
+    },
+    {
+      "epoch": 1.7770382695507487,
+      "grad_norm": 0.20020431280136108,
+      "learning_rate": 7.820003216133284e-07,
+      "loss": 0.4982,
+      "step": 6408
+    },
+    {
+      "epoch": 1.7773155851358846,
+      "grad_norm": 0.21346312761306763,
+      "learning_rate": 7.800791226101112e-07,
+      "loss": 0.5011,
+      "step": 6409
+    },
+    {
+      "epoch": 1.7775929007210205,
+      "grad_norm": 0.20786842703819275,
+      "learning_rate": 7.781602104739674e-07,
+      "loss": 0.4825,
+      "step": 6410
+    },
+    {
+      "epoch": 1.7778702163061564,
+      "grad_norm": 0.21008695662021637,
+      "learning_rate": 7.762435855793352e-07,
+      "loss": 0.5017,
+      "step": 6411
+    },
+    {
+      "epoch": 1.7781475318912923,
+      "grad_norm": 0.22787503898143768,
+      "learning_rate": 7.743292483001944e-07,
+      "loss": 0.5009,
+      "step": 6412
+    },
+    {
+      "epoch": 1.7784248474764282,
+      "grad_norm": 0.21107859909534454,
+      "learning_rate": 7.724171990100809e-07,
+      "loss": 0.5056,
+      "step": 6413
+    },
+    {
+      "epoch": 1.778702163061564,
+      "grad_norm": 0.20022447407245636,
+      "learning_rate": 7.705074380820881e-07,
+      "loss": 0.4997,
+      "step": 6414
+    },
+    {
+      "epoch": 1.7789794786467,
+      "grad_norm": 0.20294740796089172,
+      "learning_rate": 7.685999658888637e-07,
+      "loss": 0.4944,
+      "step": 6415
+    },
+    {
+      "epoch": 1.7792567942318358,
+      "grad_norm": 0.21547819674015045,
+      "learning_rate": 7.66694782802603e-07,
+      "loss": 0.5007,
+      "step": 6416
+    },
+    {
+      "epoch": 1.7795341098169717,
+      "grad_norm": 0.20150557160377502,
+      "learning_rate": 7.647918891950559e-07,
+      "loss": 0.5125,
+      "step": 6417
+    },
+    {
+      "epoch": 1.7798114254021076,
+      "grad_norm": 0.22688019275665283,
+      "learning_rate": 7.628912854375308e-07,
+      "loss": 0.4976,
+      "step": 6418
+    },
+    {
+      "epoch": 1.7800887409872435,
+      "grad_norm": 0.20131336152553558,
+      "learning_rate": 7.609929719008854e-07,
+      "loss": 0.4804,
+      "step": 6419
+    },
+    {
+      "epoch": 1.7803660565723793,
+      "grad_norm": 0.20517846941947937,
+      "learning_rate": 7.590969489555289e-07,
+      "loss": 0.5062,
+      "step": 6420
+    },
+    {
+      "epoch": 1.7806433721575152,
+      "grad_norm": 0.21819156408309937,
+      "learning_rate": 7.572032169714277e-07,
+      "loss": 0.4709,
+      "step": 6421
+    },
+    {
+      "epoch": 1.780920687742651,
+      "grad_norm": 0.21262036263942719,
+      "learning_rate": 7.553117763180989e-07,
+      "loss": 0.4938,
+      "step": 6422
+    },
+    {
+      "epoch": 1.781198003327787,
+      "grad_norm": 0.2089860737323761,
+      "learning_rate": 7.534226273646111e-07,
+      "loss": 0.542,
+      "step": 6423
+    },
+    {
+      "epoch": 1.7814753189129229,
+      "grad_norm": 0.20706579089164734,
+      "learning_rate": 7.515357704795902e-07,
+      "loss": 0.5062,
+      "step": 6424
+    },
+    {
+      "epoch": 1.7817526344980588,
+      "grad_norm": 0.21182586252689362,
+      "learning_rate": 7.496512060312086e-07,
+      "loss": 0.4642,
+      "step": 6425
+    },
+    {
+      "epoch": 1.7820299500831946,
+      "grad_norm": 0.21045905351638794,
+      "learning_rate": 7.477689343871983e-07,
+      "loss": 0.5155,
+      "step": 6426
+    },
+    {
+      "epoch": 1.7823072656683305,
+      "grad_norm": 0.21174542605876923,
+      "learning_rate": 7.458889559148363e-07,
+      "loss": 0.5125,
+      "step": 6427
+    },
+    {
+      "epoch": 1.7825845812534664,
+      "grad_norm": 0.21431586146354675,
+      "learning_rate": 7.440112709809599e-07,
+      "loss": 0.5094,
+      "step": 6428
+    },
+    {
+      "epoch": 1.7828618968386023,
+      "grad_norm": 0.21488359570503235,
+      "learning_rate": 7.421358799519524e-07,
+      "loss": 0.5196,
+      "step": 6429
+    },
+    {
+      "epoch": 1.7831392124237382,
+      "grad_norm": 0.21055498719215393,
+      "learning_rate": 7.402627831937528e-07,
+      "loss": 0.5338,
+      "step": 6430
+    },
+    {
+      "epoch": 1.783416528008874,
+      "grad_norm": 0.19541572034358978,
+      "learning_rate": 7.383919810718537e-07,
+      "loss": 0.4689,
+      "step": 6431
+    },
+    {
+      "epoch": 1.78369384359401,
+      "grad_norm": 0.22601917386054993,
+      "learning_rate": 7.365234739512977e-07,
+      "loss": 0.4881,
+      "step": 6432
+    },
+    {
+      "epoch": 1.7839711591791458,
+      "grad_norm": 0.20422977209091187,
+      "learning_rate": 7.34657262196678e-07,
+      "loss": 0.489,
+      "step": 6433
+    },
+    {
+      "epoch": 1.7842484747642817,
+      "grad_norm": 0.21061384677886963,
+      "learning_rate": 7.327933461721434e-07,
+      "loss": 0.523,
+      "step": 6434
+    },
+    {
+      "epoch": 1.7845257903494176,
+      "grad_norm": 0.23255345225334167,
+      "learning_rate": 7.309317262413948e-07,
+      "loss": 0.4948,
+      "step": 6435
+    },
+    {
+      "epoch": 1.7848031059345535,
+      "grad_norm": 0.20327569544315338,
+      "learning_rate": 7.290724027676791e-07,
+      "loss": 0.4884,
+      "step": 6436
+    },
+    {
+      "epoch": 1.7850804215196894,
+      "grad_norm": 0.20791566371917725,
+      "learning_rate": 7.272153761138031e-07,
+      "loss": 0.487,
+      "step": 6437
+    },
+    {
+      "epoch": 1.7853577371048253,
+      "grad_norm": 0.19935283064842224,
+      "learning_rate": 7.253606466421231e-07,
+      "loss": 0.4907,
+      "step": 6438
+    },
+    {
+      "epoch": 1.7856350526899611,
+      "grad_norm": 0.2055501639842987,
+      "learning_rate": 7.235082147145453e-07,
+      "loss": 0.4908,
+      "step": 6439
+    },
+    {
+      "epoch": 1.785912368275097,
+      "grad_norm": 0.20820872485637665,
+      "learning_rate": 7.216580806925235e-07,
+      "loss": 0.4897,
+      "step": 6440
+    },
+    {
+      "epoch": 1.786189683860233,
+      "grad_norm": 0.19953136146068573,
+      "learning_rate": 7.19810244937072e-07,
+      "loss": 0.4806,
+      "step": 6441
+    },
+    {
+      "epoch": 1.7864669994453688,
+      "grad_norm": 0.20614445209503174,
+      "learning_rate": 7.179647078087548e-07,
+      "loss": 0.4992,
+      "step": 6442
+    },
+    {
+      "epoch": 1.7867443150305047,
+      "grad_norm": 0.20100446045398712,
+      "learning_rate": 7.161214696676813e-07,
+      "loss": 0.4811,
+      "step": 6443
+    },
+    {
+      "epoch": 1.7870216306156406,
+      "grad_norm": 0.21023845672607422,
+      "learning_rate": 7.142805308735151e-07,
+      "loss": 0.4809,
+      "step": 6444
+    },
+    {
+      "epoch": 1.7872989462007765,
+      "grad_norm": 0.21367646753787994,
+      "learning_rate": 7.124418917854745e-07,
+      "loss": 0.4902,
+      "step": 6445
+    },
+    {
+      "epoch": 1.7875762617859123,
+      "grad_norm": 0.20011699199676514,
+      "learning_rate": 7.106055527623279e-07,
+      "loss": 0.4909,
+      "step": 6446
+    },
+    {
+      "epoch": 1.7878535773710482,
+      "grad_norm": 0.2133239507675171,
+      "learning_rate": 7.087715141623916e-07,
+      "loss": 0.4953,
+      "step": 6447
+    },
+    {
+      "epoch": 1.788130892956184,
+      "grad_norm": 0.20663748681545258,
+      "learning_rate": 7.06939776343532e-07,
+      "loss": 0.4885,
+      "step": 6448
+    },
+    {
+      "epoch": 1.78840820854132,
+      "grad_norm": 0.2234223186969757,
+      "learning_rate": 7.051103396631772e-07,
+      "loss": 0.468,
+      "step": 6449
+    },
+    {
+      "epoch": 1.7886855241264559,
+      "grad_norm": 0.2111656367778778,
+      "learning_rate": 7.032832044782959e-07,
+      "loss": 0.4909,
+      "step": 6450
+    },
+    {
+      "epoch": 1.7889628397115918,
+      "grad_norm": 0.2099202424287796,
+      "learning_rate": 7.014583711454053e-07,
+      "loss": 0.4802,
+      "step": 6451
+    },
+    {
+      "epoch": 1.7892401552967276,
+      "grad_norm": 0.19917812943458557,
+      "learning_rate": 6.996358400205849e-07,
+      "loss": 0.4913,
+      "step": 6452
+    },
+    {
+      "epoch": 1.7895174708818635,
+      "grad_norm": 0.20266632735729218,
+      "learning_rate": 6.978156114594583e-07,
+      "loss": 0.5082,
+      "step": 6453
+    },
+    {
+      "epoch": 1.7897947864669994,
+      "grad_norm": 0.21638800203800201,
+      "learning_rate": 6.959976858171985e-07,
+      "loss": 0.5163,
+      "step": 6454
+    },
+    {
+      "epoch": 1.7900721020521353,
+      "grad_norm": 0.20662005245685577,
+      "learning_rate": 6.941820634485299e-07,
+      "loss": 0.501,
+      "step": 6455
+    },
+    {
+      "epoch": 1.7903494176372712,
+      "grad_norm": 0.2111278474330902,
+      "learning_rate": 6.92368744707729e-07,
+      "loss": 0.5009,
+      "step": 6456
+    },
+    {
+      "epoch": 1.790626733222407,
+      "grad_norm": 0.21369148790836334,
+      "learning_rate": 6.905577299486266e-07,
+      "loss": 0.4807,
+      "step": 6457
+    },
+    {
+      "epoch": 1.790904048807543,
+      "grad_norm": 0.2006935179233551,
+      "learning_rate": 6.887490195245941e-07,
+      "loss": 0.5229,
+      "step": 6458
+    },
+    {
+      "epoch": 1.7911813643926788,
+      "grad_norm": 0.20024555921554565,
+      "learning_rate": 6.869426137885604e-07,
+      "loss": 0.5033,
+      "step": 6459
+    },
+    {
+      "epoch": 1.7914586799778147,
+      "grad_norm": 0.21740752458572388,
+      "learning_rate": 6.85138513093006e-07,
+      "loss": 0.4924,
+      "step": 6460
+    },
+    {
+      "epoch": 1.7917359955629506,
+      "grad_norm": 0.21370796859264374,
+      "learning_rate": 6.833367177899564e-07,
+      "loss": 0.4914,
+      "step": 6461
+    },
+    {
+      "epoch": 1.7920133111480865,
+      "grad_norm": 0.203489288687706,
+      "learning_rate": 6.815372282309884e-07,
+      "loss": 0.4905,
+      "step": 6462
+    },
+    {
+      "epoch": 1.7922906267332224,
+      "grad_norm": 0.20067156851291656,
+      "learning_rate": 6.797400447672311e-07,
+      "loss": 0.4508,
+      "step": 6463
+    },
+    {
+      "epoch": 1.7925679423183583,
+      "grad_norm": 0.2093934714794159,
+      "learning_rate": 6.779451677493636e-07,
+      "loss": 0.5265,
+      "step": 6464
+    },
+    {
+      "epoch": 1.7928452579034941,
+      "grad_norm": 0.20720593631267548,
+      "learning_rate": 6.761525975276129e-07,
+      "loss": 0.4972,
+      "step": 6465
+    },
+    {
+      "epoch": 1.79312257348863,
+      "grad_norm": 0.21246635913848877,
+      "learning_rate": 6.7436233445176e-07,
+      "loss": 0.5076,
+      "step": 6466
+    },
+    {
+      "epoch": 1.793399889073766,
+      "grad_norm": 0.2040356993675232,
+      "learning_rate": 6.72574378871127e-07,
+      "loss": 0.4878,
+      "step": 6467
+    },
+    {
+      "epoch": 1.7936772046589018,
+      "grad_norm": 0.20411434769630432,
+      "learning_rate": 6.707887311345959e-07,
+      "loss": 0.4903,
+      "step": 6468
+    },
+    {
+      "epoch": 1.7939545202440377,
+      "grad_norm": 0.21358831226825714,
+      "learning_rate": 6.690053915905908e-07,
+      "loss": 0.5239,
+      "step": 6469
+    },
+    {
+      "epoch": 1.7942318358291736,
+      "grad_norm": 0.20531177520751953,
+      "learning_rate": 6.672243605870918e-07,
+      "loss": 0.4836,
+      "step": 6470
+    },
+    {
+      "epoch": 1.7945091514143094,
+      "grad_norm": 0.203311026096344,
+      "learning_rate": 6.654456384716221e-07,
+      "loss": 0.4853,
+      "step": 6471
+    },
+    {
+      "epoch": 1.7947864669994453,
+      "grad_norm": 0.20684535801410675,
+      "learning_rate": 6.636692255912583e-07,
+      "loss": 0.4996,
+      "step": 6472
+    },
+    {
+      "epoch": 1.7950637825845812,
+      "grad_norm": 0.21285724639892578,
+      "learning_rate": 6.618951222926287e-07,
+      "loss": 0.5002,
+      "step": 6473
+    },
+    {
+      "epoch": 1.795341098169717,
+      "grad_norm": 0.201505646109581,
+      "learning_rate": 6.601233289219036e-07,
+      "loss": 0.5027,
+      "step": 6474
+    },
+    {
+      "epoch": 1.795618413754853,
+      "grad_norm": 0.24583852291107178,
+      "learning_rate": 6.583538458248106e-07,
+      "loss": 0.4928,
+      "step": 6475
+    },
+    {
+      "epoch": 1.7958957293399889,
+      "grad_norm": 0.22444890439510345,
+      "learning_rate": 6.565866733466181e-07,
+      "loss": 0.4982,
+      "step": 6476
+    },
+    {
+      "epoch": 1.7961730449251248,
+      "grad_norm": 0.21000193059444427,
+      "learning_rate": 6.548218118321542e-07,
+      "loss": 0.4952,
+      "step": 6477
+    },
+    {
+      "epoch": 1.7964503605102606,
+      "grad_norm": 0.20615287125110626,
+      "learning_rate": 6.530592616257839e-07,
+      "loss": 0.4686,
+      "step": 6478
+    },
+    {
+      "epoch": 1.7967276760953965,
+      "grad_norm": 0.19822682440280914,
+      "learning_rate": 6.512990230714306e-07,
+      "loss": 0.5017,
+      "step": 6479
+    },
+    {
+      "epoch": 1.7970049916805324,
+      "grad_norm": 0.20360395312309265,
+      "learning_rate": 6.495410965125653e-07,
+      "loss": 0.4922,
+      "step": 6480
+    },
+    {
+      "epoch": 1.7972823072656683,
+      "grad_norm": 0.19985002279281616,
+      "learning_rate": 6.477854822922042e-07,
+      "loss": 0.4931,
+      "step": 6481
+    },
+    {
+      "epoch": 1.7975596228508042,
+      "grad_norm": 0.19747690856456757,
+      "learning_rate": 6.460321807529118e-07,
+      "loss": 0.4757,
+      "step": 6482
+    },
+    {
+      "epoch": 1.79783693843594,
+      "grad_norm": 0.2075447291135788,
+      "learning_rate": 6.44281192236805e-07,
+      "loss": 0.5065,
+      "step": 6483
+    },
+    {
+      "epoch": 1.798114254021076,
+      "grad_norm": 0.2010498195886612,
+      "learning_rate": 6.42532517085552e-07,
+      "loss": 0.4978,
+      "step": 6484
+    },
+    {
+      "epoch": 1.7983915696062118,
+      "grad_norm": 0.2031354308128357,
+      "learning_rate": 6.407861556403633e-07,
+      "loss": 0.5098,
+      "step": 6485
+    },
+    {
+      "epoch": 1.7986688851913477,
+      "grad_norm": 0.21396131813526154,
+      "learning_rate": 6.390421082419939e-07,
+      "loss": 0.5098,
+      "step": 6486
+    },
+    {
+      "epoch": 1.7989462007764836,
+      "grad_norm": 0.20978130400180817,
+      "learning_rate": 6.373003752307649e-07,
+      "loss": 0.4845,
+      "step": 6487
+    },
+    {
+      "epoch": 1.7992235163616195,
+      "grad_norm": 0.20987875759601593,
+      "learning_rate": 6.355609569465279e-07,
+      "loss": 0.5111,
+      "step": 6488
+    },
+    {
+      "epoch": 1.7995008319467554,
+      "grad_norm": 0.20756718516349792,
+      "learning_rate": 6.338238537286892e-07,
+      "loss": 0.5202,
+      "step": 6489
+    },
+    {
+      "epoch": 1.7997781475318912,
+      "grad_norm": 0.2095729559659958,
+      "learning_rate": 6.320890659162041e-07,
+      "loss": 0.4798,
+      "step": 6490
+    },
+    {
+      "epoch": 1.8000554631170271,
+      "grad_norm": 0.21260963380336761,
+      "learning_rate": 6.303565938475794e-07,
+      "loss": 0.4852,
+      "step": 6491
+    },
+    {
+      "epoch": 1.800332778702163,
+      "grad_norm": 0.20586569607257843,
+      "learning_rate": 6.286264378608631e-07,
+      "loss": 0.4981,
+      "step": 6492
+    },
+    {
+      "epoch": 1.800610094287299,
+      "grad_norm": 0.20685313642024994,
+      "learning_rate": 6.268985982936529e-07,
+      "loss": 0.4848,
+      "step": 6493
+    },
+    {
+      "epoch": 1.8008874098724348,
+      "grad_norm": 0.21318146586418152,
+      "learning_rate": 6.251730754830987e-07,
+      "loss": 0.4873,
+      "step": 6494
+    },
+    {
+      "epoch": 1.8011647254575707,
+      "grad_norm": 0.21374082565307617,
+      "learning_rate": 6.234498697658964e-07,
+      "loss": 0.4889,
+      "step": 6495
+    },
+    {
+      "epoch": 1.8014420410427066,
+      "grad_norm": 0.20280690491199493,
+      "learning_rate": 6.217289814782867e-07,
+      "loss": 0.5096,
+      "step": 6496
+    },
+    {
+      "epoch": 1.8017193566278424,
+      "grad_norm": 0.2248666137456894,
+      "learning_rate": 6.200104109560622e-07,
+      "loss": 0.5,
+      "step": 6497
+    },
+    {
+      "epoch": 1.8019966722129783,
+      "grad_norm": 0.25136902928352356,
+      "learning_rate": 6.182941585345603e-07,
+      "loss": 0.5054,
+      "step": 6498
+    },
+    {
+      "epoch": 1.8022739877981142,
+      "grad_norm": 0.2106061577796936,
+      "learning_rate": 6.165802245486684e-07,
+      "loss": 0.503,
+      "step": 6499
+    },
+    {
+      "epoch": 1.80255130338325,
+      "grad_norm": 0.19975021481513977,
+      "learning_rate": 6.148686093328193e-07,
+      "loss": 0.4912,
+      "step": 6500
+    },
+    {
+      "epoch": 1.802828618968386,
+      "grad_norm": 0.2124665230512619,
+      "learning_rate": 6.131593132209942e-07,
+      "loss": 0.5096,
+      "step": 6501
+    },
+    {
+      "epoch": 1.8031059345535219,
+      "grad_norm": 0.20581713318824768,
+      "learning_rate": 6.11452336546725e-07,
+      "loss": 0.4901,
+      "step": 6502
+    },
+    {
+      "epoch": 1.8033832501386577,
+      "grad_norm": 0.20466797053813934,
+      "learning_rate": 6.097476796430856e-07,
+      "loss": 0.4827,
+      "step": 6503
+    },
+    {
+      "epoch": 1.8036605657237936,
+      "grad_norm": 0.20932024717330933,
+      "learning_rate": 6.080453428427003e-07,
+      "loss": 0.4881,
+      "step": 6504
+    },
+    {
+      "epoch": 1.8039378813089295,
+      "grad_norm": 0.20006607472896576,
+      "learning_rate": 6.063453264777391e-07,
+      "loss": 0.4892,
+      "step": 6505
+    },
+    {
+      "epoch": 1.8042151968940654,
+      "grad_norm": 0.21569404006004333,
+      "learning_rate": 6.046476308799217e-07,
+      "loss": 0.4704,
+      "step": 6506
+    },
+    {
+      "epoch": 1.8044925124792013,
+      "grad_norm": 0.22104816138744354,
+      "learning_rate": 6.029522563805123e-07,
+      "loss": 0.4961,
+      "step": 6507
+    },
+    {
+      "epoch": 1.8047698280643372,
+      "grad_norm": 0.21142521500587463,
+      "learning_rate": 6.012592033103253e-07,
+      "loss": 0.5161,
+      "step": 6508
+    },
