diff --git "a/checkpoint-16000/trainer_state.json" "b/checkpoint-16000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-16000/trainer_state.json"
@@ -0,0 +1,112033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.8379185572339325,
+  "eval_steps": 500,
+  "global_step": 16000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00011486990982712079,
+      "grad_norm": 1.0950555801391602,
+      "learning_rate": 0.0001,
+      "loss": 2.0384,
+      "step": 1
+    },
+    {
+      "epoch": 0.00022973981965424158,
+      "grad_norm": 0.6104637384414673,
+      "learning_rate": 0.0001,
+      "loss": 2.0026,
+      "step": 2
+    },
+    {
+      "epoch": 0.00034460972948136237,
+      "grad_norm": 1.6990306377410889,
+      "learning_rate": 0.0001,
+      "loss": 2.2147,
+      "step": 3
+    },
+    {
+      "epoch": 0.00045947963930848316,
+      "grad_norm": 0.5962863564491272,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 4
+    },
+    {
+      "epoch": 0.0005743495491356039,
+      "grad_norm": 0.5962501764297485,
+      "learning_rate": 0.0001,
+      "loss": 1.9908,
+      "step": 5
+    },
+    {
+      "epoch": 0.0006892194589627247,
+      "grad_norm": 0.46677908301353455,
+      "learning_rate": 0.0001,
+      "loss": 1.8843,
+      "step": 6
+    },
+    {
+      "epoch": 0.0008040893687898455,
+      "grad_norm": 0.5024643540382385,
+      "learning_rate": 0.0001,
+      "loss": 2.0395,
+      "step": 7
+    },
+    {
+      "epoch": 0.0009189592786169663,
+      "grad_norm": 0.3814510703086853,
+      "learning_rate": 0.0001,
+      "loss": 1.7637,
+      "step": 8
+    },
+    {
+      "epoch": 0.0010338291884440872,
+      "grad_norm": 0.4189926087856293,
+      "learning_rate": 0.0001,
+      "loss": 1.9592,
+      "step": 9
+    },
+    {
+      "epoch": 0.0011486990982712078,
+      "grad_norm": 0.44599318504333496,
+      "learning_rate": 0.0001,
+      "loss": 1.7664,
+      "step": 10
+    },
+    {
+      "epoch": 0.0012635690080983286,
+      "grad_norm": 0.3737109899520874,
+      "learning_rate": 0.0001,
+      "loss": 1.7794,
+      "step": 11
+    },
+    {
+      "epoch": 0.0013784389179254495,
+      "grad_norm": 0.39896202087402344,
+      "learning_rate": 0.0001,
+      "loss": 1.8159,
+      "step": 12
+    },
+    {
+      "epoch": 0.0014933088277525703,
+      "grad_norm": 0.3495546281337738,
+      "learning_rate": 0.0001,
+      "loss": 1.739,
+      "step": 13
+    },
+    {
+      "epoch": 0.001608178737579691,
+      "grad_norm": 0.4069090187549591,
+      "learning_rate": 0.0001,
+      "loss": 1.9259,
+      "step": 14
+    },
+    {
+      "epoch": 0.0017230486474068118,
+      "grad_norm": 0.4087446331977844,
+      "learning_rate": 0.0001,
+      "loss": 1.8746,
+      "step": 15
+    },
+    {
+      "epoch": 0.0018379185572339326,
+      "grad_norm": 0.33482518792152405,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 16
+    },
+    {
+      "epoch": 0.0019527884670610533,
+      "grad_norm": 0.37447136640548706,
+      "learning_rate": 0.0001,
+      "loss": 2.0974,
+      "step": 17
+    },
+    {
+      "epoch": 0.0020676583768881743,
+      "grad_norm": 0.359221875667572,
+      "learning_rate": 0.0001,
+      "loss": 1.7629,
+      "step": 18
+    },
+    {
+      "epoch": 0.0021825282867152947,
+      "grad_norm": 0.38664090633392334,
+      "learning_rate": 0.0001,
+      "loss": 1.9719,
+      "step": 19
+    },
+    {
+      "epoch": 0.0022973981965424156,
+      "grad_norm": 0.326967716217041,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 20
+    },
+    {
+      "epoch": 0.0024122681063695364,
+      "grad_norm": 0.3577702045440674,
+      "learning_rate": 0.0001,
+      "loss": 1.9068,
+      "step": 21
+    },
+    {
+      "epoch": 0.0025271380161966573,
+      "grad_norm": 0.3471507728099823,
+      "learning_rate": 0.0001,
+      "loss": 1.8111,
+      "step": 22
+    },
+    {
+      "epoch": 0.002642007926023778,
+      "grad_norm": 0.3563218414783478,
+      "learning_rate": 0.0001,
+      "loss": 1.862,
+      "step": 23
+    },
+    {
+      "epoch": 0.002756877835850899,
+      "grad_norm": 0.38289228081703186,
+      "learning_rate": 0.0001,
+      "loss": 1.9097,
+      "step": 24
+    },
+    {
+      "epoch": 0.0028717477456780198,
+      "grad_norm": 0.37357401847839355,
+      "learning_rate": 0.0001,
+      "loss": 1.9897,
+      "step": 25
+    },
+    {
+      "epoch": 0.0029866176555051406,
+      "grad_norm": 0.3501838445663452,
+      "learning_rate": 0.0001,
+      "loss": 1.9073,
+      "step": 26
+    },
+    {
+      "epoch": 0.003101487565332261,
+      "grad_norm": 0.3973883390426636,
+      "learning_rate": 0.0001,
+      "loss": 1.961,
+      "step": 27
+    },
+    {
+      "epoch": 0.003216357475159382,
+      "grad_norm": 0.44121477007865906,
+      "learning_rate": 0.0001,
+      "loss": 1.9053,
+      "step": 28
+    },
+    {
+      "epoch": 0.0033312273849865027,
+      "grad_norm": 0.35977938771247864,
+      "learning_rate": 0.0001,
+      "loss": 1.8474,
+      "step": 29
+    },
+    {
+      "epoch": 0.0034460972948136236,
+      "grad_norm": 0.3839881420135498,
+      "learning_rate": 0.0001,
+      "loss": 1.9948,
+      "step": 30
+    },
+    {
+      "epoch": 0.0035609672046407444,
+      "grad_norm": 0.3463224768638611,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 31
+    },
+    {
+      "epoch": 0.0036758371144678652,
+      "grad_norm": 0.3363831639289856,
+      "learning_rate": 0.0001,
+      "loss": 1.7479,
+      "step": 32
+    },
+    {
+      "epoch": 0.003790707024294986,
+      "grad_norm": 0.3735474944114685,
+      "learning_rate": 0.0001,
+      "loss": 1.7217,
+      "step": 33
+    },
+    {
+      "epoch": 0.0039055769341221065,
+      "grad_norm": 0.3522309958934784,
+      "learning_rate": 0.0001,
+      "loss": 1.9002,
+      "step": 34
+    },
+    {
+      "epoch": 0.004020446843949228,
+      "grad_norm": 0.35659080743789673,
+      "learning_rate": 0.0001,
+      "loss": 1.9188,
+      "step": 35
+    },
+    {
+      "epoch": 0.004135316753776349,
+      "grad_norm": 0.3676978647708893,
+      "learning_rate": 0.0001,
+      "loss": 2.0147,
+      "step": 36
+    },
+    {
+      "epoch": 0.0042501866636034695,
+      "grad_norm": 0.3312574326992035,
+      "learning_rate": 0.0001,
+      "loss": 1.6601,
+      "step": 37
+    },
+    {
+      "epoch": 0.0043650565734305894,
+      "grad_norm": 0.31456542015075684,
+      "learning_rate": 0.0001,
+      "loss": 1.7814,
+      "step": 38
+    },
+    {
+      "epoch": 0.00447992648325771,
+      "grad_norm": 0.34201157093048096,
+      "learning_rate": 0.0001,
+      "loss": 1.8276,
+      "step": 39
+    },
+    {
+      "epoch": 0.004594796393084831,
+      "grad_norm": 0.3848654329776764,
+      "learning_rate": 0.0001,
+      "loss": 1.9418,
+      "step": 40
+    },
+    {
+      "epoch": 0.004709666302911952,
+      "grad_norm": 0.333026260137558,
+      "learning_rate": 0.0001,
+      "loss": 1.8563,
+      "step": 41
+    },
+    {
+      "epoch": 0.004824536212739073,
+      "grad_norm": 0.3494780659675598,
+      "learning_rate": 0.0001,
+      "loss": 1.8389,
+      "step": 42
+    },
+    {
+      "epoch": 0.004939406122566194,
+      "grad_norm": 0.3298174738883972,
+      "learning_rate": 0.0001,
+      "loss": 1.8113,
+      "step": 43
+    },
+    {
+      "epoch": 0.0050542760323933145,
+      "grad_norm": 0.33735814690589905,
+      "learning_rate": 0.0001,
+      "loss": 1.7348,
+      "step": 44
+    },
+    {
+      "epoch": 0.005169145942220435,
+      "grad_norm": 0.3232858180999756,
+      "learning_rate": 0.0001,
+      "loss": 1.7709,
+      "step": 45
+    },
+    {
+      "epoch": 0.005284015852047556,
+      "grad_norm": 0.34412050247192383,
+      "learning_rate": 0.0001,
+      "loss": 1.8392,
+      "step": 46
+    },
+    {
+      "epoch": 0.005398885761874677,
+      "grad_norm": 0.36024415493011475,
+      "learning_rate": 0.0001,
+      "loss": 1.8496,
+      "step": 47
+    },
+    {
+      "epoch": 0.005513755671701798,
+      "grad_norm": 0.36573317646980286,
+      "learning_rate": 0.0001,
+      "loss": 2.0788,
+      "step": 48
+    },
+    {
+      "epoch": 0.005628625581528919,
+      "grad_norm": 0.36239179968833923,
+      "learning_rate": 0.0001,
+      "loss": 1.8717,
+      "step": 49
+    },
+    {
+      "epoch": 0.0057434954913560396,
+      "grad_norm": 0.35284945368766785,
+      "learning_rate": 0.0001,
+      "loss": 2.0221,
+      "step": 50
+    },
+    {
+      "epoch": 0.00585836540118316,
+      "grad_norm": 0.33570870757102966,
+      "learning_rate": 0.0001,
+      "loss": 1.9219,
+      "step": 51
+    },
+    {
+      "epoch": 0.005973235311010281,
+      "grad_norm": 0.36989328265190125,
+      "learning_rate": 0.0001,
+      "loss": 1.848,
+      "step": 52
+    },
+    {
+      "epoch": 0.006088105220837401,
+      "grad_norm": 0.3059888482093811,
+      "learning_rate": 0.0001,
+      "loss": 1.7479,
+      "step": 53
+    },
+    {
+      "epoch": 0.006202975130664522,
+      "grad_norm": 0.3483046293258667,
+      "learning_rate": 0.0001,
+      "loss": 1.9025,
+      "step": 54
+    },
+    {
+      "epoch": 0.006317845040491643,
+      "grad_norm": 0.33112019300460815,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 55
+    },
+    {
+      "epoch": 0.006432714950318764,
+      "grad_norm": 0.3339669406414032,
+      "learning_rate": 0.0001,
+      "loss": 1.6784,
+      "step": 56
+    },
+    {
+      "epoch": 0.006547584860145885,
+      "grad_norm": 0.3434010446071625,
+      "learning_rate": 0.0001,
+      "loss": 1.9298,
+      "step": 57
+    },
+    {
+      "epoch": 0.0066624547699730054,
+      "grad_norm": 0.3272426724433899,
+      "learning_rate": 0.0001,
+      "loss": 1.7955,
+      "step": 58
+    },
+    {
+      "epoch": 0.006777324679800126,
+      "grad_norm": 0.3431703746318817,
+      "learning_rate": 0.0001,
+      "loss": 1.835,
+      "step": 59
+    },
+    {
+      "epoch": 0.006892194589627247,
+      "grad_norm": 0.36123043298721313,
+      "learning_rate": 0.0001,
+      "loss": 1.8674,
+      "step": 60
+    },
+    {
+      "epoch": 0.007007064499454368,
+      "grad_norm": 0.3534761071205139,
+      "learning_rate": 0.0001,
+      "loss": 1.8935,
+      "step": 61
+    },
+    {
+      "epoch": 0.007121934409281489,
+      "grad_norm": 0.3585789203643799,
+      "learning_rate": 0.0001,
+      "loss": 1.8238,
+      "step": 62
+    },
+    {
+      "epoch": 0.00723680431910861,
+      "grad_norm": 0.3434857726097107,
+      "learning_rate": 0.0001,
+      "loss": 1.7009,
+      "step": 63
+    },
+    {
+      "epoch": 0.0073516742289357305,
+      "grad_norm": 0.3498717248439789,
+      "learning_rate": 0.0001,
+      "loss": 1.854,
+      "step": 64
+    },
+    {
+      "epoch": 0.007466544138762851,
+      "grad_norm": 0.37057748436927795,
+      "learning_rate": 0.0001,
+      "loss": 1.9863,
+      "step": 65
+    },
+    {
+      "epoch": 0.007581414048589972,
+      "grad_norm": 0.36728590726852417,
+      "learning_rate": 0.0001,
+      "loss": 1.8498,
+      "step": 66
+    },
+    {
+      "epoch": 0.007696283958417093,
+      "grad_norm": 0.34580233693122864,
+      "learning_rate": 0.0001,
+      "loss": 2.0121,
+      "step": 67
+    },
+    {
+      "epoch": 0.007811153868244213,
+      "grad_norm": 0.3278733491897583,
+      "learning_rate": 0.0001,
+      "loss": 1.8673,
+      "step": 68
+    },
+    {
+      "epoch": 0.007926023778071334,
+      "grad_norm": 0.33521461486816406,
+      "learning_rate": 0.0001,
+      "loss": 1.7693,
+      "step": 69
+    },
+    {
+      "epoch": 0.008040893687898456,
+      "grad_norm": 0.34356561303138733,
+      "learning_rate": 0.0001,
+      "loss": 1.9444,
+      "step": 70
+    },
+    {
+      "epoch": 0.008155763597725576,
+      "grad_norm": 0.38897210359573364,
+      "learning_rate": 0.0001,
+      "loss": 1.9214,
+      "step": 71
+    },
+    {
+      "epoch": 0.008270633507552697,
+      "grad_norm": 0.34927767515182495,
+      "learning_rate": 0.0001,
+      "loss": 1.8995,
+      "step": 72
+    },
+    {
+      "epoch": 0.008385503417379817,
+      "grad_norm": 0.3214438259601593,
+      "learning_rate": 0.0001,
+      "loss": 1.6437,
+      "step": 73
+    },
+    {
+      "epoch": 0.008500373327206939,
+      "grad_norm": 0.32360997796058655,
+      "learning_rate": 0.0001,
+      "loss": 1.7942,
+      "step": 74
+    },
+    {
+      "epoch": 0.008615243237034059,
+      "grad_norm": 0.34416717290878296,
+      "learning_rate": 0.0001,
+      "loss": 1.8455,
+      "step": 75
+    },
+    {
+      "epoch": 0.008730113146861179,
+      "grad_norm": 0.3457936644554138,
+      "learning_rate": 0.0001,
+      "loss": 1.9531,
+      "step": 76
+    },
+    {
+      "epoch": 0.0088449830566883,
+      "grad_norm": 0.3502795398235321,
+      "learning_rate": 0.0001,
+      "loss": 1.7687,
+      "step": 77
+    },
+    {
+      "epoch": 0.00895985296651542,
+      "grad_norm": 0.348290354013443,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 78
+    },
+    {
+      "epoch": 0.009074722876342542,
+      "grad_norm": 0.32894769310951233,
+      "learning_rate": 0.0001,
+      "loss": 1.7323,
+      "step": 79
+    },
+    {
+      "epoch": 0.009189592786169662,
+      "grad_norm": 0.32729727029800415,
+      "learning_rate": 0.0001,
+      "loss": 1.7761,
+      "step": 80
+    },
+    {
+      "epoch": 0.009304462695996784,
+      "grad_norm": 0.3347381055355072,
+      "learning_rate": 0.0001,
+      "loss": 1.9055,
+      "step": 81
+    },
+    {
+      "epoch": 0.009419332605823904,
+      "grad_norm": 0.35440585017204285,
+      "learning_rate": 0.0001,
+      "loss": 1.9095,
+      "step": 82
+    },
+    {
+      "epoch": 0.009534202515651026,
+      "grad_norm": 0.33549365401268005,
+      "learning_rate": 0.0001,
+      "loss": 1.7806,
+      "step": 83
+    },
+    {
+      "epoch": 0.009649072425478146,
+      "grad_norm": 0.37921905517578125,
+      "learning_rate": 0.0001,
+      "loss": 2.0155,
+      "step": 84
+    },
+    {
+      "epoch": 0.009763942335305267,
+      "grad_norm": 0.3329186737537384,
+      "learning_rate": 0.0001,
+      "loss": 1.8357,
+      "step": 85
+    },
+    {
+      "epoch": 0.009878812245132387,
+      "grad_norm": 0.36343327164649963,
+      "learning_rate": 0.0001,
+      "loss": 1.9286,
+      "step": 86
+    },
+    {
+      "epoch": 0.009993682154959509,
+      "grad_norm": 0.3510076403617859,
+      "learning_rate": 0.0001,
+      "loss": 1.8747,
+      "step": 87
+    },
+    {
+      "epoch": 0.010108552064786629,
+      "grad_norm": 0.32831454277038574,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 88
+    },
+    {
+      "epoch": 0.01022342197461375,
+      "grad_norm": 0.3502967655658722,
+      "learning_rate": 0.0001,
+      "loss": 1.9506,
+      "step": 89
+    },
+    {
+      "epoch": 0.01033829188444087,
+      "grad_norm": 0.34648793935775757,
+      "learning_rate": 0.0001,
+      "loss": 1.8406,
+      "step": 90
+    },
+    {
+      "epoch": 0.01045316179426799,
+      "grad_norm": 0.35201138257980347,
+      "learning_rate": 0.0001,
+      "loss": 1.8329,
+      "step": 91
+    },
+    {
+      "epoch": 0.010568031704095112,
+      "grad_norm": 0.3722350299358368,
+      "learning_rate": 0.0001,
+      "loss": 1.9775,
+      "step": 92
+    },
+    {
+      "epoch": 0.010682901613922232,
+      "grad_norm": 0.3540225327014923,
+      "learning_rate": 0.0001,
+      "loss": 1.8165,
+      "step": 93
+    },
+    {
+      "epoch": 0.010797771523749354,
+      "grad_norm": 0.33188265562057495,
+      "learning_rate": 0.0001,
+      "loss": 1.9455,
+      "step": 94
+    },
+    {
+      "epoch": 0.010912641433576474,
+      "grad_norm": 0.32253599166870117,
+      "learning_rate": 0.0001,
+      "loss": 1.7694,
+      "step": 95
+    },
+    {
+      "epoch": 0.011027511343403596,
+      "grad_norm": 0.34879428148269653,
+      "learning_rate": 0.0001,
+      "loss": 1.9055,
+      "step": 96
+    },
+    {
+      "epoch": 0.011142381253230716,
+      "grad_norm": 0.3185734450817108,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 97
+    },
+    {
+      "epoch": 0.011257251163057837,
+      "grad_norm": 0.34270620346069336,
+      "learning_rate": 0.0001,
+      "loss": 1.8956,
+      "step": 98
+    },
+    {
+      "epoch": 0.011372121072884957,
+      "grad_norm": 0.3397463858127594,
+      "learning_rate": 0.0001,
+      "loss": 1.8395,
+      "step": 99
+    },
+    {
+      "epoch": 0.011486990982712079,
+      "grad_norm": 0.587956428527832,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 100
+    },
+    {
+      "epoch": 0.011601860892539199,
+      "grad_norm": 0.34749558568000793,
+      "learning_rate": 0.0001,
+      "loss": 1.9738,
+      "step": 101
+    },
+    {
+      "epoch": 0.01171673080236632,
+      "grad_norm": 0.31300491094589233,
+      "learning_rate": 0.0001,
+      "loss": 1.8049,
+      "step": 102
+    },
+    {
+      "epoch": 0.01183160071219344,
+      "grad_norm": 0.3428112864494324,
+      "learning_rate": 0.0001,
+      "loss": 1.8066,
+      "step": 103
+    },
+    {
+      "epoch": 0.011946470622020562,
+      "grad_norm": 0.3374922573566437,
+      "learning_rate": 0.0001,
+      "loss": 1.8526,
+      "step": 104
+    },
+    {
+      "epoch": 0.012061340531847682,
+      "grad_norm": 0.32223325967788696,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 105
+    },
+    {
+      "epoch": 0.012176210441674802,
+      "grad_norm": 0.3365132808685303,
+      "learning_rate": 0.0001,
+      "loss": 1.7933,
+      "step": 106
+    },
+    {
+      "epoch": 0.012291080351501924,
+      "grad_norm": 0.31686636805534363,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 107
+    },
+    {
+      "epoch": 0.012405950261329044,
+      "grad_norm": 0.3461359441280365,
+      "learning_rate": 0.0001,
+      "loss": 1.563,
+      "step": 108
+    },
+    {
+      "epoch": 0.012520820171156166,
+      "grad_norm": 0.31762605905532837,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 109
+    },
+    {
+      "epoch": 0.012635690080983286,
+      "grad_norm": 0.3150866627693176,
+      "learning_rate": 0.0001,
+      "loss": 1.6591,
+      "step": 110
+    },
+    {
+      "epoch": 0.012750559990810408,
+      "grad_norm": 0.3498469889163971,
+      "learning_rate": 0.0001,
+      "loss": 1.8973,
+      "step": 111
+    },
+    {
+      "epoch": 0.012865429900637528,
+      "grad_norm": 0.3512935936450958,
+      "learning_rate": 0.0001,
+      "loss": 1.9445,
+      "step": 112
+    },
+    {
+      "epoch": 0.01298029981046465,
+      "grad_norm": 0.3678135573863983,
+      "learning_rate": 0.0001,
+      "loss": 1.7957,
+      "step": 113
+    },
+    {
+      "epoch": 0.01309516972029177,
+      "grad_norm": 0.3330201208591461,
+      "learning_rate": 0.0001,
+      "loss": 1.7507,
+      "step": 114
+    },
+    {
+      "epoch": 0.013210039630118891,
+      "grad_norm": 0.33755484223365784,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 115
+    },
+    {
+      "epoch": 0.013324909539946011,
+      "grad_norm": 0.34242576360702515,
+      "learning_rate": 0.0001,
+      "loss": 1.8006,
+      "step": 116
+    },
+    {
+      "epoch": 0.013439779449773133,
+      "grad_norm": 0.35527321696281433,
+      "learning_rate": 0.0001,
+      "loss": 1.8855,
+      "step": 117
+    },
+    {
+      "epoch": 0.013554649359600253,
+      "grad_norm": 0.3338663578033447,
+      "learning_rate": 0.0001,
+      "loss": 1.7004,
+      "step": 118
+    },
+    {
+      "epoch": 0.013669519269427374,
+      "grad_norm": 0.3488336503505707,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 119
+    },
+    {
+      "epoch": 0.013784389179254494,
+      "grad_norm": 0.3300139605998993,
+      "learning_rate": 0.0001,
+      "loss": 1.7849,
+      "step": 120
+    },
+    {
+      "epoch": 0.013899259089081614,
+      "grad_norm": 0.29988691210746765,
+      "learning_rate": 0.0001,
+      "loss": 1.476,
+      "step": 121
+    },
+    {
+      "epoch": 0.014014128998908736,
+      "grad_norm": 0.3436170220375061,
+      "learning_rate": 0.0001,
+      "loss": 1.7939,
+      "step": 122
+    },
+    {
+      "epoch": 0.014128998908735856,
+      "grad_norm": 0.3263620436191559,
+      "learning_rate": 0.0001,
+      "loss": 1.7402,
+      "step": 123
+    },
+    {
+      "epoch": 0.014243868818562978,
+      "grad_norm": 0.3402661383152008,
+      "learning_rate": 0.0001,
+      "loss": 1.7146,
+      "step": 124
+    },
+    {
+      "epoch": 0.014358738728390098,
+      "grad_norm": 0.3439052700996399,
+      "learning_rate": 0.0001,
+      "loss": 1.7521,
+      "step": 125
+    },
+    {
+      "epoch": 0.01447360863821722,
+      "grad_norm": 0.3327753245830536,
+      "learning_rate": 0.0001,
+      "loss": 1.8139,
+      "step": 126
+    },
+    {
+      "epoch": 0.01458847854804434,
+      "grad_norm": 0.3490423560142517,
+      "learning_rate": 0.0001,
+      "loss": 1.9259,
+      "step": 127
+    },
+    {
+      "epoch": 0.014703348457871461,
+      "grad_norm": 0.33834782242774963,
+      "learning_rate": 0.0001,
+      "loss": 1.8205,
+      "step": 128
+    },
+    {
+      "epoch": 0.014818218367698581,
+      "grad_norm": 0.34202057123184204,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 129
+    },
+    {
+      "epoch": 0.014933088277525703,
+      "grad_norm": 0.31165608763694763,
+      "learning_rate": 0.0001,
+      "loss": 1.7411,
+      "step": 130
+    },
+    {
+      "epoch": 0.015047958187352823,
+      "grad_norm": 0.3511159420013428,
+      "learning_rate": 0.0001,
+      "loss": 1.8655,
+      "step": 131
+    },
+    {
+      "epoch": 0.015162828097179944,
+      "grad_norm": 0.3437829613685608,
+      "learning_rate": 0.0001,
+      "loss": 1.9517,
+      "step": 132
+    },
+    {
+      "epoch": 0.015277698007007064,
+      "grad_norm": 0.3470558822154999,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 133
+    },
+    {
+      "epoch": 0.015392567916834186,
+      "grad_norm": 0.3070822060108185,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 134
+    },
+    {
+      "epoch": 0.015507437826661306,
+      "grad_norm": 0.3400954306125641,
+      "learning_rate": 0.0001,
+      "loss": 1.8433,
+      "step": 135
+    },
+    {
+      "epoch": 0.015622307736488426,
+      "grad_norm": 0.37735289335250854,
+      "learning_rate": 0.0001,
+      "loss": 1.742,
+      "step": 136
+    },
+    {
+      "epoch": 0.015737177646315548,
+      "grad_norm": 0.33449429273605347,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 137
+    },
+    {
+      "epoch": 0.015852047556142668,
+      "grad_norm": 0.3531495928764343,
+      "learning_rate": 0.0001,
+      "loss": 1.9361,
+      "step": 138
+    },
+    {
+      "epoch": 0.015966917465969788,
+      "grad_norm": 0.3543091118335724,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 139
+    },
+    {
+      "epoch": 0.01608178737579691,
+      "grad_norm": 0.3521219789981842,
+      "learning_rate": 0.0001,
+      "loss": 1.7885,
+      "step": 140
+    },
+    {
+      "epoch": 0.01619665728562403,
+      "grad_norm": 0.3496558368206024,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 141
+    },
+    {
+      "epoch": 0.01631152719545115,
+      "grad_norm": 0.35304608941078186,
+      "learning_rate": 0.0001,
+      "loss": 1.774,
+      "step": 142
+    },
+    {
+      "epoch": 0.01642639710527827,
+      "grad_norm": 0.3590501546859741,
+      "learning_rate": 0.0001,
+      "loss": 1.8743,
+      "step": 143
+    },
+    {
+      "epoch": 0.016541267015105394,
+      "grad_norm": 0.30004453659057617,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 144
+    },
+    {
+      "epoch": 0.016656136924932514,
+      "grad_norm": 0.3431392014026642,
+      "learning_rate": 0.0001,
+      "loss": 1.7894,
+      "step": 145
+    },
+    {
+      "epoch": 0.016771006834759634,
+      "grad_norm": 0.38148149847984314,
+      "learning_rate": 0.0001,
+      "loss": 1.9799,
+      "step": 146
+    },
+    {
+      "epoch": 0.016885876744586754,
+      "grad_norm": 0.35888367891311646,
+      "learning_rate": 0.0001,
+      "loss": 2.0782,
+      "step": 147
+    },
+    {
+      "epoch": 0.017000746654413878,
+      "grad_norm": 0.3419305980205536,
+      "learning_rate": 0.0001,
+      "loss": 1.8875,
+      "step": 148
+    },
+    {
+      "epoch": 0.017115616564240998,
+      "grad_norm": 0.33760684728622437,
+      "learning_rate": 0.0001,
+      "loss": 1.7646,
+      "step": 149
+    },
+    {
+      "epoch": 0.017230486474068118,
+      "grad_norm": 0.37631165981292725,
+      "learning_rate": 0.0001,
+      "loss": 1.7915,
+      "step": 150
+    },
+    {
+      "epoch": 0.017345356383895238,
+      "grad_norm": 0.3630012273788452,
+      "learning_rate": 0.0001,
+      "loss": 1.9777,
+      "step": 151
+    },
+    {
+      "epoch": 0.017460226293722358,
+      "grad_norm": 0.3230808973312378,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 152
+    },
+    {
+      "epoch": 0.01757509620354948,
+      "grad_norm": 0.38511770963668823,
+      "learning_rate": 0.0001,
+      "loss": 1.9186,
+      "step": 153
+    },
+    {
+      "epoch": 0.0176899661133766,
+      "grad_norm": 0.32792919874191284,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 154
+    },
+    {
+      "epoch": 0.01780483602320372,
+      "grad_norm": 0.3158191442489624,
+      "learning_rate": 0.0001,
+      "loss": 1.6447,
+      "step": 155
+    },
+    {
+      "epoch": 0.01791970593303084,
+      "grad_norm": 0.36141133308410645,
+      "learning_rate": 0.0001,
+      "loss": 1.8623,
+      "step": 156
+    },
+    {
+      "epoch": 0.018034575842857965,
+      "grad_norm": 0.3470099866390228,
+      "learning_rate": 0.0001,
+      "loss": 1.7607,
+      "step": 157
+    },
+    {
+      "epoch": 0.018149445752685085,
+      "grad_norm": 0.31708627939224243,
+      "learning_rate": 0.0001,
+      "loss": 1.8103,
+      "step": 158
+    },
+    {
+      "epoch": 0.018264315662512205,
+      "grad_norm": 0.3248085677623749,
+      "learning_rate": 0.0001,
+      "loss": 1.8378,
+      "step": 159
+    },
+    {
+      "epoch": 0.018379185572339325,
+      "grad_norm": 0.32634925842285156,
+      "learning_rate": 0.0001,
+      "loss": 1.6819,
+      "step": 160
+    },
+    {
+      "epoch": 0.018494055482166448,
+      "grad_norm": 0.33397799730300903,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 161
+    },
+    {
+      "epoch": 0.018608925391993568,
+      "grad_norm": 0.3550772964954376,
+      "learning_rate": 0.0001,
+      "loss": 1.9381,
+      "step": 162
+    },
+    {
+      "epoch": 0.018723795301820688,
+      "grad_norm": 0.3159865438938141,
+      "learning_rate": 0.0001,
+      "loss": 1.6508,
+      "step": 163
+    },
+    {
+      "epoch": 0.018838665211647808,
+      "grad_norm": 0.3258025348186493,
+      "learning_rate": 0.0001,
+      "loss": 1.6776,
+      "step": 164
+    },
+    {
+      "epoch": 0.01895353512147493,
+      "grad_norm": 0.3488035798072815,
+      "learning_rate": 0.0001,
+      "loss": 2.0627,
+      "step": 165
+    },
+    {
+      "epoch": 0.01906840503130205,
+      "grad_norm": 0.3233996033668518,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 166
+    },
+    {
+      "epoch": 0.01918327494112917,
+      "grad_norm": 0.34753552079200745,
+      "learning_rate": 0.0001,
+      "loss": 1.9096,
+      "step": 167
+    },
+    {
+      "epoch": 0.01929814485095629,
+      "grad_norm": 0.3238770067691803,
+      "learning_rate": 0.0001,
+      "loss": 1.6883,
+      "step": 168
+    },
+    {
+      "epoch": 0.01941301476078341,
+      "grad_norm": 0.3520447611808777,
+      "learning_rate": 0.0001,
+      "loss": 1.8742,
+      "step": 169
+    },
+    {
+      "epoch": 0.019527884670610535,
+      "grad_norm": 0.35114407539367676,
+      "learning_rate": 0.0001,
+      "loss": 1.9098,
+      "step": 170
+    },
+    {
+      "epoch": 0.019642754580437655,
+      "grad_norm": 0.3472774922847748,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 171
+    },
+    {
+      "epoch": 0.019757624490264775,
+      "grad_norm": 0.33552825450897217,
+      "learning_rate": 0.0001,
+      "loss": 1.8656,
+      "step": 172
+    },
+    {
+      "epoch": 0.019872494400091895,
+      "grad_norm": 0.3363969027996063,
+      "learning_rate": 0.0001,
+      "loss": 1.8223,
+      "step": 173
+    },
+    {
+      "epoch": 0.019987364309919018,
+      "grad_norm": 0.3562895655632019,
+      "learning_rate": 0.0001,
+      "loss": 2.0164,
+      "step": 174
+    },
+    {
+      "epoch": 0.020102234219746138,
+      "grad_norm": 0.3177125155925751,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 175
+    },
+    {
+      "epoch": 0.020217104129573258,
+      "grad_norm": 0.33565396070480347,
+      "learning_rate": 0.0001,
+      "loss": 1.8483,
+      "step": 176
+    },
+    {
+      "epoch": 0.020331974039400378,
+      "grad_norm": 0.3205985128879547,
+      "learning_rate": 0.0001,
+      "loss": 1.9118,
+      "step": 177
+    },
+    {
+      "epoch": 0.0204468439492275,
+      "grad_norm": 0.32828885316848755,
+      "learning_rate": 0.0001,
+      "loss": 1.7084,
+      "step": 178
+    },
+    {
+      "epoch": 0.02056171385905462,
+      "grad_norm": 0.3119982182979584,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 179
+    },
+    {
+      "epoch": 0.02067658376888174,
+      "grad_norm": 0.34377679228782654,
+      "learning_rate": 0.0001,
+      "loss": 1.8481,
+      "step": 180
+    },
+    {
+      "epoch": 0.02079145367870886,
+      "grad_norm": 0.3235953152179718,
+      "learning_rate": 0.0001,
+      "loss": 1.8979,
+      "step": 181
+    },
+    {
+      "epoch": 0.02090632358853598,
+      "grad_norm": 0.35601556301116943,
+      "learning_rate": 0.0001,
+      "loss": 1.8469,
+      "step": 182
+    },
+    {
+      "epoch": 0.021021193498363105,
+      "grad_norm": 0.37358132004737854,
+      "learning_rate": 0.0001,
+      "loss": 2.0467,
+      "step": 183
+    },
+    {
+      "epoch": 0.021136063408190225,
+      "grad_norm": 0.31928551197052,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 184
+    },
+    {
+      "epoch": 0.021250933318017345,
+      "grad_norm": 0.33927685022354126,
+      "learning_rate": 0.0001,
+      "loss": 1.8849,
+      "step": 185
+    },
+    {
+      "epoch": 0.021365803227844465,
+      "grad_norm": 0.3599051237106323,
+      "learning_rate": 0.0001,
+      "loss": 1.9486,
+      "step": 186
+    },
+    {
+      "epoch": 0.021480673137671588,
+      "grad_norm": 0.32173585891723633,
+      "learning_rate": 0.0001,
+      "loss": 1.7674,
+      "step": 187
+    },
+    {
+      "epoch": 0.021595543047498708,
+      "grad_norm": 0.3629266023635864,
+      "learning_rate": 0.0001,
+      "loss": 1.7596,
+      "step": 188
+    },
+    {
+      "epoch": 0.021710412957325828,
+      "grad_norm": 0.31827229261398315,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 189
+    },
+    {
+      "epoch": 0.021825282867152948,
+      "grad_norm": 0.3418801426887512,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 190
+    },
+    {
+      "epoch": 0.02194015277698007,
+      "grad_norm": 0.3373776376247406,
+      "learning_rate": 0.0001,
+      "loss": 1.7593,
+      "step": 191
+    },
+    {
+      "epoch": 0.02205502268680719,
+      "grad_norm": 0.32532092928886414,
+      "learning_rate": 0.0001,
+      "loss": 1.8815,
+      "step": 192
+    },
+    {
+      "epoch": 0.02216989259663431,
+      "grad_norm": 0.3122539520263672,
+      "learning_rate": 0.0001,
+      "loss": 1.7756,
+      "step": 193
+    },
+    {
+      "epoch": 0.02228476250646143,
+      "grad_norm": 0.3185436725616455,
+      "learning_rate": 0.0001,
+      "loss": 1.6333,
+      "step": 194
+    },
+    {
+      "epoch": 0.022399632416288555,
+      "grad_norm": 0.3428870439529419,
+      "learning_rate": 0.0001,
+      "loss": 1.7126,
+      "step": 195
+    },
+    {
+      "epoch": 0.022514502326115675,
+      "grad_norm": 0.345682293176651,
+      "learning_rate": 0.0001,
+      "loss": 1.7324,
+      "step": 196
+    },
+    {
+      "epoch": 0.022629372235942795,
+      "grad_norm": 0.33158332109451294,
+      "learning_rate": 0.0001,
+      "loss": 1.78,
+      "step": 197
+    },
+    {
+      "epoch": 0.022744242145769915,
+      "grad_norm": 0.3336940109729767,
+      "learning_rate": 0.0001,
+      "loss": 1.7551,
+      "step": 198
+    },
+    {
+      "epoch": 0.022859112055597035,
+      "grad_norm": 0.3249858319759369,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 199
+    },
+    {
+      "epoch": 0.022973981965424158,
+      "grad_norm": 0.3251815736293793,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 200
+    },
+    {
+      "epoch": 0.023088851875251278,
+      "grad_norm": 0.3444681763648987,
+      "learning_rate": 0.0001,
+      "loss": 1.6374,
+      "step": 201
+    },
+    {
+      "epoch": 0.023203721785078398,
+      "grad_norm": 0.33435139060020447,
+      "learning_rate": 0.0001,
+      "loss": 1.7646,
+      "step": 202
+    },
+    {
+      "epoch": 0.023318591694905518,
+      "grad_norm": 0.33999738097190857,
+      "learning_rate": 0.0001,
+      "loss": 1.8351,
+      "step": 203
+    },
+    {
+      "epoch": 0.02343346160473264,
+      "grad_norm": 0.33522528409957886,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 204
+    },
+    {
+      "epoch": 0.02354833151455976,
+      "grad_norm": 0.343474805355072,
+      "learning_rate": 0.0001,
+      "loss": 1.7712,
+      "step": 205
+    },
+    {
+      "epoch": 0.02366320142438688,
+      "grad_norm": 0.35150644183158875,
+      "learning_rate": 0.0001,
+      "loss": 1.9298,
+      "step": 206
+    },
+    {
+      "epoch": 0.023778071334214,
+      "grad_norm": 0.3332938253879547,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 207
+    },
+    {
+      "epoch": 0.023892941244041125,
+      "grad_norm": 0.34478330612182617,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 208
+    },
+    {
+      "epoch": 0.024007811153868245,
+      "grad_norm": 0.36427655816078186,
+      "learning_rate": 0.0001,
+      "loss": 1.7768,
+      "step": 209
+    },
+    {
+      "epoch": 0.024122681063695365,
+      "grad_norm": 0.33923599123954773,
+      "learning_rate": 0.0001,
+      "loss": 1.8095,
+      "step": 210
+    },
+    {
+      "epoch": 0.024237550973522485,
+      "grad_norm": 0.342052161693573,
+      "learning_rate": 0.0001,
+      "loss": 1.6883,
+      "step": 211
+    },
+    {
+      "epoch": 0.024352420883349605,
+      "grad_norm": 0.3708147704601288,
+      "learning_rate": 0.0001,
+      "loss": 1.9331,
+      "step": 212
+    },
+    {
+      "epoch": 0.02446729079317673,
+      "grad_norm": 0.3495383858680725,
+      "learning_rate": 0.0001,
+      "loss": 1.8306,
+      "step": 213
+    },
+    {
+      "epoch": 0.02458216070300385,
+      "grad_norm": 0.34303730726242065,
+      "learning_rate": 0.0001,
+      "loss": 1.848,
+      "step": 214
+    },
+    {
+      "epoch": 0.02469703061283097,
+      "grad_norm": 0.29449400305747986,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 215
+    },
+    {
+      "epoch": 0.024811900522658088,
+      "grad_norm": 0.31254515051841736,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 216
+    },
+    {
+      "epoch": 0.02492677043248521,
+      "grad_norm": 0.3257744014263153,
+      "learning_rate": 0.0001,
+      "loss": 1.5715,
+      "step": 217
+    },
+    {
+      "epoch": 0.02504164034231233,
+      "grad_norm": 0.37795010209083557,
+      "learning_rate": 0.0001,
+      "loss": 1.8833,
+      "step": 218
+    },
+    {
+      "epoch": 0.02515651025213945,
+      "grad_norm": 0.3558560311794281,
+      "learning_rate": 0.0001,
+      "loss": 1.7896,
+      "step": 219
+    },
+    {
+      "epoch": 0.02527138016196657,
+      "grad_norm": 0.3125206530094147,
+      "learning_rate": 0.0001,
+      "loss": 1.7799,
+      "step": 220
+    },
+    {
+      "epoch": 0.025386250071793695,
+      "grad_norm": 0.37031280994415283,
+      "learning_rate": 0.0001,
+      "loss": 1.7588,
+      "step": 221
+    },
+    {
+      "epoch": 0.025501119981620815,
+      "grad_norm": 0.32977885007858276,
+      "learning_rate": 0.0001,
+      "loss": 1.8772,
+      "step": 222
+    },
+    {
+      "epoch": 0.025615989891447935,
+      "grad_norm": 0.3385184407234192,
+      "learning_rate": 0.0001,
+      "loss": 1.7859,
+      "step": 223
+    },
+    {
+      "epoch": 0.025730859801275055,
+      "grad_norm": 0.3234337866306305,
+      "learning_rate": 0.0001,
+      "loss": 1.8324,
+      "step": 224
+    },
+    {
+      "epoch": 0.02584572971110218,
+      "grad_norm": 0.33667194843292236,
+      "learning_rate": 0.0001,
+      "loss": 1.8861,
+      "step": 225
+    },
+    {
+      "epoch": 0.0259605996209293,
+      "grad_norm": 0.3473112881183624,
+      "learning_rate": 0.0001,
+      "loss": 1.8953,
+      "step": 226
+    },
+    {
+      "epoch": 0.02607546953075642,
+      "grad_norm": 0.36767181754112244,
+      "learning_rate": 0.0001,
+      "loss": 1.8615,
+      "step": 227
+    },
+    {
+      "epoch": 0.02619033944058354,
+      "grad_norm": 0.2963101267814636,
+      "learning_rate": 0.0001,
+      "loss": 1.6676,
+      "step": 228
+    },
+    {
+      "epoch": 0.02630520935041066,
+      "grad_norm": 0.35011404752731323,
+      "learning_rate": 0.0001,
+      "loss": 1.9064,
+      "step": 229
+    },
+    {
+      "epoch": 0.026420079260237782,
+      "grad_norm": 0.3166825771331787,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 230
+    },
+    {
+      "epoch": 0.026534949170064902,
+      "grad_norm": 0.33136868476867676,
+      "learning_rate": 0.0001,
+      "loss": 1.9791,
+      "step": 231
+    },
+    {
+      "epoch": 0.026649819079892022,
+      "grad_norm": 0.3448893129825592,
+      "learning_rate": 0.0001,
+      "loss": 1.9482,
+      "step": 232
+    },
+    {
+      "epoch": 0.02676468898971914,
+      "grad_norm": 0.33245259523391724,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 233
+    },
+    {
+      "epoch": 0.026879558899546265,
+      "grad_norm": 0.29518410563468933,
+      "learning_rate": 0.0001,
+      "loss": 1.5645,
+      "step": 234
+    },
+    {
+      "epoch": 0.026994428809373385,
+      "grad_norm": 0.33720827102661133,
+      "learning_rate": 0.0001,
+      "loss": 1.7339,
+      "step": 235
+    },
+    {
+      "epoch": 0.027109298719200505,
+      "grad_norm": 0.3211282789707184,
+      "learning_rate": 0.0001,
+      "loss": 1.7116,
+      "step": 236
+    },
+    {
+      "epoch": 0.027224168629027625,
+      "grad_norm": 0.3239465355873108,
+      "learning_rate": 0.0001,
+      "loss": 1.7349,
+      "step": 237
+    },
+    {
+      "epoch": 0.02733903853885475,
+      "grad_norm": 0.32968419790267944,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 238
+    },
+    {
+      "epoch": 0.02745390844868187,
+      "grad_norm": 0.3076919615268707,
+      "learning_rate": 0.0001,
+      "loss": 1.7016,
+      "step": 239
+    },
+    {
+      "epoch": 0.02756877835850899,
+      "grad_norm": 0.3537105619907379,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 240
+    },
+    {
+      "epoch": 0.02768364826833611,
+      "grad_norm": 0.3480488955974579,
+      "learning_rate": 0.0001,
+      "loss": 1.8703,
+      "step": 241
+    },
+    {
+      "epoch": 0.02779851817816323,
+      "grad_norm": 0.3412216901779175,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 242
+    },
+    {
+      "epoch": 0.027913388087990352,
+      "grad_norm": 0.3362070918083191,
+      "learning_rate": 0.0001,
+      "loss": 1.7042,
+      "step": 243
+    },
+    {
+      "epoch": 0.028028257997817472,
+      "grad_norm": 0.3288329839706421,
+      "learning_rate": 0.0001,
+      "loss": 1.7347,
+      "step": 244
+    },
+    {
+      "epoch": 0.028143127907644592,
+      "grad_norm": 0.33331453800201416,
+      "learning_rate": 0.0001,
+      "loss": 1.7895,
+      "step": 245
+    },
+    {
+      "epoch": 0.028257997817471712,
+      "grad_norm": 0.34005260467529297,
+      "learning_rate": 0.0001,
+      "loss": 1.8154,
+      "step": 246
+    },
+    {
+      "epoch": 0.028372867727298835,
+      "grad_norm": 0.33353686332702637,
+      "learning_rate": 0.0001,
+      "loss": 1.7256,
+      "step": 247
+    },
+    {
+      "epoch": 0.028487737637125955,