+    {
+      "epoch": 1.805047143649473,
+      "grad_norm": 0.20408941805362701,
+      "learning_rate": 5.995684719997174e-07,
+      "loss": 0.512,
+      "step": 6509
+    },
+    {
+      "epoch": 1.805324459234609,
+      "grad_norm": 0.20980273187160492,
+      "learning_rate": 5.978800627785968e-07,
+      "loss": 0.5006,
+      "step": 6510
+    },
+    {
+      "epoch": 1.8056017748197448,
+      "grad_norm": 0.21785345673561096,
+      "learning_rate": 5.961939759764181e-07,
+      "loss": 0.4957,
+      "step": 6511
+    },
+    {
+      "epoch": 1.8058790904048807,
+      "grad_norm": 0.20818081498146057,
+      "learning_rate": 5.945102119221793e-07,
+      "loss": 0.5021,
+      "step": 6512
+    },
+    {
+      "epoch": 1.8061564059900166,
+      "grad_norm": 0.21114230155944824,
+      "learning_rate": 5.928287709444285e-07,
+      "loss": 0.4915,
+      "step": 6513
+    },
+    {
+      "epoch": 1.8064337215751525,
+      "grad_norm": 0.20450453460216522,
+      "learning_rate": 5.911496533712577e-07,
+      "loss": 0.5232,
+      "step": 6514
+    },
+    {
+      "epoch": 1.8067110371602884,
+      "grad_norm": 0.20697475969791412,
+      "learning_rate": 5.894728595303101e-07,
+      "loss": 0.4895,
+      "step": 6515
+    },
+    {
+      "epoch": 1.8069883527454242,
+      "grad_norm": 0.20897066593170166,
+      "learning_rate": 5.877983897487699e-07,
+      "loss": 0.4692,
+      "step": 6516
+    },
+    {
+      "epoch": 1.8072656683305601,
+      "grad_norm": 0.2387371063232422,
+      "learning_rate": 5.861262443533716e-07,
+      "loss": 0.4805,
+      "step": 6517
+    },
+    {
+      "epoch": 1.807542983915696,
+      "grad_norm": 0.2114652842283249,
+      "learning_rate": 5.844564236703972e-07,
+      "loss": 0.5114,
+      "step": 6518
+    },
+    {
+      "epoch": 1.807820299500832,
+      "grad_norm": 0.19332511723041534,
+      "learning_rate": 5.827889280256696e-07,
+      "loss": 0.4668,
+      "step": 6519
+    },
+    {
+      "epoch": 1.8080976150859678,
+      "grad_norm": 0.22820791602134705,
+      "learning_rate": 5.811237577445616e-07,
+      "loss": 0.4955,
+      "step": 6520
+    },
+    {
+      "epoch": 1.8083749306711037,
+      "grad_norm": 0.2099064439535141,
+      "learning_rate": 5.79460913151994e-07,
+      "loss": 0.529,
+      "step": 6521
+    },
+    {
+      "epoch": 1.8086522462562395,
+      "grad_norm": 0.22128312289714813,
+      "learning_rate": 5.778003945724322e-07,
+      "loss": 0.4778,
+      "step": 6522
+    },
+    {
+      "epoch": 1.8089295618413754,
+      "grad_norm": 0.20738999545574188,
+      "learning_rate": 5.761422023298851e-07,
+      "loss": 0.4965,
+      "step": 6523
+    },
+    {
+      "epoch": 1.8092068774265113,
+      "grad_norm": 0.23788759112358093,
+      "learning_rate": 5.744863367479092e-07,
+      "loss": 0.5113,
+      "step": 6524
+    },
+    {
+      "epoch": 1.8094841930116472,
+      "grad_norm": 0.21730081737041473,
+      "learning_rate": 5.728327981496112e-07,
+      "loss": 0.4981,
+      "step": 6525
+    },
+    {
+      "epoch": 1.809761508596783,
+      "grad_norm": 0.21804404258728027,
+      "learning_rate": 5.711815868576401e-07,
+      "loss": 0.4849,
+      "step": 6526
+    },
+    {
+      "epoch": 1.810038824181919,
+      "grad_norm": 0.21189890801906586,
+      "learning_rate": 5.695327031941866e-07,
+      "loss": 0.4909,
+      "step": 6527
+    },
+    {
+      "epoch": 1.8103161397670549,
+      "grad_norm": 0.2018709033727646,
+      "learning_rate": 5.678861474809949e-07,
+      "loss": 0.4799,
+      "step": 6528
+    },
+    {
+      "epoch": 1.8105934553521907,
+      "grad_norm": 0.22132818400859833,
+      "learning_rate": 5.662419200393537e-07,
+      "loss": 0.4723,
+      "step": 6529
+    },
+    {
+      "epoch": 1.8108707709373266,
+      "grad_norm": 0.2106301486492157,
+      "learning_rate": 5.646000211900925e-07,
+      "loss": 0.4924,
+      "step": 6530
+    },
+    {
+      "epoch": 1.8111480865224625,
+      "grad_norm": 0.20907257497310638,
+      "learning_rate": 5.629604512535871e-07,
+      "loss": 0.497,
+      "step": 6531
+    },
+    {
+      "epoch": 1.8114254021075984,
+      "grad_norm": 0.20625951886177063,
+      "learning_rate": 5.613232105497649e-07,
+      "loss": 0.504,
+      "step": 6532
+    },
+    {
+      "epoch": 1.8117027176927343,
+      "grad_norm": 0.2099820077419281,
+      "learning_rate": 5.596882993980937e-07,
+      "loss": 0.4957,
+      "step": 6533
+    },
+    {
+      "epoch": 1.8119800332778702,
+      "grad_norm": 0.21241778135299683,
+      "learning_rate": 5.580557181175893e-07,
+      "loss": 0.4908,
+      "step": 6534
+    },
+    {
+      "epoch": 1.812257348863006,
+      "grad_norm": 0.20039701461791992,
+      "learning_rate": 5.564254670268068e-07,
+      "loss": 0.4982,
+      "step": 6535
+    },
+    {
+      "epoch": 1.8125346644481422,
+      "grad_norm": 0.22225303947925568,
+      "learning_rate": 5.547975464438568e-07,
+      "loss": 0.5102,
+      "step": 6536
+    },
+    {
+      "epoch": 1.812811980033278,
+      "grad_norm": 0.21028605103492737,
+      "learning_rate": 5.53171956686388e-07,
+      "loss": 0.5236,
+      "step": 6537
+    },
+    {
+      "epoch": 1.813089295618414,
+      "grad_norm": 0.2042209506034851,
+      "learning_rate": 5.51548698071594e-07,
+      "loss": 0.4895,
+      "step": 6538
+    },
+    {
+      "epoch": 1.8133666112035498,
+      "grad_norm": 0.21611033380031586,
+      "learning_rate": 5.499277709162171e-07,
+      "loss": 0.4968,
+      "step": 6539
+    },
+    {
+      "epoch": 1.8136439267886857,
+      "grad_norm": 0.20384977757930756,
+      "learning_rate": 5.483091755365461e-07,
+      "loss": 0.492,
+      "step": 6540
+    },
+    {
+      "epoch": 1.8139212423738216,
+      "grad_norm": 0.1995922178030014,
+      "learning_rate": 5.466929122484075e-07,
+      "loss": 0.4917,
+      "step": 6541
+    },
+    {
+      "epoch": 1.8141985579589575,
+      "grad_norm": 0.20135626196861267,
+      "learning_rate": 5.450789813671781e-07,
+      "loss": 0.5002,
+      "step": 6542
+    },
+    {
+      "epoch": 1.8144758735440933,
+      "grad_norm": 0.21459290385246277,
+      "learning_rate": 5.434673832077783e-07,
+      "loss": 0.4684,
+      "step": 6543
+    },
+    {
+      "epoch": 1.8147531891292292,
+      "grad_norm": 0.2142772525548935,
+      "learning_rate": 5.41858118084676e-07,
+      "loss": 0.4991,
+      "step": 6544
+    },
+    {
+      "epoch": 1.8150305047143651,
+      "grad_norm": 0.20316624641418457,
+      "learning_rate": 5.402511863118798e-07,
+      "loss": 0.4978,
+      "step": 6545
+    },
+    {
+      "epoch": 1.815307820299501,
+      "grad_norm": 0.23183675110340118,
+      "learning_rate": 5.386465882029443e-07,
+      "loss": 0.4948,
+      "step": 6546
+    },
+    {
+      "epoch": 1.8155851358846369,
+      "grad_norm": 0.2099410742521286,
+      "learning_rate": 5.370443240709691e-07,
+      "loss": 0.5037,
+      "step": 6547
+    },
+    {
+      "epoch": 1.8158624514697728,
+      "grad_norm": 0.2161746770143509,
+      "learning_rate": 5.354443942285986e-07,
+      "loss": 0.5388,
+      "step": 6548
+    },
+    {
+      "epoch": 1.8161397670549086,
+      "grad_norm": 0.2113848328590393,
+      "learning_rate": 5.338467989880233e-07,
+      "loss": 0.5012,
+      "step": 6549
+    },
+    {
+      "epoch": 1.8164170826400445,
+      "grad_norm": 0.20695488154888153,
+      "learning_rate": 5.322515386609731e-07,
+      "loss": 0.5092,
+      "step": 6550
+    },
+    {
+      "epoch": 1.8166943982251804,
+      "grad_norm": 0.22753605246543884,
+      "learning_rate": 5.30658613558728e-07,
+      "loss": 0.4784,
+      "step": 6551
+    },
+    {
+      "epoch": 1.8169717138103163,
+      "grad_norm": 0.20156748592853546,
+      "learning_rate": 5.290680239921089e-07,
+      "loss": 0.4905,
+      "step": 6552
+    },
+    {
+      "epoch": 1.8172490293954522,
+      "grad_norm": 0.20030784606933594,
+      "learning_rate": 5.274797702714829e-07,
+      "loss": 0.4789,
+      "step": 6553
+    },
+    {
+      "epoch": 1.817526344980588,
+      "grad_norm": 0.2246222347021103,
+      "learning_rate": 5.258938527067575e-07,
+      "loss": 0.5038,
+      "step": 6554
+    },
+    {
+      "epoch": 1.817803660565724,
+      "grad_norm": 0.21242398023605347,
+      "learning_rate": 5.243102716073908e-07,
+      "loss": 0.4957,
+      "step": 6555
+    },
+    {
+      "epoch": 1.8180809761508598,
+      "grad_norm": 0.20436997711658478,
+      "learning_rate": 5.227290272823801e-07,
+      "loss": 0.4759,
+      "step": 6556
+    },
+    {
+      "epoch": 1.8183582917359957,
+      "grad_norm": 0.20299552381038666,
+      "learning_rate": 5.211501200402688e-07,
+      "loss": 0.5054,
+      "step": 6557
+    },
+    {
+      "epoch": 1.8186356073211316,
+      "grad_norm": 0.2033611238002777,
+      "learning_rate": 5.195735501891424e-07,
+      "loss": 0.4932,
+      "step": 6558
+    },
+    {
+      "epoch": 1.8189129229062675,
+      "grad_norm": 0.2847311794757843,
+      "learning_rate": 5.1799931803663e-07,
+      "loss": 0.4986,
+      "step": 6559
+    },
+    {
+      "epoch": 1.8191902384914034,
+      "grad_norm": 0.2158021777868271,
+      "learning_rate": 5.164274238899091e-07,
+      "loss": 0.4769,
+      "step": 6560
+    },
+    {
+      "epoch": 1.8194675540765393,
+      "grad_norm": 0.21182480454444885,
+      "learning_rate": 5.148578680556987e-07,
+      "loss": 0.4827,
+      "step": 6561
+    },
+    {
+      "epoch": 1.8197448696616751,
+      "grad_norm": 0.20851145684719086,
+      "learning_rate": 5.132906508402535e-07,
+      "loss": 0.512,
+      "step": 6562
+    },
+    {
+      "epoch": 1.820022185246811,
+      "grad_norm": 0.20874373614788055,
+      "learning_rate": 5.117257725493874e-07,
+      "loss": 0.4721,
+      "step": 6563
+    },
+    {
+      "epoch": 1.820299500831947,
+      "grad_norm": 0.203582301735878,
+      "learning_rate": 5.101632334884476e-07,
+      "loss": 0.467,
+      "step": 6564
+    },
+    {
+      "epoch": 1.8205768164170828,
+      "grad_norm": 0.21535137295722961,
+      "learning_rate": 5.086030339623237e-07,
+      "loss": 0.477,
+      "step": 6565
+    },
+    {
+      "epoch": 1.8208541320022187,
+      "grad_norm": 0.2023378312587738,
+      "learning_rate": 5.070451742754528e-07,
+      "loss": 0.4867,
+      "step": 6566
+    },
+    {
+      "epoch": 1.8211314475873546,
+      "grad_norm": 0.20345695316791534,
+      "learning_rate": 5.054896547318181e-07,
+      "loss": 0.4803,
+      "step": 6567
+    },
+    {
+      "epoch": 1.8214087631724905,
+      "grad_norm": 0.2040456086397171,
+      "learning_rate": 5.039364756349405e-07,
+      "loss": 0.5076,
+      "step": 6568
+    },
+    {
+      "epoch": 1.8216860787576263,
+      "grad_norm": 0.20643432438373566,
+      "learning_rate": 5.023856372878846e-07,
+      "loss": 0.51,
+      "step": 6569
+    },
+    {
+      "epoch": 1.8219633943427622,
+      "grad_norm": 0.1972525417804718,
+      "learning_rate": 5.008371399932613e-07,
+      "loss": 0.4995,
+      "step": 6570
+    },
+    {
+      "epoch": 1.822240709927898,
+      "grad_norm": 0.27110210061073303,
+      "learning_rate": 4.992909840532259e-07,
+      "loss": 0.505,
+      "step": 6571
+    },
+    {
+      "epoch": 1.822518025513034,
+      "grad_norm": 0.20463772118091583,
+      "learning_rate": 4.977471697694719e-07,
+      "loss": 0.5256,
+      "step": 6572
+    },
+    {
+      "epoch": 1.8227953410981699,
+      "grad_norm": 0.20936624705791473,
+      "learning_rate": 4.962056974432374e-07,
+      "loss": 0.4791,
+      "step": 6573
+    },
+    {
+      "epoch": 1.8230726566833058,
+      "grad_norm": 0.21320702135562897,
+      "learning_rate": 4.946665673753056e-07,
+      "loss": 0.5057,
+      "step": 6574
+    },
+    {
+      "epoch": 1.8233499722684416,
+      "grad_norm": 0.20866161584854126,
+      "learning_rate": 4.931297798660043e-07,
+      "loss": 0.4935,
+      "step": 6575
+    },
+    {
+      "epoch": 1.8236272878535775,
+      "grad_norm": 0.21233926713466644,
+      "learning_rate": 4.915953352151961e-07,
+      "loss": 0.5105,
+      "step": 6576
+    },
+    {
+      "epoch": 1.8239046034387134,
+      "grad_norm": 0.20736192166805267,
+      "learning_rate": 4.900632337222947e-07,
+      "loss": 0.5026,
+      "step": 6577
+    },
+    {
+      "epoch": 1.8241819190238493,
+      "grad_norm": 0.21084974706172943,
+      "learning_rate": 4.885334756862564e-07,
+      "loss": 0.4946,
+      "step": 6578
+    },
+    {
+      "epoch": 1.8244592346089852,
+      "grad_norm": 0.20611052215099335,
+      "learning_rate": 4.870060614055733e-07,
+      "loss": 0.5124,
+      "step": 6579
+    },
+    {
+      "epoch": 1.824736550194121,
+      "grad_norm": 0.19793611764907837,
+      "learning_rate": 4.85480991178286e-07,
+      "loss": 0.4918,
+      "step": 6580
+    },
+    {
+      "epoch": 1.825013865779257,
+      "grad_norm": 0.20998556911945343,
+      "learning_rate": 4.839582653019745e-07,
+      "loss": 0.4959,
+      "step": 6581
+    },
+    {
+      "epoch": 1.8252911813643928,
+      "grad_norm": 0.2145131677389145,
+      "learning_rate": 4.824378840737664e-07,
+      "loss": 0.4858,
+      "step": 6582
+    },
+    {
+      "epoch": 1.8255684969495287,
+      "grad_norm": 0.20194010436534882,
+      "learning_rate": 4.809198477903259e-07,
+      "loss": 0.4685,
+      "step": 6583
+    },
+    {
+      "epoch": 1.8258458125346646,
+      "grad_norm": 0.2031271904706955,
+      "learning_rate": 4.794041567478632e-07,
+      "loss": 0.4833,
+      "step": 6584
+    },
+    {
+      "epoch": 1.8261231281198005,
+      "grad_norm": 0.2092730849981308,
+      "learning_rate": 4.778908112421279e-07,
+      "loss": 0.4969,
+      "step": 6585
+    },
+    {
+      "epoch": 1.8264004437049364,
+      "grad_norm": 0.21413682401180267,
+      "learning_rate": 4.7637981156841563e-07,
+      "loss": 0.4964,
+      "step": 6586
+    },
+    {
+      "epoch": 1.8266777592900723,
+      "grad_norm": 0.2138873189687729,
+      "learning_rate": 4.7487115802156147e-07,
+      "loss": 0.484,
+      "step": 6587
+    },
+    {
+      "epoch": 1.8269550748752081,
+      "grad_norm": 0.23256392776966095,
+      "learning_rate": 4.733648508959465e-07,
+      "loss": 0.4868,
+      "step": 6588
+    },
+    {
+      "epoch": 1.827232390460344,
+      "grad_norm": 0.20839428901672363,
+      "learning_rate": 4.718608904854857e-07,
+      "loss": 0.4998,
+      "step": 6589
+    },
+    {
+      "epoch": 1.82750970604548,
+      "grad_norm": 0.21245619654655457,
+      "learning_rate": 4.703592770836457e-07,
+      "loss": 0.4976,
+      "step": 6590
+    },
+    {
+      "epoch": 1.8277870216306158,
+      "grad_norm": 0.21361759305000305,
+      "learning_rate": 4.6886001098343094e-07,
+      "loss": 0.5143,
+      "step": 6591
+    },
+    {
+      "epoch": 1.8280643372157517,
+      "grad_norm": 0.20406369864940643,
+      "learning_rate": 4.673630924773853e-07,
+      "loss": 0.4561,
+      "step": 6592
+    },
+    {
+      "epoch": 1.8283416528008876,
+      "grad_norm": 0.20464986562728882,
+      "learning_rate": 4.6586852185760144e-07,
+      "loss": 0.4884,
+      "step": 6593
+    },
+    {
+      "epoch": 1.8286189683860234,
+      "grad_norm": 0.21941900253295898,
+      "learning_rate": 4.643762994157058e-07,
+      "loss": 0.5124,
+      "step": 6594
+    },
+    {
+      "epoch": 1.8288962839711593,
+      "grad_norm": 0.21227768063545227,
+      "learning_rate": 4.628864254428725e-07,
+      "loss": 0.4913,
+      "step": 6595
+    },
+    {
+      "epoch": 1.8291735995562952,
+      "grad_norm": 0.20707818865776062,
+      "learning_rate": 4.613989002298133e-07,
+      "loss": 0.4637,
+      "step": 6596
+    },
+    {
+      "epoch": 1.829450915141431,
+      "grad_norm": 0.21716617047786713,
+      "learning_rate": 4.599137240667864e-07,
+      "loss": 0.4596,
+      "step": 6597
+    },
+    {
+      "epoch": 1.829728230726567,
+      "grad_norm": 0.21836552023887634,
+      "learning_rate": 4.5843089724358913e-07,
+      "loss": 0.4907,
+      "step": 6598
+    },
+    {
+      "epoch": 1.8300055463117029,
+      "grad_norm": 0.20570078492164612,
+      "learning_rate": 4.5695042004955943e-07,
+      "loss": 0.4854,
+      "step": 6599
+    },
+    {
+      "epoch": 1.8302828618968388,
+      "grad_norm": 0.21633854508399963,
+      "learning_rate": 4.554722927735747e-07,
+      "loss": 0.517,
+      "step": 6600
+    },
+    {
+      "epoch": 1.8305601774819746,
+      "grad_norm": 0.20889052748680115,
+      "learning_rate": 4.53996515704061e-07,
+      "loss": 0.5119,
+      "step": 6601
+    },
+    {
+      "epoch": 1.8308374930671105,
+      "grad_norm": 0.21190080046653748,
+      "learning_rate": 4.5252308912897973e-07,
+      "loss": 0.492,
+      "step": 6602
+    },
+    {
+      "epoch": 1.8311148086522464,
+      "grad_norm": 0.19486699998378754,
+      "learning_rate": 4.5105201333583565e-07,
+      "loss": 0.4603,
+      "step": 6603
+    },
+    {
+      "epoch": 1.8313921242373823,
+      "grad_norm": 0.20805886387825012,
+      "learning_rate": 4.495832886116741e-07,
+      "loss": 0.5107,
+      "step": 6604
+    },
+    {
+      "epoch": 1.8316694398225182,
+      "grad_norm": 0.2246580719947815,
+      "learning_rate": 4.481169152430839e-07,
+      "loss": 0.5245,
+      "step": 6605
+    },
+    {
+      "epoch": 1.831946755407654,
+      "grad_norm": 0.2717232406139374,
+      "learning_rate": 4.466528935161918e-07,
+      "loss": 0.5234,
+      "step": 6606
+    },
+    {
+      "epoch": 1.83222407099279,
+      "grad_norm": 0.2078229784965515,
+      "learning_rate": 4.451912237166664e-07,
+      "loss": 0.4893,
+      "step": 6607
+    },
+    {
+      "epoch": 1.8325013865779258,
+      "grad_norm": 0.20676632225513458,
+      "learning_rate": 4.4373190612971986e-07,
+      "loss": 0.5072,
+      "step": 6608
+    },
+    {
+      "epoch": 1.8327787021630617,
+      "grad_norm": 0.21076683700084686,
+      "learning_rate": 4.4227494104010503e-07,
+      "loss": 0.5058,
+      "step": 6609
+    },
+    {
+      "epoch": 1.8330560177481976,
+      "grad_norm": 0.20716801285743713,
+      "learning_rate": 4.408203287321111e-07,
+      "loss": 0.4752,
+      "step": 6610
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 0.20303776860237122,
+      "learning_rate": 4.3936806948957354e-07,
+      "loss": 0.4991,
+      "step": 6611
+    },
+    {
+      "epoch": 1.8336106489184694,
+      "grad_norm": 0.2012963891029358,
+      "learning_rate": 4.379181635958643e-07,
+      "loss": 0.486,
+      "step": 6612
+    },
+    {
+      "epoch": 1.8338879645036053,
+      "grad_norm": 0.20308934152126312,
+      "learning_rate": 4.3647061133390286e-07,
+      "loss": 0.506,
+      "step": 6613
+    },
+    {
+      "epoch": 1.8341652800887411,
+      "grad_norm": 0.20220200717449188,
+      "learning_rate": 4.3502541298613977e-07,
+      "loss": 0.4809,
+      "step": 6614
+    },
+    {
+      "epoch": 1.834442595673877,
+      "grad_norm": 0.20746108889579773,
+      "learning_rate": 4.335825688345743e-07,
+      "loss": 0.4813,
+      "step": 6615
+    },
+    {
+      "epoch": 1.834719911259013,
+      "grad_norm": 0.22002127766609192,
+      "learning_rate": 4.321420791607453e-07,
+      "loss": 0.5233,
+      "step": 6616
+    },
+    {
+      "epoch": 1.8349972268441488,
+      "grad_norm": 0.20879393815994263,
+      "learning_rate": 4.307039442457278e-07,
+      "loss": 0.4956,
+      "step": 6617
+    },
+    {
+      "epoch": 1.8352745424292847,
+      "grad_norm": 0.21325801312923431,
+      "learning_rate": 4.2926816437014047e-07,
+      "loss": 0.4913,
+      "step": 6618
+    },
+    {
+      "epoch": 1.8355518580144206,
+      "grad_norm": 0.19979670643806458,
+      "learning_rate": 4.278347398141411e-07,
+      "loss": 0.5124,
+      "step": 6619
+    },
+    {
+      "epoch": 1.8358291735995564,
+      "grad_norm": 0.21641233563423157,
+      "learning_rate": 4.264036708574323e-07,
+      "loss": 0.505,
+      "step": 6620
+    },
+    {
+      "epoch": 1.8361064891846923,
+      "grad_norm": 0.2334747463464737,
+      "learning_rate": 4.249749577792492e-07,
+      "loss": 0.5114,
+      "step": 6621
+    },
+    {
+      "epoch": 1.8363838047698282,
+      "grad_norm": 0.19688747823238373,
+      "learning_rate": 4.2354860085837414e-07,
+      "loss": 0.5072,
+      "step": 6622
+    },
+    {
+      "epoch": 1.836661120354964,
+      "grad_norm": 0.21270081400871277,
+      "learning_rate": 4.2212460037312636e-07,
+      "loss": 0.5024,
+      "step": 6623
+    },
+    {
+      "epoch": 1.8369384359401,
+      "grad_norm": 0.21791474521160126,
+      "learning_rate": 4.2070295660136817e-07,
+      "loss": 0.5123,
+      "step": 6624
+    },
+    {
+      "epoch": 1.8372157515252359,
+      "grad_norm": 0.1985998898744583,
+      "learning_rate": 4.192836698204958e-07,
+      "loss": 0.4741,
+      "step": 6625
+    },
+    {
+      "epoch": 1.8374930671103717,
+      "grad_norm": 0.2119298279285431,
+      "learning_rate": 4.1786674030745295e-07,
+      "loss": 0.4941,
+      "step": 6626
+    },
+    {
+      "epoch": 1.8377703826955076,
+      "grad_norm": 0.20587483048439026,
+      "learning_rate": 4.164521683387185e-07,
+      "loss": 0.5002,
+      "step": 6627
+    },
+    {
+      "epoch": 1.8380476982806435,
+      "grad_norm": 0.19921058416366577,
+      "learning_rate": 4.1503995419031325e-07,
+      "loss": 0.4908,
+      "step": 6628
+    },
+    {
+      "epoch": 1.8383250138657794,
+      "grad_norm": 0.2109578251838684,
+      "learning_rate": 4.1363009813780026e-07,
+      "loss": 0.4987,
+      "step": 6629
+    },
+    {
+      "epoch": 1.8386023294509153,
+      "grad_norm": 0.21179182827472687,
+      "learning_rate": 4.122226004562746e-07,
+      "loss": 0.5024,
+      "step": 6630
+    },
+    {
+      "epoch": 1.8388796450360512,
+      "grad_norm": 0.20858225226402283,
+      "learning_rate": 4.108174614203819e-07,
+      "loss": 0.517,
+      "step": 6631
+    },
+    {
+      "epoch": 1.839156960621187,
+      "grad_norm": 0.2038223147392273,
+      "learning_rate": 4.094146813042973e-07,
+      "loss": 0.448,
+      "step": 6632
+    },
+    {
+      "epoch": 1.839434276206323,
+      "grad_norm": 0.21669189631938934,
+      "learning_rate": 4.0801426038174357e-07,
+      "loss": 0.4757,
+      "step": 6633
+    },
+    {
+      "epoch": 1.8397115917914588,
+      "grad_norm": 0.204112246632576,
+      "learning_rate": 4.06616198925977e-07,
+      "loss": 0.4729,
+      "step": 6634
+    },
+    {
+      "epoch": 1.8399889073765947,
+      "grad_norm": 0.22220736742019653,
+      "learning_rate": 4.052204972097989e-07,
+      "loss": 0.5062,
+      "step": 6635
+    },
+    {
+      "epoch": 1.8402662229617306,
+      "grad_norm": 0.2096226066350937,
+      "learning_rate": 4.0382715550554837e-07,
+      "loss": 0.4841,
+      "step": 6636
+    },
+    {
+      "epoch": 1.8405435385468665,
+      "grad_norm": 0.20454815030097961,
+      "learning_rate": 4.024361740851024e-07,
+      "loss": 0.4883,
+      "step": 6637
+    },
+    {
+      "epoch": 1.8408208541320024,
+      "grad_norm": 0.20967929065227509,
+      "learning_rate": 4.010475532198757e-07,
+      "loss": 0.5155,
+      "step": 6638
+    },
+    {
+      "epoch": 1.8410981697171382,
+      "grad_norm": 0.22011229395866394,
+      "learning_rate": 3.996612931808266e-07,
+      "loss": 0.5232,
+      "step": 6639
+    },
+    {
+      "epoch": 1.8413754853022741,
+      "grad_norm": 0.22266606986522675,
+      "learning_rate": 3.9827739423845265e-07,
+      "loss": 0.5071,
+      "step": 6640
+    },
+    {
+      "epoch": 1.84165280088741,
+      "grad_norm": 0.20759738981723785,
+      "learning_rate": 3.9689585666278784e-07,
+      "loss": 0.4825,
+      "step": 6641
+    },
+    {
+      "epoch": 1.841930116472546,
+      "grad_norm": 0.2150697559118271,
+      "learning_rate": 3.9551668072340675e-07,
+      "loss": 0.511,
+      "step": 6642
+    },
+    {
+      "epoch": 1.8422074320576818,
+      "grad_norm": 0.21445181965827942,
+      "learning_rate": 3.9413986668942473e-07,
+      "loss": 0.4903,
+      "step": 6643
+    },
+    {
+      "epoch": 1.8424847476428177,
+      "grad_norm": 0.2114182561635971,
+      "learning_rate": 3.9276541482949347e-07,
+      "loss": 0.51,
+      "step": 6644
+    },
+    {
+      "epoch": 1.8427620632279536,
+      "grad_norm": 0.2046855241060257,
+      "learning_rate": 3.913933254118041e-07,
+      "loss": 0.5055,
+      "step": 6645
+    },
+    {
+      "epoch": 1.8430393788130894,
+      "grad_norm": 0.20602397620677948,
+      "learning_rate": 3.9002359870408817e-07,
+      "loss": 0.5024,
+      "step": 6646
+    },
+    {
+      "epoch": 1.8433166943982253,
+      "grad_norm": 0.21433816850185394,
+      "learning_rate": 3.886562349736167e-07,
+      "loss": 0.4955,
+      "step": 6647
+    },
+    {
+      "epoch": 1.8435940099833612,
+      "grad_norm": 0.21116018295288086,
+      "learning_rate": 3.872912344871985e-07,
+      "loss": 0.4696,
+      "step": 6648
+    },
+    {
+      "epoch": 1.843871325568497,
+      "grad_norm": 0.20629142224788666,
+      "learning_rate": 3.8592859751117873e-07,
+      "loss": 0.4636,
+      "step": 6649
+    },
+    {
+      "epoch": 1.844148641153633,
+      "grad_norm": 0.2071436494588852,
+      "learning_rate": 3.845683243114462e-07,
+      "loss": 0.5121,
+      "step": 6650
+    },
+    {
+      "epoch": 1.8444259567387689,
+      "grad_norm": 0.20660892128944397,
+      "learning_rate": 3.8321041515342744e-07,
+      "loss": 0.5133,
+      "step": 6651
+    },
+    {
+      "epoch": 1.8447032723239047,
+      "grad_norm": 0.19560420513153076,
+      "learning_rate": 3.818548703020841e-07,
+      "loss": 0.5027,
+      "step": 6652
+    },
+    {
+      "epoch": 1.8449805879090406,
+      "grad_norm": 0.20409871637821198,
+      "learning_rate": 3.805016900219172e-07,
+      "loss": 0.507,
+      "step": 6653
+    },
+    {
+      "epoch": 1.8452579034941765,
+      "grad_norm": 0.21331243216991425,
+      "learning_rate": 3.791508745769737e-07,
+      "loss": 0.5273,
+      "step": 6654
+    },
+    {
+      "epoch": 1.8455352190793124,
+      "grad_norm": 0.2074180394411087,
+      "learning_rate": 3.7780242423083036e-07,
+      "loss": 0.4921,
+      "step": 6655
+    },
+    {
+      "epoch": 1.8458125346644483,
+      "grad_norm": 0.22154366970062256,
+      "learning_rate": 3.7645633924660446e-07,
+      "loss": 0.5068,
+      "step": 6656
+    },
+    {
+      "epoch": 1.8460898502495842,
+      "grad_norm": 0.20785929262638092,
+      "learning_rate": 3.7511261988695393e-07,
+      "loss": 0.4947,
+      "step": 6657
+    },
+    {
+      "epoch": 1.84636716583472,
+      "grad_norm": 0.22468653321266174,
+      "learning_rate": 3.737712664140747e-07,
+      "loss": 0.5096,
+      "step": 6658
+    },
+    {
+      "epoch": 1.846644481419856,
+      "grad_norm": 0.2090657502412796,
+      "learning_rate": 3.724322790897003e-07,
+      "loss": 0.5105,
+      "step": 6659
+    },
+    {
+      "epoch": 1.8469217970049918,
+      "grad_norm": 0.20280075073242188,
+      "learning_rate": 3.71095658175101e-07,
+      "loss": 0.4872,
+      "step": 6660
+    },
+    {
+      "epoch": 1.8471991125901277,
+      "grad_norm": 0.2233634740114212,
+      "learning_rate": 3.697614039310876e-07,
+      "loss": 0.506,
+      "step": 6661
+    },
+    {
+      "epoch": 1.8474764281752636,
+      "grad_norm": 0.2086162120103836,
+      "learning_rate": 3.684295166180102e-07,
+      "loss": 0.5089,
+      "step": 6662
+    },
+    {
+      "epoch": 1.8477537437603995,
+      "grad_norm": 0.20018382370471954,
+      "learning_rate": 3.6709999649575386e-07,
+      "loss": 0.5051,
+      "step": 6663
+    },
+    {
+      "epoch": 1.8480310593455354,
+      "grad_norm": 0.21848781406879425,
+      "learning_rate": 3.6577284382374316e-07,
+      "loss": 0.497,
+      "step": 6664
+    },
+    {
+      "epoch": 1.8483083749306712,
+      "grad_norm": 0.24338483810424805,
+      "learning_rate": 3.644480588609403e-07,
+      "loss": 0.5335,
+      "step": 6665
+    },
+    {
+      "epoch": 1.8485856905158071,
+      "grad_norm": 0.21215735375881195,
+      "learning_rate": 3.6312564186584826e-07,
+      "loss": 0.5265,
+      "step": 6666
+    },
+    {
+      "epoch": 1.848863006100943,
+      "grad_norm": 0.22368744015693665,
+      "learning_rate": 3.6180559309650086e-07,
+      "loss": 0.5033,
+      "step": 6667
+    },
+    {
+      "epoch": 1.849140321686079,
+      "grad_norm": 0.20563702285289764,
+      "learning_rate": 3.6048791281047963e-07,
+      "loss": 0.4938,
+      "step": 6668
+    },
+    {
+      "epoch": 1.8494176372712148,
+      "grad_norm": 0.21737602353096008,
+      "learning_rate": 3.5917260126489687e-07,
+      "loss": 0.5033,
+      "step": 6669
+    },
+    {
+      "epoch": 1.8496949528563507,
+      "grad_norm": 0.32653746008872986,
+      "learning_rate": 3.578596587164043e-07,
+      "loss": 0.517,
+      "step": 6670
+    },
+    {
+      "epoch": 1.8499722684414865,
+      "grad_norm": 0.45970603823661804,
+      "learning_rate": 3.565490854211928e-07,
+      "loss": 0.5046,
+      "step": 6671
+    },
+    {
+      "epoch": 1.8502495840266224,
+      "grad_norm": 0.20332348346710205,
+      "learning_rate": 3.552408816349884e-07,
+      "loss": 0.4842,
+      "step": 6672
+    },
+    {
+      "epoch": 1.8505268996117583,
+      "grad_norm": 0.20179514586925507,
+      "learning_rate": 3.5393504761305903e-07,
+      "loss": 0.4844,
+      "step": 6673
+    },
+    {
+      "epoch": 1.8508042151968942,
+      "grad_norm": 0.20190241932868958,
+      "learning_rate": 3.5263158361020373e-07,
+      "loss": 0.4805,
+      "step": 6674
+    },
+    {
+      "epoch": 1.85108153078203,
+      "grad_norm": 0.2159728854894638,
+      "learning_rate": 3.513304898807676e-07,
+      "loss": 0.5239,
+      "step": 6675
+    },
+    {
+      "epoch": 1.851358846367166,
+      "grad_norm": 0.20452584326267242,
+      "learning_rate": 3.5003176667862265e-07,
+      "loss": 0.4694,
+      "step": 6676
+    },
+    {
+      "epoch": 1.8516361619523019,
+      "grad_norm": 0.19299055635929108,
+      "learning_rate": 3.487354142571883e-07,
+      "loss": 0.5076,
+      "step": 6677
+    },
+    {
+      "epoch": 1.8519134775374377,
+      "grad_norm": 0.20924656093120575,
+      "learning_rate": 3.474414328694178e-07,
+      "loss": 0.502,
+      "step": 6678
+    },
+    {
+      "epoch": 1.8521907931225736,
+      "grad_norm": 1.066011905670166,
+      "learning_rate": 3.4614982276779953e-07,
+      "loss": 0.4964,
+      "step": 6679
+    },
+    {
+      "epoch": 1.8524681087077095,
+      "grad_norm": 0.21032501757144928,
+      "learning_rate": 3.448605842043581e-07,
+      "loss": 0.4718,
+      "step": 6680
+    },
+    {
+      "epoch": 1.8527454242928454,
+      "grad_norm": 0.216018944978714,
+      "learning_rate": 3.435737174306633e-07,
+      "loss": 0.5178,
+      "step": 6681
+    },
+    {
+      "epoch": 1.8530227398779813,
+      "grad_norm": 0.2162523716688156,
+      "learning_rate": 3.4228922269781556e-07,
+      "loss": 0.4962,
+      "step": 6682
+    },
+    {
+      "epoch": 1.8533000554631172,
+      "grad_norm": 0.20311126112937927,
+      "learning_rate": 3.410071002564519e-07,
+      "loss": 0.5093,
+      "step": 6683
+    },
+    {
+      "epoch": 1.853577371048253,
+      "grad_norm": 0.21295960247516632,
+      "learning_rate": 3.397273503567486e-07,
+      "loss": 0.4753,
+      "step": 6684
+    },
+    {
+      "epoch": 1.853854686633389,
+      "grad_norm": 0.21265694499015808,
+      "learning_rate": 3.3844997324842113e-07,
+      "loss": 0.4974,
+      "step": 6685
+    },
+    {
+      "epoch": 1.8541320022185248,
+      "grad_norm": 0.21514546871185303,
+      "learning_rate": 3.3717496918071746e-07,
+      "loss": 0.4879,
+      "step": 6686
+    },
+    {
+      "epoch": 1.8544093178036607,
+      "grad_norm": 0.23401249945163727,
+      "learning_rate": 3.3590233840242455e-07,
+      "loss": 0.5049,
+      "step": 6687
+    },
+    {
+      "epoch": 1.8546866333887966,
+      "grad_norm": 0.20478412508964539,
+      "learning_rate": 3.346320811618675e-07,
+      "loss": 0.5053,
+      "step": 6688
+    },
+    {
+      "epoch": 1.8549639489739325,
+      "grad_norm": 0.21184539794921875,
+      "learning_rate": 3.333641977069077e-07,
+      "loss": 0.528,
+      "step": 6689
+    },
+    {
+      "epoch": 1.8552412645590683,
+      "grad_norm": 0.1997302621603012,
+      "learning_rate": 3.320986882849417e-07,
+      "loss": 0.4981,
+      "step": 6690
+    },
+    {
+      "epoch": 1.8555185801442042,
+      "grad_norm": 0.21800538897514343,
+      "learning_rate": 3.308355531429011e-07,
+      "loss": 0.5069,
+      "step": 6691
+    },
+    {
+      "epoch": 1.8557958957293401,
+      "grad_norm": 0.2156432867050171,
+      "learning_rate": 3.295747925272638e-07,
+      "loss": 0.5039,
+      "step": 6692
+    },
+    {
+      "epoch": 1.856073211314476,
+      "grad_norm": 0.2117234319448471,
+      "learning_rate": 3.28316406684033e-07,
+      "loss": 0.4791,
+      "step": 6693
+    },
+    {
+      "epoch": 1.8563505268996119,
+      "grad_norm": 0.48360639810562134,
+      "learning_rate": 3.2706039585875257e-07,
+      "loss": 0.4949,
+      "step": 6694
+    },
+    {
+      "epoch": 1.8566278424847478,
+      "grad_norm": 0.20971214771270752,
+      "learning_rate": 3.2580676029650566e-07,
+      "loss": 0.5,
+      "step": 6695
+    },
+    {
+      "epoch": 1.8569051580698837,
+      "grad_norm": 0.20673950016498566,
+      "learning_rate": 3.24555500241909e-07,
+      "loss": 0.5199,
+      "step": 6696
+    },
+    {
+      "epoch": 1.8571824736550195,
+      "grad_norm": 0.20716223120689392,
+      "learning_rate": 3.233066159391174e-07,
+      "loss": 0.5073,
+      "step": 6697
+    },
+    {
+      "epoch": 1.8574597892401554,
+      "grad_norm": 0.2109098881483078,
+      "learning_rate": 3.2206010763181916e-07,
+      "loss": 0.5055,
+      "step": 6698
+    },
+    {
+      "epoch": 1.8577371048252913,
+      "grad_norm": 0.20620295405387878,
+      "learning_rate": 3.20815975563242e-07,
+      "loss": 0.5266,
+      "step": 6699
+    },
+    {
+      "epoch": 1.8580144204104272,
+      "grad_norm": 0.22776873409748077,
+      "learning_rate": 3.1957421997615004e-07,
+      "loss": 0.5076,
+      "step": 6700
+    },
+    {
+      "epoch": 1.858291735995563,
+      "grad_norm": 0.2032913714647293,
+      "learning_rate": 3.1833484111284115e-07,
+      "loss": 0.5065,
+      "step": 6701
+    },
+    {
+      "epoch": 1.858569051580699,
+      "grad_norm": 0.2074134647846222,
+      "learning_rate": 3.1709783921515383e-07,
+      "loss": 0.5079,
+      "step": 6702
+    },
+    {
+      "epoch": 1.8588463671658348,
+      "grad_norm": 0.20687542855739594,
+      "learning_rate": 3.158632145244561e-07,
+      "loss": 0.5004,
+      "step": 6703
+    },
+    {
+      "epoch": 1.8591236827509707,
+      "grad_norm": 0.21574708819389343,
+      "learning_rate": 3.1463096728165944e-07,
+      "loss": 0.5162,
+      "step": 6704
+    },
+    {
+      "epoch": 1.8594009983361066,
+      "grad_norm": 0.2129552662372589,
+      "learning_rate": 3.134010977272048e-07,
+      "loss": 0.505,
+      "step": 6705
+    },
+    {
+      "epoch": 1.8596783139212425,
+      "grad_norm": 0.2198726236820221,
+      "learning_rate": 3.121736061010738e-07,
+      "loss": 0.5004,
+      "step": 6706
+    },
+    {
+      "epoch": 1.8599556295063784,
+      "grad_norm": 0.21903812885284424,
+      "learning_rate": 3.109484926427847e-07,
+      "loss": 0.4909,
+      "step": 6707
+    },
+    {
+      "epoch": 1.8602329450915143,
+      "grad_norm": 0.2174917459487915,
+      "learning_rate": 3.0972575759138503e-07,
+      "loss": 0.4957,
+      "step": 6708
+    },
+    {
+      "epoch": 1.8605102606766502,
+      "grad_norm": 0.2153944969177246,
+      "learning_rate": 3.085054011854674e-07,
+      "loss": 0.4964,
+      "step": 6709
+    },
+    {
+      "epoch": 1.860787576261786,
+      "grad_norm": 0.22760117053985596,
+      "learning_rate": 3.0728742366315233e-07,
+      "loss": 0.4961,
+      "step": 6710
+    },