+      "grad_norm": 0.30739593505859375,
+      "learning_rate": 0.0001,
+      "loss": 1.779,
+      "step": 248
+    },
+    {
+      "epoch": 0.028602607546953075,
+      "grad_norm": 0.333033949136734,
+      "learning_rate": 0.0001,
+      "loss": 1.9319,
+      "step": 249
+    },
+    {
+      "epoch": 0.028717477456780195,
+      "grad_norm": 0.3508240878582001,
+      "learning_rate": 0.0001,
+      "loss": 1.993,
+      "step": 250
+    },
+    {
+      "epoch": 0.02883234736660732,
+      "grad_norm": 0.32476675510406494,
+      "learning_rate": 0.0001,
+      "loss": 1.8096,
+      "step": 251
+    },
+    {
+      "epoch": 0.02894721727643444,
+      "grad_norm": 0.31880107522010803,
+      "learning_rate": 0.0001,
+      "loss": 1.6591,
+      "step": 252
+    },
+    {
+      "epoch": 0.02906208718626156,
+      "grad_norm": 0.33040839433670044,
+      "learning_rate": 0.0001,
+      "loss": 1.7228,
+      "step": 253
+    },
+    {
+      "epoch": 0.02917695709608868,
+      "grad_norm": 0.3504570424556732,
+      "learning_rate": 0.0001,
+      "loss": 1.9012,
+      "step": 254
+    },
+    {
+      "epoch": 0.029291827005915802,
+      "grad_norm": 0.3304044008255005,
+      "learning_rate": 0.0001,
+      "loss": 1.8114,
+      "step": 255
+    },
+    {
+      "epoch": 0.029406696915742922,
+      "grad_norm": 0.3044085204601288,
+      "learning_rate": 0.0001,
+      "loss": 1.4707,
+      "step": 256
+    },
+    {
+      "epoch": 0.029521566825570042,
+      "grad_norm": 0.314557284116745,
+      "learning_rate": 0.0001,
+      "loss": 1.6985,
+      "step": 257
+    },
+    {
+      "epoch": 0.029636436735397162,
+      "grad_norm": 0.3400304317474365,
+      "learning_rate": 0.0001,
+      "loss": 1.7826,
+      "step": 258
+    },
+    {
+      "epoch": 0.029751306645224282,
+      "grad_norm": 0.3435501158237457,
+      "learning_rate": 0.0001,
+      "loss": 1.8779,
+      "step": 259
+    },
+    {
+      "epoch": 0.029866176555051405,
+      "grad_norm": 0.35256657004356384,
+      "learning_rate": 0.0001,
+      "loss": 1.9058,
+      "step": 260
+    },
+    {
+      "epoch": 0.029981046464878525,
+      "grad_norm": 0.32517293095588684,
+      "learning_rate": 0.0001,
+      "loss": 1.7505,
+      "step": 261
+    },
+    {
+      "epoch": 0.030095916374705645,
+      "grad_norm": 0.32231926918029785,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 262
+    },
+    {
+      "epoch": 0.030210786284532765,
+      "grad_norm": 0.312837153673172,
+      "learning_rate": 0.0001,
+      "loss": 1.7485,
+      "step": 263
+    },
+    {
+      "epoch": 0.03032565619435989,
+      "grad_norm": 0.33711639046669006,
+      "learning_rate": 0.0001,
+      "loss": 1.8586,
+      "step": 264
+    },
+    {
+      "epoch": 0.03044052610418701,
+      "grad_norm": 0.33530449867248535,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 265
+    },
+    {
+      "epoch": 0.03055539601401413,
+      "grad_norm": 0.3420933783054352,
+      "learning_rate": 0.0001,
+      "loss": 1.8716,
+      "step": 266
+    },
+    {
+      "epoch": 0.03067026592384125,
+      "grad_norm": 0.3286440372467041,
+      "learning_rate": 0.0001,
+      "loss": 1.6937,
+      "step": 267
+    },
+    {
+      "epoch": 0.030785135833668372,
+      "grad_norm": 0.331714391708374,
+      "learning_rate": 0.0001,
+      "loss": 1.8212,
+      "step": 268
+    },
+    {
+      "epoch": 0.030900005743495492,
+      "grad_norm": 0.32993122935295105,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 269
+    },
+    {
+      "epoch": 0.031014875653322612,
+      "grad_norm": 0.32213708758354187,
+      "learning_rate": 0.0001,
+      "loss": 1.8358,
+      "step": 270
+    },
+    {
+      "epoch": 0.031129745563149732,
+      "grad_norm": 0.33745911717414856,
+      "learning_rate": 0.0001,
+      "loss": 1.7545,
+      "step": 271
+    },
+    {
+      "epoch": 0.031244615472976852,
+      "grad_norm": 0.32847779989242554,
+      "learning_rate": 0.0001,
+      "loss": 1.7301,
+      "step": 272
+    },
+    {
+      "epoch": 0.031359485382803975,
+      "grad_norm": 0.3093877136707306,
+      "learning_rate": 0.0001,
+      "loss": 1.684,
+      "step": 273
+    },
+    {
+      "epoch": 0.031474355292631095,
+      "grad_norm": 0.32528170943260193,
+      "learning_rate": 0.0001,
+      "loss": 1.7503,
+      "step": 274
+    },
+    {
+      "epoch": 0.031589225202458215,
+      "grad_norm": 0.3486316204071045,
+      "learning_rate": 0.0001,
+      "loss": 1.8535,
+      "step": 275
+    },
+    {
+      "epoch": 0.031704095112285335,
+      "grad_norm": 0.3452807068824768,
+      "learning_rate": 0.0001,
+      "loss": 1.791,
+      "step": 276
+    },
+    {
+      "epoch": 0.031818965022112455,
+      "grad_norm": 0.3220309615135193,
+      "learning_rate": 0.0001,
+      "loss": 1.7711,
+      "step": 277
+    },
+    {
+      "epoch": 0.031933834931939575,
+      "grad_norm": 0.3608686327934265,
+      "learning_rate": 0.0001,
+      "loss": 1.853,
+      "step": 278
+    },
+    {
+      "epoch": 0.0320487048417667,
+      "grad_norm": 0.3356455862522125,
+      "learning_rate": 0.0001,
+      "loss": 1.8301,
+      "step": 279
+    },
+    {
+      "epoch": 0.03216357475159382,
+      "grad_norm": 0.36832770705223083,
+      "learning_rate": 0.0001,
+      "loss": 1.9413,
+      "step": 280
+    },
+    {
+      "epoch": 0.03227844466142094,
+      "grad_norm": 0.31815385818481445,
+      "learning_rate": 0.0001,
+      "loss": 1.7599,
+      "step": 281
+    },
+    {
+      "epoch": 0.03239331457124806,
+      "grad_norm": 0.3182365894317627,
+      "learning_rate": 0.0001,
+      "loss": 1.8303,
+      "step": 282
+    },
+    {
+      "epoch": 0.03250818448107518,
+      "grad_norm": 0.3294694721698761,
+      "learning_rate": 0.0001,
+      "loss": 1.7363,
+      "step": 283
+    },
+    {
+      "epoch": 0.0326230543909023,
+      "grad_norm": 0.3316207826137543,
+      "learning_rate": 0.0001,
+      "loss": 1.8191,
+      "step": 284
+    },
+    {
+      "epoch": 0.03273792430072942,
+      "grad_norm": 0.33212903141975403,
+      "learning_rate": 0.0001,
+      "loss": 1.8683,
+      "step": 285
+    },
+    {
+      "epoch": 0.03285279421055654,
+      "grad_norm": 0.3195394277572632,
+      "learning_rate": 0.0001,
+      "loss": 1.726,
+      "step": 286
+    },
+    {
+      "epoch": 0.03296766412038366,
+      "grad_norm": 0.30052995681762695,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 287
+    },
+    {
+      "epoch": 0.03308253403021079,
+      "grad_norm": 0.34062662720680237,
+      "learning_rate": 0.0001,
+      "loss": 1.9087,
+      "step": 288
+    },
+    {
+      "epoch": 0.03319740394003791,
+      "grad_norm": 0.34171825647354126,
+      "learning_rate": 0.0001,
+      "loss": 1.8431,
+      "step": 289
+    },
+    {
+      "epoch": 0.03331227384986503,
+      "grad_norm": 0.3276672065258026,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 290
+    },
+    {
+      "epoch": 0.03342714375969215,
+      "grad_norm": 0.36738014221191406,
+      "learning_rate": 0.0001,
+      "loss": 2.1369,
+      "step": 291
+    },
+    {
+      "epoch": 0.03354201366951927,
+      "grad_norm": 0.33889520168304443,
+      "learning_rate": 0.0001,
+      "loss": 1.6107,
+      "step": 292
+    },
+    {
+      "epoch": 0.03365688357934639,
+      "grad_norm": 0.33499544858932495,
+      "learning_rate": 0.0001,
+      "loss": 1.9176,
+      "step": 293
+    },
+    {
+      "epoch": 0.03377175348917351,
+      "grad_norm": 0.32459279894828796,
+      "learning_rate": 0.0001,
+      "loss": 1.7669,
+      "step": 294
+    },
+    {
+      "epoch": 0.03388662339900063,
+      "grad_norm": 0.3338513672351837,
+      "learning_rate": 0.0001,
+      "loss": 1.8755,
+      "step": 295
+    },
+    {
+      "epoch": 0.034001493308827756,
+      "grad_norm": 0.3502792418003082,
+      "learning_rate": 0.0001,
+      "loss": 1.7341,
+      "step": 296
+    },
+    {
+      "epoch": 0.034116363218654876,
+      "grad_norm": 0.34327709674835205,
+      "learning_rate": 0.0001,
+      "loss": 1.8455,
+      "step": 297
+    },
+    {
+      "epoch": 0.034231233128481996,
+      "grad_norm": 0.34909358620643616,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 298
+    },
+    {
+      "epoch": 0.034346103038309116,
+      "grad_norm": 0.3473874032497406,
+      "learning_rate": 0.0001,
+      "loss": 1.7983,
+      "step": 299
+    },
+    {
+      "epoch": 0.034460972948136236,
+      "grad_norm": 0.3394108712673187,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 300
+    },
+    {
+      "epoch": 0.034575842857963356,
+      "grad_norm": 0.34042516350746155,
+      "learning_rate": 0.0001,
+      "loss": 1.9465,
+      "step": 301
+    },
+    {
+      "epoch": 0.034690712767790476,
+      "grad_norm": 0.3370528221130371,
+      "learning_rate": 0.0001,
+      "loss": 1.7753,
+      "step": 302
+    },
+    {
+      "epoch": 0.034805582677617596,
+      "grad_norm": 0.32830432057380676,
+      "learning_rate": 0.0001,
+      "loss": 1.9652,
+      "step": 303
+    },
+    {
+      "epoch": 0.034920452587444716,
+      "grad_norm": 0.33252227306365967,
+      "learning_rate": 0.0001,
+      "loss": 1.8031,
+      "step": 304
+    },
+    {
+      "epoch": 0.03503532249727184,
+      "grad_norm": 0.32804980874061584,
+      "learning_rate": 0.0001,
+      "loss": 1.8358,
+      "step": 305
+    },
+    {
+      "epoch": 0.03515019240709896,
+      "grad_norm": 0.29810619354248047,
+      "learning_rate": 0.0001,
+      "loss": 1.6119,
+      "step": 306
+    },
+    {
+      "epoch": 0.03526506231692608,
+      "grad_norm": 0.3392501473426819,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 307
+    },
+    {
+      "epoch": 0.0353799322267532,
+      "grad_norm": 0.31986865401268005,
+      "learning_rate": 0.0001,
+      "loss": 1.7034,
+      "step": 308
+    },
+    {
+      "epoch": 0.03549480213658032,
+      "grad_norm": 0.33063703775405884,
+      "learning_rate": 0.0001,
+      "loss": 1.7634,
+      "step": 309
+    },
+    {
+      "epoch": 0.03560967204640744,
+      "grad_norm": 0.3045586049556732,
+      "learning_rate": 0.0001,
+      "loss": 1.6945,
+      "step": 310
+    },
+    {
+      "epoch": 0.03572454195623456,
+      "grad_norm": 0.3407602608203888,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 311
+    },
+    {
+      "epoch": 0.03583941186606168,
+      "grad_norm": 0.3090314567089081,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 312
+    },
+    {
+      "epoch": 0.03595428177588881,
+      "grad_norm": 0.3623706102371216,
+      "learning_rate": 0.0001,
+      "loss": 1.8877,
+      "step": 313
+    },
+    {
+      "epoch": 0.03606915168571593,
+      "grad_norm": 0.3450353145599365,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 314
+    },
+    {
+      "epoch": 0.03618402159554305,
+      "grad_norm": 0.3811194598674774,
+      "learning_rate": 0.0001,
+      "loss": 2.0295,
+      "step": 315
+    },
+    {
+      "epoch": 0.03629889150537017,
+      "grad_norm": 0.3349141776561737,
+      "learning_rate": 0.0001,
+      "loss": 1.8964,
+      "step": 316
+    },
+    {
+      "epoch": 0.03641376141519729,
+      "grad_norm": 0.3339914381504059,
+      "learning_rate": 0.0001,
+      "loss": 1.6882,
+      "step": 317
+    },
+    {
+      "epoch": 0.03652863132502441,
+      "grad_norm": 0.32706162333488464,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 318
+    },
+    {
+      "epoch": 0.03664350123485153,
+      "grad_norm": 0.35195019841194153,
+      "learning_rate": 0.0001,
+      "loss": 1.8871,
+      "step": 319
+    },
+    {
+      "epoch": 0.03675837114467865,
+      "grad_norm": 0.2981555163860321,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 320
+    },
+    {
+      "epoch": 0.03687324105450577,
+      "grad_norm": 0.3194637894630432,
+      "learning_rate": 0.0001,
+      "loss": 1.7356,
+      "step": 321
+    },
+    {
+      "epoch": 0.036988110964332896,
+      "grad_norm": 0.3623133599758148,
+      "learning_rate": 0.0001,
+      "loss": 1.8527,
+      "step": 322
+    },
+    {
+      "epoch": 0.037102980874160016,
+      "grad_norm": 0.3471209704875946,
+      "learning_rate": 0.0001,
+      "loss": 1.7521,
+      "step": 323
+    },
+    {
+      "epoch": 0.037217850783987136,
+      "grad_norm": 0.35352832078933716,
+      "learning_rate": 0.0001,
+      "loss": 1.7211,
+      "step": 324
+    },
+    {
+      "epoch": 0.037332720693814256,
+      "grad_norm": 0.32329848408699036,
+      "learning_rate": 0.0001,
+      "loss": 1.7678,
+      "step": 325
+    },
+    {
+      "epoch": 0.037447590603641376,
+      "grad_norm": 0.32614487409591675,
+      "learning_rate": 0.0001,
+      "loss": 1.7791,
+      "step": 326
+    },
+    {
+      "epoch": 0.037562460513468496,
+      "grad_norm": 0.3770069479942322,
+      "learning_rate": 0.0001,
+      "loss": 1.7185,
+      "step": 327
+    },
+    {
+      "epoch": 0.037677330423295616,
+      "grad_norm": 0.32778894901275635,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 328
+    },
+    {
+      "epoch": 0.037792200333122736,
+      "grad_norm": 0.37443724274635315,
+      "learning_rate": 0.0001,
+      "loss": 2.0193,
+      "step": 329
+    },
+    {
+      "epoch": 0.03790707024294986,
+      "grad_norm": 0.3752276599407196,
+      "learning_rate": 0.0001,
+      "loss": 1.9978,
+      "step": 330
+    },
+    {
+      "epoch": 0.03802194015277698,
+      "grad_norm": 0.33359140157699585,
+      "learning_rate": 0.0001,
+      "loss": 1.7836,
+      "step": 331
+    },
+    {
+      "epoch": 0.0381368100626041,
+      "grad_norm": 0.3710048794746399,
+      "learning_rate": 0.0001,
+      "loss": 1.9855,
+      "step": 332
+    },
+    {
+      "epoch": 0.03825167997243122,
+      "grad_norm": 0.3335336446762085,
+      "learning_rate": 0.0001,
+      "loss": 1.7433,
+      "step": 333
+    },
+    {
+      "epoch": 0.03836654988225834,
+      "grad_norm": 0.33409109711647034,
+      "learning_rate": 0.0001,
+      "loss": 1.7517,
+      "step": 334
+    },
+    {
+      "epoch": 0.03848141979208546,
+      "grad_norm": 0.3645201623439789,
+      "learning_rate": 0.0001,
+      "loss": 1.8764,
+      "step": 335
+    },
+    {
+      "epoch": 0.03859628970191258,
+      "grad_norm": 0.33294662833213806,
+      "learning_rate": 0.0001,
+      "loss": 1.867,
+      "step": 336
+    },
+    {
+      "epoch": 0.0387111596117397,
+      "grad_norm": 0.32880741357803345,
+      "learning_rate": 0.0001,
+      "loss": 1.8049,
+      "step": 337
+    },
+    {
+      "epoch": 0.03882602952156682,
+      "grad_norm": 0.353669136762619,
+      "learning_rate": 0.0001,
+      "loss": 1.7027,
+      "step": 338
+    },
+    {
+      "epoch": 0.03894089943139395,
+      "grad_norm": 0.3438865840435028,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 339
+    },
+    {
+      "epoch": 0.03905576934122107,
+      "grad_norm": 0.3039886951446533,
+      "learning_rate": 0.0001,
+      "loss": 1.7142,
+      "step": 340
+    },
+    {
+      "epoch": 0.03917063925104819,
+      "grad_norm": 0.34314149618148804,
+      "learning_rate": 0.0001,
+      "loss": 1.7391,
+      "step": 341
+    },
+    {
+      "epoch": 0.03928550916087531,
+      "grad_norm": 0.3602879047393799,
+      "learning_rate": 0.0001,
+      "loss": 1.8054,
+      "step": 342
+    },
+    {
+      "epoch": 0.03940037907070243,
+      "grad_norm": 0.33207011222839355,
+      "learning_rate": 0.0001,
+      "loss": 1.7549,
+      "step": 343
+    },
+    {
+      "epoch": 0.03951524898052955,
+      "grad_norm": 0.36213555932044983,
+      "learning_rate": 0.0001,
+      "loss": 1.8795,
+      "step": 344
+    },
+    {
+      "epoch": 0.03963011889035667,
+      "grad_norm": 0.37275567650794983,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 345
+    },
+    {
+      "epoch": 0.03974498880018379,
+      "grad_norm": 0.3421449661254883,
+      "learning_rate": 0.0001,
+      "loss": 1.7917,
+      "step": 346
+    },
+    {
+      "epoch": 0.03985985871001091,
+      "grad_norm": 0.33920395374298096,
+      "learning_rate": 0.0001,
+      "loss": 1.8476,
+      "step": 347
+    },
+    {
+      "epoch": 0.039974728619838036,
+      "grad_norm": 0.3364729881286621,
+      "learning_rate": 0.0001,
+      "loss": 1.8673,
+      "step": 348
+    },
+    {
+      "epoch": 0.040089598529665156,
+      "grad_norm": 0.3258533179759979,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 349
+    },
+    {
+      "epoch": 0.040204468439492276,
+      "grad_norm": 0.31070706248283386,
+      "learning_rate": 0.0001,
+      "loss": 1.482,
+      "step": 350
+    },
+    {
+      "epoch": 0.040319338349319396,
+      "grad_norm": 0.3752513527870178,
+      "learning_rate": 0.0001,
+      "loss": 1.9127,
+      "step": 351
+    },
+    {
+      "epoch": 0.040434208259146516,
+      "grad_norm": 0.34795695543289185,
+      "learning_rate": 0.0001,
+      "loss": 1.7637,
+      "step": 352
+    },
+    {
+      "epoch": 0.040549078168973636,
+      "grad_norm": 0.3222607374191284,
+      "learning_rate": 0.0001,
+      "loss": 1.8052,
+      "step": 353
+    },
+    {
+      "epoch": 0.040663948078800756,
+      "grad_norm": 0.3587421476840973,
+      "learning_rate": 0.0001,
+      "loss": 1.8354,
+      "step": 354
+    },
+    {
+      "epoch": 0.040778817988627876,
+      "grad_norm": 0.3381020724773407,
+      "learning_rate": 0.0001,
+      "loss": 1.9156,
+      "step": 355
+    },
+    {
+      "epoch": 0.040893687898455,
+      "grad_norm": 0.3503625690937042,
+      "learning_rate": 0.0001,
+      "loss": 1.8307,
+      "step": 356
+    },
+    {
+      "epoch": 0.04100855780828212,
+      "grad_norm": 0.34590432047843933,
+      "learning_rate": 0.0001,
+      "loss": 1.7548,
+      "step": 357
+    },
+    {
+      "epoch": 0.04112342771810924,
+      "grad_norm": 0.38469988107681274,
+      "learning_rate": 0.0001,
+      "loss": 1.8686,
+      "step": 358
+    },
+    {
+      "epoch": 0.04123829762793636,
+      "grad_norm": 0.34001877903938293,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 359
+    },
+    {
+      "epoch": 0.04135316753776348,
+      "grad_norm": 0.35807985067367554,
+      "learning_rate": 0.0001,
+      "loss": 1.8079,
+      "step": 360
+    },
+    {
+      "epoch": 0.0414680374475906,
+      "grad_norm": 0.3272436857223511,
+      "learning_rate": 0.0001,
+      "loss": 1.8085,
+      "step": 361
+    },
+    {
+      "epoch": 0.04158290735741772,
+      "grad_norm": 0.31507229804992676,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 362
+    },
+    {
+      "epoch": 0.04169777726724484,
+      "grad_norm": 0.3048648238182068,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 363
+    },
+    {
+      "epoch": 0.04181264717707196,
+      "grad_norm": 0.3657607436180115,
+      "learning_rate": 0.0001,
+      "loss": 1.7309,
+      "step": 364
+    },
+    {
+      "epoch": 0.04192751708689909,
+      "grad_norm": 0.3520914316177368,
+      "learning_rate": 0.0001,
+      "loss": 1.746,
+      "step": 365
+    },
+    {
+      "epoch": 0.04204238699672621,
+      "grad_norm": 0.3329162299633026,
+      "learning_rate": 0.0001,
+      "loss": 1.7818,
+      "step": 366
+    },
+    {
+      "epoch": 0.04215725690655333,
+      "grad_norm": 0.342009574174881,
+      "learning_rate": 0.0001,
+      "loss": 1.6875,
+      "step": 367
+    },
+    {
+      "epoch": 0.04227212681638045,
+      "grad_norm": 0.3568592667579651,
+      "learning_rate": 0.0001,
+      "loss": 1.8474,
+      "step": 368
+    },
+    {
+      "epoch": 0.04238699672620757,
+      "grad_norm": 0.32136744260787964,
+      "learning_rate": 0.0001,
+      "loss": 1.7908,
+      "step": 369
+    },
+    {
+      "epoch": 0.04250186663603469,
+      "grad_norm": 0.34790685772895813,
+      "learning_rate": 0.0001,
+      "loss": 1.8666,
+      "step": 370
+    },
+    {
+      "epoch": 0.04261673654586181,
+      "grad_norm": 0.3257257044315338,
+      "learning_rate": 0.0001,
+      "loss": 1.8399,
+      "step": 371
+    },
+    {
+      "epoch": 0.04273160645568893,
+      "grad_norm": 0.319909006357193,
+      "learning_rate": 0.0001,
+      "loss": 1.5395,
+      "step": 372
+    },
+    {
+      "epoch": 0.042846476365516056,
+      "grad_norm": 0.34518831968307495,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 373
+    },
+    {
+      "epoch": 0.042961346275343176,
+      "grad_norm": 0.350404292345047,
+      "learning_rate": 0.0001,
+      "loss": 1.7933,
+      "step": 374
+    },
+    {
+      "epoch": 0.043076216185170296,
+      "grad_norm": 0.3150002956390381,
+      "learning_rate": 0.0001,
+      "loss": 1.4682,
+      "step": 375
+    },
+    {
+      "epoch": 0.043191086094997416,
+      "grad_norm": 0.3526088297367096,
+      "learning_rate": 0.0001,
+      "loss": 1.7383,
+      "step": 376
+    },
+    {
+      "epoch": 0.043305956004824536,
+      "grad_norm": 0.3494514226913452,
+      "learning_rate": 0.0001,
+      "loss": 1.7906,
+      "step": 377
+    },
+    {
+      "epoch": 0.043420825914651656,
+      "grad_norm": 0.3323955833911896,
+      "learning_rate": 0.0001,
+      "loss": 1.8454,
+      "step": 378
+    },
+    {
+      "epoch": 0.043535695824478776,
+      "grad_norm": 0.31533968448638916,
+      "learning_rate": 0.0001,
+      "loss": 1.538,
+      "step": 379
+    },
+    {
+      "epoch": 0.043650565734305896,
+      "grad_norm": 0.3485229015350342,
+      "learning_rate": 0.0001,
+      "loss": 1.704,
+      "step": 380
+    },
+    {
+      "epoch": 0.043765435644133016,
+      "grad_norm": 0.3400436043739319,
+      "learning_rate": 0.0001,
+      "loss": 1.861,
+      "step": 381
+    },
+    {
+      "epoch": 0.04388030555396014,
+      "grad_norm": 0.32537248730659485,
+      "learning_rate": 0.0001,
+      "loss": 1.926,
+      "step": 382
+    },
+    {
+      "epoch": 0.04399517546378726,
+      "grad_norm": 0.35941800475120544,
+      "learning_rate": 0.0001,
+      "loss": 1.8136,
+      "step": 383
+    },
+    {
+      "epoch": 0.04411004537361438,
+      "grad_norm": 0.3239869773387909,
+      "learning_rate": 0.0001,
+      "loss": 1.5941,
+      "step": 384
+    },
+    {
+      "epoch": 0.0442249152834415,
+      "grad_norm": 0.34292441606521606,
+      "learning_rate": 0.0001,
+      "loss": 2.0001,
+      "step": 385
+    },
+    {
+      "epoch": 0.04433978519326862,
+      "grad_norm": 0.35059481859207153,
+      "learning_rate": 0.0001,
+      "loss": 1.8536,
+      "step": 386
+    },
+    {
+      "epoch": 0.04445465510309574,
+      "grad_norm": 0.35030195116996765,
+      "learning_rate": 0.0001,
+      "loss": 1.7471,
+      "step": 387
+    },
+    {
+      "epoch": 0.04456952501292286,
+      "grad_norm": 0.34561511874198914,
+      "learning_rate": 0.0001,
+      "loss": 1.6975,
+      "step": 388
+    },
+    {
+      "epoch": 0.04468439492274998,
+      "grad_norm": 0.3485510051250458,
+      "learning_rate": 0.0001,
+      "loss": 1.7986,
+      "step": 389
+    },
+    {
+      "epoch": 0.04479926483257711,
+      "grad_norm": 0.3528231978416443,
+      "learning_rate": 0.0001,
+      "loss": 1.7736,
+      "step": 390
+    },
+    {
+      "epoch": 0.04491413474240423,
+      "grad_norm": 0.3497932255268097,
+      "learning_rate": 0.0001,
+      "loss": 1.8213,
+      "step": 391
+    },
+    {
+      "epoch": 0.04502900465223135,
+      "grad_norm": 0.34640979766845703,
+      "learning_rate": 0.0001,
+      "loss": 2.0511,
+      "step": 392
+    },
+    {
+      "epoch": 0.04514387456205847,
+      "grad_norm": 0.3179776668548584,
+      "learning_rate": 0.0001,
+      "loss": 1.6705,
+      "step": 393
+    },
+    {
+      "epoch": 0.04525874447188559,
+      "grad_norm": 0.3492574691772461,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 394
+    },
+    {
+      "epoch": 0.04537361438171271,
+      "grad_norm": 0.32157397270202637,
+      "learning_rate": 0.0001,
+      "loss": 1.8084,
+      "step": 395
+    },
+    {
+      "epoch": 0.04548848429153983,
+      "grad_norm": 0.32758328318595886,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 396
+    },
+    {
+      "epoch": 0.04560335420136695,
+      "grad_norm": 0.34034964442253113,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 397
+    },
+    {
+      "epoch": 0.04571822411119407,
+      "grad_norm": 0.3422950506210327,
+      "learning_rate": 0.0001,
+      "loss": 1.9178,
+      "step": 398
+    },
+    {
+      "epoch": 0.045833094021021197,
+      "grad_norm": 0.3140711784362793,
+      "learning_rate": 0.0001,
+      "loss": 1.6569,
+      "step": 399
+    },
+    {
+      "epoch": 0.045947963930848316,
+      "grad_norm": 0.33444762229919434,
+      "learning_rate": 0.0001,
+      "loss": 1.7443,
+      "step": 400
+    },
+    {
+      "epoch": 0.046062833840675436,
+      "grad_norm": 0.3292236924171448,
+      "learning_rate": 0.0001,
+      "loss": 1.786,
+      "step": 401
+    },
+    {
+      "epoch": 0.046177703750502556,
+      "grad_norm": 0.3406231999397278,
+      "learning_rate": 0.0001,
+      "loss": 1.8477,
+      "step": 402
+    },
+    {
+      "epoch": 0.046292573660329676,
+      "grad_norm": 0.35850241780281067,
+      "learning_rate": 0.0001,
+      "loss": 1.9219,
+      "step": 403
+    },
+    {
+      "epoch": 0.046407443570156796,
+      "grad_norm": 0.39436420798301697,
+      "learning_rate": 0.0001,
+      "loss": 1.9564,
+      "step": 404
+    },
+    {
+      "epoch": 0.046522313479983916,
+      "grad_norm": 0.3540903627872467,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 405
+    },
+    {
+      "epoch": 0.046637183389811036,
+      "grad_norm": 0.4025273323059082,
+      "learning_rate": 0.0001,
+      "loss": 1.7531,
+      "step": 406
+    },
+    {
+      "epoch": 0.04675205329963816,
+      "grad_norm": 0.3019482493400574,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 407
+    },
+    {
+      "epoch": 0.04686692320946528,
+      "grad_norm": 0.3409338593482971,
+      "learning_rate": 0.0001,
+      "loss": 1.8024,
+      "step": 408
+    },
+    {
+      "epoch": 0.0469817931192924,
+      "grad_norm": 0.3111829161643982,
+      "learning_rate": 0.0001,
+      "loss": 1.486,
+      "step": 409
+    },
+    {
+      "epoch": 0.04709666302911952,
+      "grad_norm": 0.38814571499824524,
+      "learning_rate": 0.0001,
+      "loss": 2.0234,
+      "step": 410
+    },
+    {
+      "epoch": 0.04721153293894664,
+      "grad_norm": 0.3604361116886139,
+      "learning_rate": 0.0001,
+      "loss": 1.7389,
+      "step": 411
+    },
+    {
+      "epoch": 0.04732640284877376,
+      "grad_norm": 0.36267879605293274,
+      "learning_rate": 0.0001,
+      "loss": 1.9424,
+      "step": 412
+    },
+    {
+      "epoch": 0.04744127275860088,
+      "grad_norm": 0.3534083366394043,
+      "learning_rate": 0.0001,
+      "loss": 1.9132,
+      "step": 413
+    },
+    {
+      "epoch": 0.047556142668428,
+      "grad_norm": 0.31778329610824585,
+      "learning_rate": 0.0001,
+      "loss": 1.7196,
+      "step": 414
+    },
+    {
+      "epoch": 0.04767101257825512,
+      "grad_norm": 0.3307456970214844,
+      "learning_rate": 0.0001,
+      "loss": 1.7706,
+      "step": 415
+    },
+    {
+      "epoch": 0.04778588248808225,
+      "grad_norm": 0.33666694164276123,
+      "learning_rate": 0.0001,
+      "loss": 1.8634,
+      "step": 416
+    },
+    {
+      "epoch": 0.04790075239790937,
+      "grad_norm": 0.352847158908844,
+      "learning_rate": 0.0001,
+      "loss": 1.8828,
+      "step": 417
+    },
+    {
+      "epoch": 0.04801562230773649,
+      "grad_norm": 0.33423542976379395,
+      "learning_rate": 0.0001,
+      "loss": 1.9455,
+      "step": 418
+    },
+    {
+      "epoch": 0.04813049221756361,
+      "grad_norm": 0.35909709334373474,
+      "learning_rate": 0.0001,
+      "loss": 1.8293,
+      "step": 419
+    },
+    {
+      "epoch": 0.04824536212739073,
+      "grad_norm": 0.35591983795166016,
+      "learning_rate": 0.0001,
+      "loss": 1.8619,
+      "step": 420
+    },
+    {
+      "epoch": 0.04836023203721785,
+      "grad_norm": 0.33195993304252625,
+      "learning_rate": 0.0001,
+      "loss": 1.6672,
+      "step": 421
+    },
+    {
+      "epoch": 0.04847510194704497,
+      "grad_norm": 0.3257507383823395,
+      "learning_rate": 0.0001,
+      "loss": 1.7567,
+      "step": 422
+    },
+    {
+      "epoch": 0.04858997185687209,
+      "grad_norm": 0.31000784039497375,
+      "learning_rate": 0.0001,
+      "loss": 1.7718,
+      "step": 423
+    },
+    {
+      "epoch": 0.04870484176669921,
+      "grad_norm": 0.3311057686805725,
+      "learning_rate": 0.0001,
+      "loss": 1.8224,
+      "step": 424
+    },
+    {
+      "epoch": 0.04881971167652634,
+      "grad_norm": 0.32574066519737244,
+      "learning_rate": 0.0001,
+      "loss": 1.7199,
+      "step": 425
+    },
+    {
+      "epoch": 0.04893458158635346,
+      "grad_norm": 0.30064114928245544,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 426
+    },
+    {
+      "epoch": 0.04904945149618058,
+      "grad_norm": 0.331144779920578,
+      "learning_rate": 0.0001,
+      "loss": 1.8779,
+      "step": 427
+    },
+    {
+      "epoch": 0.0491643214060077,
+      "grad_norm": 0.3030913472175598,
+      "learning_rate": 0.0001,
+      "loss": 1.5864,
+      "step": 428
+    },
+    {
+      "epoch": 0.04927919131583482,
+      "grad_norm": 0.3529425263404846,
+      "learning_rate": 0.0001,
+      "loss": 1.743,
+      "step": 429
+    },
+    {
+      "epoch": 0.04939406122566194,
+      "grad_norm": 0.32856279611587524,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 430
+    },
+    {
+      "epoch": 0.04950893113548906,
+      "grad_norm": 0.32099583745002747,
+      "learning_rate": 0.0001,
+      "loss": 1.6706,
+      "step": 431
+    },
+    {
+      "epoch": 0.049623801045316177,
+      "grad_norm": 0.3427245616912842,
+      "learning_rate": 0.0001,
+      "loss": 1.8355,
+      "step": 432
+    },
+    {
+      "epoch": 0.0497386709551433,
+      "grad_norm": 0.34210020303726196,
+      "learning_rate": 0.0001,
+      "loss": 1.7778,
+      "step": 433
+    },
+    {
+      "epoch": 0.04985354086497042,
+      "grad_norm": 0.3216181695461273,
+      "learning_rate": 0.0001,
+      "loss": 1.8209,
+      "step": 434
+    },
+    {
+      "epoch": 0.04996841077479754,
+      "grad_norm": 0.33420485258102417,
+      "learning_rate": 0.0001,
+      "loss": 1.8087,
+      "step": 435
+    },
+    {
+      "epoch": 0.05008328068462466,
+      "grad_norm": 0.3496338725090027,
+      "learning_rate": 0.0001,
+      "loss": 1.737,
+      "step": 436
+    },
+    {
+      "epoch": 0.05019815059445178,
+      "grad_norm": 0.3511291444301605,
+      "learning_rate": 0.0001,
+      "loss": 2.0355,
+      "step": 437
+    },
+    {
+      "epoch": 0.0503130205042789,
+      "grad_norm": 0.34182974696159363,
+      "learning_rate": 0.0001,
+      "loss": 1.7236,
+      "step": 438
+    },
+    {
+      "epoch": 0.05042789041410602,
+      "grad_norm": 0.3205072581768036,
+      "learning_rate": 0.0001,
+      "loss": 1.7691,
+      "step": 439
+    },
+    {
+      "epoch": 0.05054276032393314,
+      "grad_norm": 0.3407754600048065,
+      "learning_rate": 0.0001,
+      "loss": 1.9003,
+      "step": 440
+    },
+    {
+      "epoch": 0.05065763023376026,
+      "grad_norm": 0.31289535760879517,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 441
+    },
+    {
+      "epoch": 0.05077250014358739,
+      "grad_norm": 0.2934773564338684,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 442
+    },
+    {
+      "epoch": 0.05088737005341451,
+      "grad_norm": 0.3349979817867279,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 443
+    },
+    {
+      "epoch": 0.05100223996324163,
+      "grad_norm": 0.3626347482204437,
+      "learning_rate": 0.0001,
+      "loss": 1.8299,
+      "step": 444
+    },
+    {
+      "epoch": 0.05111710987306875,
+      "grad_norm": 0.3510279059410095,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 445
+    },
+    {
+      "epoch": 0.05123197978289587,
+      "grad_norm": 0.3967926502227783,
+      "learning_rate": 0.0001,
+      "loss": 2.0913,
+      "step": 446
+    },
+    {
+      "epoch": 0.05134684969272299,
+      "grad_norm": 0.31338948011398315,
+      "learning_rate": 0.0001,
+      "loss": 1.5693,
+      "step": 447
+    },
+    {
+      "epoch": 0.05146171960255011,
+      "grad_norm": 0.32318705320358276,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 448
+    },
+    {
+      "epoch": 0.05157658951237723,
+      "grad_norm": 0.31790030002593994,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 449
+    },
+    {
+      "epoch": 0.05169145942220436,
+      "grad_norm": 0.3432879149913788,
+      "learning_rate": 0.0001,
+      "loss": 1.6502,
+      "step": 450
+    },
+    {
+      "epoch": 0.05180632933203148,
+      "grad_norm": 0.38444429636001587,
+      "learning_rate": 0.0001,
+      "loss": 1.844,
+      "step": 451
+    },
+    {
+      "epoch": 0.0519211992418586,
+      "grad_norm": 0.33826589584350586,
+      "learning_rate": 0.0001,
+      "loss": 1.7821,
+      "step": 452
+    },
+    {
+      "epoch": 0.05203606915168572,
+      "grad_norm": 0.340986967086792,
+      "learning_rate": 0.0001,
+      "loss": 1.7472,
+      "step": 453
+    },
+    {
+      "epoch": 0.05215093906151284,
+      "grad_norm": 0.34421804547309875,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 454
+    },
+    {
+      "epoch": 0.05226580897133996,
+      "grad_norm": 0.33938467502593994,
+      "learning_rate": 0.0001,
+      "loss": 1.8119,
+      "step": 455
+    },
+    {
+      "epoch": 0.05238067888116708,
+      "grad_norm": 0.3173467516899109,
+      "learning_rate": 0.0001,
+      "loss": 1.6541,
+      "step": 456
+    },
+    {
+      "epoch": 0.0524955487909942,
+      "grad_norm": 0.32488059997558594,
+      "learning_rate": 0.0001,
+      "loss": 1.5981,
+      "step": 457
+    },
+    {
+      "epoch": 0.05261041870082132,
+      "grad_norm": 0.35664987564086914,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 458
+    },
+    {
+      "epoch": 0.052725288610648444,
+      "grad_norm": 0.3289016783237457,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 459
+    },
+    {
+      "epoch": 0.052840158520475564,
+      "grad_norm": 0.3202899992465973,
+      "learning_rate": 0.0001,
+      "loss": 1.6696,
+      "step": 460
+    },
+    {
+      "epoch": 0.052955028430302684,
+      "grad_norm": 0.34299010038375854,
+      "learning_rate": 0.0001,
+      "loss": 1.8016,
+      "step": 461
+    },
+    {
+      "epoch": 0.053069898340129804,
+      "grad_norm": 0.34644559025764465,
+      "learning_rate": 0.0001,
+      "loss": 1.7947,
+      "step": 462
+    },
+    {
+      "epoch": 0.053184768249956924,
+      "grad_norm": 0.32393816113471985,
+      "learning_rate": 0.0001,
+      "loss": 1.7503,
+      "step": 463
+    },
+    {
+      "epoch": 0.053299638159784044,
+      "grad_norm": 0.36531350016593933,
+      "learning_rate": 0.0001,
+      "loss": 1.6858,
+      "step": 464
+    },
+    {
+      "epoch": 0.053414508069611163,
+      "grad_norm": 0.29397326707839966,
+      "learning_rate": 0.0001,
+      "loss": 1.6449,
+      "step": 465
+    },
+    {
+      "epoch": 0.05352937797943828,
+      "grad_norm": 0.3278639614582062,
+      "learning_rate": 0.0001,
+      "loss": 1.8286,
+      "step": 466
+    },
+    {
+      "epoch": 0.05364424788926541,
+      "grad_norm": 0.3598267734050751,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 467
+    },
+    {
+      "epoch": 0.05375911779909253,
+      "grad_norm": 0.36339178681373596,
+      "learning_rate": 0.0001,
+      "loss": 1.9749,
+      "step": 468
+    },
+    {
+      "epoch": 0.05387398770891965,
+      "grad_norm": 0.31908461451530457,
+      "learning_rate": 0.0001,
+      "loss": 1.6836,
+      "step": 469
+    },
+    {
+      "epoch": 0.05398885761874677,
+      "grad_norm": 0.3400243818759918,
+      "learning_rate": 0.0001,
+      "loss": 1.7625,
+      "step": 470
+    },
+    {
+      "epoch": 0.05410372752857389,
+      "grad_norm": 0.3101711869239807,
+      "learning_rate": 0.0001,
+      "loss": 1.5419,
+      "step": 471
+    },
+    {
+      "epoch": 0.05421859743840101,
+      "grad_norm": 0.33695968985557556,
+      "learning_rate": 0.0001,
+      "loss": 1.8077,
+      "step": 472
+    },
+    {
+      "epoch": 0.05433346734822813,
+      "grad_norm": 0.3501220941543579,
+      "learning_rate": 0.0001,
+      "loss": 1.8684,
+      "step": 473
+    },
+    {
+      "epoch": 0.05444833725805525,
+      "grad_norm": 0.3185228705406189,
+      "learning_rate": 0.0001,
+      "loss": 1.7575,
+      "step": 474
+    },
+    {
+      "epoch": 0.05456320716788237,
+      "grad_norm": 0.33037081360816956,
+      "learning_rate": 0.0001,
+      "loss": 1.8286,
+      "step": 475
+    },
+    {
+      "epoch": 0.0546780770777095,
+      "grad_norm": 0.3522806763648987,
+      "learning_rate": 0.0001,
+      "loss": 1.8658,
+      "step": 476
+    },
+    {
+      "epoch": 0.05479294698753662,
+      "grad_norm": 0.35064372420310974,
+      "learning_rate": 0.0001,
+      "loss": 1.7537,
+      "step": 477
+    },
+    {
+      "epoch": 0.05490781689736374,
+      "grad_norm": 0.3410029411315918,
+      "learning_rate": 0.0001,
+      "loss": 1.8334,
+      "step": 478
+    },
+    {
+      "epoch": 0.05502268680719086,
+      "grad_norm": 0.34402716159820557,
+      "learning_rate": 0.0001,
+      "loss": 1.8506,
+      "step": 479
+    },
+    {
+      "epoch": 0.05513755671701798,
+      "grad_norm": 0.34367635846138,
+      "learning_rate": 0.0001,
+      "loss": 1.832,
+      "step": 480
+    },
+    {
+      "epoch": 0.0552524266268451,
+      "grad_norm": 0.33249610662460327,
+      "learning_rate": 0.0001,
+      "loss": 1.7356,
+      "step": 481
+    },
+    {
+      "epoch": 0.05536729653667222,
+      "grad_norm": 0.33778467774391174,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 482
+    },
+    {
+      "epoch": 0.05548216644649934,
+      "grad_norm": 0.31293541193008423,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 483
+    },
+    {
+      "epoch": 0.05559703635632646,
+      "grad_norm": 0.3448321521282196,
+      "learning_rate": 0.0001,
+      "loss": 1.8017,
+      "step": 484
+    },
+    {
+      "epoch": 0.055711906266153584,
+      "grad_norm": 0.36519747972488403,
+      "learning_rate": 0.0001,
+      "loss": 2.0294,
+      "step": 485
+    },
+    {
+      "epoch": 0.055826776175980704,
+      "grad_norm": 0.36726245284080505,
+      "learning_rate": 0.0001,
+      "loss": 1.9747,
+      "step": 486
+    },
+    {
+      "epoch": 0.055941646085807824,
+      "grad_norm": 0.35360264778137207,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 487
+    },
+    {
+      "epoch": 0.056056515995634944,
+      "grad_norm": 0.31348568201065063,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 488
+    },
+    {
+      "epoch": 0.056171385905462064,
+      "grad_norm": 0.3481610119342804,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 489
+    },
+    {
+      "epoch": 0.056286255815289184,
+      "grad_norm": 0.30713188648223877,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 490
+    },
+    {
+      "epoch": 0.056401125725116304,
+      "grad_norm": 0.33684420585632324,
+      "learning_rate": 0.0001,
+      "loss": 1.7113,
+      "step": 491
+    },
+    {
+      "epoch": 0.056515995634943424,
+      "grad_norm": 0.36983436346054077,
+      "learning_rate": 0.0001,
+      "loss": 1.8795,
+      "step": 492
+    },
+    {
+      "epoch": 0.05663086554477055,
+      "grad_norm": 0.32992857694625854,
+      "learning_rate": 0.0001,
+      "loss": 1.8728,
+      "step": 493
+    },
+    {
+      "epoch": 0.05674573545459767,
+      "grad_norm": 0.3567913770675659,
+      "learning_rate": 0.0001,
+      "loss": 1.7194,
+      "step": 494
+    },
+    {
+      "epoch": 0.05686060536442479,
+      "grad_norm": 0.3176042437553406,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 495
+    },
+    {
+      "epoch": 0.05697547527425191,
+      "grad_norm": 0.30702850222587585,
+      "learning_rate": 0.0001,
+      "loss": 1.6685,
+      "step": 496
+    },
+    {
+      "epoch": 0.05709034518407903,
+      "grad_norm": 0.33594316244125366,
+      "learning_rate": 0.0001,