+    {
+      "epoch": 1.861064891846922,
+      "grad_norm": 0.20543061196804047,
+      "learning_rate": 3.0607182526210115e-07,
+      "loss": 0.4923,
+      "step": 6711
+    },
+    {
+      "epoch": 1.8613422074320578,
+      "grad_norm": 0.2019956409931183,
+      "learning_rate": 3.0485860621950887e-07,
+      "loss": 0.5033,
+      "step": 6712
+    },
+    {
+      "epoch": 1.8616195230171937,
+      "grad_norm": 0.2030191272497177,
+      "learning_rate": 3.036477667721069e-07,
+      "loss": 0.4962,
+      "step": 6713
+    },
+    {
+      "epoch": 1.8618968386023296,
+      "grad_norm": 0.20309413969516754,
+      "learning_rate": 3.024393071561604e-07,
+      "loss": 0.5137,
+      "step": 6714
+    },
+    {
+      "epoch": 1.8621741541874655,
+      "grad_norm": 0.20694345235824585,
+      "learning_rate": 3.0123322760747394e-07,
+      "loss": 0.4906,
+      "step": 6715
+    },
+    {
+      "epoch": 1.8624514697726013,
+      "grad_norm": 0.21089065074920654,
+      "learning_rate": 3.000295283613869e-07,
+      "loss": 0.4833,
+      "step": 6716
+    },
+    {
+      "epoch": 1.8627287853577372,
+      "grad_norm": 0.20231840014457703,
+      "learning_rate": 2.9882820965276975e-07,
+      "loss": 0.5002,
+      "step": 6717
+    },
+    {
+      "epoch": 1.8630061009428731,
+      "grad_norm": 0.20089836418628693,
+      "learning_rate": 2.9762927171603226e-07,
+      "loss": 0.4721,
+      "step": 6718
+    },
+    {
+      "epoch": 1.863283416528009,
+      "grad_norm": 0.25077909231185913,
+      "learning_rate": 2.9643271478511925e-07,
+      "loss": 0.5162,
+      "step": 6719
+    },
+    {
+      "epoch": 1.8635607321131449,
+      "grad_norm": 0.22070035338401794,
+      "learning_rate": 2.952385390935133e-07,
+      "loss": 0.482,
+      "step": 6720
+    },
+    {
+      "epoch": 1.8638380476982808,
+      "grad_norm": 0.21285070478916168,
+      "learning_rate": 2.94046744874224e-07,
+      "loss": 0.5105,
+      "step": 6721
+    },
+    {
+      "epoch": 1.8641153632834166,
+      "grad_norm": 0.20452368259429932,
+      "learning_rate": 2.928573323598069e-07,
+      "loss": 0.4942,
+      "step": 6722
+    },
+    {
+      "epoch": 1.8643926788685525,
+      "grad_norm": 0.21610909700393677,
+      "learning_rate": 2.9167030178234853e-07,
+      "loss": 0.5066,
+      "step": 6723
+    },
+    {
+      "epoch": 1.8646699944536884,
+      "grad_norm": 0.2100083827972412,
+      "learning_rate": 2.904856533734665e-07,
+      "loss": 0.486,
+      "step": 6724
+    },
+    {
+      "epoch": 1.8649473100388243,
+      "grad_norm": 0.211506187915802,
+      "learning_rate": 2.893033873643175e-07,
+      "loss": 0.4986,
+      "step": 6725
+    },
+    {
+      "epoch": 1.8652246256239602,
+      "grad_norm": 0.2029467523097992,
+      "learning_rate": 2.881235039855934e-07,
+      "loss": 0.4879,
+      "step": 6726
+    },
+    {
+      "epoch": 1.865501941209096,
+      "grad_norm": 0.21474742889404297,
+      "learning_rate": 2.8694600346752255e-07,
+      "loss": 0.5127,
+      "step": 6727
+    },
+    {
+      "epoch": 1.865779256794232,
+      "grad_norm": 0.20987072587013245,
+      "learning_rate": 2.857708860398656e-07,
+      "loss": 0.5121,
+      "step": 6728
+    },
+    {
+      "epoch": 1.8660565723793678,
+      "grad_norm": 0.2003088742494583,
+      "learning_rate": 2.845981519319169e-07,
+      "loss": 0.4814,
+      "step": 6729
+    },
+    {
+      "epoch": 1.8663338879645037,
+      "grad_norm": 0.20292720198631287,
+      "learning_rate": 2.834278013725114e-07,
+      "loss": 0.4883,
+      "step": 6730
+    },
+    {
+      "epoch": 1.8666112035496396,
+      "grad_norm": 0.2072405070066452,
+      "learning_rate": 2.8225983459001374e-07,
+      "loss": 0.5012,
+      "step": 6731
+    },
+    {
+      "epoch": 1.8668885191347755,
+      "grad_norm": 0.22093196213245392,
+      "learning_rate": 2.81094251812325e-07,
+      "loss": 0.5124,
+      "step": 6732
+    },
+    {
+      "epoch": 1.8671658347199114,
+      "grad_norm": 0.21116647124290466,
+      "learning_rate": 2.7993105326687975e-07,
+      "loss": 0.4803,
+      "step": 6733
+    },
+    {
+      "epoch": 1.8674431503050473,
+      "grad_norm": 0.20909181237220764,
+      "learning_rate": 2.7877023918065346e-07,
+      "loss": 0.5075,
+      "step": 6734
+    },
+    {
+      "epoch": 1.8677204658901831,
+      "grad_norm": 0.2065388560295105,
+      "learning_rate": 2.776118097801483e-07,
+      "loss": 0.4884,
+      "step": 6735
+    },
+    {
+      "epoch": 1.867997781475319,
+      "grad_norm": 0.2007947862148285,
+      "learning_rate": 2.764557652914029e-07,
+      "loss": 0.4811,
+      "step": 6736
+    },
+    {
+      "epoch": 1.868275097060455,
+      "grad_norm": 0.21440117061138153,
+      "learning_rate": 2.753021059399952e-07,
+      "loss": 0.4785,
+      "step": 6737
+    },
+    {
+      "epoch": 1.8685524126455908,
+      "grad_norm": 0.2125922590494156,
+      "learning_rate": 2.7415083195103525e-07,
+      "loss": 0.473,
+      "step": 6738
+    },
+    {
+      "epoch": 1.8688297282307267,
+      "grad_norm": 0.21997934579849243,
+      "learning_rate": 2.730019435491657e-07,
+      "loss": 0.4944,
+      "step": 6739
+    },
+    {
+      "epoch": 1.8691070438158626,
+      "grad_norm": 0.19915248453617096,
+      "learning_rate": 2.7185544095856413e-07,
+      "loss": 0.4901,
+      "step": 6740
+    },
+    {
+      "epoch": 1.8693843594009985,
+      "grad_norm": 0.21164058148860931,
+      "learning_rate": 2.7071132440294464e-07,
+      "loss": 0.4929,
+      "step": 6741
+    },
+    {
+      "epoch": 1.8696616749861343,
+      "grad_norm": 0.1981211006641388,
+      "learning_rate": 2.695695941055551e-07,
+      "loss": 0.4817,
+      "step": 6742
+    },
+    {
+      "epoch": 1.8699389905712702,
+      "grad_norm": 0.21673041582107544,
+      "learning_rate": 2.68430250289177e-07,
+      "loss": 0.5048,
+      "step": 6743
+    },
+    {
+      "epoch": 1.870216306156406,
+      "grad_norm": 0.19340240955352783,
+      "learning_rate": 2.6729329317612545e-07,
+      "loss": 0.5013,
+      "step": 6744
+    },
+    {
+      "epoch": 1.870493621741542,
+      "grad_norm": 0.20828621089458466,
+      "learning_rate": 2.661587229882537e-07,
+      "loss": 0.4879,
+      "step": 6745
+    },
+    {
+      "epoch": 1.8707709373266779,
+      "grad_norm": 0.2040943205356598,
+      "learning_rate": 2.650265399469429e-07,
+      "loss": 0.5113,
+      "step": 6746
+    },
+    {
+      "epoch": 1.8710482529118138,
+      "grad_norm": 0.2092759907245636,
+      "learning_rate": 2.6389674427311494e-07,
+      "loss": 0.5067,
+      "step": 6747
+    },
+    {
+      "epoch": 1.8713255684969496,
+      "grad_norm": 0.20813149213790894,
+      "learning_rate": 2.6276933618721995e-07,
+      "loss": 0.5074,
+      "step": 6748
+    },
+    {
+      "epoch": 1.8716028840820855,
+      "grad_norm": 0.2012152075767517,
+      "learning_rate": 2.6164431590924856e-07,
+      "loss": 0.5058,
+      "step": 6749
+    },
+    {
+      "epoch": 1.8718801996672214,
+      "grad_norm": 0.2121763676404953,
+      "learning_rate": 2.605216836587182e-07,
+      "loss": 0.508,
+      "step": 6750
+    },
+    {
+      "epoch": 1.8721575152523573,
+      "grad_norm": 0.20884345471858978,
+      "learning_rate": 2.5940143965468843e-07,
+      "loss": 0.5022,
+      "step": 6751
+    },
+    {
+      "epoch": 1.8724348308374932,
+      "grad_norm": 0.21974632143974304,
+      "learning_rate": 2.582835841157441e-07,
+      "loss": 0.4794,
+      "step": 6752
+    },
+    {
+      "epoch": 1.872712146422629,
+      "grad_norm": 0.22024495899677277,
+      "learning_rate": 2.5716811726001213e-07,
+      "loss": 0.5149,
+      "step": 6753
+    },
+    {
+      "epoch": 1.872989462007765,
+      "grad_norm": 0.20395393669605255,
+      "learning_rate": 2.560550393051475e-07,
+      "loss": 0.4796,
+      "step": 6754
+    },
+    {
+      "epoch": 1.8732667775929008,
+      "grad_norm": 0.2120208889245987,
+      "learning_rate": 2.5494435046834324e-07,
+      "loss": 0.5186,
+      "step": 6755
+    },
+    {
+      "epoch": 1.8735440931780367,
+      "grad_norm": 0.20500154793262482,
+      "learning_rate": 2.538360509663218e-07,
+      "loss": 0.4926,
+      "step": 6756
+    },
+    {
+      "epoch": 1.8738214087631726,
+      "grad_norm": 0.20315538346767426,
+      "learning_rate": 2.527301410153421e-07,
+      "loss": 0.4664,
+      "step": 6757
+    },
+    {
+      "epoch": 1.8740987243483085,
+      "grad_norm": 0.19243794679641724,
+      "learning_rate": 2.516266208311982e-07,
+      "loss": 0.4876,
+      "step": 6758
+    },
+    {
+      "epoch": 1.8743760399334444,
+      "grad_norm": 0.2194959968328476,
+      "learning_rate": 2.505254906292151e-07,
+      "loss": 0.5032,
+      "step": 6759
+    },
+    {
+      "epoch": 1.8746533555185803,
+      "grad_norm": 0.21170787513256073,
+      "learning_rate": 2.4942675062425147e-07,
+      "loss": 0.4849,
+      "step": 6760
+    },
+    {
+      "epoch": 1.8749306711037161,
+      "grad_norm": 0.2141827791929245,
+      "learning_rate": 2.483304010307025e-07,
+      "loss": 0.5182,
+      "step": 6761
+    },
+    {
+      "epoch": 1.875207986688852,
+      "grad_norm": 0.21987125277519226,
+      "learning_rate": 2.4723644206249424e-07,
+      "loss": 0.4889,
+      "step": 6762
+    },
+    {
+      "epoch": 1.875485302273988,
+      "grad_norm": 0.2075456976890564,
+      "learning_rate": 2.4614487393308657e-07,
+      "loss": 0.5282,
+      "step": 6763
+    },
+    {
+      "epoch": 1.8757626178591238,
+      "grad_norm": 0.20293202996253967,
+      "learning_rate": 2.450556968554743e-07,
+      "loss": 0.4726,
+      "step": 6764
+    },
+    {
+      "epoch": 1.8760399334442597,
+      "grad_norm": 0.2271379679441452,
+      "learning_rate": 2.4396891104218335e-07,
+      "loss": 0.5321,
+      "step": 6765
+    },
+    {
+      "epoch": 1.8763172490293956,
+      "grad_norm": 0.206784188747406,
+      "learning_rate": 2.4288451670527736e-07,
+      "loss": 0.5207,
+      "step": 6766
+    },
+    {
+      "epoch": 1.8765945646145314,
+      "grad_norm": 0.21522942185401917,
+      "learning_rate": 2.41802514056344e-07,
+      "loss": 0.5015,
+      "step": 6767
+    },
+    {
+      "epoch": 1.8768718801996673,
+      "grad_norm": 0.2123783975839615,
+      "learning_rate": 2.4072290330651867e-07,
+      "loss": 0.488,
+      "step": 6768
+    },
+    {
+      "epoch": 1.8771491957848032,
+      "grad_norm": 0.2079179435968399,
+      "learning_rate": 2.396456846664577e-07,
+      "loss": 0.4738,
+      "step": 6769
+    },
+    {
+      "epoch": 1.877426511369939,
+      "grad_norm": 0.21037521958351135,
+      "learning_rate": 2.3857085834635557e-07,
+      "loss": 0.5059,
+      "step": 6770
+    },
+    {
+      "epoch": 1.877703826955075,
+      "grad_norm": 0.21396882832050323,
+      "learning_rate": 2.374984245559375e-07,
+      "loss": 0.5156,
+      "step": 6771
+    },
+    {
+      "epoch": 1.8779811425402109,
+      "grad_norm": 0.21056626737117767,
+      "learning_rate": 2.3642838350446812e-07,
+      "loss": 0.5154,
+      "step": 6772
+    },
+    {
+      "epoch": 1.8782584581253468,
+      "grad_norm": 0.2141093611717224,
+      "learning_rate": 2.3536073540073727e-07,
+      "loss": 0.5053,
+      "step": 6773
+    },
+    {
+      "epoch": 1.8785357737104826,
+      "grad_norm": 0.2114919126033783,
+      "learning_rate": 2.342954804530728e-07,
+      "loss": 0.5151,
+      "step": 6774
+    },
+    {
+      "epoch": 1.8788130892956185,
+      "grad_norm": 0.2190243899822235,
+      "learning_rate": 2.3323261886933344e-07,
+      "loss": 0.5099,
+      "step": 6775
+    },
+    {
+      "epoch": 1.8790904048807544,
+      "grad_norm": 0.22468531131744385,
+      "learning_rate": 2.3217215085691164e-07,
+      "loss": 0.5139,
+      "step": 6776
+    },
+    {
+      "epoch": 1.8793677204658903,
+      "grad_norm": 0.20503219962120056,
+      "learning_rate": 2.311140766227349e-07,
+      "loss": 0.5068,
+      "step": 6777
+    },
+    {
+      "epoch": 1.8796450360510262,
+      "grad_norm": 0.22132523357868195,
+      "learning_rate": 2.300583963732575e-07,
+      "loss": 0.5136,
+      "step": 6778
+    },
+    {
+      "epoch": 1.879922351636162,
+      "grad_norm": 0.21017663180828094,
+      "learning_rate": 2.2900511031447303e-07,
+      "loss": 0.4947,
+      "step": 6779
+    },
+    {
+      "epoch": 1.880199667221298,
+      "grad_norm": 0.20506809651851654,
+      "learning_rate": 2.2795421865190737e-07,
+      "loss": 0.5036,
+      "step": 6780
+    },
+    {
+      "epoch": 1.8804769828064338,
+      "grad_norm": 0.2145530879497528,
+      "learning_rate": 2.2690572159061453e-07,
+      "loss": 0.5016,
+      "step": 6781
+    },
+    {
+      "epoch": 1.8807542983915697,
+      "grad_norm": 0.20361502468585968,
+      "learning_rate": 2.2585961933518362e-07,
+      "loss": 0.4867,
+      "step": 6782
+    },
+    {
+      "epoch": 1.8810316139767056,
+      "grad_norm": 0.2013719379901886,
+      "learning_rate": 2.248159120897403e-07,
+      "loss": 0.4996,
+      "step": 6783
+    },
+    {
+      "epoch": 1.8813089295618415,
+      "grad_norm": 0.21445105969905853,
+      "learning_rate": 2.2377460005793688e-07,
+      "loss": 0.4938,
+      "step": 6784
+    },
+    {
+      "epoch": 1.8815862451469774,
+      "grad_norm": 0.2540777027606964,
+      "learning_rate": 2.2273568344295948e-07,
+      "loss": 0.4668,
+      "step": 6785
+    },
+    {
+      "epoch": 1.8818635607321132,
+      "grad_norm": 0.21935346722602844,
+      "learning_rate": 2.2169916244753204e-07,
+      "loss": 0.4713,
+      "step": 6786
+    },
+    {
+      "epoch": 1.8821408763172491,
+      "grad_norm": 0.20928525924682617,
+      "learning_rate": 2.2066503727390526e-07,
+      "loss": 0.492,
+      "step": 6787
+    },
+    {
+      "epoch": 1.882418191902385,
+      "grad_norm": 0.21874678134918213,
+      "learning_rate": 2.1963330812386356e-07,
+      "loss": 0.4754,
+      "step": 6788
+    },
+    {
+      "epoch": 1.882695507487521,
+      "grad_norm": 0.22388552129268646,
+      "learning_rate": 2.186039751987265e-07,
+      "loss": 0.5055,
+      "step": 6789
+    },
+    {
+      "epoch": 1.8829728230726568,
+      "grad_norm": 0.20376282930374146,
+      "learning_rate": 2.175770386993431e-07,
+      "loss": 0.4799,
+      "step": 6790
+    },
+    {
+      "epoch": 1.8832501386577927,
+      "grad_norm": 0.22341391444206238,
+      "learning_rate": 2.1655249882609612e-07,
+      "loss": 0.5089,
+      "step": 6791
+    },
+    {
+      "epoch": 1.8835274542429286,
+      "grad_norm": 0.20708690583705902,
+      "learning_rate": 2.1553035577890068e-07,
+      "loss": 0.4717,
+      "step": 6792
+    },
+    {
+      "epoch": 1.8838047698280644,
+      "grad_norm": 0.20336076617240906,
+      "learning_rate": 2.1451060975720417e-07,
+      "loss": 0.4888,
+      "step": 6793
+    },
+    {
+      "epoch": 1.8840820854132003,
+      "grad_norm": 0.23210936784744263,
+      "learning_rate": 2.134932609599849e-07,
+      "loss": 0.4947,
+      "step": 6794
+    },
+    {
+      "epoch": 1.8843594009983362,
+      "grad_norm": 0.2158088982105255,
+      "learning_rate": 2.1247830958575493e-07,
+      "loss": 0.5283,
+      "step": 6795
+    },
+    {
+      "epoch": 1.884636716583472,
+      "grad_norm": 0.21517014503479004,
+      "learning_rate": 2.1146575583255862e-07,
+      "loss": 0.4747,
+      "step": 6796
+    },
+    {
+      "epoch": 1.884914032168608,
+      "grad_norm": 0.20993436872959137,
+      "learning_rate": 2.1045559989797125e-07,
+      "loss": 0.4945,
+      "step": 6797
+    },
+    {
+      "epoch": 1.8851913477537439,
+      "grad_norm": 0.210999995470047,
+      "learning_rate": 2.0944784197910321e-07,
+      "loss": 0.4993,
+      "step": 6798
+    },
+    {
+      "epoch": 1.8854686633388797,
+      "grad_norm": 0.2413627654314041,
+      "learning_rate": 2.0844248227259305e-07,
+      "loss": 0.4893,
+      "step": 6799
+    },
+    {
+      "epoch": 1.8857459789240156,
+      "grad_norm": 0.20545423030853271,
+      "learning_rate": 2.07439520974613e-07,
+      "loss": 0.5073,
+      "step": 6800
+    },
+    {
+      "epoch": 1.8860232945091515,
+      "grad_norm": 0.20195883512496948,
+      "learning_rate": 2.0643895828086625e-07,
+      "loss": 0.4633,
+      "step": 6801
+    },
+    {
+      "epoch": 1.8863006100942874,
+      "grad_norm": 0.21203799545764923,
+      "learning_rate": 2.0544079438659242e-07,
+      "loss": 0.5121,
+      "step": 6802
+    },
+    {
+      "epoch": 1.8865779256794233,
+      "grad_norm": 0.2213236391544342,
+      "learning_rate": 2.0444502948655658e-07,
+      "loss": 0.4984,
+      "step": 6803
+    },
+    {
+      "epoch": 1.8868552412645592,
+      "grad_norm": 0.2082570642232895,
+      "learning_rate": 2.0345166377506165e-07,
+      "loss": 0.5068,
+      "step": 6804
+    },
+    {
+      "epoch": 1.887132556849695,
+      "grad_norm": 0.20740516483783722,
+      "learning_rate": 2.024606974459359e-07,
+      "loss": 0.4746,
+      "step": 6805
+    },
+    {
+      "epoch": 1.887409872434831,
+      "grad_norm": 0.20049674808979034,
+      "learning_rate": 2.0147213069254683e-07,
+      "loss": 0.478,
+      "step": 6806
+    },
+    {
+      "epoch": 1.8876871880199668,
+      "grad_norm": 0.1955636441707611,
+      "learning_rate": 2.0048596370778744e-07,
+      "loss": 0.4785,
+      "step": 6807
+    },
+    {
+      "epoch": 1.8879645036051027,
+      "grad_norm": 0.21246939897537231,
+      "learning_rate": 1.995021966840871e-07,
+      "loss": 0.4925,
+      "step": 6808
+    },
+    {
+      "epoch": 1.8882418191902386,
+      "grad_norm": 0.20627954602241516,
+      "learning_rate": 1.9852082981340198e-07,
+      "loss": 0.4882,
+      "step": 6809
+    },
+    {
+      "epoch": 1.8885191347753745,
+      "grad_norm": 0.20891375839710236,
+      "learning_rate": 1.9754186328722614e-07,
+      "loss": 0.5049,
+      "step": 6810
+    },
+    {
+      "epoch": 1.8887964503605104,
+      "grad_norm": 0.2203957885503769,
+      "learning_rate": 1.9656529729658036e-07,
+      "loss": 0.4642,
+      "step": 6811
+    },
+    {
+      "epoch": 1.8890737659456462,
+      "grad_norm": 0.2100536823272705,
+      "learning_rate": 1.955911320320164e-07,
+      "loss": 0.4995,
+      "step": 6812
+    },
+    {
+      "epoch": 1.8893510815307821,
+      "grad_norm": 0.20311182737350464,
+      "learning_rate": 1.946193676836225e-07,
+      "loss": 0.525,
+      "step": 6813
+    },
+    {
+      "epoch": 1.889628397115918,
+      "grad_norm": 0.2031848132610321,
+      "learning_rate": 1.936500044410164e-07,
+      "loss": 0.519,
+      "step": 6814
+    },
+    {
+      "epoch": 1.889905712701054,
+      "grad_norm": 0.19998584687709808,
+      "learning_rate": 1.9268304249334402e-07,
+      "loss": 0.4888,
+      "step": 6815
+    },
+    {
+      "epoch": 1.8901830282861898,
+      "grad_norm": 0.20453502237796783,
+      "learning_rate": 1.9171848202928637e-07,
+      "loss": 0.4936,
+      "step": 6816
+    },
+    {
+      "epoch": 1.8904603438713257,
+      "grad_norm": 0.1982734203338623,
+      "learning_rate": 1.9075632323705405e-07,
+      "loss": 0.4986,
+      "step": 6817
+    },
+    {
+      "epoch": 1.8907376594564616,
+      "grad_norm": 0.21546244621276855,
+      "learning_rate": 1.897965663043913e-07,
+      "loss": 0.5138,
+      "step": 6818
+    },
+    {
+      "epoch": 1.8910149750415974,
+      "grad_norm": 0.20453289151191711,
+      "learning_rate": 1.8883921141857053e-07,
+      "loss": 0.4685,
+      "step": 6819
+    },
+    {
+      "epoch": 1.8912922906267333,
+      "grad_norm": 0.2022780179977417,
+      "learning_rate": 1.8788425876639647e-07,
+      "loss": 0.4743,
+      "step": 6820
+    },
+    {
+      "epoch": 1.8915696062118692,
+      "grad_norm": 0.20572780072689056,
+      "learning_rate": 1.8693170853420898e-07,
+      "loss": 0.4792,
+      "step": 6821
+    },
+    {
+      "epoch": 1.891846921797005,
+      "grad_norm": 0.2061588168144226,
+      "learning_rate": 1.8598156090787328e-07,
+      "loss": 0.491,
+      "step": 6822
+    },
+    {
+      "epoch": 1.892124237382141,
+      "grad_norm": 0.21832877397537231,
+      "learning_rate": 1.8503381607278825e-07,
+      "loss": 0.5126,
+      "step": 6823
+    },
+    {
+      "epoch": 1.8924015529672769,
+      "grad_norm": 0.21814782917499542,
+      "learning_rate": 1.8408847421388376e-07,
+      "loss": 0.4833,
+      "step": 6824
+    },
+    {
+      "epoch": 1.8926788685524127,
+      "grad_norm": 0.2065282016992569,
+      "learning_rate": 1.8314553551562336e-07,
+      "loss": 0.5193,
+      "step": 6825
+    },
+    {
+      "epoch": 1.8929561841375486,
+      "grad_norm": 0.21190817654132843,
+      "learning_rate": 1.8220500016199736e-07,
+      "loss": 0.5011,
+      "step": 6826
+    },
+    {
+      "epoch": 1.8932334997226845,
+      "grad_norm": 0.3020632266998291,
+      "learning_rate": 1.8126686833652984e-07,
+      "loss": 0.4953,
+      "step": 6827
+    },
+    {
+      "epoch": 1.8935108153078204,
+      "grad_norm": 0.2020588517189026,
+      "learning_rate": 1.8033114022227432e-07,
+      "loss": 0.4876,
+      "step": 6828
+    },
+    {
+      "epoch": 1.8937881308929563,
+      "grad_norm": 0.21000345051288605,
+      "learning_rate": 1.793978160018181e-07,
+      "loss": 0.4933,
+      "step": 6829
+    },
+    {
+      "epoch": 1.8940654464780922,
+      "grad_norm": 0.20526744425296783,
+      "learning_rate": 1.784668958572752e-07,
+      "loss": 0.5048,
+      "step": 6830
+    },
+    {
+      "epoch": 1.894342762063228,
+      "grad_norm": 0.20586876571178436,
+      "learning_rate": 1.7753837997029343e-07,
+      "loss": 0.4861,
+      "step": 6831
+    },
+    {
+      "epoch": 1.894620077648364,
+      "grad_norm": 0.22010831534862518,
+      "learning_rate": 1.7661226852205143e-07,
+      "loss": 0.5143,
+      "step": 6832
+    },
+    {
+      "epoch": 1.8948973932334998,
+      "grad_norm": 0.21001818776130676,
+      "learning_rate": 1.7568856169325743e-07,
+      "loss": 0.4987,
+      "step": 6833
+    },
+    {
+      "epoch": 1.8951747088186357,
+      "grad_norm": 0.1980331391096115,
+      "learning_rate": 1.7476725966415335e-07,
+      "loss": 0.4786,
+      "step": 6834
+    },
+    {
+      "epoch": 1.8954520244037716,
+      "grad_norm": 0.20589183270931244,
+      "learning_rate": 1.7384836261450655e-07,
+      "loss": 0.5138,
+      "step": 6835
+    },
+    {
+      "epoch": 1.8957293399889075,
+      "grad_norm": 0.20626741647720337,
+      "learning_rate": 1.7293187072361938e-07,
+      "loss": 0.489,
+      "step": 6836
+    },
+    {
+      "epoch": 1.8960066555740434,
+      "grad_norm": 0.20527756214141846,
+      "learning_rate": 1.7201778417032383e-07,
+      "loss": 0.4767,
+      "step": 6837
+    },
+    {
+      "epoch": 1.8962839711591792,
+      "grad_norm": 0.21390627324581146,
+      "learning_rate": 1.7110610313298274e-07,
+      "loss": 0.4862,
+      "step": 6838
+    },
+    {
+      "epoch": 1.8965612867443151,
+      "grad_norm": 0.205204576253891,
+      "learning_rate": 1.701968277894872e-07,
+      "loss": 0.487,
+      "step": 6839
+    },
+    {
+      "epoch": 1.896838602329451,
+      "grad_norm": 0.20338958501815796,
+      "learning_rate": 1.692899583172633e-07,
+      "loss": 0.492,
+      "step": 6840
+    },
+    {
+      "epoch": 1.897115917914587,
+      "grad_norm": 0.2076566219329834,
+      "learning_rate": 1.6838549489326533e-07,
+      "loss": 0.5127,
+      "step": 6841
+    },
+    {
+      "epoch": 1.8973932334997228,
+      "grad_norm": 0.21550370752811432,
+      "learning_rate": 1.6748343769397713e-07,
+      "loss": 0.5049,
+      "step": 6842
+    },
+    {
+      "epoch": 1.8976705490848587,
+      "grad_norm": 0.21209217607975006,
+      "learning_rate": 1.6658378689541343e-07,
+      "loss": 0.5149,
+      "step": 6843
+    },
+    {
+      "epoch": 1.8979478646699945,
+      "grad_norm": 0.2269715517759323,
+      "learning_rate": 1.6568654267312133e-07,
+      "loss": 0.4871,
+      "step": 6844
+    },
+    {
+      "epoch": 1.8982251802551304,
+      "grad_norm": 0.2141071856021881,
+      "learning_rate": 1.6479170520217607e-07,
+      "loss": 0.4873,
+      "step": 6845
+    },
+    {
+      "epoch": 1.8985024958402663,
+      "grad_norm": 0.21018719673156738,
+      "learning_rate": 1.638992746571852e-07,
+      "loss": 0.5122,
+      "step": 6846
+    },
+    {
+      "epoch": 1.8987798114254022,
+      "grad_norm": 0.21268904209136963,
+      "learning_rate": 1.630092512122816e-07,
+      "loss": 0.5282,
+      "step": 6847
+    },
+    {
+      "epoch": 1.899057127010538,
+      "grad_norm": 0.21051792800426483,
+      "learning_rate": 1.6212163504113753e-07,
+      "loss": 0.4921,
+      "step": 6848
+    },
+    {
+      "epoch": 1.899334442595674,
+      "grad_norm": 0.20133888721466064,
+      "learning_rate": 1.6123642631694913e-07,
+      "loss": 0.4807,
+      "step": 6849
+    },
+    {
+      "epoch": 1.8996117581808099,
+      "grad_norm": 0.19789256155490875,
+      "learning_rate": 1.6035362521244213e-07,
+      "loss": 0.5019,
+      "step": 6850
+    },
+    {
+      "epoch": 1.8998890737659457,
+      "grad_norm": 0.19800089299678802,
+      "learning_rate": 1.5947323189987595e-07,
+      "loss": 0.4618,
+      "step": 6851
+    },
+    {
+      "epoch": 1.9001663893510816,
+      "grad_norm": 0.20594573020935059,
+      "learning_rate": 1.585952465510382e-07,
+      "loss": 0.4941,
+      "step": 6852
+    },
+    {
+      "epoch": 1.9004437049362175,
+      "grad_norm": 0.18927329778671265,
+      "learning_rate": 1.5771966933724736e-07,
+      "loss": 0.4925,
+      "step": 6853
+    },
+    {
+      "epoch": 1.9007210205213534,
+      "grad_norm": 0.2202519327402115,
+      "learning_rate": 1.568465004293515e-07,
+      "loss": 0.5087,
+      "step": 6854
+    },
+    {
+      "epoch": 1.9009983361064893,
+      "grad_norm": 0.2147960215806961,
+      "learning_rate": 1.5597573999772823e-07,
+      "loss": 0.516,
+      "step": 6855
+    },
+    {
+      "epoch": 1.9012756516916252,
+      "grad_norm": 0.21644139289855957,
+      "learning_rate": 1.5510738821228887e-07,
+      "loss": 0.4624,
+      "step": 6856
+    },
+    {
+      "epoch": 1.901552967276761,
+      "grad_norm": 0.2117416262626648,
+      "learning_rate": 1.542414452424687e-07,
+      "loss": 0.4926,
+      "step": 6857
+    },
+    {
+      "epoch": 1.901830282861897,
+      "grad_norm": 0.2076312005519867,
+      "learning_rate": 1.533779112572367e-07,
+      "loss": 0.514,
+      "step": 6858
+    },
+    {
+      "epoch": 1.9021075984470328,
+      "grad_norm": 0.21172620356082916,
+      "learning_rate": 1.5251678642509286e-07,
+      "loss": 0.4708,
+      "step": 6859
+    },
+    {
+      "epoch": 1.9023849140321687,
+      "grad_norm": 0.2097117304801941,
+      "learning_rate": 1.5165807091406386e-07,
+      "loss": 0.5145,
+      "step": 6860
+    },
+    {
+      "epoch": 1.9026622296173046,
+      "grad_norm": 0.21858976781368256,
+      "learning_rate": 1.5080176489170734e-07,
+      "loss": 0.4811,
+      "step": 6861
+    },
+    {
+      "epoch": 1.9029395452024405,
+      "grad_norm": 0.1994626671075821,
+      "learning_rate": 1.4994786852511322e-07,
+      "loss": 0.4825,
+      "step": 6862
+    },
+    {
+      "epoch": 1.9032168607875763,
+      "grad_norm": 0.20749947428703308,
+      "learning_rate": 1.4909638198089966e-07,
+      "loss": 0.5112,
+      "step": 6863
+    },
+    {
+      "epoch": 1.9034941763727122,
+      "grad_norm": 0.20840351283550262,
+      "learning_rate": 1.4824730542521148e-07,
+      "loss": 0.4832,
+      "step": 6864
+    },
+    {
+      "epoch": 1.9037714919578481,
+      "grad_norm": 0.20934903621673584,
+      "learning_rate": 1.474006390237273e-07,
+      "loss": 0.5057,
+      "step": 6865
+    },
+    {
+      "epoch": 1.904048807542984,
+      "grad_norm": 0.21367427706718445,
+      "learning_rate": 1.4655638294165525e-07,
+      "loss": 0.5018,
+      "step": 6866
+    },
+    {
+      "epoch": 1.9043261231281199,
+      "grad_norm": 0.21489764750003815,
+      "learning_rate": 1.4571453734373157e-07,
+      "loss": 0.5019,
+      "step": 6867
+    },
+    {
+      "epoch": 1.9046034387132558,
+      "grad_norm": 0.20891396701335907,
+      "learning_rate": 1.4487510239422076e-07,
+      "loss": 0.4948,
+      "step": 6868
+    },
+    {
+      "epoch": 1.9048807542983917,
+      "grad_norm": 0.20342187583446503,
+      "learning_rate": 1.440380782569209e-07,
+      "loss": 0.4613,
+      "step": 6869
+    },
+    {
+      "epoch": 1.9051580698835275,
+      "grad_norm": 0.20930233597755432,
+      "learning_rate": 1.432034650951569e-07,
+      "loss": 0.5102,
+      "step": 6870
+    },
+    {
+      "epoch": 1.9054353854686634,
+      "grad_norm": 0.19923752546310425,
+      "learning_rate": 1.4237126307178467e-07,
+      "loss": 0.4857,
+      "step": 6871
+    },
+    {
+      "epoch": 1.9057127010537993,
+      "grad_norm": 0.2359122633934021,
+      "learning_rate": 1.4154147234918814e-07,
+      "loss": 0.4877,
+      "step": 6872
+    },
+    {
+      "epoch": 1.9059900166389352,
+      "grad_norm": 0.19613344967365265,
+      "learning_rate": 1.4071409308928086e-07,
+      "loss": 0.4993,
+      "step": 6873
+    },
+    {
+      "epoch": 1.906267332224071,
+      "grad_norm": 0.22320273518562317,
+      "learning_rate": 1.398891254535073e-07,
+      "loss": 0.5018,
+      "step": 6874
+    },
+    {
+      "epoch": 1.906544647809207,
+      "grad_norm": 0.21227893233299255,
+      "learning_rate": 1.3906656960284008e-07,
+      "loss": 0.5012,
+      "step": 6875
+    },
+    {
+      "epoch": 1.9068219633943428,
+      "grad_norm": 0.2211824506521225,
+      "learning_rate": 1.382464256977828e-07,
+      "loss": 0.4604,
+      "step": 6876
+    },
+    {
+      "epoch": 1.9070992789794787,
+      "grad_norm": 0.20455965399742126,
+      "learning_rate": 1.3742869389836572e-07,
+      "loss": 0.4759,
+      "step": 6877
+    },
+    {
+      "epoch": 1.9073765945646146,
+      "grad_norm": 0.20828378200531006,
+      "learning_rate": 1.3661337436415012e-07,
+      "loss": 0.5047,
+      "step": 6878
+    },
+    {
+      "epoch": 1.9076539101497505,
+      "grad_norm": 0.2089402675628662,
+      "learning_rate": 1.358004672542282e-07,
+      "loss": 0.4886,
+      "step": 6879
+    },
+    {
+      "epoch": 1.9079312257348864,
+      "grad_norm": 0.23213088512420654,
+      "learning_rate": 1.3498997272721886e-07,
+      "loss": 0.513,
+      "step": 6880
+    },
+    {
+      "epoch": 1.9082085413200223,
+      "grad_norm": 0.22040383517742157,
+      "learning_rate": 1.3418189094126926e-07,
+      "loss": 0.5072,
+      "step": 6881
+    },
+    {
+      "epoch": 1.9084858569051582,
+      "grad_norm": 0.22452102601528168,
+      "learning_rate": 1.333762220540588e-07,
+      "loss": 0.5167,
+      "step": 6882
+    },
+    {
+      "epoch": 1.908763172490294,
+      "grad_norm": 0.20027916133403778,
+      "learning_rate": 1.325729662227951e-07,
+      "loss": 0.4924,
+      "step": 6883
+    },
+    {
+      "epoch": 1.90904048807543,
+      "grad_norm": 0.20974531769752502,
+      "learning_rate": 1.317721236042152e-07,
+      "loss": 0.4752,
+      "step": 6884
+    },
+    {
+      "epoch": 1.9093178036605658,
+      "grad_norm": 0.22228215634822845,
+      "learning_rate": 1.3097369435458167e-07,
+      "loss": 0.4951,
+      "step": 6885
+    },
+    {
+      "epoch": 1.9095951192457017,
+      "grad_norm": 0.21499434113502502,
+      "learning_rate": 1.3017767862969487e-07,
+      "loss": 0.507,
+      "step": 6886
+    },
+    {
+      "epoch": 1.9098724348308376,
+      "grad_norm": 0.1968400925397873,
+      "learning_rate": 1.293840765848736e-07,
+      "loss": 0.4698,
+      "step": 6887
+    },
+    {
+      "epoch": 1.9101497504159735,
+      "grad_norm": 0.20211312174797058,
+      "learning_rate": 1.285928883749718e-07,
+      "loss": 0.4916,
+      "step": 6888
+    },
+    {
+      "epoch": 1.9104270660011093,
+      "grad_norm": 0.20148654282093048,
+      "learning_rate": 1.2780411415437148e-07,
+      "loss": 0.4794,
+      "step": 6889
+    },
+    {
+      "epoch": 1.9107043815862452,
+      "grad_norm": 0.20846286416053772,
+      "learning_rate": 1.2701775407698567e-07,
+      "loss": 0.4847,
+      "step": 6890
+    },
+    {
+      "epoch": 1.910981697171381,
+      "grad_norm": 0.2036297619342804,
+      "learning_rate": 1.2623380829624997e-07,
+      "loss": 0.4813,
+      "step": 6891
+    },
+    {
+      "epoch": 1.911259012756517,
+      "grad_norm": 0.21009089052677155,
+      "learning_rate": 1.2545227696513644e-07,
+      "loss": 0.5258,
+      "step": 6892
+    },
+    {
+      "epoch": 1.9115363283416529,
+      "grad_norm": 0.21154698729515076,
+      "learning_rate": 1.2467316023613978e-07,
+      "loss": 0.5118,
+      "step": 6893
+    },
+    {
+      "epoch": 1.9118136439267888,
+      "grad_norm": 0.21194148063659668,
+      "learning_rate": 1.2389645826128836e-07,
+      "loss": 0.4956,
+      "step": 6894
+    },
+    {
+      "epoch": 1.9120909595119246,
+      "grad_norm": 0.21466988325119019,
+      "learning_rate": 1.2312217119213737e-07,
+      "loss": 0.5134,
+      "step": 6895
+    },
+    {
+      "epoch": 1.9123682750970605,
+      "grad_norm": 0.22028987109661102,
+      "learning_rate": 1.223502991797687e-07,
+      "loss": 0.5057,
+      "step": 6896
+    },
+    {
+      "epoch": 1.9126455906821964,
+      "grad_norm": 0.21367661654949188,
+      "learning_rate": 1.215808423747966e-07,
+      "loss": 0.5005,
+      "step": 6897
+    },
+    {
+      "epoch": 1.9129229062673323,
+      "grad_norm": 0.2160566747188568,
+      "learning_rate": 1.208138009273621e-07,
+      "loss": 0.5011,
+      "step": 6898
+    },
+    {
+      "epoch": 1.9132002218524682,
+      "grad_norm": 0.21020996570587158,
+      "learning_rate": 1.2004917498713576e-07,
+      "loss": 0.5011,
+      "step": 6899
+    },
+    {
+      "epoch": 1.913477537437604,
+      "grad_norm": 0.20518991351127625,
+      "learning_rate": 1.1928696470331486e-07,
+      "loss": 0.4979,
+      "step": 6900
+    },
+    {
+      "epoch": 1.91375485302274,
+      "grad_norm": 0.20846955478191376,
+      "learning_rate": 1.1852717022463045e-07,
+      "loss": 0.5073,
+      "step": 6901
+    },
+    {
+      "epoch": 1.9140321686078758,
+      "grad_norm": 0.21253100037574768,
+      "learning_rate": 1.1776979169933478e-07,
+      "loss": 0.5105,
+      "step": 6902
+    },
+    {
+      "epoch": 1.9143094841930117,
+      "grad_norm": 0.22121791541576385,
+      "learning_rate": 1.1701482927521241e-07,
+      "loss": 0.4747,
+      "step": 6903
+    },
+    {
+      "epoch": 1.9145867997781476,
+      "grad_norm": 0.20756427943706512,
+      "learning_rate": 1.1626228309957881e-07,
+      "loss": 0.493,
+      "step": 6904
+    },
+    {
+      "epoch": 1.9148641153632835,
+      "grad_norm": 0.22023040056228638,
+      "learning_rate": 1.1551215331927489e-07,
+      "loss": 0.5037,
+      "step": 6905
+    },
+    {
+      "epoch": 1.9151414309484194,
+      "grad_norm": 0.2466723471879959,
+      "learning_rate": 1.1476444008067105e-07,
+      "loss": 0.5177,
+      "step": 6906
+    },
+    {
+      "epoch": 1.9154187465335553,
+      "grad_norm": 0.21187765896320343,
+      "learning_rate": 1.1401914352966447e-07,
+      "loss": 0.4916,
+      "step": 6907
+    },
+    {
+      "epoch": 1.9156960621186911,
+      "grad_norm": 0.21347801387310028,
+      "learning_rate": 1.1327626381168466e-07,
+      "loss": 0.511,
+      "step": 6908
+    },
+    {
+      "epoch": 1.915973377703827,
+      "grad_norm": 0.20320340991020203,
+      "learning_rate": 1.125358010716851e-07,
+      "loss": 0.5196,
+      "step": 6909
+    },
+    {
+      "epoch": 1.916250693288963,
+      "grad_norm": 0.211409792304039,
+      "learning_rate": 1.1179775545415022e-07,
+      "loss": 0.4988,
+      "step": 6910
+    },
+    {
+      "epoch": 1.9165280088740988,
+      "grad_norm": 0.20658671855926514,
+      "learning_rate": 1.1106212710309261e-07,
+      "loss": 0.499,
+      "step": 6911
+    },
+    {
+      "epoch": 1.9168053244592347,
+      "grad_norm": 0.21737776696681976,