+      "loss": 1.7385,
+      "step": 497
+    },
+    {
+      "epoch": 0.05720521509390615,
+      "grad_norm": 0.3079899549484253,
+      "learning_rate": 0.0001,
+      "loss": 1.6913,
+      "step": 498
+    },
+    {
+      "epoch": 0.05732008500373327,
+      "grad_norm": 0.3289260268211365,
+      "learning_rate": 0.0001,
+      "loss": 1.7654,
+      "step": 499
+    },
+    {
+      "epoch": 0.05743495491356039,
+      "grad_norm": 0.33289793133735657,
+      "learning_rate": 0.0001,
+      "loss": 1.7681,
+      "step": 500
+    },
+    {
+      "epoch": 0.05754982482338751,
+      "grad_norm": 0.34192466735839844,
+      "learning_rate": 0.0001,
+      "loss": 1.753,
+      "step": 501
+    },
+    {
+      "epoch": 0.05766469473321464,
+      "grad_norm": 0.40127864480018616,
+      "learning_rate": 0.0001,
+      "loss": 2.114,
+      "step": 502
+    },
+    {
+      "epoch": 0.05777956464304176,
+      "grad_norm": 0.3597794473171234,
+      "learning_rate": 0.0001,
+      "loss": 1.8872,
+      "step": 503
+    },
+    {
+      "epoch": 0.05789443455286888,
+      "grad_norm": 0.3503691256046295,
+      "learning_rate": 0.0001,
+      "loss": 1.764,
+      "step": 504
+    },
+    {
+      "epoch": 0.058009304462696,
+      "grad_norm": 0.33984240889549255,
+      "learning_rate": 0.0001,
+      "loss": 1.6724,
+      "step": 505
+    },
+    {
+      "epoch": 0.05812417437252312,
+      "grad_norm": 0.3279857337474823,
+      "learning_rate": 0.0001,
+      "loss": 1.6906,
+      "step": 506
+    },
+    {
+      "epoch": 0.05823904428235024,
+      "grad_norm": 0.3238902986049652,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 507
+    },
+    {
+      "epoch": 0.05835391419217736,
+      "grad_norm": 0.35532209277153015,
+      "learning_rate": 0.0001,
+      "loss": 1.8577,
+      "step": 508
+    },
+    {
+      "epoch": 0.05846878410200448,
+      "grad_norm": 0.3488325774669647,
+      "learning_rate": 0.0001,
+      "loss": 1.7485,
+      "step": 509
+    },
+    {
+      "epoch": 0.058583654011831604,
+      "grad_norm": 0.32330816984176636,
+      "learning_rate": 0.0001,
+      "loss": 1.5808,
+      "step": 510
+    },
+    {
+      "epoch": 0.058698523921658724,
+      "grad_norm": 0.32216721773147583,
+      "learning_rate": 0.0001,
+      "loss": 1.6325,
+      "step": 511
+    },
+    {
+      "epoch": 0.058813393831485844,
+      "grad_norm": 0.3667677640914917,
+      "learning_rate": 0.0001,
+      "loss": 1.916,
+      "step": 512
+    },
+    {
+      "epoch": 0.058928263741312964,
+      "grad_norm": 0.3243674635887146,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 513
+    },
+    {
+      "epoch": 0.059043133651140084,
+      "grad_norm": 0.3738958537578583,
+      "learning_rate": 0.0001,
+      "loss": 2.0036,
+      "step": 514
+    },
+    {
+      "epoch": 0.059158003560967204,
+      "grad_norm": 0.33784958720207214,
+      "learning_rate": 0.0001,
+      "loss": 1.8144,
+      "step": 515
+    },
+    {
+      "epoch": 0.059272873470794324,
+      "grad_norm": 0.34589648246765137,
+      "learning_rate": 0.0001,
+      "loss": 1.7907,
+      "step": 516
+    },
+    {
+      "epoch": 0.059387743380621444,
+      "grad_norm": 0.3396931290626526,
+      "learning_rate": 0.0001,
+      "loss": 1.8062,
+      "step": 517
+    },
+    {
+      "epoch": 0.059502613290448564,
+      "grad_norm": 0.33430787920951843,
+      "learning_rate": 0.0001,
+      "loss": 1.8861,
+      "step": 518
+    },
+    {
+      "epoch": 0.05961748320027569,
+      "grad_norm": 0.33483850955963135,
+      "learning_rate": 0.0001,
+      "loss": 1.7733,
+      "step": 519
+    },
+    {
+      "epoch": 0.05973235311010281,
+      "grad_norm": 0.3303142488002777,
+      "learning_rate": 0.0001,
+      "loss": 1.6669,
+      "step": 520
+    },
+    {
+      "epoch": 0.05984722301992993,
+      "grad_norm": 0.38823017477989197,
+      "learning_rate": 0.0001,
+      "loss": 1.9658,
+      "step": 521
+    },
+    {
+      "epoch": 0.05996209292975705,
+      "grad_norm": 0.3367463946342468,
+      "learning_rate": 0.0001,
+      "loss": 1.7768,
+      "step": 522
+    },
+    {
+      "epoch": 0.06007696283958417,
+      "grad_norm": 0.3448854684829712,
+      "learning_rate": 0.0001,
+      "loss": 1.8997,
+      "step": 523
+    },
+    {
+      "epoch": 0.06019183274941129,
+      "grad_norm": 0.33760958909988403,
+      "learning_rate": 0.0001,
+      "loss": 1.7803,
+      "step": 524
+    },
+    {
+      "epoch": 0.06030670265923841,
+      "grad_norm": 0.3285827934741974,
+      "learning_rate": 0.0001,
+      "loss": 1.7758,
+      "step": 525
+    },
+    {
+      "epoch": 0.06042157256906553,
+      "grad_norm": 0.32994288206100464,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 526
+    },
+    {
+      "epoch": 0.06053644247889266,
+      "grad_norm": 0.31827473640441895,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 527
+    },
+    {
+      "epoch": 0.06065131238871978,
+      "grad_norm": 0.35536283254623413,
+      "learning_rate": 0.0001,
+      "loss": 1.8389,
+      "step": 528
+    },
+    {
+      "epoch": 0.0607661822985469,
+      "grad_norm": 0.34808462858200073,
+      "learning_rate": 0.0001,
+      "loss": 1.7249,
+      "step": 529
+    },
+    {
+      "epoch": 0.06088105220837402,
+      "grad_norm": 0.32803285121917725,
+      "learning_rate": 0.0001,
+      "loss": 1.7431,
+      "step": 530
+    },
+    {
+      "epoch": 0.06099592211820114,
+      "grad_norm": 0.3689769506454468,
+      "learning_rate": 0.0001,
+      "loss": 1.95,
+      "step": 531
+    },
+    {
+      "epoch": 0.06111079202802826,
+      "grad_norm": 0.324849396944046,
+      "learning_rate": 0.0001,
+      "loss": 1.6858,
+      "step": 532
+    },
+    {
+      "epoch": 0.06122566193785538,
+      "grad_norm": 0.367699533700943,
+      "learning_rate": 0.0001,
+      "loss": 1.9701,
+      "step": 533
+    },
+    {
+      "epoch": 0.0613405318476825,
+      "grad_norm": 0.346123069524765,
+      "learning_rate": 0.0001,
+      "loss": 1.7326,
+      "step": 534
+    },
+    {
+      "epoch": 0.06145540175750962,
+      "grad_norm": 0.33581048250198364,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 535
+    },
+    {
+      "epoch": 0.061570271667336744,
+      "grad_norm": 0.31508779525756836,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 536
+    },
+    {
+      "epoch": 0.061685141577163864,
+      "grad_norm": 0.31980377435684204,
+      "learning_rate": 0.0001,
+      "loss": 1.7605,
+      "step": 537
+    },
+    {
+      "epoch": 0.061800011486990984,
+      "grad_norm": 0.324613481760025,
+      "learning_rate": 0.0001,
+      "loss": 1.6779,
+      "step": 538
+    },
+    {
+      "epoch": 0.061914881396818104,
+      "grad_norm": 0.36573949456214905,
+      "learning_rate": 0.0001,
+      "loss": 1.8471,
+      "step": 539
+    },
+    {
+      "epoch": 0.062029751306645224,
+      "grad_norm": 0.34220999479293823,
+      "learning_rate": 0.0001,
+      "loss": 1.8383,
+      "step": 540
+    },
+    {
+      "epoch": 0.062144621216472344,
+      "grad_norm": 0.3276033103466034,
+      "learning_rate": 0.0001,
+      "loss": 1.6503,
+      "step": 541
+    },
+    {
+      "epoch": 0.062259491126299464,
+      "grad_norm": 0.320403516292572,
+      "learning_rate": 0.0001,
+      "loss": 1.4869,
+      "step": 542
+    },
+    {
+      "epoch": 0.062374361036126584,
+      "grad_norm": 0.333486407995224,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 543
+    },
+    {
+      "epoch": 0.062489230945953704,
+      "grad_norm": 0.3390301764011383,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 544
+    },
+    {
+      "epoch": 0.06260410085578083,
+      "grad_norm": 0.3051248788833618,
+      "learning_rate": 0.0001,
+      "loss": 1.632,
+      "step": 545
+    },
+    {
+      "epoch": 0.06271897076560795,
+      "grad_norm": 0.4015922248363495,
+      "learning_rate": 0.0001,
+      "loss": 1.9027,
+      "step": 546
+    },
+    {
+      "epoch": 0.06283384067543507,
+      "grad_norm": 0.35976481437683105,
+      "learning_rate": 0.0001,
+      "loss": 1.8451,
+      "step": 547
+    },
+    {
+      "epoch": 0.06294871058526219,
+      "grad_norm": 0.3470692038536072,
+      "learning_rate": 0.0001,
+      "loss": 1.7992,
+      "step": 548
+    },
+    {
+      "epoch": 0.06306358049508931,
+      "grad_norm": 0.324569433927536,
+      "learning_rate": 0.0001,
+      "loss": 1.7567,
+      "step": 549
+    },
+    {
+      "epoch": 0.06317845040491643,
+      "grad_norm": 0.36248597502708435,
+      "learning_rate": 0.0001,
+      "loss": 1.8184,
+      "step": 550
+    },
+    {
+      "epoch": 0.06329332031474355,
+      "grad_norm": 0.33431920409202576,
+      "learning_rate": 0.0001,
+      "loss": 1.7591,
+      "step": 551
+    },
+    {
+      "epoch": 0.06340819022457067,
+      "grad_norm": 0.3006725013256073,
+      "learning_rate": 0.0001,
+      "loss": 1.5713,
+      "step": 552
+    },
+    {
+      "epoch": 0.06352306013439779,
+      "grad_norm": 0.3330213725566864,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 553
+    },
+    {
+      "epoch": 0.06363793004422491,
+      "grad_norm": 0.34222641587257385,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 554
+    },
+    {
+      "epoch": 0.06375279995405203,
+      "grad_norm": 0.34243908524513245,
+      "learning_rate": 0.0001,
+      "loss": 1.7752,
+      "step": 555
+    },
+    {
+      "epoch": 0.06386766986387915,
+      "grad_norm": 0.3517223596572876,
+      "learning_rate": 0.0001,
+      "loss": 1.91,
+      "step": 556
+    },
+    {
+      "epoch": 0.06398253977370628,
+      "grad_norm": 0.38396772742271423,
+      "learning_rate": 0.0001,
+      "loss": 2.1684,
+      "step": 557
+    },
+    {
+      "epoch": 0.0640974096835334,
+      "grad_norm": 0.34429916739463806,
+      "learning_rate": 0.0001,
+      "loss": 1.9133,
+      "step": 558
+    },
+    {
+      "epoch": 0.06421227959336052,
+      "grad_norm": 0.31630486249923706,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 559
+    },
+    {
+      "epoch": 0.06432714950318764,
+      "grad_norm": 0.3668649196624756,
+      "learning_rate": 0.0001,
+      "loss": 1.7739,
+      "step": 560
+    },
+    {
+      "epoch": 0.06444201941301476,
+      "grad_norm": 0.33105143904685974,
+      "learning_rate": 0.0001,
+      "loss": 1.6974,
+      "step": 561
+    },
+    {
+      "epoch": 0.06455688932284188,
+      "grad_norm": 0.3978722393512726,
+      "learning_rate": 0.0001,
+      "loss": 1.8365,
+      "step": 562
+    },
+    {
+      "epoch": 0.064671759232669,
+      "grad_norm": 0.3352854549884796,
+      "learning_rate": 0.0001,
+      "loss": 1.7237,
+      "step": 563
+    },
+    {
+      "epoch": 0.06478662914249612,
+      "grad_norm": 0.3484468460083008,
+      "learning_rate": 0.0001,
+      "loss": 1.8297,
+      "step": 564
+    },
+    {
+      "epoch": 0.06490149905232324,
+      "grad_norm": 0.3346973955631256,
+      "learning_rate": 0.0001,
+      "loss": 1.7631,
+      "step": 565
+    },
+    {
+      "epoch": 0.06501636896215036,
+      "grad_norm": 0.3363039195537567,
+      "learning_rate": 0.0001,
+      "loss": 1.6818,
+      "step": 566
+    },
+    {
+      "epoch": 0.06513123887197748,
+      "grad_norm": 0.34610244631767273,
+      "learning_rate": 0.0001,
+      "loss": 1.8759,
+      "step": 567
+    },
+    {
+      "epoch": 0.0652461087818046,
+      "grad_norm": 0.36252561211586,
+      "learning_rate": 0.0001,
+      "loss": 1.8793,
+      "step": 568
+    },
+    {
+      "epoch": 0.06536097869163172,
+      "grad_norm": 0.3493739068508148,
+      "learning_rate": 0.0001,
+      "loss": 1.7335,
+      "step": 569
+    },
+    {
+      "epoch": 0.06547584860145884,
+      "grad_norm": 0.3322302997112274,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 570
+    },
+    {
+      "epoch": 0.06559071851128596,
+      "grad_norm": 0.31431615352630615,
+      "learning_rate": 0.0001,
+      "loss": 1.5679,
+      "step": 571
+    },
+    {
+      "epoch": 0.06570558842111308,
+      "grad_norm": 0.3374696671962738,
+      "learning_rate": 0.0001,
+      "loss": 1.7067,
+      "step": 572
+    },
+    {
+      "epoch": 0.0658204583309402,
+      "grad_norm": 0.343352735042572,
+      "learning_rate": 0.0001,
+      "loss": 1.8222,
+      "step": 573
+    },
+    {
+      "epoch": 0.06593532824076732,
+      "grad_norm": 0.33851170539855957,
+      "learning_rate": 0.0001,
+      "loss": 1.6893,
+      "step": 574
+    },
+    {
+      "epoch": 0.06605019815059446,
+      "grad_norm": 0.36369964480400085,
+      "learning_rate": 0.0001,
+      "loss": 1.8065,
+      "step": 575
+    },
+    {
+      "epoch": 0.06616506806042158,
+      "grad_norm": 0.31349602341651917,
+      "learning_rate": 0.0001,
+      "loss": 1.5028,
+      "step": 576
+    },
+    {
+      "epoch": 0.0662799379702487,
+      "grad_norm": 0.3367163836956024,
+      "learning_rate": 0.0001,
+      "loss": 1.8033,
+      "step": 577
+    },
+    {
+      "epoch": 0.06639480788007582,
+      "grad_norm": 0.3456117510795593,
+      "learning_rate": 0.0001,
+      "loss": 1.711,
+      "step": 578
+    },
+    {
+      "epoch": 0.06650967778990294,
+      "grad_norm": 0.31135809421539307,
+      "learning_rate": 0.0001,
+      "loss": 1.3939,
+      "step": 579
+    },
+    {
+      "epoch": 0.06662454769973006,
+      "grad_norm": 0.327361136674881,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 580
+    },
+    {
+      "epoch": 0.06673941760955718,
+      "grad_norm": 0.345680296421051,
+      "learning_rate": 0.0001,
+      "loss": 1.7006,
+      "step": 581
+    },
+    {
+      "epoch": 0.0668542875193843,
+      "grad_norm": 0.33879801630973816,
+      "learning_rate": 0.0001,
+      "loss": 1.8236,
+      "step": 582
+    },
+    {
+      "epoch": 0.06696915742921142,
+      "grad_norm": 0.3614217936992645,
+      "learning_rate": 0.0001,
+      "loss": 1.8137,
+      "step": 583
+    },
+    {
+      "epoch": 0.06708402733903854,
+      "grad_norm": 0.34495973587036133,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 584
+    },
+    {
+      "epoch": 0.06719889724886566,
+      "grad_norm": 0.31370049715042114,
+      "learning_rate": 0.0001,
+      "loss": 1.7482,
+      "step": 585
+    },
+    {
+      "epoch": 0.06731376715869278,
+      "grad_norm": 0.3325652480125427,
+      "learning_rate": 0.0001,
+      "loss": 1.88,
+      "step": 586
+    },
+    {
+      "epoch": 0.0674286370685199,
+      "grad_norm": 0.351700097322464,
+      "learning_rate": 0.0001,
+      "loss": 1.8817,
+      "step": 587
+    },
+    {
+      "epoch": 0.06754350697834702,
+      "grad_norm": 0.3200022280216217,
+      "learning_rate": 0.0001,
+      "loss": 1.7419,
+      "step": 588
+    },
+    {
+      "epoch": 0.06765837688817414,
+      "grad_norm": 0.36215293407440186,
+      "learning_rate": 0.0001,
+      "loss": 1.9603,
+      "step": 589
+    },
+    {
+      "epoch": 0.06777324679800126,
+      "grad_norm": 0.3392893970012665,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 590
+    },
+    {
+      "epoch": 0.06788811670782838,
+      "grad_norm": 0.3454267084598541,
+      "learning_rate": 0.0001,
+      "loss": 1.6759,
+      "step": 591
+    },
+    {
+      "epoch": 0.06800298661765551,
+      "grad_norm": 0.35802924633026123,
+      "learning_rate": 0.0001,
+      "loss": 1.8607,
+      "step": 592
+    },
+    {
+      "epoch": 0.06811785652748263,
+      "grad_norm": 0.2966287434101105,
+      "learning_rate": 0.0001,
+      "loss": 1.594,
+      "step": 593
+    },
+    {
+      "epoch": 0.06823272643730975,
+      "grad_norm": 0.36141011118888855,
+      "learning_rate": 0.0001,
+      "loss": 1.7596,
+      "step": 594
+    },
+    {
+      "epoch": 0.06834759634713687,
+      "grad_norm": 0.36919906735420227,
+      "learning_rate": 0.0001,
+      "loss": 1.9334,
+      "step": 595
+    },
+    {
+      "epoch": 0.06846246625696399,
+      "grad_norm": 0.34811851382255554,
+      "learning_rate": 0.0001,
+      "loss": 1.92,
+      "step": 596
+    },
+    {
+      "epoch": 0.06857733616679111,
+      "grad_norm": 0.3520393669605255,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 597
+    },
+    {
+      "epoch": 0.06869220607661823,
+      "grad_norm": 0.3605727553367615,
+      "learning_rate": 0.0001,
+      "loss": 1.8132,
+      "step": 598
+    },
+    {
+      "epoch": 0.06880707598644535,
+      "grad_norm": 0.3948690593242645,
+      "learning_rate": 0.0001,
+      "loss": 1.7842,
+      "step": 599
+    },
+    {
+      "epoch": 0.06892194589627247,
+      "grad_norm": 0.34386035799980164,
+      "learning_rate": 0.0001,
+      "loss": 1.6827,
+      "step": 600
+    },
+    {
+      "epoch": 0.06903681580609959,
+      "grad_norm": 0.37180081009864807,
+      "learning_rate": 0.0001,
+      "loss": 1.7982,
+      "step": 601
+    },
+    {
+      "epoch": 0.06915168571592671,
+      "grad_norm": 0.3451867401599884,
+      "learning_rate": 0.0001,
+      "loss": 1.8339,
+      "step": 602
+    },
+    {
+      "epoch": 0.06926655562575383,
+      "grad_norm": 0.3325120210647583,
+      "learning_rate": 0.0001,
+      "loss": 1.7506,
+      "step": 603
+    },
+    {
+      "epoch": 0.06938142553558095,
+      "grad_norm": 0.3503422737121582,
+      "learning_rate": 0.0001,
+      "loss": 1.8154,
+      "step": 604
+    },
+    {
+      "epoch": 0.06949629544540807,
+      "grad_norm": 0.36198514699935913,
+      "learning_rate": 0.0001,
+      "loss": 1.7698,
+      "step": 605
+    },
+    {
+      "epoch": 0.06961116535523519,
+      "grad_norm": 0.35194844007492065,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 606
+    },
+    {
+      "epoch": 0.06972603526506231,
+      "grad_norm": 0.3265458643436432,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 607
+    },
+    {
+      "epoch": 0.06984090517488943,
+      "grad_norm": 0.3669833838939667,
+      "learning_rate": 0.0001,
+      "loss": 1.8176,
+      "step": 608
+    },
+    {
+      "epoch": 0.06995577508471656,
+      "grad_norm": 0.35220587253570557,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 609
+    },
+    {
+      "epoch": 0.07007064499454368,
+      "grad_norm": 0.3398061692714691,
+      "learning_rate": 0.0001,
+      "loss": 1.4962,
+      "step": 610
+    },
+    {
+      "epoch": 0.0701855149043708,
+      "grad_norm": 0.36075279116630554,
+      "learning_rate": 0.0001,
+      "loss": 1.8924,
+      "step": 611
+    },
+    {
+      "epoch": 0.07030038481419792,
+      "grad_norm": 0.34688082337379456,
+      "learning_rate": 0.0001,
+      "loss": 1.5272,
+      "step": 612
+    },
+    {
+      "epoch": 0.07041525472402504,
+      "grad_norm": 0.33742544054985046,
+      "learning_rate": 0.0001,
+      "loss": 1.75,
+      "step": 613
+    },
+    {
+      "epoch": 0.07053012463385216,
+      "grad_norm": 0.3314124345779419,
+      "learning_rate": 0.0001,
+      "loss": 1.7574,
+      "step": 614
+    },
+    {
+      "epoch": 0.07064499454367928,
+      "grad_norm": 0.36540499329566956,
+      "learning_rate": 0.0001,
+      "loss": 1.926,
+      "step": 615
+    },
+    {
+      "epoch": 0.0707598644535064,
+      "grad_norm": 0.3602568507194519,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 616
+    },
+    {
+      "epoch": 0.07087473436333352,
+      "grad_norm": 0.3252723515033722,
+      "learning_rate": 0.0001,
+      "loss": 1.8458,
+      "step": 617
+    },
+    {
+      "epoch": 0.07098960427316064,
+      "grad_norm": 0.3487424850463867,
+      "learning_rate": 0.0001,
+      "loss": 1.8981,
+      "step": 618
+    },
+    {
+      "epoch": 0.07110447418298776,
+      "grad_norm": 0.33129236102104187,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 619
+    },
+    {
+      "epoch": 0.07121934409281488,
+      "grad_norm": 0.33063703775405884,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 620
+    },
+    {
+      "epoch": 0.071334214002642,
+      "grad_norm": 0.32881104946136475,
+      "learning_rate": 0.0001,
+      "loss": 1.7321,
+      "step": 621
+    },
+    {
+      "epoch": 0.07144908391246912,
+      "grad_norm": 0.3607088029384613,
+      "learning_rate": 0.0001,
+      "loss": 1.698,
+      "step": 622
+    },
+    {
+      "epoch": 0.07156395382229624,
+      "grad_norm": 0.3350905179977417,
+      "learning_rate": 0.0001,
+      "loss": 1.7254,
+      "step": 623
+    },
+    {
+      "epoch": 0.07167882373212336,
+      "grad_norm": 0.35111361742019653,
+      "learning_rate": 0.0001,
+      "loss": 1.9324,
+      "step": 624
+    },
+    {
+      "epoch": 0.07179369364195048,
+      "grad_norm": 0.3444902002811432,
+      "learning_rate": 0.0001,
+      "loss": 1.8281,
+      "step": 625
+    },
+    {
+      "epoch": 0.07190856355177762,
+      "grad_norm": 0.34226348996162415,
+      "learning_rate": 0.0001,
+      "loss": 1.7824,
+      "step": 626
+    },
+    {
+      "epoch": 0.07202343346160474,
+      "grad_norm": 0.33303242921829224,
+      "learning_rate": 0.0001,
+      "loss": 1.8413,
+      "step": 627
+    },
+    {
+      "epoch": 0.07213830337143186,
+      "grad_norm": 0.3243139088153839,
+      "learning_rate": 0.0001,
+      "loss": 1.8731,
+      "step": 628
+    },
+    {
+      "epoch": 0.07225317328125898,
+      "grad_norm": 0.3446861505508423,
+      "learning_rate": 0.0001,
+      "loss": 1.7179,
+      "step": 629
+    },
+    {
+      "epoch": 0.0723680431910861,
+      "grad_norm": 0.337261438369751,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 630
+    },
+    {
+      "epoch": 0.07248291310091322,
+      "grad_norm": 0.32150766253471375,
+      "learning_rate": 0.0001,
+      "loss": 1.8097,
+      "step": 631
+    },
+    {
+      "epoch": 0.07259778301074034,
+      "grad_norm": 0.3300226032733917,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 632
+    },
+    {
+      "epoch": 0.07271265292056746,
+      "grad_norm": 0.3479246199131012,
+      "learning_rate": 0.0001,
+      "loss": 1.7969,
+      "step": 633
+    },
+    {
+      "epoch": 0.07282752283039458,
+      "grad_norm": 0.33039695024490356,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 634
+    },
+    {
+      "epoch": 0.0729423927402217,
+      "grad_norm": 0.3294956088066101,
+      "learning_rate": 0.0001,
+      "loss": 1.7817,
+      "step": 635
+    },
+    {
+      "epoch": 0.07305726265004882,
+      "grad_norm": 0.3452272415161133,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 636
+    },
+    {
+      "epoch": 0.07317213255987594,
+      "grad_norm": 0.33393406867980957,
+      "learning_rate": 0.0001,
+      "loss": 1.8335,
+      "step": 637
+    },
+    {
+      "epoch": 0.07328700246970306,
+      "grad_norm": 0.3169970214366913,
+      "learning_rate": 0.0001,
+      "loss": 1.6022,
+      "step": 638
+    },
+    {
+      "epoch": 0.07340187237953018,
+      "grad_norm": 0.3217456340789795,
+      "learning_rate": 0.0001,
+      "loss": 1.579,
+      "step": 639
+    },
+    {
+      "epoch": 0.0735167422893573,
+      "grad_norm": 0.34844133257865906,
+      "learning_rate": 0.0001,
+      "loss": 1.8499,
+      "step": 640
+    },
+    {
+      "epoch": 0.07363161219918442,
+      "grad_norm": 0.33645370602607727,
+      "learning_rate": 0.0001,
+      "loss": 1.862,
+      "step": 641
+    },
+    {
+      "epoch": 0.07374648210901154,
+      "grad_norm": 0.3201218843460083,
+      "learning_rate": 0.0001,
+      "loss": 1.7561,
+      "step": 642
+    },
+    {
+      "epoch": 0.07386135201883867,
+      "grad_norm": 0.34113234281539917,
+      "learning_rate": 0.0001,
+      "loss": 1.8113,
+      "step": 643
+    },
+    {
+      "epoch": 0.07397622192866579,
+      "grad_norm": 0.33081957697868347,
+      "learning_rate": 0.0001,
+      "loss": 1.7435,
+      "step": 644
+    },
+    {
+      "epoch": 0.07409109183849291,
+      "grad_norm": 0.3413662314414978,
+      "learning_rate": 0.0001,
+      "loss": 1.8755,
+      "step": 645
+    },
+    {
+      "epoch": 0.07420596174832003,
+      "grad_norm": 0.3311666250228882,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 646
+    },
+    {
+      "epoch": 0.07432083165814715,
+      "grad_norm": 0.33269715309143066,
+      "learning_rate": 0.0001,
+      "loss": 1.7321,
+      "step": 647
+    },
+    {
+      "epoch": 0.07443570156797427,
+      "grad_norm": 0.34695979952812195,
+      "learning_rate": 0.0001,
+      "loss": 1.7628,
+      "step": 648
+    },
+    {
+      "epoch": 0.07455057147780139,
+      "grad_norm": 0.3331931233406067,
+      "learning_rate": 0.0001,
+      "loss": 1.7498,
+      "step": 649
+    },
+    {
+      "epoch": 0.07466544138762851,
+      "grad_norm": 0.3264698088169098,
+      "learning_rate": 0.0001,
+      "loss": 1.6114,
+      "step": 650
+    },
+    {
+      "epoch": 0.07478031129745563,
+      "grad_norm": 0.3626859486103058,
+      "learning_rate": 0.0001,
+      "loss": 1.8154,
+      "step": 651
+    },
+    {
+      "epoch": 0.07489518120728275,
+      "grad_norm": 0.32509714365005493,
+      "learning_rate": 0.0001,
+      "loss": 1.4671,
+      "step": 652
+    },
+    {
+      "epoch": 0.07501005111710987,
+      "grad_norm": 0.3228186070919037,
+      "learning_rate": 0.0001,
+      "loss": 1.8735,
+      "step": 653
+    },
+    {
+      "epoch": 0.07512492102693699,
+      "grad_norm": 0.3465333878993988,
+      "learning_rate": 0.0001,
+      "loss": 1.872,
+      "step": 654
+    },
+    {
+      "epoch": 0.07523979093676411,
+      "grad_norm": 0.3378332555294037,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 655
+    },
+    {
+      "epoch": 0.07535466084659123,
+      "grad_norm": 0.3364262282848358,
+      "learning_rate": 0.0001,
+      "loss": 1.8611,
+      "step": 656
+    },
+    {
+      "epoch": 0.07546953075641835,
+      "grad_norm": 0.34226563572883606,
+      "learning_rate": 0.0001,
+      "loss": 1.7843,
+      "step": 657
+    },
+    {
+      "epoch": 0.07558440066624547,
+      "grad_norm": 0.3533295691013336,
+      "learning_rate": 0.0001,
+      "loss": 1.7962,
+      "step": 658
+    },
+    {
+      "epoch": 0.07569927057607259,
+      "grad_norm": 0.3422401249408722,
+      "learning_rate": 0.0001,
+      "loss": 1.8801,
+      "step": 659
+    },
+    {
+      "epoch": 0.07581414048589973,
+      "grad_norm": 0.3540160357952118,
+      "learning_rate": 0.0001,
+      "loss": 1.8129,
+      "step": 660
+    },
+    {
+      "epoch": 0.07592901039572685,
+      "grad_norm": 0.334587424993515,
+      "learning_rate": 0.0001,
+      "loss": 1.8578,
+      "step": 661
+    },
+    {
+      "epoch": 0.07604388030555397,
+      "grad_norm": 0.32655155658721924,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 662
+    },
+    {
+      "epoch": 0.07615875021538109,
+      "grad_norm": 0.36004751920700073,
+      "learning_rate": 0.0001,
+      "loss": 1.7735,
+      "step": 663
+    },
+    {
+      "epoch": 0.0762736201252082,
+      "grad_norm": 0.32442474365234375,
+      "learning_rate": 0.0001,
+      "loss": 1.5389,
+      "step": 664
+    },
+    {
+      "epoch": 0.07638849003503533,
+      "grad_norm": 0.344626784324646,
+      "learning_rate": 0.0001,
+      "loss": 1.6952,
+      "step": 665
+    },
+    {
+      "epoch": 0.07650335994486245,
+      "grad_norm": 0.31557947397232056,
+      "learning_rate": 0.0001,
+      "loss": 1.5595,
+      "step": 666
+    },
+    {
+      "epoch": 0.07661822985468957,
+      "grad_norm": 0.3274221420288086,
+      "learning_rate": 0.0001,
+      "loss": 1.6471,
+      "step": 667
+    },
+    {
+      "epoch": 0.07673309976451669,
+      "grad_norm": 0.3906736969947815,
+      "learning_rate": 0.0001,
+      "loss": 1.6702,
+      "step": 668
+    },
+    {
+      "epoch": 0.0768479696743438,
+      "grad_norm": 0.35045820474624634,
+      "learning_rate": 0.0001,
+      "loss": 1.8738,
+      "step": 669
+    },
+    {
+      "epoch": 0.07696283958417093,
+      "grad_norm": 0.3959348201751709,
+      "learning_rate": 0.0001,
+      "loss": 2.094,
+      "step": 670
+    },
+    {
+      "epoch": 0.07707770949399805,
+      "grad_norm": 0.3369539976119995,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 671
+    },
+    {
+      "epoch": 0.07719257940382517,
+      "grad_norm": 0.34965980052948,
+      "learning_rate": 0.0001,
+      "loss": 1.8694,
+      "step": 672
+    },
+    {
+      "epoch": 0.07730744931365229,
+      "grad_norm": 0.33249253034591675,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 673
+    },
+    {
+      "epoch": 0.0774223192234794,
+      "grad_norm": 0.32257145643234253,
+      "learning_rate": 0.0001,
+      "loss": 1.7565,
+      "step": 674
+    },
+    {
+      "epoch": 0.07753718913330652,
+      "grad_norm": 0.33610349893569946,
+      "learning_rate": 0.0001,
+      "loss": 1.8517,
+      "step": 675
+    },
+    {
+      "epoch": 0.07765205904313364,
+      "grad_norm": 0.3666530251502991,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 676
+    },
+    {
+      "epoch": 0.07776692895296078,
+      "grad_norm": 0.3422529697418213,
+      "learning_rate": 0.0001,
+      "loss": 1.6884,
+      "step": 677
+    },
+    {
+      "epoch": 0.0778817988627879,
+      "grad_norm": 0.3361228108406067,
+      "learning_rate": 0.0001,
+      "loss": 1.7452,
+      "step": 678
+    },
+    {
+      "epoch": 0.07799666877261502,
+      "grad_norm": 0.3532163202762604,
+      "learning_rate": 0.0001,
+      "loss": 1.6878,
+      "step": 679
+    },
+    {
+      "epoch": 0.07811153868244214,
+      "grad_norm": 0.3304887115955353,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 680
+    },
+    {
+      "epoch": 0.07822640859226926,
+      "grad_norm": 0.3171667456626892,
+      "learning_rate": 0.0001,
+      "loss": 1.5896,
+      "step": 681
+    },
+    {
+      "epoch": 0.07834127850209638,
+      "grad_norm": 0.36117199063301086,
+      "learning_rate": 0.0001,
+      "loss": 1.7566,
+      "step": 682
+    },
+    {
+      "epoch": 0.0784561484119235,
+      "grad_norm": 0.37346988916397095,
+      "learning_rate": 0.0001,
+      "loss": 1.8458,
+      "step": 683
+    },
+    {
+      "epoch": 0.07857101832175062,
+      "grad_norm": 0.34928634762763977,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 684
+    },
+    {
+      "epoch": 0.07868588823157774,
+      "grad_norm": 0.34768396615982056,
+      "learning_rate": 0.0001,
+      "loss": 1.8074,
+      "step": 685
+    },
+    {
+      "epoch": 0.07880075814140486,
+      "grad_norm": 0.3412458300590515,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 686
+    },
+    {
+      "epoch": 0.07891562805123198,
+      "grad_norm": 0.3672271966934204,
+      "learning_rate": 0.0001,
+      "loss": 1.8826,
+      "step": 687
+    },
+    {
+      "epoch": 0.0790304979610591,
+      "grad_norm": 0.33574315905570984,
+      "learning_rate": 0.0001,
+      "loss": 1.6836,
+      "step": 688
+    },
+    {
+      "epoch": 0.07914536787088622,
+      "grad_norm": 0.3387349247932434,
+      "learning_rate": 0.0001,
+      "loss": 1.6531,
+      "step": 689
+    },
+    {
+      "epoch": 0.07926023778071334,
+      "grad_norm": 0.3550173342227936,
+      "learning_rate": 0.0001,
+      "loss": 1.7975,
+      "step": 690
+    },
+    {
+      "epoch": 0.07937510769054046,
+      "grad_norm": 0.380522221326828,
+      "learning_rate": 0.0001,
+      "loss": 1.9264,
+      "step": 691
+    },
+    {
+      "epoch": 0.07948997760036758,
+      "grad_norm": 0.374406099319458,
+      "learning_rate": 0.0001,
+      "loss": 1.8652,
+      "step": 692
+    },
+    {
+      "epoch": 0.0796048475101947,
+      "grad_norm": 0.34742456674575806,
+      "learning_rate": 0.0001,
+      "loss": 1.9069,
+      "step": 693
+    },
+    {
+      "epoch": 0.07971971742002182,
+      "grad_norm": 0.35284706950187683,
+      "learning_rate": 0.0001,
+      "loss": 1.8065,
+      "step": 694
+    },
+    {
+      "epoch": 0.07983458732984895,
+      "grad_norm": 0.3327619135379791,
+      "learning_rate": 0.0001,
+      "loss": 1.7721,
+      "step": 695
+    },
+    {
+      "epoch": 0.07994945723967607,
+      "grad_norm": 0.3591189384460449,
+      "learning_rate": 0.0001,
+      "loss": 1.6802,
+      "step": 696
+    },
+    {
+      "epoch": 0.08006432714950319,
+      "grad_norm": 0.34091916680336,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 697
+    },
+    {
+      "epoch": 0.08017919705933031,
+      "grad_norm": 0.352022260427475,
+      "learning_rate": 0.0001,
+      "loss": 1.7479,
+      "step": 698
+    },
+    {
+      "epoch": 0.08029406696915743,
+      "grad_norm": 0.33692196011543274,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 699
+    },
+    {
+      "epoch": 0.08040893687898455,
+      "grad_norm": 0.370638370513916,
+      "learning_rate": 0.0001,
+      "loss": 2.0285,
+      "step": 700
+    },
+    {
+      "epoch": 0.08052380678881167,
+      "grad_norm": 0.3345963656902313,
+      "learning_rate": 0.0001,
+      "loss": 1.7766,
+      "step": 701
+    },
+    {
+      "epoch": 0.08063867669863879,
+      "grad_norm": 0.34501156210899353,
+      "learning_rate": 0.0001,
+      "loss": 1.8179,
+      "step": 702
+    },
+    {
+      "epoch": 0.08075354660846591,
+      "grad_norm": 0.33231601119041443,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 703
+    },
+    {
+      "epoch": 0.08086841651829303,
+      "grad_norm": 0.34279513359069824,
+      "learning_rate": 0.0001,
+      "loss": 1.695,
+      "step": 704
+    },
+    {
+      "epoch": 0.08098328642812015,
+      "grad_norm": 0.3368370234966278,
+      "learning_rate": 0.0001,
+      "loss": 1.7584,
+      "step": 705
+    },
+    {
+      "epoch": 0.08109815633794727,
+      "grad_norm": 0.34584423899650574,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 706
+    },
+    {
+      "epoch": 0.08121302624777439,
+      "grad_norm": 0.35114485025405884,
+      "learning_rate": 0.0001,
+      "loss": 1.7287,
+      "step": 707
+    },
+    {
+      "epoch": 0.08132789615760151,
+      "grad_norm": 0.31173431873321533,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 708
+    },
+    {
+      "epoch": 0.08144276606742863,
+      "grad_norm": 0.3620467782020569,
+      "learning_rate": 0.0001,
+      "loss": 1.8973,
+      "step": 709
+    },
+    {
+      "epoch": 0.08155763597725575,
+      "grad_norm": 0.3193514049053192,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 710
+    },
+    {
+      "epoch": 0.08167250588708287,
+      "grad_norm": 0.34377214312553406,
+      "learning_rate": 0.0001,
+      "loss": 1.8458,
+      "step": 711
+    },
+    {
+      "epoch": 0.08178737579691,
+      "grad_norm": 0.3406418263912201,
+      "learning_rate": 0.0001,
+      "loss": 1.6495,
+      "step": 712
+    },
+    {
+      "epoch": 0.08190224570673713,
+      "grad_norm": 0.33058011531829834,
+      "learning_rate": 0.0001,
+      "loss": 1.7232,
+      "step": 713
+    },
+    {
+      "epoch": 0.08201711561656425,
+      "grad_norm": 0.32958585023880005,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 714
+    },
+    {
+      "epoch": 0.08213198552639137,
+      "grad_norm": 0.3345566689968109,
+      "learning_rate": 0.0001,
+      "loss": 1.7372,
+      "step": 715
+    },
+    {
+      "epoch": 0.08224685543621849,
+      "grad_norm": 0.3363969027996063,
+      "learning_rate": 0.0001,
+      "loss": 1.5248,
+      "step": 716
+    },
+    {
+      "epoch": 0.0823617253460456,
+      "grad_norm": 0.3310002386569977,
+      "learning_rate": 0.0001,
+      "loss": 1.4364,
+      "step": 717
+    },
+    {
+      "epoch": 0.08247659525587273,
+      "grad_norm": 0.35177963972091675,
+      "learning_rate": 0.0001,
+      "loss": 1.8578,
+      "step": 718
+    },
+    {
+      "epoch": 0.08259146516569985,
+      "grad_norm": 0.37990766763687134,
+      "learning_rate": 0.0001,
+      "loss": 1.8266,
+      "step": 719
+    },
+    {
+      "epoch": 0.08270633507552697,
+      "grad_norm": 0.3506127893924713,
+      "learning_rate": 0.0001,
+      "loss": 1.7454,
+      "step": 720
+    },
+    {
+      "epoch": 0.08282120498535409,
+      "grad_norm": 0.3488253653049469,
+      "learning_rate": 0.0001,
+      "loss": 1.8147,
+      "step": 721
+    },
+    {
+      "epoch": 0.0829360748951812,
+      "grad_norm": 0.34471482038497925,
+      "learning_rate": 0.0001,
+      "loss": 1.7288,
+      "step": 722
+    },
+    {
+      "epoch": 0.08305094480500833,
+      "grad_norm": 0.33807575702667236,
+      "learning_rate": 0.0001,
+      "loss": 1.8128,
+      "step": 723
+    },
+    {
+      "epoch": 0.08316581471483545,
+      "grad_norm": 0.3196840286254883,
+      "learning_rate": 0.0001,
+      "loss": 1.6008,
+      "step": 724
+    },
+    {
+      "epoch": 0.08328068462466257,
+      "grad_norm": 0.36051392555236816,
+      "learning_rate": 0.0001,
+      "loss": 1.7202,
+      "step": 725
+    },
+    {
+      "epoch": 0.08339555453448969,
+      "grad_norm": 0.3275487720966339,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 726
+    },
+    {
+      "epoch": 0.0835104244443168,
+      "grad_norm": 0.3423649072647095,
+      "learning_rate": 0.0001,
+      "loss": 1.7988,
+      "step": 727
+    },
+    {
+      "epoch": 0.08362529435414393,
+      "grad_norm": 0.3507118225097656,
+      "learning_rate": 0.0001,
+      "loss": 1.5723,
+      "step": 728
+    },
+    {
+      "epoch": 0.08374016426397106,
+      "grad_norm": 0.3463688790798187,
+      "learning_rate": 0.0001,
+      "loss": 1.8294,
+      "step": 729
+    },
+    {
+      "epoch": 0.08385503417379818,
+      "grad_norm": 0.3360377848148346,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 730
+    },
+    {
+      "epoch": 0.0839699040836253,
+      "grad_norm": 0.3324314057826996,
+      "learning_rate": 0.0001,
+      "loss": 1.7487,
+      "step": 731
+    },
+    {
+      "epoch": 0.08408477399345242,
+      "grad_norm": 0.3390614688396454,
+      "learning_rate": 0.0001,
+      "loss": 1.7351,
+      "step": 732
+    },
+    {
+      "epoch": 0.08419964390327954,
+      "grad_norm": 0.35863199830055237,
+      "learning_rate": 0.0001,
+      "loss": 1.908,
+      "step": 733
+    },
+    {
+      "epoch": 0.08431451381310666,
+      "grad_norm": 0.3417515158653259,
+      "learning_rate": 0.0001,
+      "loss": 1.8832,
+      "step": 734
+    },
+    {
+      "epoch": 0.08442938372293378,
+      "grad_norm": 0.31135502457618713,
+      "learning_rate": 0.0001,
+      "loss": 1.432,
+      "step": 735
+    },
+    {
+      "epoch": 0.0845442536327609,
+      "grad_norm": 0.33590710163116455,
+      "learning_rate": 0.0001,
+      "loss": 1.7915,
+      "step": 736
+    },
+    {
+      "epoch": 0.08465912354258802,
+      "grad_norm": 0.33955830335617065,
+      "learning_rate": 0.0001,
+      "loss": 1.7333,
+      "step": 737
+    },
+    {
+      "epoch": 0.08477399345241514,
+      "grad_norm": 0.3825172483921051,
+      "learning_rate": 0.0001,
+      "loss": 1.7999,
+      "step": 738
+    },
+    {
+      "epoch": 0.08488886336224226,
+      "grad_norm": 0.33067587018013,
+      "learning_rate": 0.0001,
+      "loss": 1.7567,
+      "step": 739
+    },
+    {
+      "epoch": 0.08500373327206938,
+      "grad_norm": 0.344756156206131,
+      "learning_rate": 0.0001,
+      "loss": 1.7844,
+      "step": 740
+    },
+    {
+      "epoch": 0.0851186031818965,
+      "grad_norm": 0.3478997051715851,
+      "learning_rate": 0.0001,
+      "loss": 1.8903,
+      "step": 741
+    },
+    {
+      "epoch": 0.08523347309172362,
+      "grad_norm": 0.3651295006275177,
+      "learning_rate": 0.0001,
+      "loss": 1.8882,
+      "step": 742
+    },
+    {
+      "epoch": 0.08534834300155074,
+      "grad_norm": 0.30588778853416443,
+      "learning_rate": 0.0001,
+      "loss": 1.6636,
+      "step": 743
+    },
+    {
+      "epoch": 0.08546321291137786,
+      "grad_norm": 0.3360552489757538,
+      "learning_rate": 0.0001,
+      "loss": 1.865,
+      "step": 744
+    },
+    {
+      "epoch": 0.08557808282120498,
+      "grad_norm": 0.3278788626194,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 745
+    },
+    {
+      "epoch": 0.08569295273103211,
+      "grad_norm": 0.3437139391899109,
+      "learning_rate": 0.0001,
+      "loss": 1.8981,
+      "step": 746
+    },
+    {
+      "epoch": 0.08580782264085923,
+      "grad_norm": 0.34554487466812134,
+      "learning_rate": 0.0001,
+      "loss": 1.7921,
+      "step": 747
+    },