+      "learning_rate": 1.1032891616205299e-07,
+      "loss": 0.5227,
+      "step": 6912
+    },
+    {
+      "epoch": 1.9170826400443706,
+      "grad_norm": 0.20637080073356628,
+      "learning_rate": 1.0959812277410025e-07,
+      "loss": 0.4793,
+      "step": 6913
+    },
+    {
+      "epoch": 1.9173599556295065,
+      "grad_norm": 0.20913554728031158,
+      "learning_rate": 1.0886974708183007e-07,
+      "loss": 0.5227,
+      "step": 6914
+    },
+    {
+      "epoch": 1.9176372712146423,
+      "grad_norm": 0.2132870852947235,
+      "learning_rate": 1.081437892273704e-07,
+      "loss": 0.4907,
+      "step": 6915
+    },
+    {
+      "epoch": 1.9179145867997782,
+      "grad_norm": 0.20474767684936523,
+      "learning_rate": 1.0742024935237322e-07,
+      "loss": 0.5073,
+      "step": 6916
+    },
+    {
+      "epoch": 1.918191902384914,
+      "grad_norm": 0.2082609385251999,
+      "learning_rate": 1.0669912759802004e-07,
+      "loss": 0.5381,
+      "step": 6917
+    },
+    {
+      "epoch": 1.91846921797005,
+      "grad_norm": 0.21400731801986694,
+      "learning_rate": 1.059804241050219e-07,
+      "loss": 0.4944,
+      "step": 6918
+    },
+    {
+      "epoch": 1.9187465335551859,
+      "grad_norm": 0.22101816534996033,
+      "learning_rate": 1.0526413901361526e-07,
+      "loss": 0.5161,
+      "step": 6919
+    },
+    {
+      "epoch": 1.9190238491403218,
+      "grad_norm": 0.21787089109420776,
+      "learning_rate": 1.0455027246356746e-07,
+      "loss": 0.5062,
+      "step": 6920
+    },
+    {
+      "epoch": 1.9193011647254576,
+      "grad_norm": 0.2018042355775833,
+      "learning_rate": 1.0383882459417404e-07,
+      "loss": 0.4633,
+      "step": 6921
+    },
+    {
+      "epoch": 1.9195784803105935,
+      "grad_norm": 0.20412792265415192,
+      "learning_rate": 1.031297955442559e-07,
+      "loss": 0.5062,
+      "step": 6922
+    },
+    {
+      "epoch": 1.9198557958957294,
+      "grad_norm": 0.214492067694664,
+      "learning_rate": 1.0242318545216207e-07,
+      "loss": 0.5178,
+      "step": 6923
+    },
+    {
+      "epoch": 1.9201331114808653,
+      "grad_norm": 0.21588201820850372,
+      "learning_rate": 1.0171899445577393e-07,
+      "loss": 0.5249,
+      "step": 6924
+    },
+    {
+      "epoch": 1.9204104270660012,
+      "grad_norm": 0.19817836582660675,
+      "learning_rate": 1.0101722269249547e-07,
+      "loss": 0.4964,
+      "step": 6925
+    },
+    {
+      "epoch": 1.920687742651137,
+      "grad_norm": 0.20311301946640015,
+      "learning_rate": 1.003178702992616e-07,
+      "loss": 0.4804,
+      "step": 6926
+    },
+    {
+      "epoch": 1.920965058236273,
+      "grad_norm": 0.20916207134723663,
+      "learning_rate": 9.962093741253537e-08,
+      "loss": 0.5013,
+      "step": 6927
+    },
+    {
+      "epoch": 1.9212423738214088,
+      "grad_norm": 0.21295547485351562,
+      "learning_rate": 9.892642416830522e-08,
+      "loss": 0.492,
+      "step": 6928
+    },
+    {
+      "epoch": 1.9215196894065447,
+      "grad_norm": 0.21146497130393982,
+      "learning_rate": 9.823433070209053e-08,
+      "loss": 0.5107,
+      "step": 6929
+    },
+    {
+      "epoch": 1.9217970049916806,
+      "grad_norm": 0.2107129544019699,
+      "learning_rate": 9.754465714893607e-08,
+      "loss": 0.5023,
+      "step": 6930
+    },
+    {
+      "epoch": 1.9220743205768165,
+      "grad_norm": 0.21256358921527863,
+      "learning_rate": 9.685740364341611e-08,
+      "loss": 0.4751,
+      "step": 6931
+    },
+    {
+      "epoch": 1.9223516361619524,
+      "grad_norm": 0.21741509437561035,
+      "learning_rate": 9.617257031963173e-08,
+      "loss": 0.5264,
+      "step": 6932
+    },
+    {
+      "epoch": 1.9226289517470883,
+      "grad_norm": 0.38610130548477173,
+      "learning_rate": 9.549015731121353e-08,
+      "loss": 0.5083,
+      "step": 6933
+    },
+    {
+      "epoch": 1.9229062673322241,
+      "grad_norm": 0.2073136270046234,
+      "learning_rate": 9.481016475131472e-08,
+      "loss": 0.4922,
+      "step": 6934
+    },
+    {
+      "epoch": 1.92318358291736,
+      "grad_norm": 0.21415603160858154,
+      "learning_rate": 9.41325927726222e-08,
+      "loss": 0.5014,
+      "step": 6935
+    },
+    {
+      "epoch": 1.923460898502496,
+      "grad_norm": 0.23321330547332764,
+      "learning_rate": 9.345744150734969e-08,
+      "loss": 0.5141,
+      "step": 6936
+    },
+    {
+      "epoch": 1.9237382140876318,
+      "grad_norm": 0.2120734453201294,
+      "learning_rate": 9.278471108723347e-08,
+      "loss": 0.4874,
+      "step": 6937
+    },
+    {
+      "epoch": 1.9240155296727677,
+      "grad_norm": 0.2397221326828003,
+      "learning_rate": 9.211440164354351e-08,
+      "loss": 0.4939,
+      "step": 6938
+    },
+    {
+      "epoch": 1.9242928452579036,
+      "grad_norm": 0.2101098895072937,
+      "learning_rate": 9.144651330707659e-08,
+      "loss": 0.5111,
+      "step": 6939
+    },
+    {
+      "epoch": 1.9245701608430394,
+      "grad_norm": 0.20636944472789764,
+      "learning_rate": 9.078104620815209e-08,
+      "loss": 0.5037,
+      "step": 6940
+    },
+    {
+      "epoch": 1.9248474764281753,
+      "grad_norm": 0.22542667388916016,
+      "learning_rate": 9.011800047662028e-08,
+      "loss": 0.5024,
+      "step": 6941
+    },
+    {
+      "epoch": 1.9251247920133112,
+      "grad_norm": 0.19407561421394348,
+      "learning_rate": 8.945737624186101e-08,
+      "loss": 0.4894,
+      "step": 6942
+    },
+    {
+      "epoch": 1.925402107598447,
+      "grad_norm": 0.21263688802719116,
+      "learning_rate": 8.87991736327795e-08,
+      "loss": 0.513,
+      "step": 6943
+    },
+    {
+      "epoch": 1.925679423183583,
+      "grad_norm": 0.20905403792858124,
+      "learning_rate": 8.814339277780636e-08,
+      "loss": 0.512,
+      "step": 6944
+    },
+    {
+      "epoch": 1.9259567387687189,
+      "grad_norm": 0.20334313809871674,
+      "learning_rate": 8.74900338049045e-08,
+      "loss": 0.5043,
+      "step": 6945
+    },
+    {
+      "epoch": 1.9262340543538548,
+      "grad_norm": 0.2149227261543274,
+      "learning_rate": 8.683909684155944e-08,
+      "loss": 0.4717,
+      "step": 6946
+    },
+    {
+      "epoch": 1.9265113699389906,
+      "grad_norm": 0.2087196409702301,
+      "learning_rate": 8.619058201478763e-08,
+      "loss": 0.5005,
+      "step": 6947
+    },
+    {
+      "epoch": 1.9267886855241265,
+      "grad_norm": 0.2073797583580017,
+      "learning_rate": 8.554448945113091e-08,
+      "loss": 0.4941,
+      "step": 6948
+    },
+    {
+      "epoch": 1.9270660011092624,
+      "grad_norm": 0.21620948612689972,
+      "learning_rate": 8.490081927665927e-08,
+      "loss": 0.4743,
+      "step": 6949
+    },
+    {
+      "epoch": 1.9273433166943983,
+      "grad_norm": 0.211389422416687,
+      "learning_rate": 8.425957161696946e-08,
+      "loss": 0.4886,
+      "step": 6950
+    },
+    {
+      "epoch": 1.9276206322795342,
+      "grad_norm": 0.20867092907428741,
+      "learning_rate": 8.3620746597185e-08,
+      "loss": 0.4856,
+      "step": 6951
+    },
+    {
+      "epoch": 1.92789794786467,
+      "grad_norm": 0.2138075828552246,
+      "learning_rate": 8.298434434196034e-08,
+      "loss": 0.479,
+      "step": 6952
+    },
+    {
+      "epoch": 1.928175263449806,
+      "grad_norm": 0.21532325446605682,
+      "learning_rate": 8.235036497547116e-08,
+      "loss": 0.4989,
+      "step": 6953
+    },
+    {
+      "epoch": 1.9284525790349418,
+      "grad_norm": 0.19918455183506012,
+      "learning_rate": 8.171880862142683e-08,
+      "loss": 0.5055,
+      "step": 6954
+    },
+    {
+      "epoch": 1.9287298946200777,
+      "grad_norm": 0.21148592233657837,
+      "learning_rate": 8.108967540305795e-08,
+      "loss": 0.4639,
+      "step": 6955
+    },
+    {
+      "epoch": 1.9290072102052136,
+      "grad_norm": 0.20538489520549774,
+      "learning_rate": 8.046296544312742e-08,
+      "loss": 0.5217,
+      "step": 6956
+    },
+    {
+      "epoch": 1.9292845257903495,
+      "grad_norm": 0.19862611591815948,
+      "learning_rate": 7.983867886391938e-08,
+      "loss": 0.4928,
+      "step": 6957
+    },
+    {
+      "epoch": 1.9295618413754854,
+      "grad_norm": 0.2005883902311325,
+      "learning_rate": 7.921681578725305e-08,
+      "loss": 0.4719,
+      "step": 6958
+    },
+    {
+      "epoch": 1.9298391569606212,
+      "grad_norm": 0.21354445815086365,
+      "learning_rate": 7.859737633446745e-08,
+      "loss": 0.5073,
+      "step": 6959
+    },
+    {
+      "epoch": 1.9301164725457571,
+      "grad_norm": 0.20679736137390137,
+      "learning_rate": 7.798036062643399e-08,
+      "loss": 0.4767,
+      "step": 6960
+    },
+    {
+      "epoch": 1.930393788130893,
+      "grad_norm": 0.2107972651720047,
+      "learning_rate": 7.736576878354523e-08,
+      "loss": 0.5154,
+      "step": 6961
+    },
+    {
+      "epoch": 1.930671103716029,
+      "grad_norm": 0.21988734602928162,
+      "learning_rate": 7.675360092572747e-08,
+      "loss": 0.4996,
+      "step": 6962
+    },
+    {
+      "epoch": 1.9309484193011648,
+      "grad_norm": 0.20176535844802856,
+      "learning_rate": 7.6143857172431e-08,
+      "loss": 0.4677,
+      "step": 6963
+    },
+    {
+      "epoch": 1.9312257348863007,
+      "grad_norm": 0.21723569929599762,
+      "learning_rate": 7.553653764263008e-08,
+      "loss": 0.5311,
+      "step": 6964
+    },
+    {
+      "epoch": 1.9315030504714366,
+      "grad_norm": 0.20594346523284912,
+      "learning_rate": 7.49316424548313e-08,
+      "loss": 0.4978,
+      "step": 6965
+    },
+    {
+      "epoch": 1.9317803660565724,
+      "grad_norm": 0.2213924676179886,
+      "learning_rate": 7.432917172706528e-08,
+      "loss": 0.5089,
+      "step": 6966
+    },
+    {
+      "epoch": 1.9320576816417083,
+      "grad_norm": 0.2052609771490097,
+      "learning_rate": 7.372912557688933e-08,
+      "loss": 0.4955,
+      "step": 6967
+    },
+    {
+      "epoch": 1.9323349972268442,
+      "grad_norm": 0.2087954580783844,
+      "learning_rate": 7.313150412138898e-08,
+      "loss": 0.4674,
+      "step": 6968
+    },
+    {
+      "epoch": 1.93261231281198,
+      "grad_norm": 0.20820866525173187,
+      "learning_rate": 7.253630747717648e-08,
+      "loss": 0.4871,
+      "step": 6969
+    },
+    {
+      "epoch": 1.932889628397116,
+      "grad_norm": 0.21151137351989746,
+      "learning_rate": 7.194353576038953e-08,
+      "loss": 0.4889,
+      "step": 6970
+    },
+    {
+      "epoch": 1.9331669439822519,
+      "grad_norm": 0.21804526448249817,
+      "learning_rate": 7.135318908669392e-08,
+      "loss": 0.4842,
+      "step": 6971
+    },
+    {
+      "epoch": 1.9334442595673877,
+      "grad_norm": 0.21454410254955292,
+      "learning_rate": 7.076526757128083e-08,
+      "loss": 0.5052,
+      "step": 6972
+    },
+    {
+      "epoch": 1.9337215751525236,
+      "grad_norm": 0.2476005256175995,
+      "learning_rate": 7.0179771328871e-08,
+      "loss": 0.5001,
+      "step": 6973
+    },
+    {
+      "epoch": 1.9339988907376595,
+      "grad_norm": 0.20526906847953796,
+      "learning_rate": 6.959670047371053e-08,
+      "loss": 0.4873,
+      "step": 6974
+    },
+    {
+      "epoch": 1.9342762063227954,
+      "grad_norm": 0.23170456290245056,
+      "learning_rate": 6.901605511957093e-08,
+      "loss": 0.4773,
+      "step": 6975
+    },
+    {
+      "epoch": 1.9345535219079313,
+      "grad_norm": 0.21698127686977386,
+      "learning_rate": 6.843783537974907e-08,
+      "loss": 0.4895,
+      "step": 6976
+    },
+    {
+      "epoch": 1.9348308374930672,
+      "grad_norm": 0.2072725147008896,
+      "learning_rate": 6.786204136707691e-08,
+      "loss": 0.5427,
+      "step": 6977
+    },
+    {
+      "epoch": 1.935108153078203,
+      "grad_norm": 0.21492376923561096,
+      "learning_rate": 6.728867319390209e-08,
+      "loss": 0.5325,
+      "step": 6978
+    },
+    {
+      "epoch": 1.935385468663339,
+      "grad_norm": 0.2331382930278778,
+      "learning_rate": 6.671773097210593e-08,
+      "loss": 0.4782,
+      "step": 6979
+    },
+    {
+      "epoch": 1.9356627842484748,
+      "grad_norm": 0.2333361804485321,
+      "learning_rate": 6.614921481309377e-08,
+      "loss": 0.4814,
+      "step": 6980
+    },
+    {
+      "epoch": 1.9359400998336107,
+      "grad_norm": 0.20312927663326263,
+      "learning_rate": 6.55831248277991e-08,
+      "loss": 0.4894,
+      "step": 6981
+    },
+    {
+      "epoch": 1.9362174154187466,
+      "grad_norm": 0.21046309173107147,
+      "learning_rate": 6.501946112668078e-08,
+      "loss": 0.4759,
+      "step": 6982
+    },
+    {
+      "epoch": 1.9364947310038825,
+      "grad_norm": 0.21393108367919922,
+      "learning_rate": 6.445822381972305e-08,
+      "loss": 0.4939,
+      "step": 6983
+    },
+    {
+      "epoch": 1.9367720465890184,
+      "grad_norm": 0.20159265398979187,
+      "learning_rate": 6.38994130164397e-08,
+      "loss": 0.508,
+      "step": 6984
+    },
+    {
+      "epoch": 1.9370493621741542,
+      "grad_norm": 0.2000395506620407,
+      "learning_rate": 6.33430288258699e-08,
+      "loss": 0.5001,
+      "step": 6985
+    },
+    {
+      "epoch": 1.9373266777592901,
+      "grad_norm": 0.21411900222301483,
+      "learning_rate": 6.27890713565768e-08,
+      "loss": 0.4783,
+      "step": 6986
+    },
+    {
+      "epoch": 1.937603993344426,
+      "grad_norm": 0.22993165254592896,
+      "learning_rate": 6.22375407166545e-08,
+      "loss": 0.491,
+      "step": 6987
+    },
+    {
+      "epoch": 1.937881308929562,
+      "grad_norm": 0.21177081763744354,
+      "learning_rate": 6.168843701371968e-08,
+      "loss": 0.4968,
+      "step": 6988
+    },
+    {
+      "epoch": 1.9381586245146978,
+      "grad_norm": 0.21067818999290466,
+      "learning_rate": 6.114176035491859e-08,
+      "loss": 0.5007,
+      "step": 6989
+    },
+    {
+      "epoch": 1.9384359400998337,
+      "grad_norm": 0.2169368863105774,
+      "learning_rate": 6.059751084692006e-08,
+      "loss": 0.4761,
+      "step": 6990
+    },
+    {
+      "epoch": 1.9387132556849695,
+      "grad_norm": 0.20408381521701813,
+      "learning_rate": 6.005568859592386e-08,
+      "loss": 0.5434,
+      "step": 6991
+    },
+    {
+      "epoch": 1.9389905712701054,
+      "grad_norm": 0.2044132947921753,
+      "learning_rate": 5.9516293707652385e-08,
+      "loss": 0.4814,
+      "step": 6992
+    },
+    {
+      "epoch": 1.9392678868552413,
+      "grad_norm": 0.2122509479522705,
+      "learning_rate": 5.897932628735614e-08,
+      "loss": 0.4915,
+      "step": 6993
+    },
+    {
+      "epoch": 1.9395452024403772,
+      "grad_norm": 0.21428446471691132,
+      "learning_rate": 5.844478643981383e-08,
+      "loss": 0.5069,
+      "step": 6994
+    },
+    {
+      "epoch": 1.939822518025513,
+      "grad_norm": 0.20293046534061432,
+      "learning_rate": 5.791267426932395e-08,
+      "loss": 0.5028,
+      "step": 6995
+    },
+    {
+      "epoch": 1.940099833610649,
+      "grad_norm": 0.20903684198856354,
+      "learning_rate": 5.7382989879720126e-08,
+      "loss": 0.4971,
+      "step": 6996
+    },
+    {
+      "epoch": 1.9403771491957849,
+      "grad_norm": 0.21450695395469666,
+      "learning_rate": 5.6855733374354404e-08,
+      "loss": 0.5173,
+      "step": 6997
+    },
+    {
+      "epoch": 1.9406544647809207,
+      "grad_norm": 0.21470417082309723,
+      "learning_rate": 5.633090485611114e-08,
+      "loss": 0.5008,
+      "step": 6998
+    },
+    {
+      "epoch": 1.9409317803660566,
+      "grad_norm": 0.20227287709712982,
+      "learning_rate": 5.580850442739732e-08,
+      "loss": 0.4925,
+      "step": 6999
+    },
+    {
+      "epoch": 1.9412090959511925,
+      "grad_norm": 0.20225927233695984,
+      "learning_rate": 5.5288532190145294e-08,
+      "loss": 0.4843,
+      "step": 7000
+    },
+    {
+      "epoch": 1.9414864115363284,
+      "grad_norm": 0.21021947264671326,
+      "learning_rate": 5.4770988245818336e-08,
+      "loss": 0.5028,
+      "step": 7001
+    },
+    {
+      "epoch": 1.9417637271214643,
+      "grad_norm": 0.21199429035186768,
+      "learning_rate": 5.4255872695400946e-08,
+      "loss": 0.5071,
+      "step": 7002
+    },
+    {
+      "epoch": 1.9420410427066002,
+      "grad_norm": 0.20802341401576996,
+      "learning_rate": 5.374318563940717e-08,
+      "loss": 0.4943,
+      "step": 7003
+    },
+    {
+      "epoch": 1.942318358291736,
+      "grad_norm": 0.20316016674041748,
+      "learning_rate": 5.323292717787504e-08,
+      "loss": 0.491,
+      "step": 7004
+    },
+    {
+      "epoch": 1.942595673876872,
+      "grad_norm": 0.20811815559864044,
+      "learning_rate": 5.272509741037074e-08,
+      "loss": 0.495,
+      "step": 7005
+    },
+    {
+      "epoch": 1.9428729894620078,
+      "grad_norm": 0.2118585854768753,
+      "learning_rate": 5.221969643598307e-08,
+      "loss": 0.49,
+      "step": 7006
+    },
+    {
+      "epoch": 1.9431503050471437,
+      "grad_norm": 0.2077036201953888,
+      "learning_rate": 5.171672435333036e-08,
+      "loss": 0.4795,
+      "step": 7007
+    },
+    {
+      "epoch": 1.9434276206322796,
+      "grad_norm": 0.21370601654052734,
+      "learning_rate": 5.121618126055633e-08,
+      "loss": 0.4795,
+      "step": 7008
+    },
+    {
+      "epoch": 1.9437049362174155,
+      "grad_norm": 0.20026057958602905,
+      "learning_rate": 5.071806725532868e-08,
+      "loss": 0.4859,
+      "step": 7009
+    },
+    {
+      "epoch": 1.9439822518025514,
+      "grad_norm": 0.20249269902706146,
+      "learning_rate": 5.022238243484467e-08,
+      "loss": 0.5196,
+      "step": 7010
+    },
+    {
+      "epoch": 1.9442595673876872,
+      "grad_norm": 0.2025049477815628,
+      "learning_rate": 4.972912689582276e-08,
+      "loss": 0.4916,
+      "step": 7011
+    },
+    {
+      "epoch": 1.9445368829728231,
+      "grad_norm": 0.21317322552204132,
+      "learning_rate": 4.923830073451374e-08,
+      "loss": 0.4984,
+      "step": 7012
+    },
+    {
+      "epoch": 1.944814198557959,
+      "grad_norm": 0.21346881985664368,