+    {
+      "epoch": 0.08592269255068635,
+      "grad_norm": 0.35071298480033875,
+      "learning_rate": 0.0001,
+      "loss": 1.8354,
+      "step": 748
+    },
+    {
+      "epoch": 0.08603756246051347,
+      "grad_norm": 0.36390817165374756,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 749
+    },
+    {
+      "epoch": 0.08615243237034059,
+      "grad_norm": 0.35594916343688965,
+      "learning_rate": 0.0001,
+      "loss": 1.4535,
+      "step": 750
+    },
+    {
+      "epoch": 0.08626730228016771,
+      "grad_norm": 0.3606272339820862,
+      "learning_rate": 0.0001,
+      "loss": 1.8661,
+      "step": 751
+    },
+    {
+      "epoch": 0.08638217218999483,
+      "grad_norm": 0.3557438552379608,
+      "learning_rate": 0.0001,
+      "loss": 1.7762,
+      "step": 752
+    },
+    {
+      "epoch": 0.08649704209982195,
+      "grad_norm": 0.37106749415397644,
+      "learning_rate": 0.0001,
+      "loss": 1.913,
+      "step": 753
+    },
+    {
+      "epoch": 0.08661191200964907,
+      "grad_norm": 0.34176573157310486,
+      "learning_rate": 0.0001,
+      "loss": 1.7081,
+      "step": 754
+    },
+    {
+      "epoch": 0.08672678191947619,
+      "grad_norm": 0.32890862226486206,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 755
+    },
+    {
+      "epoch": 0.08684165182930331,
+      "grad_norm": 0.3404117822647095,
+      "learning_rate": 0.0001,
+      "loss": 1.7522,
+      "step": 756
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 0.3754931092262268,
+      "learning_rate": 0.0001,
+      "loss": 2.0646,
+      "step": 757
+    },
+    {
+      "epoch": 0.08707139164895755,
+      "grad_norm": 0.32435300946235657,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 758
+    },
+    {
+      "epoch": 0.08718626155878467,
+      "grad_norm": 0.3234858810901642,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 759
+    },
+    {
+      "epoch": 0.08730113146861179,
+      "grad_norm": 0.34800320863723755,
+      "learning_rate": 0.0001,
+      "loss": 1.7462,
+      "step": 760
+    },
+    {
+      "epoch": 0.08741600137843891,
+      "grad_norm": 0.3534897565841675,
+      "learning_rate": 0.0001,
+      "loss": 1.7654,
+      "step": 761
+    },
+    {
+      "epoch": 0.08753087128826603,
+      "grad_norm": 0.3358789384365082,
+      "learning_rate": 0.0001,
+      "loss": 1.8423,
+      "step": 762
+    },
+    {
+      "epoch": 0.08764574119809317,
+      "grad_norm": 0.35216909646987915,
+      "learning_rate": 0.0001,
+      "loss": 1.9625,
+      "step": 763
+    },
+    {
+      "epoch": 0.08776061110792029,
+      "grad_norm": 0.32955506443977356,
+      "learning_rate": 0.0001,
+      "loss": 1.7347,
+      "step": 764
+    },
+    {
+      "epoch": 0.0878754810177474,
+      "grad_norm": 0.3332022428512573,
+      "learning_rate": 0.0001,
+      "loss": 1.7732,
+      "step": 765
+    },
+    {
+      "epoch": 0.08799035092757453,
+      "grad_norm": 0.3248218894004822,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 766
+    },
+    {
+      "epoch": 0.08810522083740165,
+      "grad_norm": 0.3316574692726135,
+      "learning_rate": 0.0001,
+      "loss": 1.8211,
+      "step": 767
+    },
+    {
+      "epoch": 0.08822009074722877,
+      "grad_norm": 0.33877885341644287,
+      "learning_rate": 0.0001,
+      "loss": 1.7671,
+      "step": 768
+    },
+    {
+      "epoch": 0.08833496065705589,
+      "grad_norm": 0.34064409136772156,
+      "learning_rate": 0.0001,
+      "loss": 1.8061,
+      "step": 769
+    },
+    {
+      "epoch": 0.088449830566883,
+      "grad_norm": 0.3360152542591095,
+      "learning_rate": 0.0001,
+      "loss": 1.762,
+      "step": 770
+    },
+    {
+      "epoch": 0.08856470047671013,
+      "grad_norm": 0.3164994716644287,
+      "learning_rate": 0.0001,
+      "loss": 1.7382,
+      "step": 771
+    },
+    {
+      "epoch": 0.08867957038653725,
+      "grad_norm": 0.3785625994205475,
+      "learning_rate": 0.0001,
+      "loss": 1.9568,
+      "step": 772
+    },
+    {
+      "epoch": 0.08879444029636437,
+      "grad_norm": 0.3452891409397125,
+      "learning_rate": 0.0001,
+      "loss": 1.7799,
+      "step": 773
+    },
+    {
+      "epoch": 0.08890931020619149,
+      "grad_norm": 0.3608962595462799,
+      "learning_rate": 0.0001,
+      "loss": 1.7845,
+      "step": 774
+    },
+    {
+      "epoch": 0.0890241801160186,
+      "grad_norm": 0.3125813901424408,
+      "learning_rate": 0.0001,
+      "loss": 1.7237,
+      "step": 775
+    },
+    {
+      "epoch": 0.08913905002584573,
+      "grad_norm": 0.34034648537635803,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 776
+    },
+    {
+      "epoch": 0.08925391993567285,
+      "grad_norm": 0.31160974502563477,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 777
+    },
+    {
+      "epoch": 0.08936878984549997,
+      "grad_norm": 0.3525000512599945,
+      "learning_rate": 0.0001,
+      "loss": 1.8185,
+      "step": 778
+    },
+    {
+      "epoch": 0.08948365975532709,
+      "grad_norm": 0.3665052056312561,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 779
+    },
+    {
+      "epoch": 0.08959852966515422,
+      "grad_norm": 0.37317225337028503,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 780
+    },
+    {
+      "epoch": 0.08971339957498134,
+      "grad_norm": 0.3269886374473572,
+      "learning_rate": 0.0001,
+      "loss": 1.6282,
+      "step": 781
+    },
+    {
+      "epoch": 0.08982826948480846,
+      "grad_norm": 0.34751588106155396,
+      "learning_rate": 0.0001,
+      "loss": 1.7788,
+      "step": 782
+    },
+    {
+      "epoch": 0.08994313939463558,
+      "grad_norm": 0.35242506861686707,
+      "learning_rate": 0.0001,
+      "loss": 1.8759,
+      "step": 783
+    },
+    {
+      "epoch": 0.0900580093044627,
+      "grad_norm": 0.3438495695590973,
+      "learning_rate": 0.0001,
+      "loss": 1.6846,
+      "step": 784
+    },
+    {
+      "epoch": 0.09017287921428982,
+      "grad_norm": 0.32642510533332825,
+      "learning_rate": 0.0001,
+      "loss": 1.6951,
+      "step": 785
+    },
+    {
+      "epoch": 0.09028774912411694,
+      "grad_norm": 0.32845741510391235,
+      "learning_rate": 0.0001,
+      "loss": 1.7464,
+      "step": 786
+    },
+    {
+      "epoch": 0.09040261903394406,
+      "grad_norm": 0.36917203664779663,
+      "learning_rate": 0.0001,
+      "loss": 1.7882,
+      "step": 787
+    },
+    {
+      "epoch": 0.09051748894377118,
+      "grad_norm": 0.3328888416290283,
+      "learning_rate": 0.0001,
+      "loss": 1.6509,
+      "step": 788
+    },
+    {
+      "epoch": 0.0906323588535983,
+      "grad_norm": 0.3314300775527954,
+      "learning_rate": 0.0001,
+      "loss": 1.7402,
+      "step": 789
+    },
+    {
+      "epoch": 0.09074722876342542,
+      "grad_norm": 0.34112605452537537,
+      "learning_rate": 0.0001,
+      "loss": 1.8362,
+      "step": 790
+    },
+    {
+      "epoch": 0.09086209867325254,
+      "grad_norm": 0.3412010669708252,
+      "learning_rate": 0.0001,
+      "loss": 1.9103,
+      "step": 791
+    },
+    {
+      "epoch": 0.09097696858307966,
+      "grad_norm": 0.34920912981033325,
+      "learning_rate": 0.0001,
+      "loss": 1.756,
+      "step": 792
+    },
+    {
+      "epoch": 0.09109183849290678,
+      "grad_norm": 0.36219510436058044,
+      "learning_rate": 0.0001,
+      "loss": 1.8899,
+      "step": 793
+    },
+    {
+      "epoch": 0.0912067084027339,
+      "grad_norm": 0.34478431940078735,
+      "learning_rate": 0.0001,
+      "loss": 1.8184,
+      "step": 794
+    },
+    {
+      "epoch": 0.09132157831256102,
+      "grad_norm": 0.32101911306381226,
+      "learning_rate": 0.0001,
+      "loss": 1.8286,
+      "step": 795
+    },
+    {
+      "epoch": 0.09143644822238814,
+      "grad_norm": 0.3403749465942383,
+      "learning_rate": 0.0001,
+      "loss": 1.7323,
+      "step": 796
+    },
+    {
+      "epoch": 0.09155131813221527,
+      "grad_norm": 0.3531983196735382,
+      "learning_rate": 0.0001,
+      "loss": 1.9426,
+      "step": 797
+    },
+    {
+      "epoch": 0.09166618804204239,
+      "grad_norm": 0.33084288239479065,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 798
+    },
+    {
+      "epoch": 0.09178105795186951,
+      "grad_norm": 0.3422669768333435,
+      "learning_rate": 0.0001,
+      "loss": 1.8586,
+      "step": 799
+    },
+    {
+      "epoch": 0.09189592786169663,
+      "grad_norm": 0.333498477935791,
+      "learning_rate": 0.0001,
+      "loss": 1.6855,
+      "step": 800
+    },
+    {
+      "epoch": 0.09201079777152375,
+      "grad_norm": 0.3705277740955353,
+      "learning_rate": 0.0001,
+      "loss": 1.7961,
+      "step": 801
+    },
+    {
+      "epoch": 0.09212566768135087,
+      "grad_norm": 0.3193943500518799,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 802
+    },
+    {
+      "epoch": 0.09224053759117799,
+      "grad_norm": 0.3290367126464844,
+      "learning_rate": 0.0001,
+      "loss": 1.6078,
+      "step": 803
+    },
+    {
+      "epoch": 0.09235540750100511,
+      "grad_norm": 0.3147204518318176,
+      "learning_rate": 0.0001,
+      "loss": 1.6357,
+      "step": 804
+    },
+    {
+      "epoch": 0.09247027741083223,
+      "grad_norm": 0.33811822533607483,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 805
+    },
+    {
+      "epoch": 0.09258514732065935,
+      "grad_norm": 0.34665969014167786,
+      "learning_rate": 0.0001,
+      "loss": 1.8406,
+      "step": 806
+    },
+    {
+      "epoch": 0.09270001723048647,
+      "grad_norm": 0.367832213640213,
+      "learning_rate": 0.0001,
+      "loss": 1.8237,
+      "step": 807
+    },
+    {
+      "epoch": 0.09281488714031359,
+      "grad_norm": 0.3354632258415222,
+      "learning_rate": 0.0001,
+      "loss": 1.7281,
+      "step": 808
+    },
+    {
+      "epoch": 0.09292975705014071,
+      "grad_norm": 0.3267354965209961,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 809
+    },
+    {
+      "epoch": 0.09304462695996783,
+      "grad_norm": 0.3412437438964844,
+      "learning_rate": 0.0001,
+      "loss": 1.8719,
+      "step": 810
+    },
+    {
+      "epoch": 0.09315949686979495,
+      "grad_norm": 0.3305417597293854,
+      "learning_rate": 0.0001,
+      "loss": 1.8327,
+      "step": 811
+    },
+    {
+      "epoch": 0.09327436677962207,
+      "grad_norm": 0.3420979678630829,
+      "learning_rate": 0.0001,
+      "loss": 1.7677,
+      "step": 812
+    },
+    {
+      "epoch": 0.09338923668944919,
+      "grad_norm": 0.3424234092235565,
+      "learning_rate": 0.0001,
+      "loss": 1.8198,
+      "step": 813
+    },
+    {
+      "epoch": 0.09350410659927633,
+      "grad_norm": 0.3489876985549927,
+      "learning_rate": 0.0001,
+      "loss": 1.8778,
+      "step": 814
+    },
+    {
+      "epoch": 0.09361897650910345,
+      "grad_norm": 0.3356654942035675,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 815
+    },
+    {
+      "epoch": 0.09373384641893057,
+      "grad_norm": 0.33615395426750183,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 816
+    },
+    {
+      "epoch": 0.09384871632875769,
+      "grad_norm": 0.3469023108482361,
+      "learning_rate": 0.0001,
+      "loss": 1.7758,
+      "step": 817
+    },
+    {
+      "epoch": 0.0939635862385848,
+      "grad_norm": 0.3822448253631592,
+      "learning_rate": 0.0001,
+      "loss": 1.8071,
+      "step": 818
+    },
+    {
+      "epoch": 0.09407845614841193,
+      "grad_norm": 0.32721102237701416,
+      "learning_rate": 0.0001,
+      "loss": 1.7461,
+      "step": 819
+    },
+    {
+      "epoch": 0.09419332605823905,
+      "grad_norm": 0.3481324017047882,
+      "learning_rate": 0.0001,
+      "loss": 1.5403,
+      "step": 820
+    },
+    {
+      "epoch": 0.09430819596806617,
+      "grad_norm": 0.34965309500694275,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 821
+    },
+    {
+      "epoch": 0.09442306587789329,
+      "grad_norm": 0.3382103443145752,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 822
+    },
+    {
+      "epoch": 0.0945379357877204,
+      "grad_norm": 0.33241525292396545,
+      "learning_rate": 0.0001,
+      "loss": 1.6408,
+      "step": 823
+    },
+    {
+      "epoch": 0.09465280569754753,
+      "grad_norm": 0.34981343150138855,
+      "learning_rate": 0.0001,
+      "loss": 1.9116,
+      "step": 824
+    },
+    {
+      "epoch": 0.09476767560737465,
+      "grad_norm": 0.3368913531303406,
+      "learning_rate": 0.0001,
+      "loss": 1.8079,
+      "step": 825
+    },
+    {
+      "epoch": 0.09488254551720177,
+      "grad_norm": 0.3350062370300293,
+      "learning_rate": 0.0001,
+      "loss": 1.7194,
+      "step": 826
+    },
+    {
+      "epoch": 0.09499741542702889,
+      "grad_norm": 0.365464448928833,
+      "learning_rate": 0.0001,
+      "loss": 1.9002,
+      "step": 827
+    },
+    {
+      "epoch": 0.095112285336856,
+      "grad_norm": 0.3544370234012604,
+      "learning_rate": 0.0001,
+      "loss": 1.9647,
+      "step": 828
+    },
+    {
+      "epoch": 0.09522715524668313,
+      "grad_norm": 0.3434012532234192,
+      "learning_rate": 0.0001,
+      "loss": 1.8148,
+      "step": 829
+    },
+    {
+      "epoch": 0.09534202515651025,
+      "grad_norm": 0.33346980810165405,
+      "learning_rate": 0.0001,
+      "loss": 1.5846,
+      "step": 830
+    },
+    {
+      "epoch": 0.09545689506633737,
+      "grad_norm": 0.3305695652961731,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 831
+    },
+    {
+      "epoch": 0.0955717649761645,
+      "grad_norm": 0.32634860277175903,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 832
+    },
+    {
+      "epoch": 0.09568663488599162,
+      "grad_norm": 0.3320368528366089,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 833
+    },
+    {
+      "epoch": 0.09580150479581874,
+      "grad_norm": 0.33831140398979187,
+      "learning_rate": 0.0001,
+      "loss": 1.9525,
+      "step": 834
+    },
+    {
+      "epoch": 0.09591637470564586,
+      "grad_norm": 0.37599530816078186,
+      "learning_rate": 0.0001,
+      "loss": 1.8062,
+      "step": 835
+    },
+    {
+      "epoch": 0.09603124461547298,
+      "grad_norm": 0.3151451647281647,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 836
+    },
+    {
+      "epoch": 0.0961461145253001,
+      "grad_norm": 0.3700610101222992,
+      "learning_rate": 0.0001,
+      "loss": 1.9294,
+      "step": 837
+    },
+    {
+      "epoch": 0.09626098443512722,
+      "grad_norm": 0.3530856668949127,
+      "learning_rate": 0.0001,
+      "loss": 1.9053,
+      "step": 838
+    },
+    {
+      "epoch": 0.09637585434495434,
+      "grad_norm": 0.3287757933139801,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 839
+    },
+    {
+      "epoch": 0.09649072425478146,
+      "grad_norm": 0.33994293212890625,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 840
+    },
+    {
+      "epoch": 0.09660559416460858,
+      "grad_norm": 0.3461272418498993,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 841
+    },
+    {
+      "epoch": 0.0967204640744357,
+      "grad_norm": 0.3444170653820038,
+      "learning_rate": 0.0001,
+      "loss": 1.6429,
+      "step": 842
+    },
+    {
+      "epoch": 0.09683533398426282,
+      "grad_norm": 0.33912792801856995,
+      "learning_rate": 0.0001,
+      "loss": 1.7618,
+      "step": 843
+    },
+    {
+      "epoch": 0.09695020389408994,
+      "grad_norm": 0.36995047330856323,
+      "learning_rate": 0.0001,
+      "loss": 1.7405,
+      "step": 844
+    },
+    {
+      "epoch": 0.09706507380391706,
+      "grad_norm": 0.318174809217453,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 845
+    },
+    {
+      "epoch": 0.09717994371374418,
+      "grad_norm": 0.340555876493454,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 846
+    },
+    {
+      "epoch": 0.0972948136235713,
+      "grad_norm": 0.3448858857154846,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 847
+    },
+    {
+      "epoch": 0.09740968353339842,
+      "grad_norm": 0.34647923707962036,
+      "learning_rate": 0.0001,
+      "loss": 1.8009,
+      "step": 848
+    },
+    {
+      "epoch": 0.09752455344322555,
+      "grad_norm": 0.3762481212615967,
+      "learning_rate": 0.0001,
+      "loss": 1.6759,
+      "step": 849
+    },
+    {
+      "epoch": 0.09763942335305267,
+      "grad_norm": 0.3799351751804352,
+      "learning_rate": 0.0001,
+      "loss": 1.9706,
+      "step": 850
+    },
+    {
+      "epoch": 0.0977542932628798,
+      "grad_norm": 0.33727753162384033,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 851
+    },
+    {
+      "epoch": 0.09786916317270691,
+      "grad_norm": 0.3406698703765869,
+      "learning_rate": 0.0001,
+      "loss": 1.7837,
+      "step": 852
+    },
+    {
+      "epoch": 0.09798403308253403,
+      "grad_norm": 0.3939476013183594,
+      "learning_rate": 0.0001,
+      "loss": 1.6971,
+      "step": 853
+    },
+    {
+      "epoch": 0.09809890299236115,
+      "grad_norm": 0.3444227874279022,
+      "learning_rate": 0.0001,
+      "loss": 1.5426,
+      "step": 854
+    },
+    {
+      "epoch": 0.09821377290218827,
+      "grad_norm": 0.38864031434059143,
+      "learning_rate": 0.0001,
+      "loss": 1.7997,
+      "step": 855
+    },
+    {
+      "epoch": 0.0983286428120154,
+      "grad_norm": 0.36266061663627625,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 856
+    },
+    {
+      "epoch": 0.09844351272184251,
+      "grad_norm": 0.3754750192165375,
+      "learning_rate": 0.0001,
+      "loss": 1.922,
+      "step": 857
+    },
+    {
+      "epoch": 0.09855838263166963,
+      "grad_norm": 0.3244558274745941,
+      "learning_rate": 0.0001,
+      "loss": 1.6094,
+      "step": 858
+    },
+    {
+      "epoch": 0.09867325254149675,
+      "grad_norm": 0.3381154537200928,
+      "learning_rate": 0.0001,
+      "loss": 1.5764,
+      "step": 859
+    },
+    {
+      "epoch": 0.09878812245132387,
+      "grad_norm": 0.37016791105270386,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 860
+    },
+    {
+      "epoch": 0.098902992361151,
+      "grad_norm": 0.3377515971660614,
+      "learning_rate": 0.0001,
+      "loss": 1.7457,
+      "step": 861
+    },
+    {
+      "epoch": 0.09901786227097811,
+      "grad_norm": 0.3582906126976013,
+      "learning_rate": 0.0001,
+      "loss": 1.7201,
+      "step": 862
+    },
+    {
+      "epoch": 0.09913273218080523,
+      "grad_norm": 0.39218735694885254,
+      "learning_rate": 0.0001,
+      "loss": 2.1357,
+      "step": 863
+    },
+    {
+      "epoch": 0.09924760209063235,
+      "grad_norm": 0.3586115837097168,
+      "learning_rate": 0.0001,
+      "loss": 1.7406,
+      "step": 864
+    },
+    {
+      "epoch": 0.09936247200045947,
+      "grad_norm": 0.3303786516189575,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 865
+    },
+    {
+      "epoch": 0.0994773419102866,
+      "grad_norm": 0.34086883068084717,
+      "learning_rate": 0.0001,
+      "loss": 1.6499,
+      "step": 866
+    },
+    {
+      "epoch": 0.09959221182011373,
+      "grad_norm": 0.3244491219520569,
+      "learning_rate": 0.0001,
+      "loss": 1.621,
+      "step": 867
+    },
+    {
+      "epoch": 0.09970708172994085,
+      "grad_norm": 0.3519229292869568,
+      "learning_rate": 0.0001,
+      "loss": 1.7324,
+      "step": 868
+    },
+    {
+      "epoch": 0.09982195163976797,
+      "grad_norm": 0.3447989523410797,
+      "learning_rate": 0.0001,
+      "loss": 1.7337,
+      "step": 869
+    },
+    {
+      "epoch": 0.09993682154959509,
+      "grad_norm": 0.3334493935108185,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 870
+    },
+    {
+      "epoch": 0.1000516914594222,
+      "grad_norm": 0.3429296612739563,
+      "learning_rate": 0.0001,
+      "loss": 1.7162,
+      "step": 871
+    },
+    {
+      "epoch": 0.10016656136924933,
+      "grad_norm": 0.3839278519153595,
+      "learning_rate": 0.0001,
+      "loss": 1.7534,
+      "step": 872
+    },
+    {
+      "epoch": 0.10028143127907645,
+      "grad_norm": 0.33588939905166626,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 873
+    },
+    {
+      "epoch": 0.10039630118890357,
+      "grad_norm": 0.34705570340156555,
+      "learning_rate": 0.0001,
+      "loss": 1.841,
+      "step": 874
+    },
+    {
+      "epoch": 0.10051117109873069,
+      "grad_norm": 0.36733031272888184,
+      "learning_rate": 0.0001,
+      "loss": 1.852,
+      "step": 875
+    },
+    {
+      "epoch": 0.1006260410085578,
+      "grad_norm": 0.3770068883895874,
+      "learning_rate": 0.0001,
+      "loss": 1.6318,
+      "step": 876
+    },
+    {
+      "epoch": 0.10074091091838493,
+      "grad_norm": 0.3212607800960541,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 877
+    },
+    {
+      "epoch": 0.10085578082821205,
+      "grad_norm": 0.36928656697273254,
+      "learning_rate": 0.0001,
+      "loss": 1.7392,
+      "step": 878
+    },
+    {
+      "epoch": 0.10097065073803917,
+      "grad_norm": 0.34793323278427124,
+      "learning_rate": 0.0001,
+      "loss": 1.7446,
+      "step": 879
+    },
+    {
+      "epoch": 0.10108552064786629,
+      "grad_norm": 0.3587128221988678,
+      "learning_rate": 0.0001,
+      "loss": 1.8248,
+      "step": 880
+    },
+    {
+      "epoch": 0.1012003905576934,
+      "grad_norm": 0.37394312024116516,
+      "learning_rate": 0.0001,
+      "loss": 1.8343,
+      "step": 881
+    },
+    {
+      "epoch": 0.10131526046752053,
+      "grad_norm": 0.3411267101764679,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 882
+    },
+    {
+      "epoch": 0.10143013037734766,
+      "grad_norm": 0.3414922058582306,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 883
+    },
+    {
+      "epoch": 0.10154500028717478,
+      "grad_norm": 0.3430386781692505,
+      "learning_rate": 0.0001,
+      "loss": 1.6358,
+      "step": 884
+    },
+    {
+      "epoch": 0.1016598701970019,
+      "grad_norm": 0.38398411870002747,
+      "learning_rate": 0.0001,
+      "loss": 1.8655,
+      "step": 885
+    },
+    {
+      "epoch": 0.10177474010682902,
+      "grad_norm": 0.35513511300086975,
+      "learning_rate": 0.0001,
+      "loss": 1.7441,
+      "step": 886
+    },
+    {
+      "epoch": 0.10188961001665614,
+      "grad_norm": 0.37063589692115784,
+      "learning_rate": 0.0001,
+      "loss": 1.8402,
+      "step": 887
+    },
+    {
+      "epoch": 0.10200447992648326,
+      "grad_norm": 0.3473655581474304,
+      "learning_rate": 0.0001,
+      "loss": 1.7719,
+      "step": 888
+    },
+    {
+      "epoch": 0.10211934983631038,
+      "grad_norm": 0.3341835141181946,
+      "learning_rate": 0.0001,
+      "loss": 1.9165,
+      "step": 889
+    },
+    {
+      "epoch": 0.1022342197461375,
+      "grad_norm": 0.33776983618736267,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 890
+    },
+    {
+      "epoch": 0.10234908965596462,
+      "grad_norm": 0.34027940034866333,
+      "learning_rate": 0.0001,
+      "loss": 1.6075,
+      "step": 891
+    },
+    {
+      "epoch": 0.10246395956579174,
+      "grad_norm": 0.3332427442073822,
+      "learning_rate": 0.0001,
+      "loss": 1.6832,
+      "step": 892
+    },
+    {
+      "epoch": 0.10257882947561886,
+      "grad_norm": 0.3453330099582672,
+      "learning_rate": 0.0001,
+      "loss": 1.8543,
+      "step": 893
+    },
+    {
+      "epoch": 0.10269369938544598,
+      "grad_norm": 0.3439124524593353,
+      "learning_rate": 0.0001,
+      "loss": 1.8294,
+      "step": 894
+    },
+    {
+      "epoch": 0.1028085692952731,
+      "grad_norm": 0.3373720645904541,
+      "learning_rate": 0.0001,
+      "loss": 1.816,
+      "step": 895
+    },
+    {
+      "epoch": 0.10292343920510022,
+      "grad_norm": 0.3701860308647156,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 896
+    },
+    {
+      "epoch": 0.10303830911492734,
+      "grad_norm": 0.3510701358318329,
+      "learning_rate": 0.0001,
+      "loss": 1.8181,
+      "step": 897
+    },
+    {
+      "epoch": 0.10315317902475446,
+      "grad_norm": 0.383064329624176,
+      "learning_rate": 0.0001,
+      "loss": 1.7649,
+      "step": 898
+    },
+    {
+      "epoch": 0.10326804893458158,
+      "grad_norm": 0.33695435523986816,
+      "learning_rate": 0.0001,
+      "loss": 1.7667,
+      "step": 899
+    },
+    {
+      "epoch": 0.10338291884440871,
+      "grad_norm": 0.32354670763015747,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 900
+    },
+    {
+      "epoch": 0.10349778875423583,
+      "grad_norm": 0.3468325138092041,
+      "learning_rate": 0.0001,
+      "loss": 1.897,
+      "step": 901
+    },
+    {
+      "epoch": 0.10361265866406295,
+      "grad_norm": 0.33674290776252747,
+      "learning_rate": 0.0001,
+      "loss": 1.7802,
+      "step": 902
+    },
+    {
+      "epoch": 0.10372752857389007,
+      "grad_norm": 0.3250616490840912,
+      "learning_rate": 0.0001,
+      "loss": 1.7262,
+      "step": 903
+    },
+    {
+      "epoch": 0.1038423984837172,
+      "grad_norm": 0.3321487009525299,
+      "learning_rate": 0.0001,
+      "loss": 1.7408,
+      "step": 904
+    },
+    {
+      "epoch": 0.10395726839354431,
+      "grad_norm": 0.3137947916984558,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 905
+    },
+    {
+      "epoch": 0.10407213830337143,
+      "grad_norm": 0.35805246233940125,
+      "learning_rate": 0.0001,
+      "loss": 1.7689,
+      "step": 906
+    },
+    {
+      "epoch": 0.10418700821319855,
+      "grad_norm": 0.35889506340026855,
+      "learning_rate": 0.0001,
+      "loss": 1.8557,
+      "step": 907
+    },
+    {
+      "epoch": 0.10430187812302567,
+      "grad_norm": 0.3665921092033386,
+      "learning_rate": 0.0001,
+      "loss": 1.8241,
+      "step": 908
+    },
+    {
+      "epoch": 0.1044167480328528,
+      "grad_norm": 0.32144057750701904,
+      "learning_rate": 0.0001,
+      "loss": 1.6686,
+      "step": 909
+    },
+    {
+      "epoch": 0.10453161794267991,
+      "grad_norm": 0.3347714841365814,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 910
+    },
+    {
+      "epoch": 0.10464648785250703,
+      "grad_norm": 0.3488980233669281,
+      "learning_rate": 0.0001,
+      "loss": 1.8368,
+      "step": 911
+    },
+    {
+      "epoch": 0.10476135776233415,
+      "grad_norm": 0.3683393895626068,
+      "learning_rate": 0.0001,
+      "loss": 1.8511,
+      "step": 912
+    },
+    {
+      "epoch": 0.10487622767216127,
+      "grad_norm": 0.36709287762641907,
+      "learning_rate": 0.0001,
+      "loss": 1.8485,
+      "step": 913
+    },
+    {
+      "epoch": 0.1049910975819884,
+      "grad_norm": 0.3464992642402649,
+      "learning_rate": 0.0001,
+      "loss": 1.8211,
+      "step": 914
+    },
+    {
+      "epoch": 0.10510596749181551,
+      "grad_norm": 0.3259120285511017,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 915
+    },
+    {
+      "epoch": 0.10522083740164263,
+      "grad_norm": 0.3253241777420044,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 916
+    },
+    {
+      "epoch": 0.10533570731146977,
+      "grad_norm": 0.3512059152126312,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 917
+    },
+    {
+      "epoch": 0.10545057722129689,
+      "grad_norm": 0.34296149015426636,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 918
+    },
+    {
+      "epoch": 0.10556544713112401,
+      "grad_norm": 0.34532561898231506,
+      "learning_rate": 0.0001,
+      "loss": 1.7067,
+      "step": 919
+    },
+    {
+      "epoch": 0.10568031704095113,
+      "grad_norm": 0.4030701816082001,
+      "learning_rate": 0.0001,
+      "loss": 1.8292,
+      "step": 920
+    },
+    {
+      "epoch": 0.10579518695077825,
+      "grad_norm": 0.3220067322254181,
+      "learning_rate": 0.0001,
+      "loss": 1.7312,
+      "step": 921
+    },
+    {
+      "epoch": 0.10591005686060537,
+      "grad_norm": 0.35036081075668335,
+      "learning_rate": 0.0001,
+      "loss": 1.9524,
+      "step": 922
+    },
+    {
+      "epoch": 0.10602492677043249,
+      "grad_norm": 0.34477829933166504,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 923
+    },
+    {
+      "epoch": 0.10613979668025961,
+      "grad_norm": 0.3312055468559265,
+      "learning_rate": 0.0001,
+      "loss": 1.7349,
+      "step": 924
+    },
+    {
+      "epoch": 0.10625466659008673,
+      "grad_norm": 0.3699291944503784,
+      "learning_rate": 0.0001,
+      "loss": 1.8016,
+      "step": 925
+    },
+    {
+      "epoch": 0.10636953649991385,
+      "grad_norm": 0.35141250491142273,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 926
+    },
+    {
+      "epoch": 0.10648440640974097,
+      "grad_norm": 0.3546350300312042,
+      "learning_rate": 0.0001,
+      "loss": 1.792,
+      "step": 927
+    },
+    {
+      "epoch": 0.10659927631956809,
+      "grad_norm": 0.32933250069618225,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 928
+    },
+    {
+      "epoch": 0.10671414622939521,
+      "grad_norm": 0.31342101097106934,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 929
+    },
+    {
+      "epoch": 0.10682901613922233,
+      "grad_norm": 0.3823557496070862,
+      "learning_rate": 0.0001,
+      "loss": 1.7735,
+      "step": 930
+    },
+    {
+      "epoch": 0.10694388604904945,
+      "grad_norm": 0.37668943405151367,
+      "learning_rate": 0.0001,
+      "loss": 1.7153,
+      "step": 931
+    },
+    {
+      "epoch": 0.10705875595887657,
+      "grad_norm": 0.3394898772239685,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 932
+    },
+    {
+      "epoch": 0.10717362586870369,
+      "grad_norm": 0.3595438301563263,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 933
+    },
+    {
+      "epoch": 0.10728849577853082,
+      "grad_norm": 0.3278176486492157,
+      "learning_rate": 0.0001,
+      "loss": 1.4859,
+      "step": 934
+    },
+    {
+      "epoch": 0.10740336568835794,
+      "grad_norm": 0.35469111800193787,
+      "learning_rate": 0.0001,
+      "loss": 1.7812,
+      "step": 935
+    },
+    {
+      "epoch": 0.10751823559818506,
+      "grad_norm": 0.35269084572792053,
+      "learning_rate": 0.0001,
+      "loss": 1.7469,
+      "step": 936
+    },
+    {
+      "epoch": 0.10763310550801218,
+      "grad_norm": 0.3482814133167267,
+      "learning_rate": 0.0001,
+      "loss": 1.4346,
+      "step": 937
+    },
+    {
+      "epoch": 0.1077479754178393,
+      "grad_norm": 0.3587512671947479,
+      "learning_rate": 0.0001,
+      "loss": 1.433,
+      "step": 938
+    },
+    {
+      "epoch": 0.10786284532766642,
+      "grad_norm": 0.3574024736881256,
+      "learning_rate": 0.0001,
+      "loss": 1.9325,
+      "step": 939
+    },
+    {
+      "epoch": 0.10797771523749354,
+      "grad_norm": 0.3392809331417084,
+      "learning_rate": 0.0001,
+      "loss": 1.7022,
+      "step": 940
+    },
+    {
+      "epoch": 0.10809258514732066,
+      "grad_norm": 0.3584175109863281,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 941
+    },
+    {
+      "epoch": 0.10820745505714778,
+      "grad_norm": 0.35646557807922363,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 942
+    },
+    {
+      "epoch": 0.1083223249669749,
+      "grad_norm": 0.3471522331237793,
+      "learning_rate": 0.0001,
+      "loss": 1.7296,
+      "step": 943
+    },
+    {
+      "epoch": 0.10843719487680202,
+      "grad_norm": 0.36520814895629883,
+      "learning_rate": 0.0001,
+      "loss": 1.7646,
+      "step": 944
+    },
+    {
+      "epoch": 0.10855206478662914,
+      "grad_norm": 0.3618639409542084,
+      "learning_rate": 0.0001,
+      "loss": 1.879,
+      "step": 945
+    },
+    {
+      "epoch": 0.10866693469645626,
+      "grad_norm": 0.3887125849723816,
+      "learning_rate": 0.0001,
+      "loss": 1.9636,
+      "step": 946
+    },
+    {
+      "epoch": 0.10878180460628338,
+      "grad_norm": 0.36584070324897766,
+      "learning_rate": 0.0001,
+      "loss": 1.8166,
+      "step": 947
+    },
+    {
+      "epoch": 0.1088966745161105,
+      "grad_norm": 0.4011528491973877,
+      "learning_rate": 0.0001,
+      "loss": 1.8934,
+      "step": 948
+    },
+    {
+      "epoch": 0.10901154442593762,
+      "grad_norm": 0.35552287101745605,
+      "learning_rate": 0.0001,
+      "loss": 1.7606,
+      "step": 949
+    },
+    {
+      "epoch": 0.10912641433576474,
+      "grad_norm": 0.3487032949924469,
+      "learning_rate": 0.0001,
+      "loss": 1.6729,
+      "step": 950
+    },
+    {
+      "epoch": 0.10924128424559186,
+      "grad_norm": 0.34331613779067993,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 951
+    },
+    {
+      "epoch": 0.109356154155419,
+      "grad_norm": 0.35324063897132874,
+      "learning_rate": 0.0001,
+      "loss": 1.7534,
+      "step": 952
+    },
+    {
+      "epoch": 0.10947102406524611,
+      "grad_norm": 0.33275148272514343,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 953
+    },
+    {
+      "epoch": 0.10958589397507323,
+      "grad_norm": 0.329111784696579,
+      "learning_rate": 0.0001,
+      "loss": 1.682,
+      "step": 954
+    },
+    {
+      "epoch": 0.10970076388490035,
+      "grad_norm": 0.3550437092781067,
+      "learning_rate": 0.0001,
+      "loss": 1.8724,
+      "step": 955
+    },
+    {
+      "epoch": 0.10981563379472747,
+      "grad_norm": 0.3454737663269043,
+      "learning_rate": 0.0001,
+      "loss": 1.7989,
+      "step": 956
+    },
+    {
+      "epoch": 0.1099305037045546,
+      "grad_norm": 0.3605286777019501,
+      "learning_rate": 0.0001,
+      "loss": 1.7102,
+      "step": 957
+    },
+    {
+      "epoch": 0.11004537361438171,
+      "grad_norm": 0.3494301438331604,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 958
+    },
+    {
+      "epoch": 0.11016024352420883,
+      "grad_norm": 0.3383633494377136,
+      "learning_rate": 0.0001,
+      "loss": 1.8941,
+      "step": 959
+    },
+    {
+      "epoch": 0.11027511343403595,
+      "grad_norm": 0.3342844247817993,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 960
+    },
+    {
+      "epoch": 0.11038998334386307,
+      "grad_norm": 0.3412388861179352,
+      "learning_rate": 0.0001,
+      "loss": 1.6875,
+      "step": 961
+    },
+    {
+      "epoch": 0.1105048532536902,
+      "grad_norm": 0.34491872787475586,
+      "learning_rate": 0.0001,
+      "loss": 1.6936,
+      "step": 962
+    },
+    {
+      "epoch": 0.11061972316351731,
+      "grad_norm": 0.3250080943107605,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 963
+    },
+    {
+      "epoch": 0.11073459307334443,
+      "grad_norm": 0.3618282973766327,
+      "learning_rate": 0.0001,
+      "loss": 1.7687,
+      "step": 964
+    },
+    {
+      "epoch": 0.11084946298317155,
+      "grad_norm": 0.37689584493637085,
+      "learning_rate": 0.0001,
+      "loss": 1.6914,
+      "step": 965
+    },
+    {
+      "epoch": 0.11096433289299867,
+      "grad_norm": 0.35625600814819336,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 966
+    },
+    {
+      "epoch": 0.1110792028028258,
+      "grad_norm": 0.36949923634529114,
+      "learning_rate": 0.0001,
+      "loss": 1.8479,
+      "step": 967
+    },
+    {
+      "epoch": 0.11119407271265291,
+      "grad_norm": 0.32159894704818726,
+      "learning_rate": 0.0001,
+      "loss": 1.7816,
+      "step": 968
+    },
+    {
+      "epoch": 0.11130894262248005,
+      "grad_norm": 0.35336339473724365,
+      "learning_rate": 0.0001,
+      "loss": 1.7206,
+      "step": 969
+    },
+    {
+      "epoch": 0.11142381253230717,
+      "grad_norm": 0.3411477506160736,
+      "learning_rate": 0.0001,
+      "loss": 1.7505,
+      "step": 970
+    },
+    {
+      "epoch": 0.11153868244213429,
+      "grad_norm": 0.37740951776504517,
+      "learning_rate": 0.0001,
+      "loss": 1.8982,
+      "step": 971
+    },
+    {
+      "epoch": 0.11165355235196141,
+      "grad_norm": 0.32587164640426636,
+      "learning_rate": 0.0001,
+      "loss": 1.7459,
+      "step": 972
+    },
+    {
+      "epoch": 0.11176842226178853,
+      "grad_norm": 0.40314409136772156,
+      "learning_rate": 0.0001,
+      "loss": 1.752,
+      "step": 973
+    },
+    {
+      "epoch": 0.11188329217161565,
+      "grad_norm": 0.3304173946380615,
+      "learning_rate": 0.0001,
+      "loss": 1.5715,
+      "step": 974
+    },
+    {
+      "epoch": 0.11199816208144277,
+      "grad_norm": 0.3680497407913208,
+      "learning_rate": 0.0001,
+      "loss": 1.6852,
+      "step": 975
+    },
+    {
+      "epoch": 0.11211303199126989,
+      "grad_norm": 0.3391849100589752,
+      "learning_rate": 0.0001,
+      "loss": 1.7026,
+      "step": 976
+    },
+    {
+      "epoch": 0.11222790190109701,
+      "grad_norm": 0.37164103984832764,
+      "learning_rate": 0.0001,
+      "loss": 1.847,
+      "step": 977
+    },
+    {
+      "epoch": 0.11234277181092413,
+      "grad_norm": 0.3532995581626892,
+      "learning_rate": 0.0001,
+      "loss": 1.8528,
+      "step": 978
+    },
+    {
+      "epoch": 0.11245764172075125,
+      "grad_norm": 0.35095518827438354,
+      "learning_rate": 0.0001,
+      "loss": 1.8725,
+      "step": 979
+    },
+    {
+      "epoch": 0.11257251163057837,
+      "grad_norm": 0.3317796289920807,
+      "learning_rate": 0.0001,
+      "loss": 1.6293,
+      "step": 980
+    },
+    {
+      "epoch": 0.11268738154040549,
+      "grad_norm": 0.3477189838886261,
+      "learning_rate": 0.0001,
+      "loss": 1.6846,
+      "step": 981
+    },
+    {
+      "epoch": 0.11280225145023261,
+      "grad_norm": 0.34023842215538025,
+      "learning_rate": 0.0001,
+      "loss": 1.6921,
+      "step": 982
+    },
+    {
+      "epoch": 0.11291712136005973,
+      "grad_norm": 0.3701896369457245,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 983
+    },
+    {
+      "epoch": 0.11303199126988685,
+      "grad_norm": 0.3711189329624176,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 984
+    },
+    {
+      "epoch": 0.11314686117971397,
+      "grad_norm": 0.35467204451560974,
+      "learning_rate": 0.0001,
+      "loss": 1.7369,
+      "step": 985
+    },
+    {
+      "epoch": 0.1132617310895411,
+      "grad_norm": 0.3388875722885132,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 986
+    },
+    {
+      "epoch": 0.11337660099936822,
+      "grad_norm": 0.34032291173934937,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 987
+    },
+    {
+      "epoch": 0.11349147090919534,
+      "grad_norm": 0.3326657712459564,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 988
+    },
+    {
+      "epoch": 0.11360634081902246,
+      "grad_norm": 0.3543054163455963,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 989
+    },
+    {
+      "epoch": 0.11372121072884958,
+      "grad_norm": 0.3578903377056122,
+      "learning_rate": 0.0001,
+      "loss": 1.7261,
+      "step": 990
+    },
+    {
+      "epoch": 0.1138360806386767,
+      "grad_norm": 0.3672734498977661,
+      "learning_rate": 0.0001,
+      "loss": 1.8798,
+      "step": 991
+    },
+    {
+      "epoch": 0.11395095054850382,
+      "grad_norm": 0.3719589114189148,
+      "learning_rate": 0.0001,
+      "loss": 1.6408,
+      "step": 992
+    },
+    {
+      "epoch": 0.11406582045833094,
+      "grad_norm": 0.3530134856700897,
+      "learning_rate": 0.0001,
+      "loss": 1.7737,
+      "step": 993
+    },
+    {
+      "epoch": 0.11418069036815806,
+      "grad_norm": 0.371442586183548,
+      "learning_rate": 0.0001,
+      "loss": 1.5202,
+      "step": 994
+    },
+    {
+      "epoch": 0.11429556027798518,
+      "grad_norm": 0.33966636657714844,
+      "learning_rate": 0.0001,
+      "loss": 1.8347,
+      "step": 995
+    },
+    {
+      "epoch": 0.1144104301878123,
+      "grad_norm": 0.3425755202770233,
+      "learning_rate": 0.0001,
+      "loss": 1.7236,
+      "step": 996
+    },
+    {
+      "epoch": 0.11452530009763942,
+      "grad_norm": 0.3630322515964508,
+      "learning_rate": 0.0001,
+      "loss": 1.6771,
+      "step": 997
+    },
+    {
+      "epoch": 0.11464017000746654,
+      "grad_norm": 0.31861042976379395,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 998
+    },
+    {
+      "epoch": 0.11475503991729366,
+      "grad_norm": 0.34319064021110535,
+      "learning_rate": 0.0001,
+      "loss": 1.7079,
+      "step": 999
+    },
+    {
+      "epoch": 0.11486990982712078,
+      "grad_norm": 0.33769530057907104,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1149847797369479,
+      "grad_norm": 0.3288537859916687,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 1001
+    },
+    {
+      "epoch": 0.11509964964677502,
+      "grad_norm": 0.41031739115715027,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 1002
+    },
+    {
+      "epoch": 0.11521451955660215,
+      "grad_norm": 0.33957231044769287,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 1003
+    },
+    {
+      "epoch": 0.11532938946642927,
+      "grad_norm": 0.35768911242485046,
+      "learning_rate": 0.0001,
+      "loss": 1.8097,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1154442593762564,
+      "grad_norm": 0.3527219593524933,
+      "learning_rate": 0.0001,
+      "loss": 1.6337,
+      "step": 1005
+    },
+    {
+      "epoch": 0.11555912928608351,
+      "grad_norm": 0.3403669595718384,
+      "learning_rate": 0.0001,
+      "loss": 1.7182,
+      "step": 1006
+    },
+    {
+      "epoch": 0.11567399919591063,
+      "grad_norm": 0.33553534746170044,
+      "learning_rate": 0.0001,
+      "loss": 1.7559,
+      "step": 1007
+    },
+    {
+      "epoch": 0.11578886910573775,
+      "grad_norm": 0.39612364768981934,
+      "learning_rate": 0.0001,
+      "loss": 2.1183,
+      "step": 1008
+    },
+    {
+      "epoch": 0.11590373901556487,
+      "grad_norm": 0.3312492072582245,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 1009
+    },
+    {
+      "epoch": 0.116018608925392,
+      "grad_norm": 0.35512349009513855,
+      "learning_rate": 0.0001,
+      "loss": 1.8648,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11613347883521911,
+      "grad_norm": 0.34850746393203735,
+      "learning_rate": 0.0001,
+      "loss": 1.7185,
+      "step": 1011
+    },
+    {
+      "epoch": 0.11624834874504623,
+      "grad_norm": 0.34042853116989136,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 1012
+    },
+    {
+      "epoch": 0.11636321865487335,
+      "grad_norm": 0.3678249716758728,
+      "learning_rate": 0.0001,
+      "loss": 1.9585,
+      "step": 1013
+    },
+    {
+      "epoch": 0.11647808856470047,
+      "grad_norm": 0.38127046823501587,
+      "learning_rate": 0.0001,
+      "loss": 1.7736,
+      "step": 1014
+    },
+    {
+      "epoch": 0.1165929584745276,
+      "grad_norm": 0.33116331696510315,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11670782838435471,
+      "grad_norm": 0.3540381193161011,
+      "learning_rate": 0.0001,
+      "loss": 1.666,
+      "step": 1016
+    },
+    {
+      "epoch": 0.11682269829418183,
+      "grad_norm": 0.34696200489997864,
+      "learning_rate": 0.0001,
+      "loss": 1.88,
+      "step": 1017
+    },
+    {
+      "epoch": 0.11693756820400895,
+      "grad_norm": 0.3833637833595276,
+      "learning_rate": 0.0001,
+      "loss": 1.9577,
+      "step": 1018
+    },
+    {
+      "epoch": 0.11705243811383607,
+      "grad_norm": 0.34638655185699463,
+      "learning_rate": 0.0001,
+      "loss": 1.7792,
+      "step": 1019
+    },
+    {
+      "epoch": 0.11716730802366321,
+      "grad_norm": 0.3427535593509674,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11728217793349033,
+      "grad_norm": 0.3441438376903534,
+      "learning_rate": 0.0001,
+      "loss": 1.7116,
+      "step": 1021
+    },
+    {
+      "epoch": 0.11739704784331745,
+      "grad_norm": 0.3629339039325714,
+      "learning_rate": 0.0001,
+      "loss": 1.7716,
+      "step": 1022
+    },
+    {
+      "epoch": 0.11751191775314457,
+      "grad_norm": 0.35050028562545776,
+      "learning_rate": 0.0001,
+      "loss": 1.7887,
+      "step": 1023
+    },
+    {
+      "epoch": 0.11762678766297169,
+      "grad_norm": 0.3770378828048706,
+      "learning_rate": 0.0001,
+      "loss": 1.9182,
+      "step": 1024
+    },
+    {
+      "epoch": 0.11774165757279881,
+      "grad_norm": 0.3678201138973236,
+      "learning_rate": 0.0001,
+      "loss": 1.7746,
+      "step": 1025
+    },
+    {
+      "epoch": 0.11785652748262593,
+      "grad_norm": 0.3541167080402374,
+      "learning_rate": 0.0001,
+      "loss": 1.7004,
+      "step": 1026
+    },
+    {
+      "epoch": 0.11797139739245305,
+      "grad_norm": 0.34430310130119324,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 1027
+    },
+    {
+      "epoch": 0.11808626730228017,
+      "grad_norm": 0.35517948865890503,
+      "learning_rate": 0.0001,
+      "loss": 1.8264,
+      "step": 1028
+    },
+    {
+      "epoch": 0.11820113721210729,
+      "grad_norm": 0.35066622495651245,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 1029
+    },
+    {
+      "epoch": 0.11831600712193441,
+      "grad_norm": 0.3344581425189972,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11843087703176153,
+      "grad_norm": 0.3826616406440735,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 1031
+    },
+    {
+      "epoch": 0.11854574694158865,
+      "grad_norm": 0.3551161587238312,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 1032
+    },
+    {
+      "epoch": 0.11866061685141577,
+      "grad_norm": 0.3458511233329773,
+      "learning_rate": 0.0001,
+      "loss": 1.7166,
+      "step": 1033
+    },
+    {
+      "epoch": 0.11877548676124289,
+      "grad_norm": 0.38753172755241394,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 1034
+    },
+    {
+      "epoch": 0.11889035667107001,
+      "grad_norm": 0.36152660846710205,
+      "learning_rate": 0.0001,
+      "loss": 1.7422,
+      "step": 1035
+    },
+    {
+      "epoch": 0.11900522658089713,
+      "grad_norm": 0.3490775525569916,
+      "learning_rate": 0.0001,
+      "loss": 1.7268,
+      "step": 1036
+    },
+    {
+      "epoch": 0.11912009649072426,
+      "grad_norm": 0.3666988015174866,
+      "learning_rate": 0.0001,
+      "loss": 1.7706,
+      "step": 1037
+    },
+    {
+      "epoch": 0.11923496640055138,
+      "grad_norm": 0.3366585671901703,
+      "learning_rate": 0.0001,
+      "loss": 1.5956,
+      "step": 1038
+    },
+    {
+      "epoch": 0.1193498363103785,
+      "grad_norm": 0.36800581216812134,
+      "learning_rate": 0.0001,
+      "loss": 1.7542,
+      "step": 1039
+    },
+    {
+      "epoch": 0.11946470622020562,
+      "grad_norm": 0.36081400513648987,
+      "learning_rate": 0.0001,
+      "loss": 1.882,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11957957613003274,
+      "grad_norm": 0.3410419821739197,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 1041
+    },
+    {
+      "epoch": 0.11969444603985986,
+      "grad_norm": 0.3386545777320862,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 1042
+    },
+    {
+      "epoch": 0.11980931594968698,
+      "grad_norm": 0.31798994541168213,
+      "learning_rate": 0.0001,
+      "loss": 1.5473,
+      "step": 1043
+    },
+    {
+      "epoch": 0.1199241858595141,
+      "grad_norm": 0.3879840672016144,
+      "learning_rate": 0.0001,
+      "loss": 1.9508,
+      "step": 1044
+    },
+    {
+      "epoch": 0.12003905576934122,
+      "grad_norm": 0.3498954176902771,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 1045
+    },
+    {
+      "epoch": 0.12015392567916834,
+      "grad_norm": 0.3412124216556549,
+      "learning_rate": 0.0001,
+      "loss": 1.6106,
+      "step": 1046
+    },
+    {
+      "epoch": 0.12026879558899546,
+      "grad_norm": 0.35735565423965454,
+      "learning_rate": 0.0001,
+      "loss": 1.7393,
+      "step": 1047
+    },
+    {
+      "epoch": 0.12038366549882258,
+      "grad_norm": 0.339915931224823,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 1048
+    },
+    {
+      "epoch": 0.1204985354086497,
+      "grad_norm": 0.3441682457923889,
+      "learning_rate": 0.0001,
+      "loss": 1.7686,
+      "step": 1049
+    },
+    {
+      "epoch": 0.12061340531847682,
+      "grad_norm": 0.36500778794288635,
+      "learning_rate": 0.0001,
+      "loss": 1.7857,
+      "step": 1050
+    },
+    {
+      "epoch": 0.12072827522830394,
+      "grad_norm": 0.34607526659965515,
+      "learning_rate": 0.0001,
+      "loss": 1.8434,
+      "step": 1051
+    },
+    {
+      "epoch": 0.12084314513813106,
+      "grad_norm": 0.33628061413764954,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 1052
+    },
+    {
+      "epoch": 0.12095801504795818,
+      "grad_norm": 0.3468761146068573,
+      "learning_rate": 0.0001,
+      "loss": 1.8219,
+      "step": 1053
+    },
+    {
+      "epoch": 0.12107288495778532,
+      "grad_norm": 0.34483084082603455,
+      "learning_rate": 0.0001,
+      "loss": 1.7694,
+      "step": 1054
+    },
+    {
+      "epoch": 0.12118775486761243,
+      "grad_norm": 0.3372325003147125,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 1055
+    },
+    {
+      "epoch": 0.12130262477743955,
+      "grad_norm": 0.33837631344795227,
+      "learning_rate": 0.0001,
+      "loss": 1.6369,
+      "step": 1056
+    },
+    {
+      "epoch": 0.12141749468726667,
+      "grad_norm": 0.3314455449581146,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 1057
+    },
+    {
+      "epoch": 0.1215323645970938,
+      "grad_norm": 0.33914583921432495,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 1058
+    },
+    {
+      "epoch": 0.12164723450692091,
+      "grad_norm": 0.33204326033592224,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 1059
+    },
+    {
+      "epoch": 0.12176210441674803,
+      "grad_norm": 0.3244878649711609,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 1060
+    },
+    {
+      "epoch": 0.12187697432657515,
+      "grad_norm": 0.37732627987861633,
+      "learning_rate": 0.0001,
+      "loss": 1.7643,
+      "step": 1061
+    },
+    {
+      "epoch": 0.12199184423640227,
+      "grad_norm": 0.3585972785949707,
+      "learning_rate": 0.0001,
+      "loss": 1.6568,
+      "step": 1062
+    },
+    {
+      "epoch": 0.1221067141462294,
+      "grad_norm": 0.36080244183540344,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 1063
+    },
+    {
+      "epoch": 0.12222158405605651,
+      "grad_norm": 0.3640299141407013,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 1064
+    },
+    {
+      "epoch": 0.12233645396588363,
+      "grad_norm": 0.3357522189617157,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 1065
+    },
+    {
+      "epoch": 0.12245132387571075,
+      "grad_norm": 0.37440431118011475,
+      "learning_rate": 0.0001,
+      "loss": 1.4352,
+      "step": 1066
+    },
+    {
+      "epoch": 0.12256619378553787,
+      "grad_norm": 0.3403536379337311,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 1067
+    },
+    {
+      "epoch": 0.122681063695365,
+      "grad_norm": 0.34183967113494873,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 1068
+    },
+    {
+      "epoch": 0.12279593360519211,
+      "grad_norm": 0.34776198863983154,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 1069
+    },
+    {
+      "epoch": 0.12291080351501923,
+      "grad_norm": 0.36906319856643677,
+      "learning_rate": 0.0001,
+      "loss": 1.8587,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12302567342484635,
+      "grad_norm": 0.33320197463035583,
+      "learning_rate": 0.0001,
+      "loss": 1.6826,
+      "step": 1071
+    },
+    {
+      "epoch": 0.12314054333467349,
+      "grad_norm": 0.35025712847709656,
+      "learning_rate": 0.0001,
+      "loss": 1.7588,
+      "step": 1072
+    },
+    {
+      "epoch": 0.12325541324450061,
+      "grad_norm": 0.33611828088760376,
+      "learning_rate": 0.0001,
+      "loss": 1.6779,
+      "step": 1073
+    },
+    {
+      "epoch": 0.12337028315432773,
+      "grad_norm": 0.34194374084472656,
+      "learning_rate": 0.0001,
+      "loss": 1.6713,
+      "step": 1074
+    },
+    {
+      "epoch": 0.12348515306415485,
+      "grad_norm": 0.3618099093437195,
+      "learning_rate": 0.0001,
+      "loss": 1.928,
+      "step": 1075
+    },
+    {
+      "epoch": 0.12360002297398197,
+      "grad_norm": 0.36288702487945557,
+      "learning_rate": 0.0001,
+      "loss": 1.825,
+      "step": 1076
+    },
+    {
+      "epoch": 0.12371489288380909,
+      "grad_norm": 0.3502649664878845,
+      "learning_rate": 0.0001,
+      "loss": 1.805,
+      "step": 1077
+    },
+    {
+      "epoch": 0.12382976279363621,
+      "grad_norm": 0.3414769470691681,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 1078
+    },
+    {
+      "epoch": 0.12394463270346333,
+      "grad_norm": 0.3354913592338562,
+      "learning_rate": 0.0001,
+      "loss": 1.8155,
+      "step": 1079
+    },
+    {
+      "epoch": 0.12405950261329045,
+      "grad_norm": 0.34553056955337524,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12417437252311757,
+      "grad_norm": 0.3307434320449829,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 1081
+    },
+    {
+      "epoch": 0.12428924243294469,
+      "grad_norm": 0.3462519347667694,
+      "learning_rate": 0.0001,
+      "loss": 1.7164,
+      "step": 1082
+    },
+    {
+      "epoch": 0.12440411234277181,
+      "grad_norm": 0.3434395492076874,
+      "learning_rate": 0.0001,
+      "loss": 1.9303,
+      "step": 1083
+    },
+    {
+      "epoch": 0.12451898225259893,
+      "grad_norm": 0.32828354835510254,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 1084
+    },
+    {
+      "epoch": 0.12463385216242605,
+      "grad_norm": 0.3581448793411255,
+      "learning_rate": 0.0001,
+      "loss": 1.7733,
+      "step": 1085
+    },
+    {
+      "epoch": 0.12474872207225317,
+      "grad_norm": 0.33728450536727905,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 1086
+    },
+    {
+      "epoch": 0.12486359198208029,
+      "grad_norm": 0.36441946029663086,
+      "learning_rate": 0.0001,
+      "loss": 1.8154,
+      "step": 1087
+    },
+    {
+      "epoch": 0.12497846189190741,
+      "grad_norm": 0.3511005640029907,
+      "learning_rate": 0.0001,
+      "loss": 1.8249,
+      "step": 1088
+    },
+    {
+      "epoch": 0.12509333180173454,
+      "grad_norm": 0.36078888177871704,
+      "learning_rate": 0.0001,
+      "loss": 1.7089,
+      "step": 1089
+    },
+    {
+      "epoch": 0.12520820171156166,
+      "grad_norm": 0.3429834246635437,
+      "learning_rate": 0.0001,
+      "loss": 1.66,
+      "step": 1090
+    },
+    {
+      "epoch": 0.12532307162138878,
+      "grad_norm": 0.3412310779094696,
+      "learning_rate": 0.0001,
+      "loss": 1.704,
+      "step": 1091
+    },
+    {
+      "epoch": 0.1254379415312159,
+      "grad_norm": 0.377566933631897,
+      "learning_rate": 0.0001,
+      "loss": 2.0321,
+      "step": 1092
+    },
+    {
+      "epoch": 0.12555281144104302,
+      "grad_norm": 0.3425246477127075,
+      "learning_rate": 0.0001,
+      "loss": 1.6953,
+      "step": 1093
+    },
+    {
+      "epoch": 0.12566768135087014,
+      "grad_norm": 0.36676526069641113,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 1094
+    },
+    {
+      "epoch": 0.12578255126069726,
+      "grad_norm": 0.3748563528060913,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 1095
+    },
+    {
+      "epoch": 0.12589742117052438,
+      "grad_norm": 0.3399527370929718,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 1096
+    },
+    {
+      "epoch": 0.1260122910803515,
+      "grad_norm": 0.32709023356437683,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 1097
+    },
+    {
+      "epoch": 0.12612716099017862,
+      "grad_norm": 0.3199503421783447,
+      "learning_rate": 0.0001,
+      "loss": 1.5727,
+      "step": 1098
+    },
+    {
+      "epoch": 0.12624203090000574,
+      "grad_norm": 0.36548885703086853,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 1099
+    },
+    {
+      "epoch": 0.12635690080983286,
+      "grad_norm": 0.35938435792922974,
+      "learning_rate": 0.0001,
+      "loss": 1.8585,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12647177071965998,
+      "grad_norm": 0.3288070857524872,
+      "learning_rate": 0.0001,
+      "loss": 1.8322,
+      "step": 1101
+    },
+    {
+      "epoch": 0.1265866406294871,
+      "grad_norm": 0.3379286527633667,
+      "learning_rate": 0.0001,
+      "loss": 1.7366,
+      "step": 1102
+    },
+    {
+      "epoch": 0.12670151053931422,
+      "grad_norm": 0.38364285230636597,
+      "learning_rate": 0.0001,
+      "loss": 1.9132,
+      "step": 1103
+    },
+    {
+      "epoch": 0.12681638044914134,
+      "grad_norm": 0.348711222410202,
+      "learning_rate": 0.0001,
+      "loss": 1.8004,
+      "step": 1104
+    },
+    {
+      "epoch": 0.12693125035896846,
+      "grad_norm": 0.3605641722679138,
+      "learning_rate": 0.0001,
+      "loss": 1.7316,
+      "step": 1105
+    },
+    {
+      "epoch": 0.12704612026879558,
+      "grad_norm": 0.3259005546569824,
+      "learning_rate": 0.0001,
+      "loss": 1.7338,
+      "step": 1106
+    },
+    {
+      "epoch": 0.1271609901786227,
+      "grad_norm": 0.374476820230484,
+      "learning_rate": 0.0001,
+      "loss": 1.9204,
+      "step": 1107
+    },
+    {
+      "epoch": 0.12727586008844982,
+      "grad_norm": 0.36476418375968933,
+      "learning_rate": 0.0001,
+      "loss": 1.7403,
+      "step": 1108
+    },
+    {
+      "epoch": 0.12739072999827694,
+      "grad_norm": 0.3357693552970886,
+      "learning_rate": 0.0001,
+      "loss": 1.6227,
+      "step": 1109
+    },
+    {
+      "epoch": 0.12750559990810406,
+      "grad_norm": 0.34735891222953796,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12762046981793118,
+      "grad_norm": 0.34894609451293945,
+      "learning_rate": 0.0001,
+      "loss": 1.6742,
+      "step": 1111
+    },
+    {
+      "epoch": 0.1277353397277583,
+      "grad_norm": 0.34470030665397644,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 1112
+    },
+    {
+      "epoch": 0.12785020963758542,
+      "grad_norm": 0.34968432784080505,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 1113
+    },
+    {
+      "epoch": 0.12796507954741257,
+      "grad_norm": 0.35253041982650757,
+      "learning_rate": 0.0001,
+      "loss": 1.5959,
+      "step": 1114
+    },
+    {
+      "epoch": 0.1280799494572397,
+      "grad_norm": 0.36112409830093384,
+      "learning_rate": 0.0001,
+      "loss": 1.7485,
+      "step": 1115
+    },
+    {
+      "epoch": 0.1281948193670668,
+      "grad_norm": 0.361447274684906,
+      "learning_rate": 0.0001,
+      "loss": 1.8199,
+      "step": 1116
+    },
+    {
+      "epoch": 0.12830968927689393,
+      "grad_norm": 0.3329083323478699,
+      "learning_rate": 0.0001,
+      "loss": 1.8049,
+      "step": 1117
+    },
+    {
+      "epoch": 0.12842455918672105,
+      "grad_norm": 0.37761425971984863,
+      "learning_rate": 0.0001,
+      "loss": 1.8367,
+      "step": 1118
+    },
+    {
+      "epoch": 0.12853942909654817,
+      "grad_norm": 0.36730146408081055,
+      "learning_rate": 0.0001,
+      "loss": 1.7953,
+      "step": 1119
+    },
+    {
+      "epoch": 0.1286542990063753,
+      "grad_norm": 0.3333646357059479,
+      "learning_rate": 0.0001,
+      "loss": 1.6046,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1287691689162024,
+      "grad_norm": 0.37335070967674255,
+      "learning_rate": 0.0001,
+      "loss": 1.9296,
+      "step": 1121
+    },
+    {
+      "epoch": 0.12888403882602953,
+      "grad_norm": 0.36722204089164734,
+      "learning_rate": 0.0001,
+      "loss": 1.7312,
+      "step": 1122
+    },
+    {
+      "epoch": 0.12899890873585665,
+      "grad_norm": 0.34996774792671204,
+      "learning_rate": 0.0001,
+      "loss": 1.7276,
+      "step": 1123
+    },
+    {
+      "epoch": 0.12911377864568377,
+      "grad_norm": 0.3631379306316376,
+      "learning_rate": 0.0001,
+      "loss": 1.7279,
+      "step": 1124
+    },
+    {
+      "epoch": 0.1292286485555109,
+      "grad_norm": 0.3820423483848572,
+      "learning_rate": 0.0001,
+      "loss": 1.7714,
+      "step": 1125
+    },
+    {
+      "epoch": 0.129343518465338,
+      "grad_norm": 0.3379541039466858,
+      "learning_rate": 0.0001,
+      "loss": 1.6302,
+      "step": 1126
+    },
+    {
+      "epoch": 0.12945838837516513,
+      "grad_norm": 0.35675719380378723,
+      "learning_rate": 0.0001,
+      "loss": 1.5973,
+      "step": 1127
+    },
+    {
+      "epoch": 0.12957325828499225,
+      "grad_norm": 0.34876230359077454,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 1128
+    },
+    {
+      "epoch": 0.12968812819481937,
+      "grad_norm": 0.35678061842918396,
+      "learning_rate": 0.0001,
+      "loss": 1.8694,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1298029981046465,
+      "grad_norm": 0.3658314347267151,
+      "learning_rate": 0.0001,
+      "loss": 1.9091,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1299178680144736,
+      "grad_norm": 0.3842300474643707,
+      "learning_rate": 0.0001,
+      "loss": 1.7441,
+      "step": 1131
+    },
+    {
+      "epoch": 0.13003273792430073,
+      "grad_norm": 0.3557139039039612,
+      "learning_rate": 0.0001,
+      "loss": 1.9989,
+      "step": 1132
+    },
+    {
+      "epoch": 0.13014760783412785,
+      "grad_norm": 0.3451981246471405,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 1133
+    },
+    {
+      "epoch": 0.13026247774395497,
+      "grad_norm": 0.34899675846099854,
+      "learning_rate": 0.0001,
+      "loss": 1.632,
+      "step": 1134
+    },
+    {
+      "epoch": 0.1303773476537821,
+      "grad_norm": 0.3958018124103546,
+      "learning_rate": 0.0001,
+      "loss": 2.2029,
+      "step": 1135
+    },
+    {
+      "epoch": 0.1304922175636092,
+      "grad_norm": 0.33789554238319397,
+      "learning_rate": 0.0001,
+      "loss": 1.677,
+      "step": 1136
+    },
+    {
+      "epoch": 0.13060708747343633,
+      "grad_norm": 0.3390919864177704,
+      "learning_rate": 0.0001,
+      "loss": 1.7488,
+      "step": 1137
+    },
+    {
+      "epoch": 0.13072195738326345,
+      "grad_norm": 0.32724031805992126,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 1138
+    },
+    {
+      "epoch": 0.13083682729309057,
+      "grad_norm": 0.35171130299568176,
+      "learning_rate": 0.0001,
+      "loss": 1.8461,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1309516972029177,
+      "grad_norm": 0.37492451071739197,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1310665671127448,
+      "grad_norm": 0.3509044051170349,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 1141
+    },
+    {
+      "epoch": 0.13118143702257193,
+      "grad_norm": 0.379950612783432,
+      "learning_rate": 0.0001,
+      "loss": 1.8488,
+      "step": 1142
+    },
+    {
+      "epoch": 0.13129630693239905,
+      "grad_norm": 0.3770541846752167,
+      "learning_rate": 0.0001,
+      "loss": 1.9027,
+      "step": 1143
+    },
+    {
+      "epoch": 0.13141117684222617,
+      "grad_norm": 0.3268395960330963,
+      "learning_rate": 0.0001,
+      "loss": 1.6642,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1315260467520533,
+      "grad_norm": 0.3465530276298523,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1316409166618804,
+      "grad_norm": 0.3526296317577362,
+      "learning_rate": 0.0001,
+      "loss": 1.9214,
+      "step": 1146
+    },
+    {
+      "epoch": 0.13175578657170753,
+      "grad_norm": 0.3607937693595886,
+      "learning_rate": 0.0001,
+      "loss": 1.9947,
+      "step": 1147
+    },
+    {
+      "epoch": 0.13187065648153465,
+      "grad_norm": 0.35963696241378784,
+      "learning_rate": 0.0001,
+      "loss": 1.7277,
+      "step": 1148
+    },
+    {
+      "epoch": 0.1319855263913618,
+      "grad_norm": 0.3521120846271515,
+      "learning_rate": 0.0001,
+      "loss": 1.7295,
+      "step": 1149
+    },
+    {
+      "epoch": 0.13210039630118892,
+      "grad_norm": 0.34942013025283813,
+      "learning_rate": 0.0001,
+      "loss": 1.6966,
+      "step": 1150
+    },
+    {
+      "epoch": 0.13221526621101604,
+      "grad_norm": 0.3466810882091522,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 1151
+    },
+    {
+      "epoch": 0.13233013612084316,
+      "grad_norm": 0.36123740673065186,
+      "learning_rate": 0.0001,
+      "loss": 1.7322,
+      "step": 1152
+    },
+    {
+      "epoch": 0.13244500603067028,
+      "grad_norm": 0.36202317476272583,
+      "learning_rate": 0.0001,
+      "loss": 1.844,
+      "step": 1153
+    },
+    {
+      "epoch": 0.1325598759404974,
+      "grad_norm": 0.3521657884120941,
+      "learning_rate": 0.0001,
+      "loss": 1.778,
+      "step": 1154
+    },
+    {
+      "epoch": 0.13267474585032452,
+      "grad_norm": 0.35501620173454285,
+      "learning_rate": 0.0001,
+      "loss": 1.7139,
+      "step": 1155
+    },
+    {
+      "epoch": 0.13278961576015164,
+      "grad_norm": 0.35365980863571167,
+      "learning_rate": 0.0001,
+      "loss": 1.5055,
+      "step": 1156
+    },
+    {
+      "epoch": 0.13290448566997876,
+      "grad_norm": 0.3492221534252167,
+      "learning_rate": 0.0001,
+      "loss": 1.7555,
+      "step": 1157
+    },
+    {
+      "epoch": 0.13301935557980588,
+      "grad_norm": 0.4094052016735077,
+      "learning_rate": 0.0001,
+      "loss": 1.7331,
+      "step": 1158
+    },
+    {
+      "epoch": 0.133134225489633,
+      "grad_norm": 0.385098934173584,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 1159
+    },
+    {
+      "epoch": 0.13324909539946012,
+      "grad_norm": 0.3582170605659485,
+      "learning_rate": 0.0001,
+      "loss": 1.8903,
+      "step": 1160
+    },
+    {
+      "epoch": 0.13336396530928724,
+      "grad_norm": 0.35824549198150635,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 1161
+    },
+    {
+      "epoch": 0.13347883521911436,
+      "grad_norm": 0.36423972249031067,
+      "learning_rate": 0.0001,
+      "loss": 1.7682,
+      "step": 1162
+    },
+    {
+      "epoch": 0.13359370512894148,
+      "grad_norm": 0.3333613872528076,
+      "learning_rate": 0.0001,
+      "loss": 1.7676,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1337085750387686,
+      "grad_norm": 0.3693676292896271,
+      "learning_rate": 0.0001,
+      "loss": 1.8687,
+      "step": 1164
+    },
+    {
+      "epoch": 0.13382344494859572,
+      "grad_norm": 0.36510157585144043,
+      "learning_rate": 0.0001,
+      "loss": 1.8022,
+      "step": 1165
+    },
+    {
+      "epoch": 0.13393831485842284,
+      "grad_norm": 0.35315313935279846,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 1166
+    },
+    {
+      "epoch": 0.13405318476824996,
+      "grad_norm": 0.3328644037246704,
+      "learning_rate": 0.0001,
+      "loss": 1.5863,
+      "step": 1167
+    },
+    {
+      "epoch": 0.13416805467807708,
+      "grad_norm": 0.34875044226646423,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1342829245879042,
+      "grad_norm": 0.3806982636451721,
+      "learning_rate": 0.0001,
+      "loss": 1.8317,
+      "step": 1169
+    },
+    {
+      "epoch": 0.13439779449773132,
+      "grad_norm": 0.35632237792015076,
+      "learning_rate": 0.0001,
+      "loss": 1.757,
+      "step": 1170
+    },
+    {
+      "epoch": 0.13451266440755844,
+      "grad_norm": 0.3340839147567749,
+      "learning_rate": 0.0001,
+      "loss": 1.6582,
+      "step": 1171
+    },
+    {
+      "epoch": 0.13462753431738556,
+      "grad_norm": 0.3517102599143982,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 1172
+    },
+    {
+      "epoch": 0.13474240422721268,
+      "grad_norm": 0.33846724033355713,
+      "learning_rate": 0.0001,
+      "loss": 1.7031,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1348572741370398,
+      "grad_norm": 0.3715790808200836,
+      "learning_rate": 0.0001,
+      "loss": 1.6864,
+      "step": 1174
+    },
+    {
+      "epoch": 0.13497214404686692,
+      "grad_norm": 0.30750709772109985,
+      "learning_rate": 0.0001,
+      "loss": 1.4213,
+      "step": 1175
+    },
+    {
+      "epoch": 0.13508701395669404,
+      "grad_norm": 0.35879287123680115,
+      "learning_rate": 0.0001,
+      "loss": 1.6197,
+      "step": 1176
+    },
+    {
+      "epoch": 0.13520188386652116,
+      "grad_norm": 0.35407108068466187,
+      "learning_rate": 0.0001,
+      "loss": 1.6578,
+      "step": 1177
+    },
+    {
+      "epoch": 0.13531675377634828,
+      "grad_norm": 0.35184237360954285,
+      "learning_rate": 0.0001,
+      "loss": 1.7593,
+      "step": 1178
+    },
+    {
+      "epoch": 0.1354316236861754,
+      "grad_norm": 0.3242950737476349,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 1179
+    },
+    {
+      "epoch": 0.13554649359600252,
+      "grad_norm": 0.3433818817138672,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 1180
+    },
+    {
+      "epoch": 0.13566136350582964,
+      "grad_norm": 0.34305769205093384,
+      "learning_rate": 0.0001,
+      "loss": 1.5726,
+      "step": 1181
+    },
+    {
+      "epoch": 0.13577623341565676,
+      "grad_norm": 0.35595643520355225,
+      "learning_rate": 0.0001,
+      "loss": 1.7659,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1358911033254839,
+      "grad_norm": 0.3613067865371704,
+      "learning_rate": 0.0001,
+      "loss": 1.8036,
+      "step": 1183
+    },
+    {
+      "epoch": 0.13600597323531102,
+      "grad_norm": 0.38018158078193665,
+      "learning_rate": 0.0001,
+      "loss": 1.9313,
+      "step": 1184
+    },
+    {
+      "epoch": 0.13612084314513814,
+      "grad_norm": 0.3490848243236542,
+      "learning_rate": 0.0001,
+      "loss": 1.7979,
+      "step": 1185
+    },
+    {
+      "epoch": 0.13623571305496526,
+      "grad_norm": 0.35665586590766907,
+      "learning_rate": 0.0001,
+      "loss": 1.7383,
+      "step": 1186
+    },
+    {
+      "epoch": 0.13635058296479238,
+      "grad_norm": 0.3440529704093933,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1364654528746195,
+      "grad_norm": 0.31870830059051514,
+      "learning_rate": 0.0001,
+      "loss": 1.5472,
+      "step": 1188
+    },
+    {
+      "epoch": 0.13658032278444662,
+      "grad_norm": 0.3264746367931366,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 1189
+    },
+    {
+      "epoch": 0.13669519269427374,
+      "grad_norm": 0.364711731672287,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13681006260410086,
+      "grad_norm": 0.34434038400650024,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 1191
+    },
+    {
+      "epoch": 0.13692493251392798,
+      "grad_norm": 0.36517781019210815,
+      "learning_rate": 0.0001,
+      "loss": 1.8602,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1370398024237551,
+      "grad_norm": 0.36987563967704773,
+      "learning_rate": 0.0001,
+      "loss": 1.7108,
+      "step": 1193
+    },
+    {
+      "epoch": 0.13715467233358222,
+      "grad_norm": 0.34185507893562317,
+      "learning_rate": 0.0001,
+      "loss": 1.7236,
+      "step": 1194
+    },
+    {
+      "epoch": 0.13726954224340934,
+      "grad_norm": 0.3375924527645111,
+      "learning_rate": 0.0001,
+      "loss": 1.7451,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13738441215323646,
+      "grad_norm": 0.3625550866127014,
+      "learning_rate": 0.0001,
+      "loss": 1.8872,
+      "step": 1196
+    },
+    {
+      "epoch": 0.13749928206306358,
+      "grad_norm": 0.36060526967048645,
+      "learning_rate": 0.0001,
+      "loss": 1.9059,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1376141519728907,
+      "grad_norm": 0.32978108525276184,
+      "learning_rate": 0.0001,
+      "loss": 1.5382,
+      "step": 1198
+    },
+    {
+      "epoch": 0.13772902188271782,
+      "grad_norm": 0.3556724488735199,
+      "learning_rate": 0.0001,
+      "loss": 1.659,
+      "step": 1199
+    },
+    {
+      "epoch": 0.13784389179254494,
+      "grad_norm": 0.36621591448783875,
+      "learning_rate": 0.0001,
+      "loss": 1.9125,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13795876170237206,
+      "grad_norm": 0.3547092080116272,
+      "learning_rate": 0.0001,
+      "loss": 1.7579,
+      "step": 1201
+    },
+    {
+      "epoch": 0.13807363161219918,
+      "grad_norm": 0.3358149826526642,
+      "learning_rate": 0.0001,
+      "loss": 1.7776,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1381885015220263,
+      "grad_norm": 0.37259694933891296,
+      "learning_rate": 0.0001,
+      "loss": 1.7184,
+      "step": 1203
+    },
+    {
+      "epoch": 0.13830337143185342,
+      "grad_norm": 0.35611268877983093,
+      "learning_rate": 0.0001,
+      "loss": 1.7256,
+      "step": 1204
+    },
+    {
+      "epoch": 0.13841824134168054,
+      "grad_norm": 0.3654508590698242,
+      "learning_rate": 0.0001,
+      "loss": 1.8226,
+      "step": 1205
+    },
+    {
+      "epoch": 0.13853311125150766,
+      "grad_norm": 0.3912622928619385,
+      "learning_rate": 0.0001,
+      "loss": 1.7948,
+      "step": 1206
+    },
+    {
+      "epoch": 0.13864798116133478,
+      "grad_norm": 0.37169212102890015,
+      "learning_rate": 0.0001,
+      "loss": 1.7112,
+      "step": 1207
+    },
+    {
+      "epoch": 0.1387628510711619,
+      "grad_norm": 0.34099262952804565,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 1208
+    },
+    {
+      "epoch": 0.13887772098098902,
+      "grad_norm": 0.34905219078063965,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 1209
+    },
+    {
+      "epoch": 0.13899259089081614,
+      "grad_norm": 0.35248732566833496,
+      "learning_rate": 0.0001,
+      "loss": 1.6179,
+      "step": 1210
+    },
+    {
+      "epoch": 0.13910746080064326,
+      "grad_norm": 0.343364417552948,
+      "learning_rate": 0.0001,
+      "loss": 1.8933,
+      "step": 1211
+    },
+    {
+      "epoch": 0.13922233071047038,
+      "grad_norm": 0.3398562967777252,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1393372006202975,
+      "grad_norm": 0.3378565013408661,
+      "learning_rate": 0.0001,
+      "loss": 1.7439,
+      "step": 1213
+    },
+    {
+      "epoch": 0.13945207053012462,
+      "grad_norm": 0.36921605467796326,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 1214
+    },
+    {
+      "epoch": 0.13956694043995174,
+      "grad_norm": 0.3712867200374603,
+      "learning_rate": 0.0001,
+      "loss": 1.8628,
+      "step": 1215
+    },
+    {
+      "epoch": 0.13968181034977886,
+      "grad_norm": 0.3491092324256897,
+      "learning_rate": 0.0001,
+      "loss": 1.7379,
+      "step": 1216
+    },
+    {
+      "epoch": 0.139796680259606,
+      "grad_norm": 0.3602731227874756,
+      "learning_rate": 0.0001,
+      "loss": 1.7088,
+      "step": 1217
+    },
+    {
+      "epoch": 0.13991155016943313,
+      "grad_norm": 0.39251574873924255,
+      "learning_rate": 0.0001,
+      "loss": 1.8268,
+      "step": 1218
+    },
+    {
+      "epoch": 0.14002642007926025,
+      "grad_norm": 0.3653351664543152,
+      "learning_rate": 0.0001,
+      "loss": 1.7597,
+      "step": 1219
+    },
+    {
+      "epoch": 0.14014128998908737,
+      "grad_norm": 0.3393062353134155,
+      "learning_rate": 0.0001,
+      "loss": 1.6635,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1402561598989145,
+      "grad_norm": 0.3772340416908264,
+      "learning_rate": 0.0001,
+      "loss": 1.7796,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1403710298087416,
+      "grad_norm": 0.3478431701660156,
+      "learning_rate": 0.0001,
+      "loss": 1.6802,
+      "step": 1222
+    },
+    {
+      "epoch": 0.14048589971856873,
+      "grad_norm": 0.33005356788635254,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 1223
+    },
+    {
+      "epoch": 0.14060076962839585,
+      "grad_norm": 0.348294734954834,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 1224
+    },
+    {
+      "epoch": 0.14071563953822297,
+      "grad_norm": 0.35383546352386475,
+      "learning_rate": 0.0001,
+      "loss": 1.5475,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1408305094480501,
+      "grad_norm": 0.37990522384643555,
+      "learning_rate": 0.0001,
+      "loss": 1.8762,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1409453793578772,
+      "grad_norm": 0.3329927623271942,
+      "learning_rate": 0.0001,
+      "loss": 1.752,
+      "step": 1227
+    },
+    {
+      "epoch": 0.14106024926770433,
+      "grad_norm": 0.3647679090499878,
+      "learning_rate": 0.0001,
+      "loss": 1.7886,
+      "step": 1228
+    },
+    {
+      "epoch": 0.14117511917753145,
+      "grad_norm": 0.3833819627761841,
+      "learning_rate": 0.0001,
+      "loss": 1.8977,
+      "step": 1229
+    },
+    {
+      "epoch": 0.14128998908735857,
+      "grad_norm": 0.3702079951763153,
+      "learning_rate": 0.0001,
+      "loss": 1.6843,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1414048589971857,
+      "grad_norm": 0.3493407368659973,
+      "learning_rate": 0.0001,
+      "loss": 1.7134,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1415197289070128,
+      "grad_norm": 0.3664112687110901,
+      "learning_rate": 0.0001,
+      "loss": 1.7641,
+      "step": 1232
+    },
+    {
+      "epoch": 0.14163459881683993,
+      "grad_norm": 0.3498665988445282,
+      "learning_rate": 0.0001,
+      "loss": 1.7903,
+      "step": 1233
+    },
+    {
+      "epoch": 0.14174946872666705,
+      "grad_norm": 0.37883898615837097,
+      "learning_rate": 0.0001,
+      "loss": 1.7601,
+      "step": 1234
+    },
+    {
+      "epoch": 0.14186433863649417,
+      "grad_norm": 0.37011733651161194,
+      "learning_rate": 0.0001,
+      "loss": 1.8304,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1419792085463213,
+      "grad_norm": 0.3556150197982788,
+      "learning_rate": 0.0001,
+      "loss": 1.7222,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1420940784561484,
+      "grad_norm": 0.3517720103263855,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 1237
+    },
+    {
+      "epoch": 0.14220894836597553,
+      "grad_norm": 0.3677484691143036,
+      "learning_rate": 0.0001,
+      "loss": 1.9557,
+      "step": 1238
+    },
+    {
+      "epoch": 0.14232381827580265,
+      "grad_norm": 0.35142794251441956,
+      "learning_rate": 0.0001,
+      "loss": 1.7057,
+      "step": 1239
+    },
+    {
+      "epoch": 0.14243868818562977,
+      "grad_norm": 0.33948856592178345,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1425535580954569,
+      "grad_norm": 0.340010404586792,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 1241
+    },
+    {
+      "epoch": 0.142668428005284,
+      "grad_norm": 0.34367939829826355,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 1242
+    },
+    {
+      "epoch": 0.14278329791511113,
+      "grad_norm": 0.38786405324935913,
+      "learning_rate": 0.0001,
+      "loss": 1.7947,
+      "step": 1243
+    },
+    {
+      "epoch": 0.14289816782493825,
+      "grad_norm": 0.35119712352752686,
+      "learning_rate": 0.0001,
+      "loss": 1.6828,
+      "step": 1244
+    },
+    {
+      "epoch": 0.14301303773476537,
+      "grad_norm": 0.3436744809150696,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1431279076445925,
+      "grad_norm": 0.3338393270969391,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1432427775544196,
+      "grad_norm": 0.38156992197036743,
+      "learning_rate": 0.0001,
+      "loss": 1.7975,
+      "step": 1247
+    },
+    {
+      "epoch": 0.14335764746424673,
+      "grad_norm": 0.34734535217285156,
+      "learning_rate": 0.0001,
+      "loss": 1.6203,
+      "step": 1248
+    },
+    {
+      "epoch": 0.14347251737407385,
+      "grad_norm": 0.34041401743888855,
+      "learning_rate": 0.0001,
+      "loss": 1.7274,
+      "step": 1249
+    },
+    {
+      "epoch": 0.14358738728390097,
+      "grad_norm": 0.37156790494918823,
+      "learning_rate": 0.0001,
+      "loss": 1.7642,
+      "step": 1250
+    },
+    {
+      "epoch": 0.14370225719372812,
+      "grad_norm": 0.3499716818332672,
+      "learning_rate": 0.0001,
+      "loss": 1.7686,
+      "step": 1251
+    },
+    {
+      "epoch": 0.14381712710355524,
+      "grad_norm": 0.32781967520713806,
+      "learning_rate": 0.0001,
+      "loss": 1.5751,
+      "step": 1252
+    },
+    {
+      "epoch": 0.14393199701338236,
+      "grad_norm": 0.3511214852333069,
+      "learning_rate": 0.0001,
+      "loss": 1.6698,
+      "step": 1253
+    },
+    {
+      "epoch": 0.14404686692320948,
+      "grad_norm": 0.3647572994232178,
+      "learning_rate": 0.0001,
+      "loss": 1.6883,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1441617368330366,
+      "grad_norm": 0.3573339879512787,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 1255