+      "learning_rate": 4.8749904046688223e-08,
+      "loss": 0.4878,
+      "step": 7013
+    },
+    {
+      "epoch": 1.945091514143095,
+      "grad_norm": 0.21879152953624725,
+      "learning_rate": 4.826393692764636e-08,
+      "loss": 0.5174,
+      "step": 7014
+    },
+    {
+      "epoch": 1.9453688297282308,
+      "grad_norm": 0.22700130939483643,
+      "learning_rate": 4.77803994722123e-08,
+      "loss": 0.4503,
+      "step": 7015
+    },
+    {
+      "epoch": 1.9456461453133667,
+      "grad_norm": 0.20824943482875824,
+      "learning_rate": 4.729929177473835e-08,
+      "loss": 0.5011,
+      "step": 7016
+    },
+    {
+      "epoch": 1.9459234608985025,
+      "grad_norm": 0.1998833268880844,
+      "learning_rate": 4.682061392910081e-08,
+      "loss": 0.4994,
+      "step": 7017
+    },
+    {
+      "epoch": 1.9462007764836384,
+      "grad_norm": 0.2063128501176834,
+      "learning_rate": 4.6344366028701346e-08,
+      "loss": 0.5316,
+      "step": 7018
+    },
+    {
+      "epoch": 1.9464780920687743,
+      "grad_norm": 0.20899119973182678,
+      "learning_rate": 4.5870548166469796e-08,
+      "loss": 0.4771,
+      "step": 7019
+    },
+    {
+      "epoch": 1.9467554076539102,
+      "grad_norm": 0.20648007094860077,
+      "learning_rate": 4.539916043485998e-08,
+      "loss": 0.4892,
+      "step": 7020
+    },
+    {
+      "epoch": 1.947032723239046,
+      "grad_norm": 0.2271704226732254,
+      "learning_rate": 4.4930202925852484e-08,
+      "loss": 0.5068,
+      "step": 7021
+    },
+    {
+      "epoch": 1.947310038824182,
+      "grad_norm": 0.2094396948814392,
+      "learning_rate": 4.446367573095328e-08,
+      "loss": 0.5027,
+      "step": 7022
+    },
+    {
+      "epoch": 1.9475873544093179,
+      "grad_norm": 0.20577369630336761,
+      "learning_rate": 4.3999578941195107e-08,
+      "loss": 0.5092,
+      "step": 7023
+    },
+    {
+      "epoch": 1.9478646699944537,
+      "grad_norm": 0.20665308833122253,
+      "learning_rate": 4.3537912647133305e-08,
+      "loss": 0.4884,
+      "step": 7024
+    },
+    {
+      "epoch": 1.9481419855795896,
+      "grad_norm": 0.21540111303329468,
+      "learning_rate": 4.3078676938852755e-08,
+      "loss": 0.5022,
+      "step": 7025
+    },
+    {
+      "epoch": 1.9484193011647255,
+      "grad_norm": 0.20871970057487488,
+      "learning_rate": 4.262187190596234e-08,
+      "loss": 0.5195,
+      "step": 7026
+    },
+    {
+      "epoch": 1.9486966167498614,
+      "grad_norm": 0.2026074379682541,
+      "learning_rate": 4.21674976375977e-08,
+      "loss": 0.5111,
+      "step": 7027
+    },
+    {
+      "epoch": 1.9489739323349973,
+      "grad_norm": 0.20567013323307037,
+      "learning_rate": 4.171555422241707e-08,
+      "loss": 0.4755,
+      "step": 7028
+    },
+    {
+      "epoch": 1.9492512479201332,
+      "grad_norm": 0.1948745846748352,
+      "learning_rate": 4.1266041748608265e-08,
+      "loss": 0.5021,
+      "step": 7029
+    },
+    {
+      "epoch": 1.949528563505269,
+      "grad_norm": 0.2238130271434784,
+      "learning_rate": 4.0818960303881656e-08,
+      "loss": 0.4984,
+      "step": 7030
+    },
+    {
+      "epoch": 1.949805879090405,
+      "grad_norm": 0.20423227548599243,
+      "learning_rate": 4.03743099754772e-08,
+      "loss": 0.495,
+      "step": 7031
+    },
+    {
+      "epoch": 1.9500831946755408,
+      "grad_norm": 0.2025814950466156,
+      "learning_rate": 3.9932090850156036e-08,
+      "loss": 0.5147,
+      "step": 7032
+    },
+    {
+      "epoch": 1.9503605102606767,
+      "grad_norm": 0.20636986196041107,
+      "learning_rate": 3.949230301420609e-08,
+      "loss": 0.496,
+      "step": 7033
+    },
+    {
+      "epoch": 1.9506378258458126,
+      "grad_norm": 0.1991264522075653,
+      "learning_rate": 3.905494655344483e-08,
+      "loss": 0.4856,
+      "step": 7034
+    },
+    {
+      "epoch": 1.9509151414309485,
+      "grad_norm": 0.21904256939888,
+      "learning_rate": 3.862002155320815e-08,
+      "loss": 0.4892,
+      "step": 7035
+    },
+    {
+      "epoch": 1.9511924570160843,
+      "grad_norm": 0.21712666749954224,
+      "learning_rate": 3.818752809836429e-08,
+      "loss": 0.5127,
+      "step": 7036
+    },
+    {
+      "epoch": 1.9514697726012202,
+      "grad_norm": 0.22889919579029083,
+      "learning_rate": 3.775746627330268e-08,
+      "loss": 0.4967,
+      "step": 7037
+    },
+    {
+      "epoch": 1.9517470881863561,
+      "grad_norm": 0.21116408705711365,
+      "learning_rate": 3.732983616193952e-08,
+      "loss": 0.4953,
+      "step": 7038
+    },
+    {
+      "epoch": 1.952024403771492,
+      "grad_norm": 0.2215505987405777,
+      "learning_rate": 3.6904637847719195e-08,
+      "loss": 0.4778,
+      "step": 7039
+    },
+    {
+      "epoch": 1.9523017193566279,
+      "grad_norm": 0.21022702753543854,
+      "learning_rate": 3.6481871413605874e-08,
+      "loss": 0.4975,
+      "step": 7040
+    },
+    {
+      "epoch": 1.9525790349417638,
+      "grad_norm": 0.20214217901229858,
+      "learning_rate": 3.606153694209608e-08,
+      "loss": 0.5121,
+      "step": 7041
+    },
+    {
+      "epoch": 1.9528563505268997,
+      "grad_norm": 0.20192945003509521,
+      "learning_rate": 3.5643634515204747e-08,
+      "loss": 0.4975,
+      "step": 7042
+    },
+    {
+      "epoch": 1.9531336661120355,
+      "grad_norm": 0.21496793627738953,
+      "learning_rate": 3.522816421447778e-08,
+      "loss": 0.5,
+      "step": 7043
+    },
+    {
+      "epoch": 1.9534109816971714,
+      "grad_norm": 0.20407256484031677,
+      "learning_rate": 3.4815126120983646e-08,
+      "loss": 0.5053,
+      "step": 7044
+    },
+    {
+      "epoch": 1.9536882972823073,
+      "grad_norm": 0.2113526612520218,
+      "learning_rate": 3.4404520315316216e-08,
+      "loss": 0.4911,
+      "step": 7045
+    },
+    {
+      "epoch": 1.9539656128674432,
+      "grad_norm": 0.2122277021408081,
+      "learning_rate": 3.399634687759751e-08,
+      "loss": 0.5139,
+      "step": 7046
+    },
+    {
+      "epoch": 1.954242928452579,
+      "grad_norm": 0.2064533829689026,
+      "learning_rate": 3.359060588747354e-08,
+      "loss": 0.5004,
+      "step": 7047
+    },
+    {
+      "epoch": 1.954520244037715,
+      "grad_norm": 0.22271350026130676,
+      "learning_rate": 3.318729742411153e-08,
+      "loss": 0.5052,
+      "step": 7048
+    },
+    {
+      "epoch": 1.9547975596228508,
+      "grad_norm": 0.20162583887577057,
+      "learning_rate": 3.278642156620965e-08,
+      "loss": 0.4936,
+      "step": 7049
+    },
+    {
+      "epoch": 1.9550748752079867,
+      "grad_norm": 0.20813512802124023,
+      "learning_rate": 3.238797839199143e-08,
+      "loss": 0.4976,
+      "step": 7050
+    },
+    {
+      "epoch": 1.9553521907931226,
+      "grad_norm": 0.20408669114112854,
+      "learning_rate": 3.1991967979200235e-08,
+      "loss": 0.4978,
+      "step": 7051
+    },
+    {
+      "epoch": 1.9556295063782585,
+      "grad_norm": 0.2017277628183365,
+      "learning_rate": 3.159839040511037e-08,
+      "loss": 0.4948,
+      "step": 7052
+    },
+    {
+      "epoch": 1.9559068219633944,
+      "grad_norm": 0.2124013453722,
+      "learning_rate": 3.120724574651873e-08,
+      "loss": 0.4853,
+      "step": 7053
+    },
+    {
+      "epoch": 1.9561841375485303,
+      "grad_norm": 0.20675784349441528,
+      "learning_rate": 3.0818534079747606e-08,
+      "loss": 0.4972,
+      "step": 7054
+    },
+    {
+      "epoch": 1.9564614531336662,
+      "grad_norm": 0.21523889899253845,
+      "learning_rate": 3.043225548064465e-08,
+      "loss": 0.5126,
+      "step": 7055
+    },
+    {
+      "epoch": 1.956738768718802,
+      "grad_norm": 0.20514823496341705,
+      "learning_rate": 3.004841002458431e-08,
+      "loss": 0.5061,
+      "step": 7056
+    },
+    {
+      "epoch": 1.957016084303938,
+      "grad_norm": 0.2089594602584839,
+      "learning_rate": 2.966699778646359e-08,
+      "loss": 0.4925,
+      "step": 7057
+    },
+    {
+      "epoch": 1.9572933998890738,
+      "grad_norm": 0.21636748313903809,
+      "learning_rate": 2.92880188407077e-08,
+      "loss": 0.4895,
+      "step": 7058
+    },
+    {
+      "epoch": 1.9575707154742097,
+      "grad_norm": 0.2242691069841385,
+      "learning_rate": 2.8911473261264423e-08,
+      "loss": 0.4819,
+      "step": 7059
+    },
+    {
+      "epoch": 1.9578480310593456,
+      "grad_norm": 0.19989970326423645,
+      "learning_rate": 2.853736112160693e-08,
+      "loss": 0.514,
+      "step": 7060
+    },
+    {
+      "epoch": 1.9581253466444815,
+      "grad_norm": 0.20352482795715332,
+      "learning_rate": 2.8165682494736556e-08,
+      "loss": 0.5165,
+      "step": 7061
+    },
+    {
+      "epoch": 1.9584026622296173,
+      "grad_norm": 0.21366307139396667,
+      "learning_rate": 2.7796437453177228e-08,
+      "loss": 0.5171,
+      "step": 7062
+    },
+    {
+      "epoch": 1.9586799778147532,
+      "grad_norm": 0.20763346552848816,
+      "learning_rate": 2.7429626068976865e-08,
+      "loss": 0.4874,
+      "step": 7063
+    },
+    {
+      "epoch": 1.958957293399889,
+      "grad_norm": 0.20051513612270355,
+      "learning_rate": 2.7065248413710166e-08,
+      "loss": 0.4858,
+      "step": 7064
+    },
+    {
+      "epoch": 1.959234608985025,
+      "grad_norm": 0.22556112706661224,
+      "learning_rate": 2.6703304558478583e-08,
+      "loss": 0.4879,
+      "step": 7065
+    },
+    {
+      "epoch": 1.9595119245701609,
+      "grad_norm": 0.20919561386108398,
+      "learning_rate": 2.634379457390618e-08,
+      "loss": 0.4919,
+      "step": 7066
+    },
+    {
+      "epoch": 1.9597892401552968,
+      "grad_norm": 0.20339058339595795,
+      "learning_rate": 2.5986718530142396e-08,
+      "loss": 0.5165,
+      "step": 7067
+    },
+    {
+      "epoch": 1.9600665557404326,
+      "grad_norm": 0.19766461849212646,
+      "learning_rate": 2.5632076496862058e-08,
+      "loss": 0.4615,
+      "step": 7068
+    },
+    {
+      "epoch": 1.9603438713255685,
+      "grad_norm": 0.20357239246368408,
+      "learning_rate": 2.527986854326675e-08,
+      "loss": 0.4637,
+      "step": 7069
+    },
+    {
+      "epoch": 1.9606211869107044,
+      "grad_norm": 0.21048687398433685,
+      "learning_rate": 2.493009473807928e-08,
+      "loss": 0.485,
+      "step": 7070
+    },
+    {
+      "epoch": 1.9608985024958403,
+      "grad_norm": 0.21230055391788483,
+      "learning_rate": 2.4582755149551995e-08,
+      "loss": 0.5122,
+      "step": 7071
+    },
+    {
+      "epoch": 1.9611758180809762,
+      "grad_norm": 0.21410781145095825,
+      "learning_rate": 2.4237849845459848e-08,
+      "loss": 0.5042,
+      "step": 7072
+    },
+    {
+      "epoch": 1.961453133666112,
+      "grad_norm": 0.2122282236814499,
+      "learning_rate": 2.3895378893100394e-08,
+      "loss": 0.5211,
+      "step": 7073
+    },
+    {
+      "epoch": 1.961730449251248,
+      "grad_norm": 0.21611341834068298,
+      "learning_rate": 2.3555342359302123e-08,
+      "loss": 0.5021,
+      "step": 7074
+    },
+    {
+      "epoch": 1.9620077648363838,
+      "grad_norm": 0.22394365072250366,
+      "learning_rate": 2.321774031041335e-08,
+      "loss": 0.5268,
+      "step": 7075
+    },
+    {
+      "epoch": 1.9622850804215197,
+      "grad_norm": 0.21854601800441742,
+      "learning_rate": 2.2882572812309156e-08,
+      "loss": 0.4732,
+      "step": 7076
+    },
+    {
+      "epoch": 1.9625623960066556,
+      "grad_norm": 0.21542960405349731,
+      "learning_rate": 2.2549839930390004e-08,
+      "loss": 0.5072,
+      "step": 7077
+    },
+    {
+      "epoch": 1.9628397115917915,
+      "grad_norm": 0.2258591204881668,
+      "learning_rate": 2.221954172958174e-08,
+      "loss": 0.517,
+      "step": 7078
+    },
+    {
+      "epoch": 1.9631170271769274,
+      "grad_norm": 0.22644679248332977,
+      "learning_rate": 2.1891678274332804e-08,
+      "loss": 0.4841,
+      "step": 7079
+    },
+    {
+      "epoch": 1.9633943427620633,
+      "grad_norm": 0.19779692590236664,
+      "learning_rate": 2.1566249628618417e-08,
+      "loss": 0.4987,
+      "step": 7080
+    },
+    {
+      "epoch": 1.9636716583471991,
+      "grad_norm": 0.21196384727954865,
+      "learning_rate": 2.124325585593917e-08,
+      "loss": 0.5227,
+      "step": 7081
+    },
+    {
+      "epoch": 1.963948973932335,
+      "grad_norm": 0.20542475581169128,
+      "learning_rate": 2.092269701931826e-08,
+      "loss": 0.4849,
+      "step": 7082
+    },
+    {
+      "epoch": 1.964226289517471,
+      "grad_norm": 0.1972956657409668,
+      "learning_rate": 2.060457318130704e-08,
+      "loss": 0.4605,
+      "step": 7083
+    },
+    {
+      "epoch": 1.9645036051026068,
+      "grad_norm": 0.2096942663192749,
+      "learning_rate": 2.028888440397947e-08,
+      "loss": 0.4854,
+      "step": 7084
+    },
+    {
+      "epoch": 1.9647809206877427,
+      "grad_norm": 0.20811447501182556,
+      "learning_rate": 1.9975630748933493e-08,
+      "loss": 0.4842,
+      "step": 7085
+    },
+    {
+      "epoch": 1.9650582362728786,
+      "grad_norm": 0.21638718247413635,
+      "learning_rate": 1.9664812277292442e-08,
+      "loss": 0.5213,
+      "step": 7086
+    },
+    {
+      "epoch": 1.9653355518580145,
+      "grad_norm": 0.22020205855369568,
+      "learning_rate": 1.935642904970919e-08,
+      "loss": 0.5084,
+      "step": 7087
+    },
+    {
+      "epoch": 1.9656128674431503,
+      "grad_norm": 0.2070709615945816,
+      "learning_rate": 1.9050481126353658e-08,
+      "loss": 0.4763,
+      "step": 7088
+    },
+    {
+      "epoch": 1.9658901830282862,
+      "grad_norm": 0.20314981043338776,
+      "learning_rate": 1.8746968566926704e-08,
+      "loss": 0.5041,
+      "step": 7089
+    },
+    {
+      "epoch": 1.966167498613422,
+      "grad_norm": 0.21381008625030518,
+      "learning_rate": 1.844589143064901e-08,
+      "loss": 0.4862,
+      "step": 7090
+    },
+    {
+      "epoch": 1.966444814198558,
+      "grad_norm": 0.2065730094909668,
+      "learning_rate": 1.814724977627219e-08,
+      "loss": 0.5071,
+      "step": 7091
+    },
+    {
+      "epoch": 1.9667221297836939,
+      "grad_norm": 0.2038286328315735,
+      "learning_rate": 1.7851043662066302e-08,
+      "loss": 0.4935,
+      "step": 7092
+    },
+    {
+      "epoch": 1.9669994453688298,
+      "grad_norm": 0.2133897840976715,
+      "learning_rate": 1.7557273145830943e-08,
+      "loss": 0.5179,
+      "step": 7093
+    },
+    {
+      "epoch": 1.9672767609539656,
+      "grad_norm": 0.20114904642105103,
+      "learning_rate": 1.726593828488693e-08,
+      "loss": 0.5107,
+      "step": 7094
+    },
+    {
+      "epoch": 1.9675540765391015,
+      "grad_norm": 0.21013131737709045,
+      "learning_rate": 1.697703913608184e-08,
+      "loss": 0.5173,
+      "step": 7095
+    },
+    {
+      "epoch": 1.9678313921242374,
+      "grad_norm": 0.2039504200220108,
+      "learning_rate": 1.669057575578864e-08,
+      "loss": 0.4999,
+      "step": 7096
+    },
+    {
+      "epoch": 1.9681087077093733,
+      "grad_norm": 0.20330029726028442,
+      "learning_rate": 1.6406548199902893e-08,
+      "loss": 0.5164,
+      "step": 7097
+    },
+    {
+      "epoch": 1.9683860232945092,
+      "grad_norm": 0.20391713082790375,
+      "learning_rate": 1.6124956523846934e-08,
+      "loss": 0.505,
+      "step": 7098
+    },
+    {
+      "epoch": 1.968663338879645,
+      "grad_norm": 0.2079223245382309,
+      "learning_rate": 1.5845800782564314e-08,
+      "loss": 0.4836,
+      "step": 7099
+    },
+    {
+      "epoch": 1.968940654464781,
+      "grad_norm": 0.20631448924541473,
+      "learning_rate": 1.5569081030529507e-08,
+      "loss": 0.4875,
+      "step": 7100
+    },
+    {
+      "epoch": 1.9692179700499168,
+      "grad_norm": 0.21116302907466888,
+      "learning_rate": 1.5294797321734057e-08,
+      "loss": 0.5115,
+      "step": 7101
+    },
+    {
+      "epoch": 1.9694952856350527,
+      "grad_norm": 0.21221213042736053,
+      "learning_rate": 1.5022949709700417e-08,
+      "loss": 0.495,
+      "step": 7102
+    },
+    {
+      "epoch": 1.9697726012201886,
+      "grad_norm": 0.20361442863941193,
+      "learning_rate": 1.4753538247472277e-08,
+      "loss": 0.4937,
+      "step": 7103
+    },
+    {
+      "epoch": 1.9700499168053245,
+      "grad_norm": 0.2115209698677063,
+      "learning_rate": 1.448656298761869e-08,
+      "loss": 0.4856,
+      "step": 7104
+    },
+    {
+      "epoch": 1.9703272323904604,
+      "grad_norm": 0.20742273330688477,
+      "learning_rate": 1.42220239822341e-08,
+      "loss": 0.4659,
+      "step": 7105
+    },
+    {
+      "epoch": 1.9706045479755963,
+      "grad_norm": 0.20554955303668976,
+      "learning_rate": 1.3959921282938327e-08,
+      "loss": 0.5062,
+      "step": 7106
+    },
+    {
+      "epoch": 1.9708818635607321,
+      "grad_norm": 0.20452530682086945,
+      "learning_rate": 1.3700254940872404e-08,
+      "loss": 0.4663,
+      "step": 7107
+    },
+    {
+      "epoch": 1.971159179145868,
+      "grad_norm": 0.2129167914390564,
+      "learning_rate": 1.3443025006705523e-08,
+      "loss": 0.5172,
+      "step": 7108
+    },
+    {
+      "epoch": 1.971436494731004,
+      "grad_norm": 0.21334435045719147,
+      "learning_rate": 1.3188231530628092e-08,
+      "loss": 0.513,
+      "step": 7109
+    },
+    {
+      "epoch": 1.9717138103161398,
+      "grad_norm": 0.20458266139030457,
+      "learning_rate": 1.2935874562360062e-08,
+      "loss": 0.4769,
+      "step": 7110
+    },
+    {
+      "epoch": 1.9719911259012757,
+      "grad_norm": 0.20730172097682953,
+      "learning_rate": 1.2685954151141211e-08,
+      "loss": 0.5029,
+      "step": 7111
+    },
+    {
+      "epoch": 1.9722684414864116,
+      "grad_norm": 0.21345804631710052,
+      "learning_rate": 1.2438470345738085e-08,
+      "loss": 0.4918,
+      "step": 7112
+    },
+    {
+      "epoch": 1.9725457570715474,
+      "grad_norm": 0.22697855532169342,
+      "learning_rate": 1.2193423194439835e-08,
+      "loss": 0.5038,
+      "step": 7113
+    },
+    {
+      "epoch": 1.9728230726566833,