+    },
+    {
+      "epoch": 0.14427660674286372,
+      "grad_norm": 0.34564316272735596,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 1256
+    },
+    {
+      "epoch": 0.14439147665269084,
+      "grad_norm": 0.3462870717048645,
+      "learning_rate": 0.0001,
+      "loss": 1.7767,
+      "step": 1257
+    },
+    {
+      "epoch": 0.14450634656251796,
+      "grad_norm": 0.3375271260738373,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 1258
+    },
+    {
+      "epoch": 0.14462121647234508,
+      "grad_norm": 0.3306554853916168,
+      "learning_rate": 0.0001,
+      "loss": 1.6813,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1447360863821722,
+      "grad_norm": 0.3348354995250702,
+      "learning_rate": 0.0001,
+      "loss": 1.5897,
+      "step": 1260
+    },
+    {
+      "epoch": 0.14485095629199932,
+      "grad_norm": 0.3619769811630249,
+      "learning_rate": 0.0001,
+      "loss": 1.6966,
+      "step": 1261
+    },
+    {
+      "epoch": 0.14496582620182644,
+      "grad_norm": 0.3743123412132263,
+      "learning_rate": 0.0001,
+      "loss": 1.8824,
+      "step": 1262
+    },
+    {
+      "epoch": 0.14508069611165356,
+      "grad_norm": 0.33532246947288513,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 1263
+    },
+    {
+      "epoch": 0.14519556602148068,
+      "grad_norm": 0.37136563658714294,
+      "learning_rate": 0.0001,
+      "loss": 1.6995,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1453104359313078,
+      "grad_norm": 0.4397648274898529,
+      "learning_rate": 0.0001,
+      "loss": 2.1078,
+      "step": 1265
+    },
+    {
+      "epoch": 0.14542530584113492,
+      "grad_norm": 0.3656454086303711,
+      "learning_rate": 0.0001,
+      "loss": 1.8501,
+      "step": 1266
+    },
+    {
+      "epoch": 0.14554017575096204,
+      "grad_norm": 0.3662125766277313,
+      "learning_rate": 0.0001,
+      "loss": 1.8535,
+      "step": 1267
+    },
+    {
+      "epoch": 0.14565504566078916,
+      "grad_norm": 0.4044576585292816,
+      "learning_rate": 0.0001,
+      "loss": 1.9077,
+      "step": 1268
+    },
+    {
+      "epoch": 0.14576991557061628,
+      "grad_norm": 0.3433517515659332,
+      "learning_rate": 0.0001,
+      "loss": 1.4109,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1458847854804434,
+      "grad_norm": 0.3514901101589203,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 1270
+    },
+    {
+      "epoch": 0.14599965539027052,
+      "grad_norm": 0.3364449143409729,
+      "learning_rate": 0.0001,
+      "loss": 1.6527,
+      "step": 1271
+    },
+    {
+      "epoch": 0.14611452530009764,
+      "grad_norm": 0.3305410146713257,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 1272
+    },
+    {
+      "epoch": 0.14622939520992476,
+      "grad_norm": 0.35502052307128906,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 1273
+    },
+    {
+      "epoch": 0.14634426511975188,
+      "grad_norm": 0.3556758165359497,
+      "learning_rate": 0.0001,
+      "loss": 1.6573,
+      "step": 1274
+    },
+    {
+      "epoch": 0.146459135029579,
+      "grad_norm": 0.3576909899711609,
+      "learning_rate": 0.0001,
+      "loss": 1.6997,
+      "step": 1275
+    },
+    {
+      "epoch": 0.14657400493940612,
+      "grad_norm": 0.34596705436706543,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 1276
+    },
+    {
+      "epoch": 0.14668887484923324,
+      "grad_norm": 0.38846471905708313,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 1277
+    },
+    {
+      "epoch": 0.14680374475906036,
+      "grad_norm": 0.37087368965148926,
+      "learning_rate": 0.0001,
+      "loss": 1.443,
+      "step": 1278
+    },
+    {
+      "epoch": 0.14691861466888748,
+      "grad_norm": 0.3498583137989044,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1470334845787146,
+      "grad_norm": 0.34810611605644226,
+      "learning_rate": 0.0001,
+      "loss": 1.6464,
+      "step": 1280
+    },
+    {
+      "epoch": 0.14714835448854172,
+      "grad_norm": 0.3536350727081299,
+      "learning_rate": 0.0001,
+      "loss": 1.8913,
+      "step": 1281
+    },
+    {
+      "epoch": 0.14726322439836884,
+      "grad_norm": 0.3864216208457947,
+      "learning_rate": 0.0001,
+      "loss": 1.9487,
+      "step": 1282
+    },
+    {
+      "epoch": 0.14737809430819596,
+      "grad_norm": 0.36202993988990784,
+      "learning_rate": 0.0001,
+      "loss": 1.7515,
+      "step": 1283
+    },
+    {
+      "epoch": 0.14749296421802308,
+      "grad_norm": 0.3679467439651489,
+      "learning_rate": 0.0001,
+      "loss": 1.8294,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1476078341278502,
+      "grad_norm": 0.3474784195423126,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 1285
+    },
+    {
+      "epoch": 0.14772270403767734,
+      "grad_norm": 0.3244973123073578,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 1286
+    },
+    {
+      "epoch": 0.14783757394750446,
+      "grad_norm": 0.36702069640159607,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 1287
+    },
+    {
+      "epoch": 0.14795244385733158,
+      "grad_norm": 0.3686244487762451,
+      "learning_rate": 0.0001,
+      "loss": 1.9021,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1480673137671587,
+      "grad_norm": 0.36901167035102844,
+      "learning_rate": 0.0001,
+      "loss": 1.8618,
+      "step": 1289
+    },
+    {
+      "epoch": 0.14818218367698582,
+      "grad_norm": 0.3515526354312897,
+      "learning_rate": 0.0001,
+      "loss": 1.6961,
+      "step": 1290
+    },
+    {
+      "epoch": 0.14829705358681294,
+      "grad_norm": 0.36879947781562805,
+      "learning_rate": 0.0001,
+      "loss": 1.7333,
+      "step": 1291
+    },
+    {
+      "epoch": 0.14841192349664006,
+      "grad_norm": 0.3325467109680176,
+      "learning_rate": 0.0001,
+      "loss": 1.5058,
+      "step": 1292
+    },
+    {
+      "epoch": 0.14852679340646718,
+      "grad_norm": 0.3682822287082672,
+      "learning_rate": 0.0001,
+      "loss": 1.8649,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1486416633162943,
+      "grad_norm": 0.3390534222126007,
+      "learning_rate": 0.0001,
+      "loss": 1.7462,
+      "step": 1294
+    },
+    {
+      "epoch": 0.14875653322612142,
+      "grad_norm": 0.3411741554737091,
+      "learning_rate": 0.0001,
+      "loss": 1.733,
+      "step": 1295
+    },
+    {
+      "epoch": 0.14887140313594854,
+      "grad_norm": 0.33508262038230896,
+      "learning_rate": 0.0001,
+      "loss": 1.7306,
+      "step": 1296
+    },
+    {
+      "epoch": 0.14898627304577566,
+      "grad_norm": 0.3365498483181,
+      "learning_rate": 0.0001,
+      "loss": 1.6963,
+      "step": 1297
+    },
+    {
+      "epoch": 0.14910114295560278,
+      "grad_norm": 0.3673759400844574,
+      "learning_rate": 0.0001,
+      "loss": 1.3957,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1492160128654299,
+      "grad_norm": 0.32368943095207214,
+      "learning_rate": 0.0001,
+      "loss": 1.4204,
+      "step": 1299
+    },
+    {
+      "epoch": 0.14933088277525702,
+      "grad_norm": 0.37173643708229065,
+      "learning_rate": 0.0001,
+      "loss": 1.7783,
+      "step": 1300
+    },
+    {
+      "epoch": 0.14944575268508414,
+      "grad_norm": 0.3451181650161743,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 1301
+    },
+    {
+      "epoch": 0.14956062259491126,
+      "grad_norm": 0.3521316945552826,
+      "learning_rate": 0.0001,
+      "loss": 1.7907,
+      "step": 1302
+    },
+    {
+      "epoch": 0.14967549250473838,
+      "grad_norm": 0.3625703454017639,
+      "learning_rate": 0.0001,
+      "loss": 1.9229,
+      "step": 1303
+    },
+    {
+      "epoch": 0.1497903624145655,
+      "grad_norm": 0.32430633902549744,
+      "learning_rate": 0.0001,
+      "loss": 1.5798,
+      "step": 1304
+    },
+    {
+      "epoch": 0.14990523232439262,
+      "grad_norm": 0.3538731038570404,
+      "learning_rate": 0.0001,
+      "loss": 1.6985,
+      "step": 1305
+    },
+    {
+      "epoch": 0.15002010223421974,
+      "grad_norm": 0.362496018409729,
+      "learning_rate": 0.0001,
+      "loss": 1.6863,
+      "step": 1306
+    },
+    {
+      "epoch": 0.15013497214404686,
+      "grad_norm": 0.33480730652809143,
+      "learning_rate": 0.0001,
+      "loss": 1.6784,
+      "step": 1307
+    },
+    {
+      "epoch": 0.15024984205387398,
+      "grad_norm": 0.3331620991230011,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 1308
+    },
+    {
+      "epoch": 0.1503647119637011,
+      "grad_norm": 0.39281463623046875,
+      "learning_rate": 0.0001,
+      "loss": 1.791,
+      "step": 1309
+    },
+    {
+      "epoch": 0.15047958187352822,
+      "grad_norm": 0.3314681351184845,
+      "learning_rate": 0.0001,
+      "loss": 1.6435,
+      "step": 1310
+    },
+    {
+      "epoch": 0.15059445178335534,
+      "grad_norm": 0.349573016166687,
+      "learning_rate": 0.0001,
+      "loss": 1.6755,
+      "step": 1311
+    },
+    {
+      "epoch": 0.15070932169318246,
+      "grad_norm": 0.35691556334495544,
+      "learning_rate": 0.0001,
+      "loss": 1.7388,
+      "step": 1312
+    },
+    {
+      "epoch": 0.15082419160300958,
+      "grad_norm": 0.3481789827346802,
+      "learning_rate": 0.0001,
+      "loss": 1.6805,
+      "step": 1313
+    },
+    {
+      "epoch": 0.1509390615128367,
+      "grad_norm": 0.36650916934013367,
+      "learning_rate": 0.0001,
+      "loss": 1.8396,
+      "step": 1314
+    },
+    {
+      "epoch": 0.15105393142266382,
+      "grad_norm": 0.3846674859523773,
+      "learning_rate": 0.0001,
+      "loss": 1.9883,
+      "step": 1315
+    },
+    {
+      "epoch": 0.15116880133249094,
+      "grad_norm": 0.3474387228488922,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 1316
+    },
+    {
+      "epoch": 0.15128367124231806,
+      "grad_norm": 0.37427589297294617,
+      "learning_rate": 0.0001,
+      "loss": 1.7977,
+      "step": 1317
+    },
+    {
+      "epoch": 0.15139854115214518,
+      "grad_norm": 0.38520902395248413,
+      "learning_rate": 0.0001,
+      "loss": 1.8338,
+      "step": 1318
+    },
+    {
+      "epoch": 0.1515134110619723,
+      "grad_norm": 0.3556143641471863,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 1319
+    },
+    {
+      "epoch": 0.15162828097179945,
+      "grad_norm": 0.36979398131370544,
+      "learning_rate": 0.0001,
+      "loss": 1.7625,
+      "step": 1320
+    },
+    {
+      "epoch": 0.15174315088162657,
+      "grad_norm": 0.35435059666633606,
+      "learning_rate": 0.0001,
+      "loss": 1.6956,
+      "step": 1321
+    },
+    {
+      "epoch": 0.1518580207914537,
+      "grad_norm": 0.35871556401252747,
+      "learning_rate": 0.0001,
+      "loss": 1.8573,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1519728907012808,
+      "grad_norm": 0.3628619611263275,
+      "learning_rate": 0.0001,
+      "loss": 1.8769,
+      "step": 1323
+    },
+    {
+      "epoch": 0.15208776061110793,
+      "grad_norm": 0.3393780291080475,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 1324
+    },
+    {
+      "epoch": 0.15220263052093505,
+      "grad_norm": 0.3389873802661896,
+      "learning_rate": 0.0001,
+      "loss": 1.6257,
+      "step": 1325
+    },
+    {
+      "epoch": 0.15231750043076217,
+      "grad_norm": 0.3557165563106537,
+      "learning_rate": 0.0001,
+      "loss": 1.8152,
+      "step": 1326
+    },
+    {
+      "epoch": 0.1524323703405893,
+      "grad_norm": 0.36443039774894714,
+      "learning_rate": 0.0001,
+      "loss": 1.7758,
+      "step": 1327
+    },
+    {
+      "epoch": 0.1525472402504164,
+      "grad_norm": 0.3473578989505768,
+      "learning_rate": 0.0001,
+      "loss": 1.6577,
+      "step": 1328
+    },
+    {
+      "epoch": 0.15266211016024353,
+      "grad_norm": 0.36054491996765137,
+      "learning_rate": 0.0001,
+      "loss": 1.7633,
+      "step": 1329
+    },
+    {
+      "epoch": 0.15277698007007065,
+      "grad_norm": 0.3966304659843445,
+      "learning_rate": 0.0001,
+      "loss": 1.8216,
+      "step": 1330
+    },
+    {
+      "epoch": 0.15289184997989777,
+      "grad_norm": 0.3621263802051544,
+      "learning_rate": 0.0001,
+      "loss": 1.8102,
+      "step": 1331
+    },
+    {
+      "epoch": 0.1530067198897249,
+      "grad_norm": 0.346164733171463,
+      "learning_rate": 0.0001,
+      "loss": 1.8498,
+      "step": 1332
+    },
+    {
+      "epoch": 0.153121589799552,
+      "grad_norm": 0.3580979108810425,
+      "learning_rate": 0.0001,
+      "loss": 1.4788,
+      "step": 1333
+    },
+    {
+      "epoch": 0.15323645970937913,
+      "grad_norm": 0.4236307740211487,
+      "learning_rate": 0.0001,
+      "loss": 2.0173,
+      "step": 1334
+    },
+    {
+      "epoch": 0.15335132961920625,
+      "grad_norm": 0.36115512251853943,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 1335
+    },
+    {
+      "epoch": 0.15346619952903337,
+      "grad_norm": 0.35453563928604126,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 1336
+    },
+    {
+      "epoch": 0.1535810694388605,
+      "grad_norm": 0.37019240856170654,
+      "learning_rate": 0.0001,
+      "loss": 1.7979,
+      "step": 1337
+    },
+    {
+      "epoch": 0.1536959393486876,
+      "grad_norm": 0.34816619753837585,
+      "learning_rate": 0.0001,
+      "loss": 1.8167,
+      "step": 1338
+    },
+    {
+      "epoch": 0.15381080925851473,
+      "grad_norm": 0.3500208556652069,
+      "learning_rate": 0.0001,
+      "loss": 1.79,
+      "step": 1339
+    },
+    {
+      "epoch": 0.15392567916834185,
+      "grad_norm": 0.34298229217529297,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 1340
+    },
+    {
+      "epoch": 0.15404054907816897,
+      "grad_norm": 0.34139198064804077,
+      "learning_rate": 0.0001,
+      "loss": 1.694,
+      "step": 1341
+    },
+    {
+      "epoch": 0.1541554189879961,
+      "grad_norm": 0.3685595989227295,
+      "learning_rate": 0.0001,
+      "loss": 1.748,
+      "step": 1342
+    },
+    {
+      "epoch": 0.1542702888978232,
+      "grad_norm": 0.33886897563934326,
+      "learning_rate": 0.0001,
+      "loss": 1.7177,
+      "step": 1343
+    },
+    {
+      "epoch": 0.15438515880765033,
+      "grad_norm": 0.3549462854862213,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 1344
+    },
+    {
+      "epoch": 0.15450002871747745,
+      "grad_norm": 0.40802568197250366,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 1345
+    },
+    {
+      "epoch": 0.15461489862730457,
+      "grad_norm": 0.37750929594039917,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 1346
+    },
+    {
+      "epoch": 0.1547297685371317,
+      "grad_norm": 0.3724265694618225,
+      "learning_rate": 0.0001,
+      "loss": 1.8619,
+      "step": 1347
+    },
+    {
+      "epoch": 0.1548446384469588,
+      "grad_norm": 0.37521660327911377,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 1348
+    },
+    {
+      "epoch": 0.15495950835678593,
+      "grad_norm": 0.35817015171051025,
+      "learning_rate": 0.0001,
+      "loss": 1.8391,
+      "step": 1349
+    },
+    {
+      "epoch": 0.15507437826661305,
+      "grad_norm": 0.35718318819999695,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 1350
+    },
+    {
+      "epoch": 0.15518924817644017,
+      "grad_norm": 0.3497890830039978,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 1351
+    },
+    {
+      "epoch": 0.1553041180862673,
+      "grad_norm": 0.3606964349746704,
+      "learning_rate": 0.0001,
+      "loss": 1.6408,
+      "step": 1352
+    },
+    {
+      "epoch": 0.1554189879960944,
+      "grad_norm": 0.367531955242157,
+      "learning_rate": 0.0001,
+      "loss": 2.0184,
+      "step": 1353
+    },
+    {
+      "epoch": 0.15553385790592156,
+      "grad_norm": 0.364165723323822,
+      "learning_rate": 0.0001,
+      "loss": 1.7831,
+      "step": 1354
+    },
+    {
+      "epoch": 0.15564872781574868,
+      "grad_norm": 0.3798482418060303,
+      "learning_rate": 0.0001,
+      "loss": 1.9703,
+      "step": 1355
+    },
+    {
+      "epoch": 0.1557635977255758,
+      "grad_norm": 0.38180306553840637,
+      "learning_rate": 0.0001,
+      "loss": 1.9124,
+      "step": 1356
+    },
+    {
+      "epoch": 0.15587846763540292,
+      "grad_norm": 0.34865236282348633,
+      "learning_rate": 0.0001,
+      "loss": 1.6616,
+      "step": 1357
+    },
+    {
+      "epoch": 0.15599333754523004,
+      "grad_norm": 0.3822656571865082,
+      "learning_rate": 0.0001,
+      "loss": 1.8847,
+      "step": 1358
+    },
+    {
+      "epoch": 0.15610820745505716,
+      "grad_norm": 0.37447991967201233,
+      "learning_rate": 0.0001,
+      "loss": 1.7326,
+      "step": 1359
+    },
+    {
+      "epoch": 0.15622307736488428,
+      "grad_norm": 0.39164337515830994,
+      "learning_rate": 0.0001,
+      "loss": 1.9483,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1563379472747114,
+      "grad_norm": 0.37871086597442627,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 1361
+    },
+    {
+      "epoch": 0.15645281718453852,
+      "grad_norm": 0.4217022657394409,
+      "learning_rate": 0.0001,
+      "loss": 1.7808,
+      "step": 1362
+    },
+    {
+      "epoch": 0.15656768709436564,
+      "grad_norm": 0.35998785495758057,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 1363
+    },
+    {
+      "epoch": 0.15668255700419276,
+      "grad_norm": 0.36455628275871277,
+      "learning_rate": 0.0001,
+      "loss": 1.5796,
+      "step": 1364
+    },
+    {
+      "epoch": 0.15679742691401988,
+      "grad_norm": 0.33390358090400696,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 1365
+    },
+    {
+      "epoch": 0.156912296823847,
+      "grad_norm": 0.35853371024131775,
+      "learning_rate": 0.0001,
+      "loss": 1.8491,
+      "step": 1366
+    },
+    {
+      "epoch": 0.15702716673367412,
+      "grad_norm": 0.39624473452568054,
+      "learning_rate": 0.0001,
+      "loss": 1.8059,
+      "step": 1367
+    },
+    {
+      "epoch": 0.15714203664350124,
+      "grad_norm": 0.341155081987381,
+      "learning_rate": 0.0001,
+      "loss": 1.6845,
+      "step": 1368
+    },
+    {
+      "epoch": 0.15725690655332836,
+      "grad_norm": 0.3553493320941925,
+      "learning_rate": 0.0001,
+      "loss": 1.7542,
+      "step": 1369
+    },
+    {
+      "epoch": 0.15737177646315548,
+      "grad_norm": 0.3464072644710541,
+      "learning_rate": 0.0001,
+      "loss": 1.6961,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1574866463729826,
+      "grad_norm": 0.32570725679397583,
+      "learning_rate": 0.0001,
+      "loss": 1.4859,
+      "step": 1371
+    },
+    {
+      "epoch": 0.15760151628280972,
+      "grad_norm": 0.3374817371368408,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 1372
+    },
+    {
+      "epoch": 0.15771638619263684,
+      "grad_norm": 0.3570788502693176,
+      "learning_rate": 0.0001,
+      "loss": 1.7114,
+      "step": 1373
+    },
+    {
+      "epoch": 0.15783125610246396,
+      "grad_norm": 0.3562948703765869,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 1374
+    },
+    {
+      "epoch": 0.15794612601229108,
+      "grad_norm": 0.3416786789894104,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1580609959221182,
+      "grad_norm": 0.36627301573753357,
+      "learning_rate": 0.0001,
+      "loss": 1.8055,
+      "step": 1376
+    },
+    {
+      "epoch": 0.15817586583194532,
+      "grad_norm": 0.38520297408103943,
+      "learning_rate": 0.0001,
+      "loss": 1.7763,
+      "step": 1377
+    },
+    {
+      "epoch": 0.15829073574177244,
+      "grad_norm": 0.35688209533691406,
+      "learning_rate": 0.0001,
+      "loss": 1.7664,
+      "step": 1378
+    },
+    {
+      "epoch": 0.15840560565159956,
+      "grad_norm": 0.3223513960838318,
+      "learning_rate": 0.0001,
+      "loss": 1.6262,
+      "step": 1379
+    },
+    {
+      "epoch": 0.15852047556142668,
+      "grad_norm": 0.3319501578807831,
+      "learning_rate": 0.0001,
+      "loss": 1.6991,
+      "step": 1380
+    },
+    {
+      "epoch": 0.1586353454712538,
+      "grad_norm": 0.3719445765018463,
+      "learning_rate": 0.0001,
+      "loss": 1.7745,
+      "step": 1381
+    },
+    {
+      "epoch": 0.15875021538108092,
+      "grad_norm": 0.41066795587539673,
+      "learning_rate": 0.0001,
+      "loss": 1.7897,
+      "step": 1382
+    },
+    {
+      "epoch": 0.15886508529090804,
+      "grad_norm": 0.38358965516090393,
+      "learning_rate": 0.0001,
+      "loss": 1.9239,
+      "step": 1383
+    },
+    {
+      "epoch": 0.15897995520073516,
+      "grad_norm": 0.35356229543685913,
+      "learning_rate": 0.0001,
+      "loss": 1.7281,
+      "step": 1384
+    },
+    {
+      "epoch": 0.15909482511056228,
+      "grad_norm": 0.36122894287109375,
+      "learning_rate": 0.0001,
+      "loss": 1.7863,
+      "step": 1385
+    },
+    {
+      "epoch": 0.1592096950203894,
+      "grad_norm": 0.35605597496032715,
+      "learning_rate": 0.0001,
+      "loss": 1.7772,
+      "step": 1386
+    },
+    {
+      "epoch": 0.15932456493021652,
+      "grad_norm": 0.3338839113712311,
+      "learning_rate": 0.0001,
+      "loss": 1.4497,
+      "step": 1387
+    },
+    {
+      "epoch": 0.15943943484004364,
+      "grad_norm": 0.34843042492866516,
+      "learning_rate": 0.0001,
+      "loss": 1.8037,
+      "step": 1388
+    },
+    {
+      "epoch": 0.15955430474987078,
+      "grad_norm": 0.3467456102371216,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 1389
+    },
+    {
+      "epoch": 0.1596691746596979,
+      "grad_norm": 0.37024548649787903,
+      "learning_rate": 0.0001,
+      "loss": 1.888,
+      "step": 1390
+    },
+    {
+      "epoch": 0.15978404456952502,
+      "grad_norm": 0.3692200481891632,
+      "learning_rate": 0.0001,
+      "loss": 1.7917,
+      "step": 1391
+    },
+    {
+      "epoch": 0.15989891447935214,
+      "grad_norm": 0.32846784591674805,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 1392
+    },
+    {
+      "epoch": 0.16001378438917926,
+      "grad_norm": 0.3437194228172302,
+      "learning_rate": 0.0001,
+      "loss": 1.7055,
+      "step": 1393
+    },
+    {
+      "epoch": 0.16012865429900638,
+      "grad_norm": 0.346202552318573,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 1394
+    },
+    {
+      "epoch": 0.1602435242088335,
+      "grad_norm": 0.3795225918292999,
+      "learning_rate": 0.0001,
+      "loss": 1.9109,
+      "step": 1395
+    },
+    {
+      "epoch": 0.16035839411866062,
+      "grad_norm": 0.33237701654434204,
+      "learning_rate": 0.0001,
+      "loss": 1.4413,
+      "step": 1396
+    },
+    {
+      "epoch": 0.16047326402848774,
+      "grad_norm": 0.38135427236557007,
+      "learning_rate": 0.0001,
+      "loss": 1.767,
+      "step": 1397
+    },
+    {
+      "epoch": 0.16058813393831486,
+      "grad_norm": 0.37453657388687134,
+      "learning_rate": 0.0001,
+      "loss": 1.7438,
+      "step": 1398
+    },
+    {
+      "epoch": 0.16070300384814198,
+      "grad_norm": 0.3387562334537506,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1608178737579691,
+      "grad_norm": 0.33894360065460205,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 1400
+    },
+    {
+      "epoch": 0.16093274366779622,
+      "grad_norm": 0.3334555923938751,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 1401
+    },
+    {
+      "epoch": 0.16104761357762334,
+      "grad_norm": 0.36844760179519653,
+      "learning_rate": 0.0001,
+      "loss": 1.7755,
+      "step": 1402
+    },
+    {
+      "epoch": 0.16116248348745046,
+      "grad_norm": 0.34594935178756714,
+      "learning_rate": 0.0001,
+      "loss": 1.8161,
+      "step": 1403
+    },
+    {
+      "epoch": 0.16127735339727758,
+      "grad_norm": 0.37070515751838684,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1613922233071047,
+      "grad_norm": 0.3500889241695404,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 1405
+    },
+    {
+      "epoch": 0.16150709321693182,
+      "grad_norm": 0.3509732186794281,
+      "learning_rate": 0.0001,
+      "loss": 1.7667,
+      "step": 1406
+    },
+    {
+      "epoch": 0.16162196312675894,
+      "grad_norm": 0.33721843361854553,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 1407
+    },
+    {
+      "epoch": 0.16173683303658606,
+      "grad_norm": 0.3420223593711853,
+      "learning_rate": 0.0001,
+      "loss": 1.3828,
+      "step": 1408
+    },
+    {
+      "epoch": 0.16185170294641318,
+      "grad_norm": 0.3664703965187073,
+      "learning_rate": 0.0001,
+      "loss": 1.8106,
+      "step": 1409
+    },
+    {
+      "epoch": 0.1619665728562403,
+      "grad_norm": 0.4004788398742676,
+      "learning_rate": 0.0001,
+      "loss": 1.8107,
+      "step": 1410
+    },
+    {
+      "epoch": 0.16208144276606742,
+      "grad_norm": 0.3599262237548828,
+      "learning_rate": 0.0001,
+      "loss": 1.7453,
+      "step": 1411
+    },
+    {
+      "epoch": 0.16219631267589454,
+      "grad_norm": 0.37064090371131897,
+      "learning_rate": 0.0001,
+      "loss": 1.7842,
+      "step": 1412
+    },
+    {
+      "epoch": 0.16231118258572166,
+      "grad_norm": 0.3801650404930115,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 1413
+    },
+    {
+      "epoch": 0.16242605249554878,
+      "grad_norm": 0.3450910449028015,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 1414
+    },
+    {
+      "epoch": 0.1625409224053759,
+      "grad_norm": 0.35267990827560425,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 1415
+    },
+    {
+      "epoch": 0.16265579231520302,
+      "grad_norm": 0.3844035267829895,
+      "learning_rate": 0.0001,
+      "loss": 1.9421,
+      "step": 1416
+    },
+    {
+      "epoch": 0.16277066222503014,
+      "grad_norm": 0.35612425208091736,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 1417
+    },
+    {
+      "epoch": 0.16288553213485726,
+      "grad_norm": 0.3794462978839874,
+      "learning_rate": 0.0001,
+      "loss": 1.8599,
+      "step": 1418
+    },
+    {
+      "epoch": 0.16300040204468438,
+      "grad_norm": 0.33937835693359375,
+      "learning_rate": 0.0001,
+      "loss": 1.6891,
+      "step": 1419
+    },
+    {
+      "epoch": 0.1631152719545115,
+      "grad_norm": 0.3379872441291809,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 1420
+    },
+    {
+      "epoch": 0.16323014186433862,
+      "grad_norm": 0.35873672366142273,
+      "learning_rate": 0.0001,
+      "loss": 1.7832,
+      "step": 1421
+    },
+    {
+      "epoch": 0.16334501177416574,
+      "grad_norm": 0.37622302770614624,
+      "learning_rate": 0.0001,
+      "loss": 1.716,
+      "step": 1422
+    },
+    {
+      "epoch": 0.1634598816839929,
+      "grad_norm": 0.34433531761169434,
+      "learning_rate": 0.0001,
+      "loss": 1.6682,
+      "step": 1423
+    },
+    {
+      "epoch": 0.16357475159382,
+      "grad_norm": 0.35809025168418884,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 1424
+    },
+    {
+      "epoch": 0.16368962150364713,
+      "grad_norm": 0.35675248503685,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 1425
+    },
+    {
+      "epoch": 0.16380449141347425,
+      "grad_norm": 0.356037437915802,
+      "learning_rate": 0.0001,
+      "loss": 1.5049,
+      "step": 1426
+    },
+    {
+      "epoch": 0.16391936132330137,
+      "grad_norm": 0.3485028147697449,
+      "learning_rate": 0.0001,
+      "loss": 1.7554,
+      "step": 1427
+    },
+    {
+      "epoch": 0.1640342312331285,
+      "grad_norm": 0.36230984330177307,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 1428
+    },
+    {
+      "epoch": 0.1641491011429556,
+      "grad_norm": 0.35187479853630066,
+      "learning_rate": 0.0001,
+      "loss": 1.7184,
+      "step": 1429
+    },
+    {
+      "epoch": 0.16426397105278273,
+      "grad_norm": 0.34478455781936646,
+      "learning_rate": 0.0001,
+      "loss": 1.7086,
+      "step": 1430
+    },
+    {
+      "epoch": 0.16437884096260985,
+      "grad_norm": 0.4025746285915375,
+      "learning_rate": 0.0001,
+      "loss": 1.8836,
+      "step": 1431
+    },
+    {
+      "epoch": 0.16449371087243697,
+      "grad_norm": 0.358385294675827,
+      "learning_rate": 0.0001,
+      "loss": 1.8405,
+      "step": 1432
+    },
+    {
+      "epoch": 0.1646085807822641,
+      "grad_norm": 0.3677537143230438,
+      "learning_rate": 0.0001,
+      "loss": 1.6616,
+      "step": 1433
+    },
+    {
+      "epoch": 0.1647234506920912,
+      "grad_norm": 0.3488091826438904,
+      "learning_rate": 0.0001,
+      "loss": 1.6911,
+      "step": 1434
+    },
+    {
+      "epoch": 0.16483832060191833,
+      "grad_norm": 0.3559654951095581,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 1435
+    },
+    {
+      "epoch": 0.16495319051174545,
+      "grad_norm": 0.4024467170238495,
+      "learning_rate": 0.0001,
+      "loss": 1.8107,
+      "step": 1436
+    },
+    {
+      "epoch": 0.16506806042157257,
+      "grad_norm": 0.3618294298648834,
+      "learning_rate": 0.0001,
+      "loss": 1.8111,
+      "step": 1437
+    },
+    {
+      "epoch": 0.1651829303313997,
+      "grad_norm": 0.3504227101802826,
+      "learning_rate": 0.0001,
+      "loss": 1.639,
+      "step": 1438
+    },
+    {
+      "epoch": 0.1652978002412268,
+      "grad_norm": 0.36997562646865845,
+      "learning_rate": 0.0001,
+      "loss": 1.8394,
+      "step": 1439
+    },
+    {
+      "epoch": 0.16541267015105393,
+      "grad_norm": 0.3473089635372162,
+      "learning_rate": 0.0001,
+      "loss": 1.7694,
+      "step": 1440
+    },
+    {
+      "epoch": 0.16552754006088105,
+      "grad_norm": 0.3754732012748718,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 1441
+    },
+    {
+      "epoch": 0.16564240997070817,
+      "grad_norm": 0.318775475025177,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 1442
+    },
+    {
+      "epoch": 0.1657572798805353,
+      "grad_norm": 0.3740909695625305,
+      "learning_rate": 0.0001,
+      "loss": 1.7752,
+      "step": 1443
+    },
+    {
+      "epoch": 0.1658721497903624,
+      "grad_norm": 0.34481796622276306,
+      "learning_rate": 0.0001,
+      "loss": 1.7427,
+      "step": 1444
+    },
+    {
+      "epoch": 0.16598701970018953,
+      "grad_norm": 0.3509489595890045,
+      "learning_rate": 0.0001,
+      "loss": 1.8072,
+      "step": 1445
+    },
+    {
+      "epoch": 0.16610188961001665,
+      "grad_norm": 0.364310085773468,
+      "learning_rate": 0.0001,
+      "loss": 1.7988,
+      "step": 1446
+    },
+    {
+      "epoch": 0.16621675951984377,
+      "grad_norm": 0.36774981021881104,
+      "learning_rate": 0.0001,
+      "loss": 1.8602,
+      "step": 1447
+    },
+    {
+      "epoch": 0.1663316294296709,
+      "grad_norm": 0.34403902292251587,
+      "learning_rate": 0.0001,
+      "loss": 1.7574,
+      "step": 1448
+    },
+    {
+      "epoch": 0.166446499339498,
+      "grad_norm": 0.3631366193294525,
+      "learning_rate": 0.0001,
+      "loss": 1.7609,
+      "step": 1449
+    },
+    {
+      "epoch": 0.16656136924932513,
+      "grad_norm": 0.4153347313404083,
+      "learning_rate": 0.0001,
+      "loss": 1.8153,
+      "step": 1450
+    },
+    {
+      "epoch": 0.16667623915915225,
+      "grad_norm": 0.3547952175140381,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 1451
+    },
+    {
+      "epoch": 0.16679110906897937,
+      "grad_norm": 0.3516184091567993,
+      "learning_rate": 0.0001,
+      "loss": 1.6451,
+      "step": 1452
+    },
+    {
+      "epoch": 0.1669059789788065,
+      "grad_norm": 0.35548651218414307,
+      "learning_rate": 0.0001,
+      "loss": 1.7025,
+      "step": 1453
+    },
+    {
+      "epoch": 0.1670208488886336,
+      "grad_norm": 0.3711240291595459,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 1454
+    },
+    {
+      "epoch": 0.16713571879846073,
+      "grad_norm": 0.3621267080307007,
+      "learning_rate": 0.0001,
+      "loss": 1.587,
+      "step": 1455
+    },
+    {
+      "epoch": 0.16725058870828785,
+      "grad_norm": 0.36165568232536316,
+      "learning_rate": 0.0001,
+      "loss": 1.7507,
+      "step": 1456
+    },
+    {
+      "epoch": 0.167365458618115,
+      "grad_norm": 0.3373228907585144,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 1457
+    },
+    {
+      "epoch": 0.16748032852794212,
+      "grad_norm": 0.3716026842594147,
+      "learning_rate": 0.0001,
+      "loss": 1.7762,
+      "step": 1458
+    },
+    {
+      "epoch": 0.16759519843776924,
+      "grad_norm": 0.3812613785266876,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 1459
+    },
+    {
+      "epoch": 0.16771006834759636,
+      "grad_norm": 0.3630736470222473,
+      "learning_rate": 0.0001,
+      "loss": 1.7054,
+      "step": 1460
+    },
+    {
+      "epoch": 0.16782493825742348,
+      "grad_norm": 0.3346702754497528,
+      "learning_rate": 0.0001,
+      "loss": 1.6076,
+      "step": 1461
+    },
+    {
+      "epoch": 0.1679398081672506,
+      "grad_norm": 0.3383999764919281,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 1462
+    },
+    {
+      "epoch": 0.16805467807707772,
+      "grad_norm": 0.3548593521118164,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 1463
+    },
+    {
+      "epoch": 0.16816954798690484,
+      "grad_norm": 0.3538789451122284,
+      "learning_rate": 0.0001,
+      "loss": 1.657,
+      "step": 1464
+    },
+    {
+      "epoch": 0.16828441789673196,
+      "grad_norm": 0.3505842387676239,
+      "learning_rate": 0.0001,
+      "loss": 1.7951,
+      "step": 1465
+    },
+    {
+      "epoch": 0.16839928780655908,
+      "grad_norm": 0.3387717604637146,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 1466
+    },
+    {
+      "epoch": 0.1685141577163862,
+      "grad_norm": 0.3535427153110504,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 1467
+    },
+    {
+      "epoch": 0.16862902762621332,
+      "grad_norm": 0.35021501779556274,
+      "learning_rate": 0.0001,
+      "loss": 1.6706,
+      "step": 1468
+    },
+    {
+      "epoch": 0.16874389753604044,
+      "grad_norm": 0.36078891158103943,
+      "learning_rate": 0.0001,
+      "loss": 1.7588,
+      "step": 1469
+    },
+    {
+      "epoch": 0.16885876744586756,
+      "grad_norm": 0.3418395221233368,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 1470
+    },
+    {
+      "epoch": 0.16897363735569468,
+      "grad_norm": 0.37558332085609436,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 1471
+    },
+    {
+      "epoch": 0.1690885072655218,
+      "grad_norm": 0.3494422137737274,
+      "learning_rate": 0.0001,
+      "loss": 1.7267,
+      "step": 1472
+    },
+    {
+      "epoch": 0.16920337717534892,
+      "grad_norm": 0.35918739438056946,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 1473
+    },
+    {
+      "epoch": 0.16931824708517604,
+      "grad_norm": 0.36204949021339417,
+      "learning_rate": 0.0001,
+      "loss": 1.8109,
+      "step": 1474
+    },
+    {
+      "epoch": 0.16943311699500316,
+      "grad_norm": 0.35251131653785706,
+      "learning_rate": 0.0001,
+      "loss": 1.6827,
+      "step": 1475
+    },
+    {
+      "epoch": 0.16954798690483028,
+      "grad_norm": 0.35120296478271484,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 1476
+    },
+    {
+      "epoch": 0.1696628568146574,
+      "grad_norm": 0.34975898265838623,
+      "learning_rate": 0.0001,
+      "loss": 1.7992,
+      "step": 1477
+    },
+    {
+      "epoch": 0.16977772672448452,
+      "grad_norm": 0.33744266629219055,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 1478
+    },
+    {
+      "epoch": 0.16989259663431164,
+      "grad_norm": 0.34000107645988464,
+      "learning_rate": 0.0001,
+      "loss": 1.6484,
+      "step": 1479
+    },
+    {
+      "epoch": 0.17000746654413876,
+      "grad_norm": 0.36184847354888916,
+      "learning_rate": 0.0001,
+      "loss": 1.7398,
+      "step": 1480
+    },
+    {
+      "epoch": 0.17012233645396588,
+      "grad_norm": 0.3685035705566406,
+      "learning_rate": 0.0001,
+      "loss": 1.8032,
+      "step": 1481
+    },
+    {
+      "epoch": 0.170237206363793,
+      "grad_norm": 0.38592687249183655,
+      "learning_rate": 0.0001,
+      "loss": 1.7185,
+      "step": 1482
+    },
+    {
+      "epoch": 0.17035207627362012,
+      "grad_norm": 0.3724033832550049,
+      "learning_rate": 0.0001,
+      "loss": 1.7696,
+      "step": 1483
+    },
+    {
+      "epoch": 0.17046694618344724,
+      "grad_norm": 0.3662974238395691,
+      "learning_rate": 0.0001,
+      "loss": 1.727,
+      "step": 1484
+    },
+    {
+      "epoch": 0.17058181609327436,
+      "grad_norm": 0.38893744349479675,
+      "learning_rate": 0.0001,
+      "loss": 1.746,
+      "step": 1485
+    },
+    {
+      "epoch": 0.17069668600310148,
+      "grad_norm": 0.368671178817749,
+      "learning_rate": 0.0001,
+      "loss": 1.7863,
+      "step": 1486
+    },
+    {
+      "epoch": 0.1708115559129286,
+      "grad_norm": 0.3590819835662842,
+      "learning_rate": 0.0001,
+      "loss": 1.8859,
+      "step": 1487
+    },
+    {
+      "epoch": 0.17092642582275572,
+      "grad_norm": 0.36079901456832886,
+      "learning_rate": 0.0001,
+      "loss": 1.7423,
+      "step": 1488
+    },
+    {
+      "epoch": 0.17104129573258284,
+      "grad_norm": 0.355546772480011,
+      "learning_rate": 0.0001,
+      "loss": 1.7369,
+      "step": 1489
+    },
+    {
+      "epoch": 0.17115616564240996,
+      "grad_norm": 0.3821921944618225,
+      "learning_rate": 0.0001,
+      "loss": 1.9179,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1712710355522371,
+      "grad_norm": 0.3505462110042572,
+      "learning_rate": 0.0001,
+      "loss": 1.871,
+      "step": 1491
+    },
+    {
+      "epoch": 0.17138590546206423,
+      "grad_norm": 0.3656969368457794,
+      "learning_rate": 0.0001,
+      "loss": 1.6599,
+      "step": 1492
+    },
+    {
+      "epoch": 0.17150077537189135,
+      "grad_norm": 0.3786547780036926,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 1493
+    },
+    {
+      "epoch": 0.17161564528171847,
+      "grad_norm": 0.37065404653549194,
+      "learning_rate": 0.0001,
+      "loss": 1.6333,
+      "step": 1494
+    },
+    {
+      "epoch": 0.17173051519154559,
+      "grad_norm": 0.3699958622455597,
+      "learning_rate": 0.0001,
+      "loss": 1.7745,
+      "step": 1495
+    },
+    {
+      "epoch": 0.1718453851013727,
+      "grad_norm": 0.3573478162288666,
+      "learning_rate": 0.0001,
+      "loss": 1.6497,
+      "step": 1496
+    },
+    {
+      "epoch": 0.17196025501119983,
+      "grad_norm": 0.3474213778972626,
+      "learning_rate": 0.0001,
+      "loss": 1.5043,
+      "step": 1497
+    },
+    {
+      "epoch": 0.17207512492102695,
+      "grad_norm": 0.3627040684223175,
+      "learning_rate": 0.0001,
+      "loss": 1.7553,
+      "step": 1498
+    },
+    {
+      "epoch": 0.17218999483085407,
+      "grad_norm": 0.34735116362571716,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 1499
+    },
+    {
+      "epoch": 0.17230486474068119,
+      "grad_norm": 0.4130633771419525,
+      "learning_rate": 0.0001,
+      "loss": 1.8155,
+      "step": 1500
+    },
+    {
+      "epoch": 0.1724197346505083,
+      "grad_norm": 0.38091927766799927,
+      "learning_rate": 0.0001,
+      "loss": 1.7231,
+      "step": 1501
+    },
+    {
+      "epoch": 0.17253460456033543,
+      "grad_norm": 0.39104804396629333,
+      "learning_rate": 0.0001,
+      "loss": 1.756,
+      "step": 1502
+    },
+    {
+      "epoch": 0.17264947447016255,
+      "grad_norm": 0.39437583088874817,
+      "learning_rate": 0.0001,
+      "loss": 1.7184,
+      "step": 1503
+    },
+    {
+      "epoch": 0.17276434437998966,
+      "grad_norm": 0.37000584602355957,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 1504
+    },
+    {
+      "epoch": 0.17287921428981678,
+      "grad_norm": 0.37976545095443726,
+      "learning_rate": 0.0001,
+      "loss": 1.8923,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1729940841996439,
+      "grad_norm": 0.36573851108551025,
+      "learning_rate": 0.0001,
+      "loss": 1.8202,
+      "step": 1506
+    },
+    {
+      "epoch": 0.17310895410947102,
+      "grad_norm": 0.37791380286216736,
+      "learning_rate": 0.0001,
+      "loss": 1.7888,
+      "step": 1507
+    },
+    {
+      "epoch": 0.17322382401929814,
+      "grad_norm": 0.3388189673423767,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 1508
+    },
+    {
+      "epoch": 0.17333869392912526,
+      "grad_norm": 0.36455753445625305,
+      "learning_rate": 0.0001,
+      "loss": 1.8167,
+      "step": 1509
+    },
+    {
+      "epoch": 0.17345356383895238,
+      "grad_norm": 0.3855915665626526,
+      "learning_rate": 0.0001,
+      "loss": 1.8208,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1735684337487795,
+      "grad_norm": 0.3601621687412262,
+      "learning_rate": 0.0001,
+      "loss": 1.8135,
+      "step": 1511
+    },
+    {
+      "epoch": 0.17368330365860662,
+      "grad_norm": 0.34166282415390015,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 1512
+    },
+    {
+      "epoch": 0.17379817356843374,
+      "grad_norm": 0.36744624376296997,
+      "learning_rate": 0.0001,
+      "loss": 1.7702,
+      "step": 1513
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 0.3522723913192749,
+      "learning_rate": 0.0001,
+      "loss": 1.7836,
+      "step": 1514
+    },
+    {
+      "epoch": 0.17402791338808798,
+      "grad_norm": 0.34597423672676086,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1741427832979151,
+      "grad_norm": 0.33401763439178467,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 1516
+    },
+    {
+      "epoch": 0.17425765320774222,
+      "grad_norm": 0.3676266074180603,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 1517
+    },
+    {
+      "epoch": 0.17437252311756934,
+      "grad_norm": 0.3546941578388214,
+      "learning_rate": 0.0001,
+      "loss": 1.8186,
+      "step": 1518
+    },
+    {
+      "epoch": 0.17448739302739646,
+      "grad_norm": 0.37473762035369873,
+      "learning_rate": 0.0001,
+      "loss": 1.7717,
+      "step": 1519
+    },
+    {
+      "epoch": 0.17460226293722358,
+      "grad_norm": 0.35405465960502625,
+      "learning_rate": 0.0001,
+      "loss": 1.683,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1747171328470507,
+      "grad_norm": 0.3613182306289673,
+      "learning_rate": 0.0001,
+      "loss": 1.7042,
+      "step": 1521
+    },
+    {
+      "epoch": 0.17483200275687782,
+      "grad_norm": 0.3764897882938385,
+      "learning_rate": 0.0001,
+      "loss": 1.834,
+      "step": 1522
+    },
+    {
+      "epoch": 0.17494687266670494,
+      "grad_norm": 0.3841586410999298,
+      "learning_rate": 0.0001,
+      "loss": 1.8038,
+      "step": 1523
+    },
+    {
+      "epoch": 0.17506174257653206,
+      "grad_norm": 0.37141889333724976,
+      "learning_rate": 0.0001,
+      "loss": 1.5864,
+      "step": 1524
+    },
+    {
+      "epoch": 0.17517661248635918,
+      "grad_norm": 0.37981778383255005,
+      "learning_rate": 0.0001,
+      "loss": 1.772,
+      "step": 1525
+    },
+    {
+      "epoch": 0.17529148239618633,
+      "grad_norm": 0.40624651312828064,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 1526
+    },
+    {
+      "epoch": 0.17540635230601345,
+      "grad_norm": 0.37434014678001404,
+      "learning_rate": 0.0001,
+      "loss": 1.7501,
+      "step": 1527
+    },
+    {
+      "epoch": 0.17552122221584057,
+      "grad_norm": 0.3867623805999756,
+      "learning_rate": 0.0001,
+      "loss": 1.6846,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1756360921256677,
+      "grad_norm": 0.384644478559494,
+      "learning_rate": 0.0001,
+      "loss": 1.7358,
+      "step": 1529
+    },
+    {
+      "epoch": 0.1757509620354948,
+      "grad_norm": 0.36406537890434265,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 1530
+    },
+    {
+      "epoch": 0.17586583194532193,
+      "grad_norm": 0.3523077070713043,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 1531
+    },
+    {
+      "epoch": 0.17598070185514905,
+      "grad_norm": 0.3456611633300781,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 1532
+    },
+    {
+      "epoch": 0.17609557176497617,
+      "grad_norm": 0.4034580588340759,
+      "learning_rate": 0.0001,
+      "loss": 1.8922,
+      "step": 1533
+    },
+    {
+      "epoch": 0.1762104416748033,
+      "grad_norm": 0.3668345510959625,
+      "learning_rate": 0.0001,
+      "loss": 1.7633,
+      "step": 1534
+    },
+    {
+      "epoch": 0.1763253115846304,
+      "grad_norm": 0.3617863059043884,
+      "learning_rate": 0.0001,
+      "loss": 1.766,
+      "step": 1535
+    },
+    {
+      "epoch": 0.17644018149445753,
+      "grad_norm": 0.4189690053462982,
+      "learning_rate": 0.0001,
+      "loss": 1.8517,
+      "step": 1536
+    },
+    {
+      "epoch": 0.17655505140428465,
+      "grad_norm": 0.36103829741477966,
+      "learning_rate": 0.0001,
+      "loss": 1.7356,
+      "step": 1537
+    },
+    {
+      "epoch": 0.17666992131411177,
+      "grad_norm": 0.3502132296562195,
+      "learning_rate": 0.0001,
+      "loss": 1.7039,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1767847912239389,
+      "grad_norm": 0.4156895577907562,
+      "learning_rate": 0.0001,
+      "loss": 1.8255,
+      "step": 1539
+    },
+    {
+      "epoch": 0.176899661133766,
+      "grad_norm": 0.35794487595558167,
+      "learning_rate": 0.0001,
+      "loss": 1.762,
+      "step": 1540
+    },
+    {
+      "epoch": 0.17701453104359313,
+      "grad_norm": 0.3665020763874054,
+      "learning_rate": 0.0001,
+      "loss": 1.7417,
+      "step": 1541
+    },
+    {
+      "epoch": 0.17712940095342025,
+      "grad_norm": 0.42144718766212463,
+      "learning_rate": 0.0001,
+      "loss": 1.9003,
+      "step": 1542
+    },
+    {
+      "epoch": 0.17724427086324737,
+      "grad_norm": 0.3615649342536926,
+      "learning_rate": 0.0001,
+      "loss": 1.4503,
+      "step": 1543
+    },
+    {
+      "epoch": 0.1773591407730745,
+      "grad_norm": 0.36150482296943665,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 1544
+    },
+    {
+      "epoch": 0.1774740106829016,
+      "grad_norm": 0.39197593927383423,
+      "learning_rate": 0.0001,
+      "loss": 1.717,
+      "step": 1545
+    },
+    {
+      "epoch": 0.17758888059272873,
+      "grad_norm": 0.40486010909080505,
+      "learning_rate": 0.0001,
+      "loss": 1.7411,
+      "step": 1546
+    },
+    {
+      "epoch": 0.17770375050255585,
+      "grad_norm": 0.3970898389816284,
+      "learning_rate": 0.0001,
+      "loss": 2.0669,
+      "step": 1547
+    },
+    {
+      "epoch": 0.17781862041238297,
+      "grad_norm": 0.3371671736240387,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 1548
+    },
+    {
+      "epoch": 0.1779334903222101,
+      "grad_norm": 0.33657070994377136,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 1549
+    },
+    {
+      "epoch": 0.1780483602320372,
+      "grad_norm": 0.34936872124671936,
+      "learning_rate": 0.0001,
+      "loss": 1.6385,
+      "step": 1550
+    },
+    {
+      "epoch": 0.17816323014186433,
+      "grad_norm": 0.353533536195755,
+      "learning_rate": 0.0001,
+      "loss": 1.6985,
+      "step": 1551
+    },
+    {
+      "epoch": 0.17827810005169145,
+      "grad_norm": 0.3584658205509186,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 1552
+    },
+    {
+      "epoch": 0.17839296996151857,
+      "grad_norm": 0.3596382141113281,
+      "learning_rate": 0.0001,
+      "loss": 1.8855,
+      "step": 1553
+    },
+    {
+      "epoch": 0.1785078398713457,
+      "grad_norm": 0.3663223683834076,
+      "learning_rate": 0.0001,
+      "loss": 1.5073,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1786227097811728,
+      "grad_norm": 0.39811477065086365,
+      "learning_rate": 0.0001,
+      "loss": 1.7019,
+      "step": 1555
+    },
+    {
+      "epoch": 0.17873757969099993,
+      "grad_norm": 0.3574909567832947,
+      "learning_rate": 0.0001,
+      "loss": 1.7678,
+      "step": 1556
+    },
+    {
+      "epoch": 0.17885244960082705,
+      "grad_norm": 0.3777164816856384,
+      "learning_rate": 0.0001,
+      "loss": 1.8767,
+      "step": 1557
+    },
+    {
+      "epoch": 0.17896731951065417,
+      "grad_norm": 0.3364102244377136,
+      "learning_rate": 0.0001,
+      "loss": 1.6808,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1790821894204813,
+      "grad_norm": 0.36213961243629456,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 1559
+    },
+    {
+      "epoch": 0.17919705933030844,
+      "grad_norm": 0.4215514361858368,
+      "learning_rate": 0.0001,
+      "loss": 1.8289,
+      "step": 1560
+    },
+    {
+      "epoch": 0.17931192924013556,
+      "grad_norm": 0.35058748722076416,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 1561
+    },
+    {
+      "epoch": 0.17942679914996268,
+      "grad_norm": 0.4215516448020935,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 1562
+    },
+    {
+      "epoch": 0.1795416690597898,
+      "grad_norm": 0.3747852146625519,
+      "learning_rate": 0.0001,
+      "loss": 1.6683,
+      "step": 1563
+    },
+    {
+      "epoch": 0.17965653896961692,
+      "grad_norm": 0.3510657846927643,
+      "learning_rate": 0.0001,
+      "loss": 1.6948,
+      "step": 1564
+    },
+    {
+      "epoch": 0.17977140887944404,
+      "grad_norm": 0.3715681731700897,
+      "learning_rate": 0.0001,
+      "loss": 1.7083,
+      "step": 1565
+    },
+    {
+      "epoch": 0.17988627878927116,
+      "grad_norm": 0.3528061509132385,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 1566
+    },
+    {
+      "epoch": 0.18000114869909828,
+      "grad_norm": 0.3377302587032318,
+      "learning_rate": 0.0001,
+      "loss": 1.7079,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1801160186089254,
+      "grad_norm": 0.40321823954582214,
+      "learning_rate": 0.0001,
+      "loss": 1.8162,
+      "step": 1568
+    },
+    {
+      "epoch": 0.18023088851875252,
+      "grad_norm": 0.3601834774017334,
+      "learning_rate": 0.0001,
+      "loss": 1.6868,
+      "step": 1569
+    },
+    {
+      "epoch": 0.18034575842857964,
+      "grad_norm": 0.3452896475791931,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 1570
+    },
+    {
+      "epoch": 0.18046062833840676,
+      "grad_norm": 0.3979194462299347,
+      "learning_rate": 0.0001,
+      "loss": 1.9325,
+      "step": 1571
+    },
+    {
+      "epoch": 0.18057549824823388,
+      "grad_norm": 0.3308473229408264,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 1572
+    },
+    {
+      "epoch": 0.180690368158061,
+      "grad_norm": 0.3808495104312897,
+      "learning_rate": 0.0001,
+      "loss": 1.8788,
+      "step": 1573
+    },
+    {
+      "epoch": 0.18080523806788812,
+      "grad_norm": 0.35744503140449524,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 1574
+    },
+    {
+      "epoch": 0.18092010797771524,
+      "grad_norm": 0.35669422149658203,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 1575
+    },
+    {
+      "epoch": 0.18103497788754236,
+      "grad_norm": 0.3452187776565552,
+      "learning_rate": 0.0001,
+      "loss": 1.6189,
+      "step": 1576
+    },
+    {
+      "epoch": 0.18114984779736948,
+      "grad_norm": 0.36027148365974426,
+      "learning_rate": 0.0001,
+      "loss": 1.5917,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1812647177071966,
+      "grad_norm": 0.32262781262397766,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 1578
+    },
+    {
+      "epoch": 0.18137958761702372,
+      "grad_norm": 0.3979918658733368,
+      "learning_rate": 0.0001,
+      "loss": 1.8699,
+      "step": 1579
+    },
+    {
+      "epoch": 0.18149445752685084,
+      "grad_norm": 0.42020371556282043,
+      "learning_rate": 0.0001,
+      "loss": 1.9041,
+      "step": 1580
+    },
+    {
+      "epoch": 0.18160932743667796,
+      "grad_norm": 0.34996458888053894,
+      "learning_rate": 0.0001,
+      "loss": 1.6857,
+      "step": 1581
+    },
+    {
+      "epoch": 0.18172419734650508,
+      "grad_norm": 0.3742469549179077,
+      "learning_rate": 0.0001,
+      "loss": 1.934,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1818390672563322,
+      "grad_norm": 0.37955376505851746,
+      "learning_rate": 0.0001,
+      "loss": 1.7681,
+      "step": 1583
+    },
+    {
+      "epoch": 0.18195393716615932,
+      "grad_norm": 0.31765666604042053,
+      "learning_rate": 0.0001,
+      "loss": 1.5858,
+      "step": 1584
+    },
+    {
+      "epoch": 0.18206880707598644,
+      "grad_norm": 0.39358144998550415,
+      "learning_rate": 0.0001,
+      "loss": 1.8361,
+      "step": 1585
+    },
+    {
+      "epoch": 0.18218367698581356,
+      "grad_norm": 0.35237935185432434,
+      "learning_rate": 0.0001,
+      "loss": 1.7351,
+      "step": 1586
+    },
+    {
+      "epoch": 0.18229854689564068,
+      "grad_norm": 0.38084107637405396,
+      "learning_rate": 0.0001,
+      "loss": 1.7978,
+      "step": 1587
+    },
+    {
+      "epoch": 0.1824134168054678,
+      "grad_norm": 0.37168824672698975,
+      "learning_rate": 0.0001,
+      "loss": 1.7858,
+      "step": 1588
+    },
+    {
+      "epoch": 0.18252828671529492,
+      "grad_norm": 0.33814021944999695,
+      "learning_rate": 0.0001,
+      "loss": 1.738,
+      "step": 1589
+    },
+    {
+      "epoch": 0.18264315662512204,
+      "grad_norm": 0.39129403233528137,
+      "learning_rate": 0.0001,
+      "loss": 1.9242,
+      "step": 1590
+    },
+    {
+      "epoch": 0.18275802653494916,
+      "grad_norm": 0.3859502971172333,
+      "learning_rate": 0.0001,
+      "loss": 1.8557,
+      "step": 1591
+    },
+    {
+      "epoch": 0.18287289644477628,
+      "grad_norm": 0.3586483895778656,
+      "learning_rate": 0.0001,
+      "loss": 1.7178,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1829877663546034,
+      "grad_norm": 0.3621407449245453,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 1593
+    },
+    {
+      "epoch": 0.18310263626443055,
+      "grad_norm": 0.3588270843029022,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 1594
+    },
+    {
+      "epoch": 0.18321750617425767,
+      "grad_norm": 0.36701640486717224,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 1595
+    },
+    {
+      "epoch": 0.18333237608408479,
+      "grad_norm": 0.3731893301010132,
+      "learning_rate": 0.0001,
+      "loss": 1.5814,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1834472459939119,
+      "grad_norm": 0.3400730788707733,
+      "learning_rate": 0.0001,
+      "loss": 1.5258,
+      "step": 1597
+    },
+    {
+      "epoch": 0.18356211590373903,
+      "grad_norm": 0.3284122943878174,
+      "learning_rate": 0.0001,
+      "loss": 1.3967,
+      "step": 1598
+    },
+    {
+      "epoch": 0.18367698581356615,
+      "grad_norm": 0.3313588798046112,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 1599
+    },
+    {
+      "epoch": 0.18379185572339327,
+      "grad_norm": 0.4088406562805176,
+      "learning_rate": 0.0001,
+      "loss": 1.8716,
+      "step": 1600
+    },
+    {
+      "epoch": 0.18390672563322039,
+      "grad_norm": 0.3627072870731354,
+      "learning_rate": 0.0001,
+      "loss": 1.7801,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1840215955430475,
+      "grad_norm": 0.3515871465206146,
+      "learning_rate": 0.0001,
+      "loss": 1.6491,
+      "step": 1602
+    },
+    {
+      "epoch": 0.18413646545287463,
+      "grad_norm": 0.4161235988140106,
+      "learning_rate": 0.0001,
+      "loss": 1.749,
+      "step": 1603
+    },
+    {
+      "epoch": 0.18425133536270175,
+      "grad_norm": 0.35696670413017273,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 1604
+    },
+    {
+      "epoch": 0.18436620527252887,
+      "grad_norm": 0.3441614508628845,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 1605
+    },
+    {
+      "epoch": 0.18448107518235599,
+      "grad_norm": 0.35174059867858887,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1845959450921831,
+      "grad_norm": 0.3795402944087982,
+      "learning_rate": 0.0001,
+      "loss": 1.7702,
+      "step": 1607
+    },
+    {
+      "epoch": 0.18471081500201023,
+      "grad_norm": 0.3490031957626343,
+      "learning_rate": 0.0001,
+      "loss": 1.6797,
+      "step": 1608
+    },
+    {
+      "epoch": 0.18482568491183735,
+      "grad_norm": 0.39439791440963745,
+      "learning_rate": 0.0001,
+      "loss": 1.8993,
+      "step": 1609
+    },
+    {
+      "epoch": 0.18494055482166447,
+      "grad_norm": 0.3577129542827606,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 1610
+    },
+    {
+      "epoch": 0.18505542473149159,
+      "grad_norm": 0.34242141246795654,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1851702946413187,
+      "grad_norm": 0.3696388304233551,
+      "learning_rate": 0.0001,
+      "loss": 1.7205,
+      "step": 1612
+    },
+    {
+      "epoch": 0.18528516455114583,
+      "grad_norm": 0.3409230411052704,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 1613
+    },
+    {
+      "epoch": 0.18540003446097295,
+      "grad_norm": 0.34659913182258606,
+      "learning_rate": 0.0001,
+      "loss": 1.6436,
+      "step": 1614
+    },
+    {
+      "epoch": 0.18551490437080007,
+      "grad_norm": 0.3633543848991394,
+      "learning_rate": 0.0001,
+      "loss": 1.8687,
+      "step": 1615
+    },
+    {
+      "epoch": 0.18562977428062719,
+      "grad_norm": 0.3456427752971649,
+      "learning_rate": 0.0001,
+      "loss": 1.6913,
+      "step": 1616
+    },
+    {
+      "epoch": 0.1857446441904543,
+      "grad_norm": 0.3466663062572479,
+      "learning_rate": 0.0001,
+      "loss": 1.7164,
+      "step": 1617
+    },
+    {
+      "epoch": 0.18585951410028143,
+      "grad_norm": 0.373751163482666,
+      "learning_rate": 0.0001,
+      "loss": 1.7618,
+      "step": 1618
+    },
+    {
+      "epoch": 0.18597438401010855,
+      "grad_norm": 0.3516460657119751,
+      "learning_rate": 0.0001,
+      "loss": 1.587,
+      "step": 1619
+    },
+    {
+      "epoch": 0.18608925391993567,
+      "grad_norm": 0.36074620485305786,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 1620
+    },
+    {
+      "epoch": 0.18620412382976279,
+      "grad_norm": 0.37144434452056885,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1863189937395899,
+      "grad_norm": 0.34920889139175415,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 1622
+    },
+    {
+      "epoch": 0.18643386364941703,
+      "grad_norm": 0.3639899790287018,
+      "learning_rate": 0.0001,
+      "loss": 1.773,
+      "step": 1623
+    },
+    {
+      "epoch": 0.18654873355924415,
+      "grad_norm": 0.38695764541625977,
+      "learning_rate": 0.0001,
+      "loss": 1.9671,
+      "step": 1624
+    },
+    {
+      "epoch": 0.18666360346907127,
+      "grad_norm": 0.3656146228313446,
+      "learning_rate": 0.0001,
+      "loss": 1.8799,
+      "step": 1625
+    },
+    {
+      "epoch": 0.18677847337889839,
+      "grad_norm": 0.3708580732345581,
+      "learning_rate": 0.0001,
+      "loss": 1.8234,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1868933432887255,
+      "grad_norm": 0.3623522222042084,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 1627
+    },
+    {
+      "epoch": 0.18700821319855265,
+      "grad_norm": 0.3340558409690857,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 1628
+    },
+    {
+      "epoch": 0.18712308310837977,
+      "grad_norm": 0.3819306790828705,
+      "learning_rate": 0.0001,
+      "loss": 1.9233,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1872379530182069,
+      "grad_norm": 0.35179227590560913,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 1630
+    },
+    {
+      "epoch": 0.187352822928034,
+      "grad_norm": 0.3724440634250641,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 1631
+    },
+    {
+      "epoch": 0.18746769283786113,
+      "grad_norm": 0.35643836855888367,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 1632
+    },
+    {
+      "epoch": 0.18758256274768825,
+      "grad_norm": 0.3557715117931366,
+      "learning_rate": 0.0001,
+      "loss": 1.8076,
+      "step": 1633
+    },
+    {
+      "epoch": 0.18769743265751537,
+      "grad_norm": 0.3418234586715698,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1878123025673425,
+      "grad_norm": 0.3681597113609314,
+      "learning_rate": 0.0001,
+      "loss": 1.7233,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1879271724771696,
+      "grad_norm": 0.34994348883628845,
+      "learning_rate": 0.0001,
+      "loss": 1.4936,
+      "step": 1636
+    },
+    {
+      "epoch": 0.18804204238699673,
+      "grad_norm": 0.35400843620300293,
+      "learning_rate": 0.0001,
+      "loss": 1.5871,
+      "step": 1637
+    },
+    {
+      "epoch": 0.18815691229682385,
+      "grad_norm": 0.3841044008731842,
+      "learning_rate": 0.0001,
+      "loss": 1.8312,
+      "step": 1638
+    },
+    {
+      "epoch": 0.18827178220665097,
+      "grad_norm": 0.3630238473415375,
+      "learning_rate": 0.0001,
+      "loss": 1.7448,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1883866521164781,
+      "grad_norm": 0.3915660083293915,
+      "learning_rate": 0.0001,
+      "loss": 1.8261,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1885015220263052,
+      "grad_norm": 0.3483685851097107,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 1641
+    },
+    {
+      "epoch": 0.18861639193613233,
+      "grad_norm": 0.35304656624794006,
+      "learning_rate": 0.0001,
+      "loss": 1.6511,
+      "step": 1642
+    },
+    {
+      "epoch": 0.18873126184595945,
+      "grad_norm": 0.3424839973449707,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 1643
+    },
+    {
+      "epoch": 0.18884613175578657,
+      "grad_norm": 0.3762650489807129,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 1644
+    },
+    {
+      "epoch": 0.1889610016656137,
+      "grad_norm": 0.3635323643684387,
+      "learning_rate": 0.0001,
+      "loss": 1.7845,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1890758715754408,
+      "grad_norm": 0.36258241534233093,
+      "learning_rate": 0.0001,
+      "loss": 1.8612,
+      "step": 1646
+    },
+    {
+      "epoch": 0.18919074148526793,
+      "grad_norm": 0.35478660464286804,
+      "learning_rate": 0.0001,
+      "loss": 1.7232,
+      "step": 1647
+    },
+    {
+      "epoch": 0.18930561139509505,
+      "grad_norm": 0.38521048426628113,
+      "learning_rate": 0.0001,
+      "loss": 1.7481,
+      "step": 1648
+    },
+    {
+      "epoch": 0.18942048130492217,
+      "grad_norm": 0.34351152181625366,
+      "learning_rate": 0.0001,
+      "loss": 1.5854,
+      "step": 1649
+    },
+    {
+      "epoch": 0.1895353512147493,
+      "grad_norm": 0.38085636496543884,
+      "learning_rate": 0.0001,
+      "loss": 1.8211,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1896502211245764,
+      "grad_norm": 0.3599552512168884,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 1651
+    },
+    {
+      "epoch": 0.18976509103440353,
+      "grad_norm": 0.3752254247665405,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 1652
+    },
+    {
+      "epoch": 0.18987996094423065,
+      "grad_norm": 0.3557283580303192,
+      "learning_rate": 0.0001,
+      "loss": 1.8026,
+      "step": 1653
+    },
+    {
+      "epoch": 0.18999483085405777,
+      "grad_norm": 0.3593176603317261,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1901097007638849,
+      "grad_norm": 0.36632347106933594,
+      "learning_rate": 0.0001,
+      "loss": 1.751,
+      "step": 1655
+    },
+    {
+      "epoch": 0.190224570673712,
+      "grad_norm": 0.36848095059394836,
+      "learning_rate": 0.0001,
+      "loss": 1.7481,
+      "step": 1656
+    },
+    {
+      "epoch": 0.19033944058353913,
+      "grad_norm": 0.35497191548347473,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 1657
+    },
+    {
+      "epoch": 0.19045431049336625,
+      "grad_norm": 0.3646465837955475,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 1658
+    },
+    {
+      "epoch": 0.19056918040319337,
+      "grad_norm": 0.3539585471153259,
+      "learning_rate": 0.0001,
+      "loss": 1.6751,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1906840503130205,
+      "grad_norm": 0.36736389994621277,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1907989202228476,
+      "grad_norm": 0.38588473200798035,
+      "learning_rate": 0.0001,
+      "loss": 1.7653,
+      "step": 1661
+    },
+    {
+      "epoch": 0.19091379013267473,
+      "grad_norm": 0.38840097188949585,
+      "learning_rate": 0.0001,
+      "loss": 1.6896,
+      "step": 1662
+    },
+    {
+      "epoch": 0.19102866004250188,
+      "grad_norm": 0.34677135944366455,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 1663
+    },
+    {
+      "epoch": 0.191143529952329,
+      "grad_norm": 0.3521466553211212,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 1664
+    },
+    {
+      "epoch": 0.19125839986215612,
+      "grad_norm": 0.34969663619995117,
+      "learning_rate": 0.0001,
+      "loss": 1.7715,
+      "step": 1665
+    },
+    {
+      "epoch": 0.19137326977198324,
+      "grad_norm": 0.3782643973827362,
+      "learning_rate": 0.0001,
+      "loss": 1.777,
+      "step": 1666
+    },
+    {
+      "epoch": 0.19148813968181036,
+      "grad_norm": 0.3731124699115753,
+      "learning_rate": 0.0001,
+      "loss": 1.7143,
+      "step": 1667
+    },
+    {
+      "epoch": 0.19160300959163748,
+      "grad_norm": 0.37945446372032166,
+      "learning_rate": 0.0001,
+      "loss": 1.7618,
+      "step": 1668
+    },
+    {
+      "epoch": 0.1917178795014646,
+      "grad_norm": 0.331589937210083,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 1669
+    },
+    {
+      "epoch": 0.19183274941129172,
+      "grad_norm": 0.3730468451976776,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 1670
+    },
+    {
+      "epoch": 0.19194761932111884,
+      "grad_norm": 0.40152809023857117,
+      "learning_rate": 0.0001,
+      "loss": 1.9885,
+      "step": 1671
+    },
+    {
+      "epoch": 0.19206248923094596,
+      "grad_norm": 0.3776914179325104,
+      "learning_rate": 0.0001,
+      "loss": 1.8431,
+      "step": 1672
+    },
+    {
+      "epoch": 0.19217735914077308,
+      "grad_norm": 0.3614851236343384,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1922922290506002,
+      "grad_norm": 0.33890727162361145,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 1674
+    },
+    {
+      "epoch": 0.19240709896042732,
+      "grad_norm": 0.3711831867694855,
+      "learning_rate": 0.0001,
+      "loss": 1.8968,
+      "step": 1675
+    },
+    {
+      "epoch": 0.19252196887025444,
+      "grad_norm": 0.407746285200119,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 1676
+    },
+    {
+      "epoch": 0.19263683878008156,
+      "grad_norm": 0.3913660943508148,
+      "learning_rate": 0.0001,
+      "loss": 1.6698,
+      "step": 1677
+    },
+    {
+      "epoch": 0.19275170868990868,
+      "grad_norm": 0.3649699091911316,
+      "learning_rate": 0.0001,
+      "loss": 1.8171,
+      "step": 1678
+    },
+    {
+      "epoch": 0.1928665785997358,
+      "grad_norm": 0.34938865900039673,
+      "learning_rate": 0.0001,
+      "loss": 1.7187,
+      "step": 1679
+    },
+    {
+      "epoch": 0.19298144850956292,
+      "grad_norm": 0.3610716462135315,
+      "learning_rate": 0.0001,
+      "loss": 1.7097,
+      "step": 1680
+    },
+    {
+      "epoch": 0.19309631841939004,
+      "grad_norm": 0.3534272015094757,
+      "learning_rate": 0.0001,
+      "loss": 1.7907,
+      "step": 1681
+    },
+    {
+      "epoch": 0.19321118832921716,
+      "grad_norm": 0.3574727177619934,
+      "learning_rate": 0.0001,
+      "loss": 1.7394,
+      "step": 1682
+    },
+    {
+      "epoch": 0.19332605823904428,
+      "grad_norm": 0.32285967469215393,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 1683
+    },
+    {
+      "epoch": 0.1934409281488714,
+      "grad_norm": 0.36247870326042175,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 1684
+    },
+    {
+      "epoch": 0.19355579805869852,
+      "grad_norm": 0.3250444233417511,
+      "learning_rate": 0.0001,
+      "loss": 1.5945,
+      "step": 1685
+    },
+    {
+      "epoch": 0.19367066796852564,
+      "grad_norm": 0.3848918676376343,
+      "learning_rate": 0.0001,
+      "loss": 1.8016,
+      "step": 1686
+    },
+    {
+      "epoch": 0.19378553787835276,
+      "grad_norm": 0.3321680426597595,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 1687
+    },
+    {
+      "epoch": 0.19390040778817988,
+      "grad_norm": 0.3534335196018219,
+      "learning_rate": 0.0001,
+      "loss": 1.8101,
+      "step": 1688
+    },
+    {
+      "epoch": 0.194015277698007,
+      "grad_norm": 0.3084717094898224,
+      "learning_rate": 0.0001,
+      "loss": 1.2239,
+      "step": 1689
+    },
+    {
+      "epoch": 0.19413014760783412,
+      "grad_norm": 0.36308553814888,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 1690
+    },
+    {
+      "epoch": 0.19424501751766124,
+      "grad_norm": 0.35767173767089844,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 1691
+    },
+    {
+      "epoch": 0.19435988742748836,
+      "grad_norm": 0.3621061444282532,
+      "learning_rate": 0.0001,
+      "loss": 1.5605,
+      "step": 1692
+    },
+    {
+      "epoch": 0.19447475733731548,
+      "grad_norm": 0.3489883840084076,
+      "learning_rate": 0.0001,
+      "loss": 1.7473,
+      "step": 1693
+    },
+    {
+      "epoch": 0.1945896272471426,
+      "grad_norm": 0.35943081974983215,
+      "learning_rate": 0.0001,
+      "loss": 1.8841,
+      "step": 1694
+    },
+    {
+      "epoch": 0.19470449715696972,
+      "grad_norm": 0.3859713673591614,
+      "learning_rate": 0.0001,
+      "loss": 1.7675,
+      "step": 1695
+    },
+    {
+      "epoch": 0.19481936706679684,
+      "grad_norm": 0.3559940457344055,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 1696
+    },
+    {
+      "epoch": 0.194934236976624,
+      "grad_norm": 0.37918147444725037,
+      "learning_rate": 0.0001,
+      "loss": 1.7507,
+      "step": 1697
+    },
+    {
+      "epoch": 0.1950491068864511,
+      "grad_norm": 0.36371055245399475,
+      "learning_rate": 0.0001,
+      "loss": 1.7621,
+      "step": 1698
+    },
+    {
+      "epoch": 0.19516397679627823,
+      "grad_norm": 0.34695690870285034,
+      "learning_rate": 0.0001,
+      "loss": 1.7555,
+      "step": 1699
+    },
+    {
+      "epoch": 0.19527884670610535,
+      "grad_norm": 0.35389262437820435,
+      "learning_rate": 0.0001,
+      "loss": 1.6632,
+      "step": 1700
+    },
+    {
+      "epoch": 0.19539371661593247,
+      "grad_norm": 0.35778507590293884,
+      "learning_rate": 0.0001,
+      "loss": 1.7939,
+      "step": 1701
+    },
+    {
+      "epoch": 0.1955085865257596,
+      "grad_norm": 0.37663915753364563,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1956234564355867,
+      "grad_norm": 0.3157816529273987,
+      "learning_rate": 0.0001,
+      "loss": 1.3818,
+      "step": 1703
+    },
+    {
+      "epoch": 0.19573832634541383,
+      "grad_norm": 0.35252466797828674,
+      "learning_rate": 0.0001,
+      "loss": 1.7476,
+      "step": 1704
+    },
+    {
+      "epoch": 0.19585319625524095,
+      "grad_norm": 0.3793637752532959,
+      "learning_rate": 0.0001,
+      "loss": 1.8395,
+      "step": 1705
+    },
+    {
+      "epoch": 0.19596806616506807,
+      "grad_norm": 0.3287891149520874,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 1706
+    },
+    {
+      "epoch": 0.1960829360748952,
+      "grad_norm": 0.3720473349094391,
+      "learning_rate": 0.0001,
+      "loss": 1.8553,
+      "step": 1707
+    },
+    {
+      "epoch": 0.1961978059847223,
+      "grad_norm": 0.3807579576969147,
+      "learning_rate": 0.0001,
+      "loss": 1.8815,
+      "step": 1708
+    },
+    {
+      "epoch": 0.19631267589454943,
+      "grad_norm": 0.3885481655597687,
+      "learning_rate": 0.0001,
+      "loss": 1.8668,
+      "step": 1709
+    },
+    {
+      "epoch": 0.19642754580437655,
+      "grad_norm": 0.3450814187526703,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 1710
+    },
+    {
+      "epoch": 0.19654241571420367,
+      "grad_norm": 0.3540419936180115,
+      "learning_rate": 0.0001,
+      "loss": 1.6886,
+      "step": 1711
+    },
+    {
+      "epoch": 0.1966572856240308,
+      "grad_norm": 0.38482004404067993,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 1712
+    },
+    {
+      "epoch": 0.1967721555338579,
+      "grad_norm": 0.37782052159309387,
+      "learning_rate": 0.0001,
+      "loss": 1.6786,
+      "step": 1713
+    },
+    {
+      "epoch": 0.19688702544368503,
+      "grad_norm": 0.3788130283355713,
+      "learning_rate": 0.0001,
+      "loss": 1.6707,
+      "step": 1714
+    },
+    {
+      "epoch": 0.19700189535351215,
+      "grad_norm": 0.3536849617958069,
+      "learning_rate": 0.0001,
+      "loss": 1.7248,
+      "step": 1715
+    },
+    {
+      "epoch": 0.19711676526333927,
+      "grad_norm": 0.4016459584236145,
+      "learning_rate": 0.0001,
+      "loss": 1.4351,
+      "step": 1716
+    },
+    {
+      "epoch": 0.1972316351731664,
+      "grad_norm": 0.34957659244537354,
+      "learning_rate": 0.0001,
+      "loss": 1.7459,
+      "step": 1717
+    },
+    {
+      "epoch": 0.1973465050829935,
+      "grad_norm": 0.36093124747276306,
+      "learning_rate": 0.0001,
+      "loss": 1.7523,
+      "step": 1718
+    },
+    {
+      "epoch": 0.19746137499282063,
+      "grad_norm": 0.36018863320350647,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 1719
+    },
+    {
+      "epoch": 0.19757624490264775,
+      "grad_norm": 0.37135326862335205,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 1720
+    },
+    {
+      "epoch": 0.19769111481247487,
+      "grad_norm": 0.3906667232513428,
+      "learning_rate": 0.0001,
+      "loss": 1.7177,
+      "step": 1721
+    },
+    {
+      "epoch": 0.197805984722302,
+      "grad_norm": 0.3404282331466675,
+      "learning_rate": 0.0001,
+      "loss": 1.4111,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1979208546321291,
+      "grad_norm": 0.39864858984947205,
+      "learning_rate": 0.0001,
+      "loss": 1.7668,
+      "step": 1723
+    },
+    {
+      "epoch": 0.19803572454195623,
+      "grad_norm": 0.38260820508003235,
+      "learning_rate": 0.0001,
+      "loss": 1.4592,
+      "step": 1724
+    },
+    {
+      "epoch": 0.19815059445178335,
+      "grad_norm": 0.35546496510505676,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 1725
+    },
+    {
+      "epoch": 0.19826546436161047,
+      "grad_norm": 0.34984710812568665,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 1726
+    },
+    {
+      "epoch": 0.1983803342714376,
+      "grad_norm": 0.38462913036346436,
+      "learning_rate": 0.0001,
+      "loss": 1.9247,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1984952041812647,
+      "grad_norm": 0.3563911020755768,
+      "learning_rate": 0.0001,
+      "loss": 1.6917,
+      "step": 1728
+    },
+    {
+      "epoch": 0.19861007409109183,
+      "grad_norm": 0.3691602647304535,
+      "learning_rate": 0.0001,
+      "loss": 1.6081,
+      "step": 1729
+    },
+    {
+      "epoch": 0.19872494400091895,
+      "grad_norm": 0.36970895528793335,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1988398139107461,
+      "grad_norm": 0.3286248445510864,
+      "learning_rate": 0.0001,
+      "loss": 1.4056,
+      "step": 1731
+    },
+    {
+      "epoch": 0.1989546838205732,
+      "grad_norm": 0.3443751037120819,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 1732
+    },
+    {
+      "epoch": 0.19906955373040033,
+      "grad_norm": 0.3585871458053589,
+      "learning_rate": 0.0001,
+      "loss": 1.8211,
+      "step": 1733
+    },
+    {
+      "epoch": 0.19918442364022745,
+      "grad_norm": 0.37067654728889465,
+      "learning_rate": 0.0001,
+      "loss": 1.6953,
+      "step": 1734
+    },
+    {
+      "epoch": 0.19929929355005457,
+      "grad_norm": 0.3535691797733307,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1994141634598817,
+      "grad_norm": 0.37914103269577026,
+      "learning_rate": 0.0001,
+      "loss": 1.6576,
+      "step": 1736
+    },
+    {
+      "epoch": 0.1995290333697088,
+      "grad_norm": 0.3962135910987854,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 1737
+    },
+    {
+      "epoch": 0.19964390327953593,
+      "grad_norm": 0.37456363439559937,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 1738
+    },
+    {
+      "epoch": 0.19975877318936305,
+      "grad_norm": 0.36762315034866333,
+      "learning_rate": 0.0001,
+      "loss": 1.7478,
+      "step": 1739
+    },
+    {
+      "epoch": 0.19987364309919017,
+      "grad_norm": 0.37871554493904114,
+      "learning_rate": 0.0001,
+      "loss": 1.7603,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1999885130090173,
+      "grad_norm": 0.39336419105529785,
+      "learning_rate": 0.0001,
+      "loss": 1.7888,
+      "step": 1741
+    },
+    {
+      "epoch": 0.2001033829188444,
+      "grad_norm": 0.366931289434433,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 1742
+    },
+    {
+      "epoch": 0.20021825282867153,
+      "grad_norm": 0.37387847900390625,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 1743
+    },
+    {
+      "epoch": 0.20033312273849865,
+      "grad_norm": 0.3491780459880829,
+      "learning_rate": 0.0001,
+      "loss": 1.8099,
+      "step": 1744
+    },
+    {
+      "epoch": 0.20044799264832577,
+      "grad_norm": 0.39339229464530945,
+      "learning_rate": 0.0001,
+      "loss": 1.7889,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2005628625581529,
+      "grad_norm": 0.3711383640766144,
+      "learning_rate": 0.0001,
+      "loss": 1.853,
+      "step": 1746
+    },
+    {
+      "epoch": 0.20067773246798,
+      "grad_norm": 0.33763977885246277,
+      "learning_rate": 0.0001,
+      "loss": 1.6071,
+      "step": 1747
+    },
+    {
+      "epoch": 0.20079260237780713,
+      "grad_norm": 0.34965288639068604,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 1748
+    },
+    {
+      "epoch": 0.20090747228763425,
+      "grad_norm": 0.3817383646965027,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 1749
+    },
+    {
+      "epoch": 0.20102234219746137,
+      "grad_norm": 0.37496039271354675,
+      "learning_rate": 0.0001,
+      "loss": 1.711,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2011372121072885,
+      "grad_norm": 0.37966370582580566,
+      "learning_rate": 0.0001,
+      "loss": 1.8306,
+      "step": 1751
+    },
+    {
+      "epoch": 0.2012520820171156,
+      "grad_norm": 0.3994872272014618,
+      "learning_rate": 0.0001,
+      "loss": 1.9473,
+      "step": 1752
+    },
+    {
+      "epoch": 0.20136695192694273,
+      "grad_norm": 0.3680518865585327,
+      "learning_rate": 0.0001,
+      "loss": 1.7061,
+      "step": 1753
+    },
+    {
+      "epoch": 0.20148182183676985,
+      "grad_norm": 0.31908658146858215,
+      "learning_rate": 0.0001,
+      "loss": 1.4684,
+      "step": 1754
+    },
+    {
+      "epoch": 0.20159669174659697,
+      "grad_norm": 0.362386554479599,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 1755
+    },
+    {
+      "epoch": 0.2017115616564241,
+      "grad_norm": 0.35823360085487366,
+      "learning_rate": 0.0001,
+      "loss": 1.7714,
+      "step": 1756
+    },
+    {
+      "epoch": 0.2018264315662512,
+      "grad_norm": 0.3523258566856384,
+      "learning_rate": 0.0001,
+      "loss": 1.8227,
+      "step": 1757
+    },
+    {
+      "epoch": 0.20194130147607833,
+      "grad_norm": 0.348457396030426,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 1758
+    },
+    {
+      "epoch": 0.20205617138590545,
+      "grad_norm": 0.35159286856651306,
+      "learning_rate": 0.0001,
+      "loss": 1.4644,
+      "step": 1759
+    },
+    {
+      "epoch": 0.20217104129573257,
+      "grad_norm": 0.38442832231521606,
+      "learning_rate": 0.0001,
+      "loss": 1.7378,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2022859112055597,
+      "grad_norm": 0.3663921356201172,
+      "learning_rate": 0.0001,
+      "loss": 1.8518,
+      "step": 1761
+    },
+    {
+      "epoch": 0.2024007811153868,
+      "grad_norm": 0.3565858006477356,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 1762
+    },
+    {
+      "epoch": 0.20251565102521393,
+      "grad_norm": 0.37562236189842224,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 1763
+    },
+    {