+      "grad_norm": 0.20521080493927002,
+      "learning_rate": 1.1950812745063766e-08,
+      "loss": 0.4882,
+      "step": 7114
+    },
+    {
+      "epoch": 1.9731003882418192,
+      "grad_norm": 0.20702117681503296,
+      "learning_rate": 1.1710639044948401e-08,
+      "loss": 0.4954,
+      "step": 7115
+    },
+    {
+      "epoch": 1.973377703826955,
+      "grad_norm": 0.21297289431095123,
+      "learning_rate": 1.1472902140959029e-08,
+      "loss": 0.5015,
+      "step": 7116
+    },
+    {
+      "epoch": 1.973655019412091,
+      "grad_norm": 0.20684416592121124,
+      "learning_rate": 1.1237602079483545e-08,
+      "loss": 0.4632,
+      "step": 7117
+    },
+    {
+      "epoch": 1.9739323349972269,
+      "grad_norm": 0.20604479312896729,
+      "learning_rate": 1.100473890643522e-08,
+      "loss": 0.495,
+      "step": 7118
+    },
+    {
+      "epoch": 1.9742096505823628,
+      "grad_norm": 0.2087058126926422,
+      "learning_rate": 1.0774312667251319e-08,
+      "loss": 0.4967,
+      "step": 7119
+    },
+    {
+      "epoch": 1.9744869661674986,
+      "grad_norm": 0.23602424561977386,
+      "learning_rate": 1.0546323406895875e-08,
+      "loss": 0.4913,
+      "step": 7120
+    },
+    {
+      "epoch": 1.9747642817526345,
+      "grad_norm": 0.20307756960391998,
+      "learning_rate": 1.0320771169854137e-08,
+      "loss": 0.4753,
+      "step": 7121
+    },
+    {
+      "epoch": 1.9750415973377704,
+      "grad_norm": 0.20893272757530212,
+      "learning_rate": 1.0097656000136735e-08,
+      "loss": 0.4845,
+      "step": 7122
+    },
+    {
+      "epoch": 1.9753189129229063,
+      "grad_norm": 0.2109754979610443,
+      "learning_rate": 9.876977941282451e-09,
+      "loss": 0.4906,
+      "step": 7123
+    },
+    {
+      "epoch": 1.9755962285080422,
+      "grad_norm": 0.20519696176052094,
+      "learning_rate": 9.658737036347121e-09,
+      "loss": 0.496,
+      "step": 7124
+    },
+    {
+      "epoch": 1.975873544093178,
+      "grad_norm": 0.21880246698856354,
+      "learning_rate": 9.442933327918902e-09,
+      "loss": 0.517,
+      "step": 7125
+    },
+    {
+      "epoch": 1.976150859678314,
+      "grad_norm": 0.21169376373291016,
+      "learning_rate": 9.229566858105777e-09,
+      "loss": 0.4915,
+      "step": 7126
+    },
+    {
+      "epoch": 1.9764281752634498,
+      "grad_norm": 0.20725256204605103,
+      "learning_rate": 9.01863766853972e-09,
+      "loss": 0.4964,
+      "step": 7127
+    },
+    {
+      "epoch": 1.9767054908485857,
+      "grad_norm": 0.22511859238147736,
+      "learning_rate": 8.810145800379477e-09,
+      "loss": 0.5046,
+      "step": 7128
+    },
+    {
+      "epoch": 1.9769828064337216,
+      "grad_norm": 0.22022607922554016,
+      "learning_rate": 8.60409129430917e-09,
+      "loss": 0.4962,
+      "step": 7129
+    },
+    {
+      "epoch": 1.9772601220188575,
+      "grad_norm": 0.21660543978214264,
+      "learning_rate": 8.400474190532747e-09,
+      "loss": 0.524,
+      "step": 7130
+    },
+    {
+      "epoch": 1.9775374376039934,
+      "grad_norm": 0.21604527533054352,
+      "learning_rate": 8.199294528783707e-09,
+      "loss": 0.499,
+      "step": 7131
+    },
+    {
+      "epoch": 1.9778147531891292,
+      "grad_norm": 0.20662792026996613,
+      "learning_rate": 8.000552348315371e-09,
+      "loss": 0.5365,
+      "step": 7132
+    },
+    {
+      "epoch": 1.9780920687742651,
+      "grad_norm": 0.20979103446006775,
+      "learning_rate": 7.804247687909216e-09,
+      "loss": 0.4771,
+      "step": 7133
+    },
+    {
+      "epoch": 1.978369384359401,
+      "grad_norm": 0.21938779950141907,
+      "learning_rate": 7.610380585867937e-09,
+      "loss": 0.498,
+      "step": 7134
+    },
+    {
+      "epoch": 1.978646699944537,
+      "grad_norm": 0.20905134081840515,
+      "learning_rate": 7.418951080020997e-09,
+      "loss": 0.5047,
+      "step": 7135
+    },
+    {
+      "epoch": 1.9789240155296728,
+      "grad_norm": 0.2120780348777771,
+      "learning_rate": 7.229959207721848e-09,
+      "loss": 0.492,
+      "step": 7136
+    },
+    {
+      "epoch": 1.9792013311148087,
+      "grad_norm": 0.21072250604629517,
+      "learning_rate": 7.043405005847936e-09,
+      "loss": 0.503,
+      "step": 7137
+    },
+    {
+      "epoch": 1.9794786466999446,
+      "grad_norm": 0.2067342847585678,
+      "learning_rate": 6.859288510799311e-09,
+      "loss": 0.5092,
+      "step": 7138
+    },
+    {
+      "epoch": 1.9797559622850804,
+      "grad_norm": 0.21522685885429382,
+      "learning_rate": 6.67760975850279e-09,
+      "loss": 0.5161,
+      "step": 7139
+    },
+    {
+      "epoch": 1.9800332778702163,
+      "grad_norm": 0.20429226756095886,
+      "learning_rate": 6.498368784409181e-09,
+      "loss": 0.5381,
+      "step": 7140
+    },
+    {
+      "epoch": 1.9803105934553522,
+      "grad_norm": 0.2123131901025772,
+      "learning_rate": 6.321565623494674e-09,
+      "loss": 0.5146,
+      "step": 7141
+    },
+    {
+      "epoch": 1.980587909040488,
+      "grad_norm": 0.21125838160514832,
+      "learning_rate": 6.147200310253898e-09,
+      "loss": 0.4832,
+      "step": 7142
+    },
+    {
+      "epoch": 1.980865224625624,
+      "grad_norm": 0.20799988508224487,
+      "learning_rate": 5.9752728787138005e-09,
+      "loss": 0.4841,
+      "step": 7143
+    },
+    {
+      "epoch": 1.9811425402107599,
+      "grad_norm": 0.22431087493896484,
+      "learning_rate": 5.805783362421158e-09,
+      "loss": 0.5086,
+      "step": 7144
+    },
+    {
+      "epoch": 1.9814198557958957,
+      "grad_norm": 0.2249828279018402,
+      "learning_rate": 5.6387317944481265e-09,
+      "loss": 0.5089,
+      "step": 7145
+    },
+    {
+      "epoch": 1.9816971713810316,
+      "grad_norm": 0.22779472172260284,
+      "learning_rate": 5.474118207389467e-09,
+      "loss": 0.5253,
+      "step": 7146
+    },
+    {
+      "epoch": 1.9819744869661675,
+      "grad_norm": 0.21490508317947388,
+      "learning_rate": 5.311942633366706e-09,
+      "loss": 0.4918,
+      "step": 7147
+    },
+    {
+      "epoch": 1.9822518025513034,
+      "grad_norm": 0.21175016462802887,
+      "learning_rate": 5.152205104023977e-09,
+      "loss": 0.4858,
+      "step": 7148
+    },
+    {
+      "epoch": 1.9825291181364393,
+      "grad_norm": 0.22119320929050446,
+      "learning_rate": 4.99490565053079e-09,
+      "loss": 0.4984,
+      "step": 7149
+    },
+    {
+      "epoch": 1.9828064337215752,
+      "grad_norm": 0.22225895524024963,
+      "learning_rate": 4.840044303582036e-09,
+      "loss": 0.5394,
+      "step": 7150
+    },
+    {
+      "epoch": 1.983083749306711,
+      "grad_norm": 0.20062761008739471,
+      "learning_rate": 4.687621093392436e-09,
+      "loss": 0.4905,
+      "step": 7151
+    },
+    {
+      "epoch": 1.983361064891847,
+      "grad_norm": 0.20917369425296783,
+      "learning_rate": 4.537636049704863e-09,
+      "loss": 0.5128,
+      "step": 7152
+    },
+    {
+      "epoch": 1.9836383804769828,
+      "grad_norm": 0.21111933887004852,
+      "learning_rate": 4.390089201786185e-09,
+      "loss": 0.4886,
+      "step": 7153
+    },
+    {
+      "epoch": 1.9839156960621187,
+      "grad_norm": 0.2019733041524887,
+      "learning_rate": 4.244980578424485e-09,
+      "loss": 0.4908,
+      "step": 7154
+    },
+    {
+      "epoch": 1.9841930116472546,
+      "grad_norm": 0.21139049530029297,
+      "learning_rate": 4.1023102079373875e-09,
+      "loss": 0.5261,
+      "step": 7155
+    },
+    {
+      "epoch": 1.9844703272323905,
+      "grad_norm": 0.2183750867843628,
+      "learning_rate": 3.962078118162349e-09,
+      "loss": 0.5286,
+      "step": 7156
+    },
+    {
+      "epoch": 1.9847476428175264,
+      "grad_norm": 0.21051201224327087,
+      "learning_rate": 3.824284336460815e-09,
+      "loss": 0.4948,
+      "step": 7157
+    },
+    {
+      "epoch": 1.9850249584026622,
+      "grad_norm": 0.21117429435253143,
+      "learning_rate": 3.6889288897223872e-09,
+      "loss": 0.4867,
+      "step": 7158
+    },
+    {
+      "epoch": 1.9853022739877981,
+      "grad_norm": 0.20643287897109985,
+      "learning_rate": 3.556011804356496e-09,
+      "loss": 0.494,
+      "step": 7159
+    },
+    {
+      "epoch": 1.985579589572934,
+      "grad_norm": 0.21045584976673126,
+      "learning_rate": 3.425533106300727e-09,
+      "loss": 0.5114,
+      "step": 7160
+    },
+    {
+      "epoch": 1.98585690515807,
+      "grad_norm": 0.20693178474903107,
+      "learning_rate": 3.297492821013881e-09,
+      "loss": 0.5092,
+      "step": 7161
+    },
+    {
+      "epoch": 1.9861342207432058,
+      "grad_norm": 0.19891990721225739,
+      "learning_rate": 3.1718909734787526e-09,
+      "loss": 0.4961,
+      "step": 7162
+    },
+    {
+      "epoch": 1.9864115363283417,
+      "grad_norm": 0.21904566884040833,
+      "learning_rate": 3.0487275882062906e-09,
+      "loss": 0.4908,
+      "step": 7163
+    },
+    {
+      "epoch": 1.9866888519134775,
+      "grad_norm": 0.21359799802303314,
+      "learning_rate": 2.9280026892272715e-09,
+      "loss": 0.4848,
+      "step": 7164
+    },
+    {
+      "epoch": 1.9869661674986134,
+      "grad_norm": 0.21357576549053192,
+      "learning_rate": 2.809716300097853e-09,
+      "loss": 0.4776,
+      "step": 7165
+    },
+    {
+      "epoch": 1.9872434830837493,
+      "grad_norm": 0.2025846391916275,
+      "learning_rate": 2.6938684439009598e-09,
+      "loss": 0.477,
+      "step": 7166
+    },
+    {
+      "epoch": 1.9875207986688852,
+      "grad_norm": 0.20797260105609894,
+      "learning_rate": 2.5804591432393442e-09,
+      "loss": 0.4981,
+      "step": 7167
+    },
+    {
+      "epoch": 1.987798114254021,
+      "grad_norm": 0.21296679973602295,
+      "learning_rate": 2.469488420242527e-09,
+      "loss": 0.4845,
+      "step": 7168
+    },
+    {
+      "epoch": 1.988075429839157,
+      "grad_norm": 0.20188210904598236,
+      "learning_rate": 2.3609562965654085e-09,
+      "loss": 0.486,
+      "step": 7169
+    },
+    {
+      "epoch": 1.9883527454242929,
+      "grad_norm": 0.2065025269985199,
+      "learning_rate": 2.2548627933841047e-09,
+      "loss": 0.4778,
+      "step": 7170
+    },
+    {
+      "epoch": 1.9886300610094287,
+      "grad_norm": 0.21647189557552338,
+      "learning_rate": 2.151207931400112e-09,
+      "loss": 0.4972,
+      "step": 7171
+    },
+    {
+      "epoch": 1.9889073765945646,
+      "grad_norm": 0.21635204553604126,
+      "learning_rate": 2.0499917308403062e-09,
+      "loss": 0.5197,
+      "step": 7172
+    },
+    {
+      "epoch": 1.9891846921797005,
+      "grad_norm": 0.20818372070789337,
+      "learning_rate": 1.951214211452779e-09,
+      "loss": 0.4838,
+      "step": 7173
+    },
+    {
+      "epoch": 1.9894620077648364,
+      "grad_norm": 0.1977744996547699,
+      "learning_rate": 1.8548753925137773e-09,
+      "loss": 0.4944,
+      "step": 7174
+    },
+    {
+      "epoch": 1.9897393233499723,
+      "grad_norm": 0.38426437973976135,
+      "learning_rate": 1.7609752928207657e-09,
+      "loss": 0.5029,
+      "step": 7175
+    },
+    {
+      "epoch": 1.9900166389351082,
+      "grad_norm": 0.22992363572120667,
+      "learning_rate": 1.6695139306965869e-09,
+      "loss": 0.5092,
+      "step": 7176
+    },
+    {
+      "epoch": 1.990293954520244,
+      "grad_norm": 0.20251336693763733,
+      "learning_rate": 1.580491323986688e-09,
+      "loss": 0.4982,
+      "step": 7177
+    },
+    {
+      "epoch": 1.99057127010538,
+      "grad_norm": 0.21631348133087158,
+      "learning_rate": 1.4939074900618965e-09,
+      "loss": 0.5016,
+      "step": 7178
+    },
+    {
+      "epoch": 1.9908485856905158,
+      "grad_norm": 0.21560998260974884,
+      "learning_rate": 1.4097624458184188e-09,
+      "loss": 0.5206,
+      "step": 7179
+    },
+    {
+      "epoch": 1.9911259012756517,
+      "grad_norm": 0.2111794650554657,
+      "learning_rate": 1.328056207673678e-09,
+      "loss": 0.5029,
+      "step": 7180
+    },
+    {
+      "epoch": 1.9914032168607876,
+      "grad_norm": 0.21028463542461395,
+      "learning_rate": 1.2487887915704766e-09,
+      "loss": 0.5018,
+      "step": 7181
+    },
+    {
+      "epoch": 1.9916805324459235,
+      "grad_norm": 0.2057926207780838,
+      "learning_rate": 1.1719602129769968e-09,
+      "loss": 0.5104,
+      "step": 7182
+    },
+    {
+      "epoch": 1.9919578480310594,
+      "grad_norm": 0.2165459394454956,
+      "learning_rate": 1.097570486885413e-09,
+      "loss": 0.5189,
+      "step": 7183
+    },
+    {
+      "epoch": 1.9922351636161952,
+      "grad_norm": 0.21551057696342468,
+      "learning_rate": 1.0256196278091156e-09,
+      "loss": 0.4608,
+      "step": 7184
+    },
+    {
+      "epoch": 1.9925124792013311,
+      "grad_norm": 0.22477814555168152,
+      "learning_rate": 9.561076497882626e-10,
+      "loss": 0.4854,
+      "step": 7185
+    },
+    {
+      "epoch": 1.992789794786467,
+      "grad_norm": 0.21325767040252686,
+      "learning_rate": 8.890345663870037e-10,
+      "loss": 0.5055,
+      "step": 7186
+    },
+    {
+      "epoch": 1.993067110371603,
+      "grad_norm": 0.20826996862888336,
+      "learning_rate": 8.244003906934806e-10,
+      "loss": 0.5067,
+      "step": 7187
+    },
+    {
+      "epoch": 1.9933444259567388,
+      "grad_norm": 0.2099534571170807,
+      "learning_rate": 7.622051353184389e-10,
+      "loss": 0.5104,
+      "step": 7188
+    },
+    {
+      "epoch": 1.9936217415418747,
+      "grad_norm": 0.21604932844638824,
+      "learning_rate": 7.024488123980044e-10,
+      "loss": 0.4922,
+      "step": 7189
+    },
+    {
+      "epoch": 1.9938990571270105,
+      "grad_norm": 0.21847958862781525,
+      "learning_rate": 6.451314335922942e-10,
+      "loss": 0.4976,
+      "step": 7190
+    },
+    {
+      "epoch": 1.9941763727121464,
+      "grad_norm": 0.21057890355587006,
+      "learning_rate": 5.902530100854175e-10,
+      "loss": 0.5255,
+      "step": 7191
+    },
+    {
+      "epoch": 1.9944536882972823,
+      "grad_norm": 0.20802158117294312,
+      "learning_rate": 5.378135525868633e-10,
+      "loss": 0.5211,
+      "step": 7192
+    },
+    {
+      "epoch": 1.9947310038824182,
+      "grad_norm": 0.2087002694606781,
+      "learning_rate": 4.878130713273365e-10,
+      "loss": 0.5145,
+      "step": 7193
+    },
+    {
+      "epoch": 1.995008319467554,
+      "grad_norm": 0.211053267121315,
+      "learning_rate": 4.402515760629222e-10,
+      "loss": 0.5037,
+      "step": 7194
+    },
+    {
+      "epoch": 1.99528563505269,
+      "grad_norm": 0.20845288038253784,
+      "learning_rate": 3.9512907607647256e-10,
+      "loss": 0.4992,
+      "step": 7195
+    },
+    {
+      "epoch": 1.9955629506378258,
+      "grad_norm": 0.2077481597661972,
+      "learning_rate": 3.524455801706683e-10,
+      "loss": 0.4787,
+      "step": 7196
+    },
+    {
+      "epoch": 1.9958402662229617,
+      "grad_norm": 0.21687790751457214,
+      "learning_rate": 3.1220109667357e-10,
+      "loss": 0.488,
+      "step": 7197
+    },
+    {
+      "epoch": 1.9961175818080976,
+      "grad_norm": 0.20425012707710266,
+      "learning_rate": 2.743956334400055e-10,
+      "loss": 0.5079,
+      "step": 7198
+    },
+    {
+      "epoch": 1.9963948973932335,
+      "grad_norm": 0.19561026990413666,
+      "learning_rate": 2.3902919784601905e-10,
+      "loss": 0.4914,
+      "step": 7199
+    },
+    {
+      "epoch": 1.9966722129783694,
+      "grad_norm": 0.20784446597099304,
+      "learning_rate": 2.0610179679164676e-10,
+      "loss": 0.4945,
+      "step": 7200
+    },
+    {
+      "epoch": 1.9969495285635053,
+      "grad_norm": 0.19664216041564941,
+      "learning_rate": 1.7561343670230436e-10,
+      "loss": 0.4889,
+      "step": 7201
+    },
+    {
+      "epoch": 1.9972268441486412,
+      "grad_norm": 0.20693275332450867,
+      "learning_rate": 1.4756412352878724e-10,
+      "loss": 0.5138,
+      "step": 7202
+    },
+    {
+      "epoch": 1.997504159733777,
+      "grad_norm": 0.20792469382286072,
+      "learning_rate": 1.2195386274171938e-10,
+      "loss": 0.5086,
+      "step": 7203
+    },
+    {
+      "epoch": 1.997781475318913,
+      "grad_norm": 0.2034110128879547,
+      "learning_rate": 9.878265933987995e-11,
+      "loss": 0.4793,
+      "step": 7204
+    },
+    {
+      "epoch": 1.9980587909040488,
+      "grad_norm": 0.2156227082014084,
+      "learning_rate": 7.805051784326445e-11,
+      "loss": 0.5048,
+      "step": 7205
+    },
+    {
+      "epoch": 1.9983361064891847,
+      "grad_norm": 0.20901308953762054,
+      "learning_rate": 5.975744230002356e-11,
+      "loss": 0.5019,
+      "step": 7206
+    },
+    {
+      "epoch": 1.9986134220743206,
+      "grad_norm": 0.20007333159446716,
+      "learning_rate": 4.3903436276748756e-11,
+      "loss": 0.4837,
+      "step": 7207
+    },
+    {
+      "epoch": 1.9988907376594565,
+      "grad_norm": 0.21206532418727875,
+      "learning_rate": 3.048850286679894e-11,
+      "loss": 0.4823,
+      "step": 7208
+    },
+    {
+      "epoch": 1.9991680532445923,
+      "grad_norm": 0.2093973159790039,
+      "learning_rate": 1.9512644690300452e-11,
+      "loss": 0.4759,
+      "step": 7209
+    },
+    {
+      "epoch": 1.9994453688297282,
+      "grad_norm": 0.2100788652896881,
+      "learning_rate": 1.0975863887208171e-11,
+      "loss": 0.5038,
+      "step": 7210
+    },
+    {
+      "epoch": 1.9997226844148641,
+      "grad_norm": 0.2026882767677307,
+      "learning_rate": 4.878162124244412e-12,
+      "loss": 0.4975,
+      "step": 7211
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.23644179105758667,
+      "learning_rate": 1.2195405907355905e-12,
+      "loss": 0.5014,
+      "step": 7212
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.82587730884552,
+      "eval_runtime": 439.0989,
+      "eval_samples_per_second": 93.403,
+      "eval_steps_per_second": 1.46,
+      "step": 7212
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 7212,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.531044471940139e+20,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}