+      "epoch": 0.20263052093504105,
+      "grad_norm": 0.3319898247718811,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 1764
+    },
+    {
+      "epoch": 0.20274539084486817,
+      "grad_norm": 0.35017985105514526,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 1765
+    },
+    {
+      "epoch": 0.20286026075469532,
+      "grad_norm": 0.3883030116558075,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 1766
+    },
+    {
+      "epoch": 0.20297513066452244,
+      "grad_norm": 0.34447354078292847,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 1767
+    },
+    {
+      "epoch": 0.20309000057434956,
+      "grad_norm": 0.3480866253376007,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 1768
+    },
+    {
+      "epoch": 0.20320487048417668,
+      "grad_norm": 0.3792140781879425,
+      "learning_rate": 0.0001,
+      "loss": 1.7436,
+      "step": 1769
+    },
+    {
+      "epoch": 0.2033197403940038,
+      "grad_norm": 0.3636226952075958,
+      "learning_rate": 0.0001,
+      "loss": 1.4727,
+      "step": 1770
+    },
+    {
+      "epoch": 0.20343461030383092,
+      "grad_norm": 0.3649579584598541,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 1771
+    },
+    {
+      "epoch": 0.20354948021365804,
+      "grad_norm": 0.34622448682785034,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 1772
+    },
+    {
+      "epoch": 0.20366435012348516,
+      "grad_norm": 0.4073683023452759,
+      "learning_rate": 0.0001,
+      "loss": 1.9036,
+      "step": 1773
+    },
+    {
+      "epoch": 0.20377922003331228,
+      "grad_norm": 0.35996997356414795,
+      "learning_rate": 0.0001,
+      "loss": 1.8344,
+      "step": 1774
+    },
+    {
+      "epoch": 0.2038940899431394,
+      "grad_norm": 0.35649874806404114,
+      "learning_rate": 0.0001,
+      "loss": 1.4739,
+      "step": 1775
+    },
+    {
+      "epoch": 0.20400895985296652,
+      "grad_norm": 0.3954346477985382,
+      "learning_rate": 0.0001,
+      "loss": 1.9226,
+      "step": 1776
+    },
+    {
+      "epoch": 0.20412382976279364,
+      "grad_norm": 0.3589356243610382,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 1777
+    },
+    {
+      "epoch": 0.20423869967262076,
+      "grad_norm": 0.36763888597488403,
+      "learning_rate": 0.0001,
+      "loss": 1.7394,
+      "step": 1778
+    },
+    {
+      "epoch": 0.20435356958244788,
+      "grad_norm": 0.38217705488204956,
+      "learning_rate": 0.0001,
+      "loss": 1.9139,
+      "step": 1779
+    },
+    {
+      "epoch": 0.204468439492275,
+      "grad_norm": 0.39067110419273376,
+      "learning_rate": 0.0001,
+      "loss": 1.8865,
+      "step": 1780
+    },
+    {
+      "epoch": 0.20458330940210212,
+      "grad_norm": 0.369056761264801,
+      "learning_rate": 0.0001,
+      "loss": 1.7165,
+      "step": 1781
+    },
+    {
+      "epoch": 0.20469817931192924,
+      "grad_norm": 0.3984009623527527,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 1782
+    },
+    {
+      "epoch": 0.20481304922175636,
+      "grad_norm": 0.3637178838253021,
+      "learning_rate": 0.0001,
+      "loss": 1.7317,
+      "step": 1783
+    },
+    {
+      "epoch": 0.20492791913158348,
+      "grad_norm": 0.3644060790538788,
+      "learning_rate": 0.0001,
+      "loss": 1.7301,
+      "step": 1784
+    },
+    {
+      "epoch": 0.2050427890414106,
+      "grad_norm": 0.36648574471473694,
+      "learning_rate": 0.0001,
+      "loss": 1.793,
+      "step": 1785
+    },
+    {
+      "epoch": 0.20515765895123772,
+      "grad_norm": 0.37490981817245483,
+      "learning_rate": 0.0001,
+      "loss": 1.7762,
+      "step": 1786
+    },
+    {
+      "epoch": 0.20527252886106484,
+      "grad_norm": 0.35913270711898804,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 1787
+    },
+    {
+      "epoch": 0.20538739877089196,
+      "grad_norm": 0.35344579815864563,
+      "learning_rate": 0.0001,
+      "loss": 1.7904,
+      "step": 1788
+    },
+    {
+      "epoch": 0.20550226868071908,
+      "grad_norm": 0.4043060839176178,
+      "learning_rate": 0.0001,
+      "loss": 1.8669,
+      "step": 1789
+    },
+    {
+      "epoch": 0.2056171385905462,
+      "grad_norm": 0.35678645968437195,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 1790
+    },
+    {
+      "epoch": 0.20573200850037332,
+      "grad_norm": 0.3554267883300781,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 1791
+    },
+    {
+      "epoch": 0.20584687841020044,
+      "grad_norm": 0.3699265122413635,
+      "learning_rate": 0.0001,
+      "loss": 1.8192,
+      "step": 1792
+    },
+    {
+      "epoch": 0.20596174832002756,
+      "grad_norm": 0.3494689464569092,
+      "learning_rate": 0.0001,
+      "loss": 1.4697,
+      "step": 1793
+    },
+    {
+      "epoch": 0.20607661822985468,
+      "grad_norm": 0.3618113398551941,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 1794
+    },
+    {
+      "epoch": 0.2061914881396818,
+      "grad_norm": 0.36774349212646484,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 1795
+    },
+    {
+      "epoch": 0.20630635804950892,
+      "grad_norm": 0.3754447400569916,
+      "learning_rate": 0.0001,
+      "loss": 1.549,
+      "step": 1796
+    },
+    {
+      "epoch": 0.20642122795933604,
+      "grad_norm": 0.4026842713356018,
+      "learning_rate": 0.0001,
+      "loss": 1.832,
+      "step": 1797
+    },
+    {
+      "epoch": 0.20653609786916316,
+      "grad_norm": 0.34931662678718567,
+      "learning_rate": 0.0001,
+      "loss": 1.6936,
+      "step": 1798
+    },
+    {
+      "epoch": 0.20665096777899028,
+      "grad_norm": 0.35631394386291504,
+      "learning_rate": 0.0001,
+      "loss": 1.7474,
+      "step": 1799
+    },
+    {
+      "epoch": 0.20676583768881743,
+      "grad_norm": 0.35016921162605286,
+      "learning_rate": 0.0001,
+      "loss": 1.8252,
+      "step": 1800
+    },
+    {
+      "epoch": 0.20688070759864455,
+      "grad_norm": 0.36527907848358154,
+      "learning_rate": 0.0001,
+      "loss": 1.7728,
+      "step": 1801
+    },
+    {
+      "epoch": 0.20699557750847167,
+      "grad_norm": 0.375942200422287,
+      "learning_rate": 0.0001,
+      "loss": 1.76,
+      "step": 1802
+    },
+    {
+      "epoch": 0.2071104474182988,
+      "grad_norm": 0.3598606586456299,
+      "learning_rate": 0.0001,
+      "loss": 1.7765,
+      "step": 1803
+    },
+    {
+      "epoch": 0.2072253173281259,
+      "grad_norm": 0.3487381041049957,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 1804
+    },
+    {
+      "epoch": 0.20734018723795303,
+      "grad_norm": 0.33189913630485535,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 1805
+    },
+    {
+      "epoch": 0.20745505714778015,
+      "grad_norm": 0.3427751660346985,
+      "learning_rate": 0.0001,
+      "loss": 1.4288,
+      "step": 1806
+    },
+    {
+      "epoch": 0.20756992705760727,
+      "grad_norm": 0.3979537785053253,
+      "learning_rate": 0.0001,
+      "loss": 1.6666,
+      "step": 1807
+    },
+    {
+      "epoch": 0.2076847969674344,
+      "grad_norm": 0.37315791845321655,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 1808
+    },
+    {
+      "epoch": 0.2077996668772615,
+      "grad_norm": 0.37909650802612305,
+      "learning_rate": 0.0001,
+      "loss": 1.7111,
+      "step": 1809
+    },
+    {
+      "epoch": 0.20791453678708863,
+      "grad_norm": 0.3696240186691284,
+      "learning_rate": 0.0001,
+      "loss": 1.7612,
+      "step": 1810
+    },
+    {
+      "epoch": 0.20802940669691575,
+      "grad_norm": 0.38484904170036316,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 1811
+    },
+    {
+      "epoch": 0.20814427660674287,
+      "grad_norm": 0.3398043215274811,
+      "learning_rate": 0.0001,
+      "loss": 1.6615,
+      "step": 1812
+    },
+    {
+      "epoch": 0.20825914651657,
+      "grad_norm": 0.3796899616718292,
+      "learning_rate": 0.0001,
+      "loss": 1.8264,
+      "step": 1813
+    },
+    {
+      "epoch": 0.2083740164263971,
+      "grad_norm": 0.38819029927253723,
+      "learning_rate": 0.0001,
+      "loss": 1.8616,
+      "step": 1814
+    },
+    {
+      "epoch": 0.20848888633622423,
+      "grad_norm": 0.4355449378490448,
+      "learning_rate": 0.0001,
+      "loss": 1.8042,
+      "step": 1815
+    },
+    {
+      "epoch": 0.20860375624605135,
+      "grad_norm": 0.37194108963012695,
+      "learning_rate": 0.0001,
+      "loss": 1.7524,
+      "step": 1816
+    },
+    {
+      "epoch": 0.20871862615587847,
+      "grad_norm": 0.3792515993118286,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 1817
+    },
+    {
+      "epoch": 0.2088334960657056,
+      "grad_norm": 0.37876373529434204,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 1818
+    },
+    {
+      "epoch": 0.2089483659755327,
+      "grad_norm": 0.34558865427970886,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 1819
+    },
+    {
+      "epoch": 0.20906323588535983,
+      "grad_norm": 0.43524423241615295,
+      "learning_rate": 0.0001,
+      "loss": 2.0707,
+      "step": 1820
+    },
+    {
+      "epoch": 0.20917810579518695,
+      "grad_norm": 0.4018253982067108,
+      "learning_rate": 0.0001,
+      "loss": 1.8651,
+      "step": 1821
+    },
+    {
+      "epoch": 0.20929297570501407,
+      "grad_norm": 0.3796786367893219,
+      "learning_rate": 0.0001,
+      "loss": 1.6014,
+      "step": 1822
+    },
+    {
+      "epoch": 0.2094078456148412,
+      "grad_norm": 0.35122066736221313,
+      "learning_rate": 0.0001,
+      "loss": 1.7398,
+      "step": 1823
+    },
+    {
+      "epoch": 0.2095227155246683,
+      "grad_norm": 0.3692324161529541,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 1824
+    },
+    {
+      "epoch": 0.20963758543449543,
+      "grad_norm": 0.39032313227653503,
+      "learning_rate": 0.0001,
+      "loss": 1.7603,
+      "step": 1825
+    },
+    {
+      "epoch": 0.20975245534432255,
+      "grad_norm": 0.34031352400779724,
+      "learning_rate": 0.0001,
+      "loss": 1.4029,
+      "step": 1826
+    },
+    {
+      "epoch": 0.20986732525414967,
+      "grad_norm": 0.40523627400398254,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 1827
+    },
+    {
+      "epoch": 0.2099821951639768,
+      "grad_norm": 0.3522171378135681,
+      "learning_rate": 0.0001,
+      "loss": 1.7032,
+      "step": 1828
+    },
+    {
+      "epoch": 0.2100970650738039,
+      "grad_norm": 0.3715449273586273,
+      "learning_rate": 0.0001,
+      "loss": 1.7735,
+      "step": 1829
+    },
+    {
+      "epoch": 0.21021193498363103,
+      "grad_norm": 0.39641836285591125,
+      "learning_rate": 0.0001,
+      "loss": 1.8061,
+      "step": 1830
+    },
+    {
+      "epoch": 0.21032680489345815,
+      "grad_norm": 0.3934457302093506,
+      "learning_rate": 0.0001,
+      "loss": 1.7792,
+      "step": 1831
+    },
+    {
+      "epoch": 0.21044167480328527,
+      "grad_norm": 0.35226938128471375,
+      "learning_rate": 0.0001,
+      "loss": 1.7576,
+      "step": 1832
+    },
+    {
+      "epoch": 0.2105565447131124,
+      "grad_norm": 0.36764612793922424,
+      "learning_rate": 0.0001,
+      "loss": 1.819,
+      "step": 1833
+    },
+    {
+      "epoch": 0.21067141462293953,
+      "grad_norm": 0.36354872584342957,
+      "learning_rate": 0.0001,
+      "loss": 1.5628,
+      "step": 1834
+    },
+    {
+      "epoch": 0.21078628453276665,
+      "grad_norm": 0.3655260503292084,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 1835
+    },
+    {
+      "epoch": 0.21090115444259377,
+      "grad_norm": 0.36574462056159973,
+      "learning_rate": 0.0001,
+      "loss": 1.8352,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2110160243524209,
+      "grad_norm": 0.3510647118091583,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 1837
+    },
+    {
+      "epoch": 0.21113089426224801,
+      "grad_norm": 0.38021084666252136,
+      "learning_rate": 0.0001,
+      "loss": 1.7853,
+      "step": 1838
+    },
+    {
+      "epoch": 0.21124576417207513,
+      "grad_norm": 0.38449275493621826,
+      "learning_rate": 0.0001,
+      "loss": 1.68,
+      "step": 1839
+    },
+    {
+      "epoch": 0.21136063408190225,
+      "grad_norm": 0.378221720457077,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 1840
+    },
+    {
+      "epoch": 0.21147550399172937,
+      "grad_norm": 0.3675908148288727,
+      "learning_rate": 0.0001,
+      "loss": 1.7717,
+      "step": 1841
+    },
+    {
+      "epoch": 0.2115903739015565,
+      "grad_norm": 0.3863251805305481,
+      "learning_rate": 0.0001,
+      "loss": 1.686,
+      "step": 1842
+    },
+    {
+      "epoch": 0.21170524381138361,
+      "grad_norm": 0.4383453130722046,
+      "learning_rate": 0.0001,
+      "loss": 1.9812,
+      "step": 1843
+    },
+    {
+      "epoch": 0.21182011372121073,
+      "grad_norm": 0.3583828806877136,
+      "learning_rate": 0.0001,
+      "loss": 1.7918,
+      "step": 1844
+    },
+    {
+      "epoch": 0.21193498363103785,
+      "grad_norm": 0.39561352133750916,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 1845
+    },
+    {
+      "epoch": 0.21204985354086497,
+      "grad_norm": 0.37109145522117615,
+      "learning_rate": 0.0001,
+      "loss": 1.7545,
+      "step": 1846
+    },
+    {
+      "epoch": 0.2121647234506921,
+      "grad_norm": 0.3461610972881317,
+      "learning_rate": 0.0001,
+      "loss": 1.4589,
+      "step": 1847
+    },
+    {
+      "epoch": 0.21227959336051921,
+      "grad_norm": 0.43045949935913086,
+      "learning_rate": 0.0001,
+      "loss": 1.7546,
+      "step": 1848
+    },
+    {
+      "epoch": 0.21239446327034633,
+      "grad_norm": 0.4053126871585846,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 1849
+    },
+    {
+      "epoch": 0.21250933318017345,
+      "grad_norm": 0.35681119561195374,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 1850
+    },
+    {
+      "epoch": 0.21262420309000057,
+      "grad_norm": 0.3934498429298401,
+      "learning_rate": 0.0001,
+      "loss": 1.5628,
+      "step": 1851
+    },
+    {
+      "epoch": 0.2127390729998277,
+      "grad_norm": 0.3668583035469055,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 1852
+    },
+    {
+      "epoch": 0.21285394290965481,
+      "grad_norm": 0.36090970039367676,
+      "learning_rate": 0.0001,
+      "loss": 1.8151,
+      "step": 1853
+    },
+    {
+      "epoch": 0.21296881281948193,
+      "grad_norm": 0.3945942521095276,
+      "learning_rate": 0.0001,
+      "loss": 1.7807,
+      "step": 1854
+    },
+    {
+      "epoch": 0.21308368272930905,
+      "grad_norm": 0.34316137433052063,
+      "learning_rate": 0.0001,
+      "loss": 1.5175,
+      "step": 1855
+    },
+    {
+      "epoch": 0.21319855263913617,
+      "grad_norm": 0.3736970126628876,
+      "learning_rate": 0.0001,
+      "loss": 1.66,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2133134225489633,
+      "grad_norm": 0.36034896969795227,
+      "learning_rate": 0.0001,
+      "loss": 1.6949,
+      "step": 1857
+    },
+    {
+      "epoch": 0.21342829245879041,
+      "grad_norm": 0.38381823897361755,
+      "learning_rate": 0.0001,
+      "loss": 1.7625,
+      "step": 1858
+    },
+    {
+      "epoch": 0.21354316236861753,
+      "grad_norm": 0.39188024401664734,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 1859
+    },
+    {
+      "epoch": 0.21365803227844465,
+      "grad_norm": 0.3275800049304962,
+      "learning_rate": 0.0001,
+      "loss": 1.5081,
+      "step": 1860
+    },
+    {
+      "epoch": 0.21377290218827177,
+      "grad_norm": 0.38419318199157715,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 1861
+    },
+    {
+      "epoch": 0.2138877720980989,
+      "grad_norm": 0.3843541443347931,
+      "learning_rate": 0.0001,
+      "loss": 1.7979,
+      "step": 1862
+    },
+    {
+      "epoch": 0.21400264200792601,
+      "grad_norm": 0.4217662513256073,
+      "learning_rate": 0.0001,
+      "loss": 1.7142,
+      "step": 1863
+    },
+    {
+      "epoch": 0.21411751191775313,
+      "grad_norm": 0.37826740741729736,
+      "learning_rate": 0.0001,
+      "loss": 1.7926,
+      "step": 1864
+    },
+    {
+      "epoch": 0.21423238182758025,
+      "grad_norm": 0.38108256459236145,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 1865
+    },
+    {
+      "epoch": 0.21434725173740737,
+      "grad_norm": 0.3516540229320526,
+      "learning_rate": 0.0001,
+      "loss": 1.571,
+      "step": 1866
+    },
+    {
+      "epoch": 0.2144621216472345,
+      "grad_norm": 0.3580091595649719,
+      "learning_rate": 0.0001,
+      "loss": 1.7167,
+      "step": 1867
+    },
+    {
+      "epoch": 0.21457699155706164,
+      "grad_norm": 0.3416488468647003,
+      "learning_rate": 0.0001,
+      "loss": 1.4516,
+      "step": 1868
+    },
+    {
+      "epoch": 0.21469186146688876,
+      "grad_norm": 0.3605569303035736,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 1869
+    },
+    {
+      "epoch": 0.21480673137671588,
+      "grad_norm": 0.3645571768283844,
+      "learning_rate": 0.0001,
+      "loss": 1.6764,
+      "step": 1870
+    },
+    {
+      "epoch": 0.214921601286543,
+      "grad_norm": 0.3980044424533844,
+      "learning_rate": 0.0001,
+      "loss": 1.7082,
+      "step": 1871
+    },
+    {
+      "epoch": 0.21503647119637012,
+      "grad_norm": 0.3569796681404114,
+      "learning_rate": 0.0001,
+      "loss": 1.572,
+      "step": 1872
+    },
+    {
+      "epoch": 0.21515134110619724,
+      "grad_norm": 0.37738198041915894,
+      "learning_rate": 0.0001,
+      "loss": 1.6913,
+      "step": 1873
+    },
+    {
+      "epoch": 0.21526621101602436,
+      "grad_norm": 0.34949570894241333,
+      "learning_rate": 0.0001,
+      "loss": 1.6524,
+      "step": 1874
+    },
+    {
+      "epoch": 0.21538108092585148,
+      "grad_norm": 0.3858441114425659,
+      "learning_rate": 0.0001,
+      "loss": 1.7056,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2154959508356786,
+      "grad_norm": 0.33417800068855286,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 1876
+    },
+    {
+      "epoch": 0.21561082074550572,
+      "grad_norm": 0.363370418548584,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 1877
+    },
+    {
+      "epoch": 0.21572569065533284,
+      "grad_norm": 0.37960943579673767,
+      "learning_rate": 0.0001,
+      "loss": 1.8329,
+      "step": 1878
+    },
+    {
+      "epoch": 0.21584056056515996,
+      "grad_norm": 0.34242677688598633,
+      "learning_rate": 0.0001,
+      "loss": 1.6365,
+      "step": 1879
+    },
+    {
+      "epoch": 0.21595543047498708,
+      "grad_norm": 0.34529823064804077,
+      "learning_rate": 0.0001,
+      "loss": 1.7954,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2160703003848142,
+      "grad_norm": 0.3592895567417145,
+      "learning_rate": 0.0001,
+      "loss": 1.7834,
+      "step": 1881
+    },
+    {
+      "epoch": 0.21618517029464132,
+      "grad_norm": 0.3785233795642853,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 1882
+    },
+    {
+      "epoch": 0.21630004020446844,
+      "grad_norm": 0.3563402593135834,
+      "learning_rate": 0.0001,
+      "loss": 1.7312,
+      "step": 1883
+    },
+    {
+      "epoch": 0.21641491011429556,
+      "grad_norm": 0.4084146320819855,
+      "learning_rate": 0.0001,
+      "loss": 1.8166,
+      "step": 1884
+    },
+    {
+      "epoch": 0.21652978002412268,
+      "grad_norm": 0.3653600811958313,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2166446499339498,
+      "grad_norm": 0.34954777359962463,
+      "learning_rate": 0.0001,
+      "loss": 1.7595,
+      "step": 1886
+    },
+    {
+      "epoch": 0.21675951984377692,
+      "grad_norm": 0.34437569975852966,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 1887
+    },
+    {
+      "epoch": 0.21687438975360404,
+      "grad_norm": 0.37032750248908997,
+      "learning_rate": 0.0001,
+      "loss": 1.6742,
+      "step": 1888
+    },
+    {
+      "epoch": 0.21698925966343116,
+      "grad_norm": 0.36316627264022827,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 1889
+    },
+    {
+      "epoch": 0.21710412957325828,
+      "grad_norm": 0.3642198145389557,
+      "learning_rate": 0.0001,
+      "loss": 1.7953,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2172189994830854,
+      "grad_norm": 0.36583060026168823,
+      "learning_rate": 0.0001,
+      "loss": 1.5196,
+      "step": 1891
+    },
+    {
+      "epoch": 0.21733386939291252,
+      "grad_norm": 0.35113057494163513,
+      "learning_rate": 0.0001,
+      "loss": 1.4707,
+      "step": 1892
+    },
+    {
+      "epoch": 0.21744873930273964,
+      "grad_norm": 0.4065643846988678,
+      "learning_rate": 0.0001,
+      "loss": 1.8893,
+      "step": 1893
+    },
+    {
+      "epoch": 0.21756360921256676,
+      "grad_norm": 0.36826783418655396,
+      "learning_rate": 0.0001,
+      "loss": 1.8333,
+      "step": 1894
+    },
+    {
+      "epoch": 0.21767847912239388,
+      "grad_norm": 0.3617238998413086,
+      "learning_rate": 0.0001,
+      "loss": 1.7314,
+      "step": 1895
+    },
+    {
+      "epoch": 0.217793349032221,
+      "grad_norm": 0.40174373984336853,
+      "learning_rate": 0.0001,
+      "loss": 1.8332,
+      "step": 1896
+    },
+    {
+      "epoch": 0.21790821894204812,
+      "grad_norm": 0.3792894184589386,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 1897
+    },
+    {
+      "epoch": 0.21802308885187524,
+      "grad_norm": 0.35536929965019226,
+      "learning_rate": 0.0001,
+      "loss": 1.69,
+      "step": 1898
+    },
+    {
+      "epoch": 0.21813795876170236,
+      "grad_norm": 0.36117202043533325,
+      "learning_rate": 0.0001,
+      "loss": 1.6592,
+      "step": 1899
+    },
+    {
+      "epoch": 0.21825282867152948,
+      "grad_norm": 0.3893747627735138,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2183676985813566,
+      "grad_norm": 0.40482643246650696,
+      "learning_rate": 0.0001,
+      "loss": 1.8143,
+      "step": 1901
+    },
+    {
+      "epoch": 0.21848256849118372,
+      "grad_norm": 0.35201117396354675,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 1902
+    },
+    {
+      "epoch": 0.21859743840101087,
+      "grad_norm": 0.359512060880661,
+      "learning_rate": 0.0001,
+      "loss": 1.7391,
+      "step": 1903
+    },
+    {
+      "epoch": 0.218712308310838,
+      "grad_norm": 0.4181397259235382,
+      "learning_rate": 0.0001,
+      "loss": 1.9794,
+      "step": 1904
+    },
+    {
+      "epoch": 0.2188271782206651,
+      "grad_norm": 0.3568892478942871,
+      "learning_rate": 0.0001,
+      "loss": 1.5442,
+      "step": 1905
+    },
+    {
+      "epoch": 0.21894204813049223,
+      "grad_norm": 0.3569229245185852,
+      "learning_rate": 0.0001,
+      "loss": 1.8519,
+      "step": 1906
+    },
+    {
+      "epoch": 0.21905691804031935,
+      "grad_norm": 0.3385428488254547,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 1907
+    },
+    {
+      "epoch": 0.21917178795014647,
+      "grad_norm": 0.35773056745529175,
+      "learning_rate": 0.0001,
+      "loss": 1.7635,
+      "step": 1908
+    },
+    {
+      "epoch": 0.2192866578599736,
+      "grad_norm": 0.37452182173728943,
+      "learning_rate": 0.0001,
+      "loss": 1.8033,
+      "step": 1909
+    },
+    {
+      "epoch": 0.2194015277698007,
+      "grad_norm": 0.38838356733322144,
+      "learning_rate": 0.0001,
+      "loss": 1.6672,
+      "step": 1910
+    },
+    {
+      "epoch": 0.21951639767962783,
+      "grad_norm": 0.3544471263885498,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 1911
+    },
+    {
+      "epoch": 0.21963126758945495,
+      "grad_norm": 0.3904416561126709,
+      "learning_rate": 0.0001,
+      "loss": 1.8598,
+      "step": 1912
+    },
+    {
+      "epoch": 0.21974613749928207,
+      "grad_norm": 0.3736543357372284,
+      "learning_rate": 0.0001,
+      "loss": 1.6786,
+      "step": 1913
+    },
+    {
+      "epoch": 0.2198610074091092,
+      "grad_norm": 0.3756701648235321,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 1914
+    },
+    {
+      "epoch": 0.2199758773189363,
+      "grad_norm": 0.334176629781723,
+      "learning_rate": 0.0001,
+      "loss": 1.3314,
+      "step": 1915
+    },
+    {
+      "epoch": 0.22009074722876343,
+      "grad_norm": 0.35989829897880554,
+      "learning_rate": 0.0001,
+      "loss": 1.6578,
+      "step": 1916
+    },
+    {
+      "epoch": 0.22020561713859055,
+      "grad_norm": 0.3670518100261688,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 1917
+    },
+    {
+      "epoch": 0.22032048704841767,
+      "grad_norm": 0.3405352234840393,
+      "learning_rate": 0.0001,
+      "loss": 1.4737,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2204353569582448,
+      "grad_norm": 0.3269241154193878,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 1919
+    },
+    {
+      "epoch": 0.2205502268680719,
+      "grad_norm": 0.34716111421585083,
+      "learning_rate": 0.0001,
+      "loss": 1.6567,
+      "step": 1920
+    },
+    {
+      "epoch": 0.22066509677789903,
+      "grad_norm": 0.3655507564544678,
+      "learning_rate": 0.0001,
+      "loss": 1.7015,
+      "step": 1921
+    },
+    {
+      "epoch": 0.22077996668772615,
+      "grad_norm": 0.37460073828697205,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 1922
+    },
+    {
+      "epoch": 0.22089483659755327,
+      "grad_norm": 0.3556302785873413,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2210097065073804,
+      "grad_norm": 0.30222126841545105,
+      "learning_rate": 0.0001,
+      "loss": 1.3695,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2211245764172075,
+      "grad_norm": 0.3618417978286743,
+      "learning_rate": 0.0001,
+      "loss": 1.6877,
+      "step": 1925
+    },
+    {
+      "epoch": 0.22123944632703463,
+      "grad_norm": 0.3906730115413666,
+      "learning_rate": 0.0001,
+      "loss": 1.7139,
+      "step": 1926
+    },
+    {
+      "epoch": 0.22135431623686175,
+      "grad_norm": 0.3969448506832123,
+      "learning_rate": 0.0001,
+      "loss": 1.5555,
+      "step": 1927
+    },
+    {
+      "epoch": 0.22146918614668887,
+      "grad_norm": 0.4062451124191284,
+      "learning_rate": 0.0001,
+      "loss": 1.8545,
+      "step": 1928
+    },
+    {
+      "epoch": 0.221584056056516,
+      "grad_norm": 0.3763776421546936,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2216989259663431,
+      "grad_norm": 0.3991961181163788,
+      "learning_rate": 0.0001,
+      "loss": 1.776,
+      "step": 1930
+    },
+    {
+      "epoch": 0.22181379587617023,
+      "grad_norm": 0.37733355164527893,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 1931
+    },
+    {
+      "epoch": 0.22192866578599735,
+      "grad_norm": 0.383766233921051,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 1932
+    },
+    {
+      "epoch": 0.22204353569582447,
+      "grad_norm": 0.35736414790153503,
+      "learning_rate": 0.0001,
+      "loss": 1.7631,
+      "step": 1933
+    },
+    {
+      "epoch": 0.2221584056056516,
+      "grad_norm": 0.3915473222732544,
+      "learning_rate": 0.0001,
+      "loss": 1.8461,
+      "step": 1934
+    },
+    {
+      "epoch": 0.2222732755154787,
+      "grad_norm": 0.4120464622974396,
+      "learning_rate": 0.0001,
+      "loss": 1.9815,
+      "step": 1935
+    },
+    {
+      "epoch": 0.22238814542530583,
+      "grad_norm": 0.36867755651474,
+      "learning_rate": 0.0001,
+      "loss": 1.6904,
+      "step": 1936
+    },
+    {
+      "epoch": 0.22250301533513298,
+      "grad_norm": 0.3585013151168823,
+      "learning_rate": 0.0001,
+      "loss": 1.5512,
+      "step": 1937
+    },
+    {
+      "epoch": 0.2226178852449601,
+      "grad_norm": 0.36579304933547974,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 1938
+    },
+    {
+      "epoch": 0.22273275515478722,
+      "grad_norm": 0.43999186158180237,
+      "learning_rate": 0.0001,
+      "loss": 1.9172,
+      "step": 1939
+    },
+    {
+      "epoch": 0.22284762506461434,
+      "grad_norm": 0.340197890996933,
+      "learning_rate": 0.0001,
+      "loss": 1.6752,
+      "step": 1940
+    },
+    {
+      "epoch": 0.22296249497444146,
+      "grad_norm": 0.3636626601219177,
+      "learning_rate": 0.0001,
+      "loss": 1.849,
+      "step": 1941
+    },
+    {
+      "epoch": 0.22307736488426858,
+      "grad_norm": 0.33925965428352356,
+      "learning_rate": 0.0001,
+      "loss": 1.6694,
+      "step": 1942
+    },
+    {
+      "epoch": 0.2231922347940957,
+      "grad_norm": 0.3498164713382721,
+      "learning_rate": 0.0001,
+      "loss": 1.5879,
+      "step": 1943
+    },
+    {
+      "epoch": 0.22330710470392282,
+      "grad_norm": 0.35146066546440125,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 1944
+    },
+    {
+      "epoch": 0.22342197461374994,
+      "grad_norm": 0.40588539838790894,
+      "learning_rate": 0.0001,
+      "loss": 2.0319,
+      "step": 1945
+    },
+    {
+      "epoch": 0.22353684452357706,
+      "grad_norm": 0.35790249705314636,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 1946
+    },
+    {
+      "epoch": 0.22365171443340418,
+      "grad_norm": 0.3292383849620819,
+      "learning_rate": 0.0001,
+      "loss": 1.372,
+      "step": 1947
+    },
+    {
+      "epoch": 0.2237665843432313,
+      "grad_norm": 0.3615871071815491,
+      "learning_rate": 0.0001,
+      "loss": 1.7117,
+      "step": 1948
+    },
+    {
+      "epoch": 0.22388145425305842,
+      "grad_norm": 0.3719077408313751,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 1949
+    },
+    {
+      "epoch": 0.22399632416288554,
+      "grad_norm": 0.4070468544960022,
+      "learning_rate": 0.0001,
+      "loss": 1.9156,
+      "step": 1950
+    },
+    {
+      "epoch": 0.22411119407271266,
+      "grad_norm": 0.39306938648223877,
+      "learning_rate": 0.0001,
+      "loss": 1.8247,
+      "step": 1951
+    },
+    {
+      "epoch": 0.22422606398253977,
+      "grad_norm": 0.380942165851593,
+      "learning_rate": 0.0001,
+      "loss": 1.8662,
+      "step": 1952
+    },
+    {
+      "epoch": 0.2243409338923669,
+      "grad_norm": 0.3670596480369568,
+      "learning_rate": 0.0001,
+      "loss": 1.7237,
+      "step": 1953
+    },
+    {
+      "epoch": 0.22445580380219401,
+      "grad_norm": 0.3629164695739746,
+      "learning_rate": 0.0001,
+      "loss": 1.6835,
+      "step": 1954
+    },
+    {
+      "epoch": 0.22457067371202113,
+      "grad_norm": 0.3525116443634033,
+      "learning_rate": 0.0001,
+      "loss": 1.5414,
+      "step": 1955
+    },
+    {
+      "epoch": 0.22468554362184825,
+      "grad_norm": 0.362248957157135,
+      "learning_rate": 0.0001,
+      "loss": 1.7471,
+      "step": 1956
+    },
+    {
+      "epoch": 0.22480041353167537,
+      "grad_norm": 0.3838635981082916,
+      "learning_rate": 0.0001,
+      "loss": 1.845,
+      "step": 1957
+    },
+    {
+      "epoch": 0.2249152834415025,
+      "grad_norm": 0.3582911193370819,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 1958
+    },
+    {
+      "epoch": 0.22503015335132961,
+      "grad_norm": 0.3515222668647766,
+      "learning_rate": 0.0001,
+      "loss": 1.7908,
+      "step": 1959
+    },
+    {
+      "epoch": 0.22514502326115673,
+      "grad_norm": 0.38744720816612244,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 1960
+    },
+    {
+      "epoch": 0.22525989317098385,
+      "grad_norm": 0.356286883354187,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 1961
+    },
+    {
+      "epoch": 0.22537476308081097,
+      "grad_norm": 0.3699129819869995,
+      "learning_rate": 0.0001,
+      "loss": 1.7541,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2254896329906381,
+      "grad_norm": 0.3715314269065857,
+      "learning_rate": 0.0001,
+      "loss": 1.7375,
+      "step": 1963
+    },
+    {
+      "epoch": 0.22560450290046521,
+      "grad_norm": 0.3639882206916809,
+      "learning_rate": 0.0001,
+      "loss": 1.6326,
+      "step": 1964
+    },
+    {
+      "epoch": 0.22571937281029233,
+      "grad_norm": 0.34050387144088745,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 1965
+    },
+    {
+      "epoch": 0.22583424272011945,
+      "grad_norm": 0.36640483140945435,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 1966
+    },
+    {
+      "epoch": 0.22594911262994657,
+      "grad_norm": 0.39753469824790955,
+      "learning_rate": 0.0001,
+      "loss": 1.8873,
+      "step": 1967
+    },
+    {
+      "epoch": 0.2260639825397737,
+      "grad_norm": 0.3590884208679199,
+      "learning_rate": 0.0001,
+      "loss": 1.647,
+      "step": 1968
+    },
+    {
+      "epoch": 0.22617885244960081,
+      "grad_norm": 0.36375731229782104,
+      "learning_rate": 0.0001,
+      "loss": 1.6994,
+      "step": 1969
+    },
+    {
+      "epoch": 0.22629372235942793,
+      "grad_norm": 0.3400304317474365,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 1970
+    },
+    {
+      "epoch": 0.22640859226925508,
+      "grad_norm": 0.36975133419036865,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 1971
+    },
+    {
+      "epoch": 0.2265234621790822,
+      "grad_norm": 0.3886123299598694,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 1972
+    },
+    {
+      "epoch": 0.22663833208890932,
+      "grad_norm": 0.37827616930007935,
+      "learning_rate": 0.0001,
+      "loss": 1.8186,
+      "step": 1973
+    },
+    {
+      "epoch": 0.22675320199873644,
+      "grad_norm": 0.3893340229988098,
+      "learning_rate": 0.0001,
+      "loss": 1.7555,
+      "step": 1974
+    },
+    {
+      "epoch": 0.22686807190856356,
+      "grad_norm": 0.3668055534362793,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 1975
+    },
+    {
+      "epoch": 0.22698294181839068,
+      "grad_norm": 0.37988904118537903,
+      "learning_rate": 0.0001,
+      "loss": 1.7357,
+      "step": 1976
+    },
+    {
+      "epoch": 0.2270978117282178,
+      "grad_norm": 0.3713071942329407,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 1977
+    },
+    {
+      "epoch": 0.22721268163804492,
+      "grad_norm": 0.3730880916118622,
+      "learning_rate": 0.0001,
+      "loss": 1.8461,
+      "step": 1978
+    },
+    {
+      "epoch": 0.22732755154787204,
+      "grad_norm": 0.38077569007873535,
+      "learning_rate": 0.0001,
+      "loss": 1.8324,
+      "step": 1979
+    },
+    {
+      "epoch": 0.22744242145769916,
+      "grad_norm": 0.34529733657836914,
+      "learning_rate": 0.0001,
+      "loss": 1.5826,
+      "step": 1980
+    },
+    {
+      "epoch": 0.22755729136752628,
+      "grad_norm": 0.3712940812110901,
+      "learning_rate": 0.0001,
+      "loss": 1.8049,
+      "step": 1981
+    },
+    {
+      "epoch": 0.2276721612773534,
+      "grad_norm": 0.39372700452804565,
+      "learning_rate": 0.0001,
+      "loss": 1.7774,
+      "step": 1982
+    },
+    {
+      "epoch": 0.22778703118718052,
+      "grad_norm": 0.3780519664287567,
+      "learning_rate": 0.0001,
+      "loss": 1.8419,
+      "step": 1983
+    },
+    {
+      "epoch": 0.22790190109700764,
+      "grad_norm": 0.36118772625923157,
+      "learning_rate": 0.0001,
+      "loss": 1.6452,
+      "step": 1984
+    },
+    {
+      "epoch": 0.22801677100683476,
+      "grad_norm": 0.36958158016204834,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 1985
+    },
+    {
+      "epoch": 0.22813164091666188,
+      "grad_norm": 0.365408331155777,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 1986
+    },
+    {
+      "epoch": 0.228246510826489,
+      "grad_norm": 0.3450736999511719,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 1987
+    },
+    {
+      "epoch": 0.22836138073631612,
+      "grad_norm": 0.37709590792655945,
+      "learning_rate": 0.0001,
+      "loss": 1.7375,
+      "step": 1988
+    },
+    {
+      "epoch": 0.22847625064614324,
+      "grad_norm": 0.3818245232105255,
+      "learning_rate": 0.0001,
+      "loss": 1.7378,
+      "step": 1989
+    },
+    {
+      "epoch": 0.22859112055597036,
+      "grad_norm": 0.375186949968338,
+      "learning_rate": 0.0001,
+      "loss": 1.7919,
+      "step": 1990
+    },
+    {
+      "epoch": 0.22870599046579748,
+      "grad_norm": 0.3831838369369507,
+      "learning_rate": 0.0001,
+      "loss": 1.9628,
+      "step": 1991
+    },
+    {
+      "epoch": 0.2288208603756246,
+      "grad_norm": 0.3555734157562256,
+      "learning_rate": 0.0001,
+      "loss": 1.7085,
+      "step": 1992
+    },
+    {
+      "epoch": 0.22893573028545172,
+      "grad_norm": 0.36431822180747986,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 1993
+    },
+    {
+      "epoch": 0.22905060019527884,
+      "grad_norm": 0.36310869455337524,
+      "learning_rate": 0.0001,
+      "loss": 1.8057,
+      "step": 1994
+    },
+    {
+      "epoch": 0.22916547010510596,
+      "grad_norm": 0.3640761375427246,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 1995
+    },
+    {
+      "epoch": 0.22928034001493308,
+      "grad_norm": 0.3574616014957428,
+      "learning_rate": 0.0001,
+      "loss": 1.6583,
+      "step": 1996
+    },
+    {
+      "epoch": 0.2293952099247602,
+      "grad_norm": 0.36930689215660095,
+      "learning_rate": 0.0001,
+      "loss": 1.7673,
+      "step": 1997
+    },
+    {
+      "epoch": 0.22951007983458732,
+      "grad_norm": 0.3680112659931183,
+      "learning_rate": 0.0001,
+      "loss": 1.6705,
+      "step": 1998
+    },
+    {
+      "epoch": 0.22962494974441444,
+      "grad_norm": 0.3612775504589081,
+      "learning_rate": 0.0001,
+      "loss": 1.827,
+      "step": 1999
+    },
+    {
+      "epoch": 0.22973981965424156,
+      "grad_norm": 0.3716758191585541,
+      "learning_rate": 0.0001,
+      "loss": 1.7205,
+      "step": 2000
+    },
+    {
+      "epoch": 0.22985468956406868,
+      "grad_norm": 0.3520873785018921,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 2001
+    },
+    {
+      "epoch": 0.2299695594738958,
+      "grad_norm": 0.3716805875301361,
+      "learning_rate": 0.0001,
+      "loss": 1.9395,
+      "step": 2002
+    },
+    {
+      "epoch": 0.23008442938372292,
+      "grad_norm": 0.35050535202026367,
+      "learning_rate": 0.0001,
+      "loss": 1.7538,
+      "step": 2003
+    },
+    {
+      "epoch": 0.23019929929355004,
+      "grad_norm": 0.35754668712615967,
+      "learning_rate": 0.0001,
+      "loss": 1.6116,
+      "step": 2004
+    },
+    {
+      "epoch": 0.23031416920337716,
+      "grad_norm": 0.35769370198249817,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 2005
+    },
+    {
+      "epoch": 0.2304290391132043,
+      "grad_norm": 0.3884750008583069,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 2006
+    },
+    {
+      "epoch": 0.23054390902303143,
+      "grad_norm": 0.35404741764068604,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 2007
+    },
+    {
+      "epoch": 0.23065877893285855,
+      "grad_norm": 0.37792080640792847,
+      "learning_rate": 0.0001,
+      "loss": 1.694,
+      "step": 2008
+    },
+    {
+      "epoch": 0.23077364884268567,
+      "grad_norm": 0.3727210462093353,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 2009
+    },
+    {
+      "epoch": 0.2308885187525128,
+      "grad_norm": 0.3520371615886688,
+      "learning_rate": 0.0001,
+      "loss": 1.6815,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2310033886623399,
+      "grad_norm": 0.3894649147987366,
+      "learning_rate": 0.0001,
+      "loss": 1.6817,
+      "step": 2011
+    },
+    {
+      "epoch": 0.23111825857216703,
+      "grad_norm": 0.3612670302391052,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 2012
+    },
+    {
+      "epoch": 0.23123312848199415,
+      "grad_norm": 0.35834938287734985,
+      "learning_rate": 0.0001,
+      "loss": 1.6607,
+      "step": 2013
+    },
+    {
+      "epoch": 0.23134799839182127,
+      "grad_norm": 0.36019662022590637,
+      "learning_rate": 0.0001,
+      "loss": 1.8426,
+      "step": 2014
+    },
+    {
+      "epoch": 0.2314628683016484,
+      "grad_norm": 0.35688281059265137,
+      "learning_rate": 0.0001,
+      "loss": 1.6774,
+      "step": 2015
+    },
+    {
+      "epoch": 0.2315777382114755,
+      "grad_norm": 0.3944391906261444,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 2016
+    },
+    {
+      "epoch": 0.23169260812130263,
+      "grad_norm": 0.4213406443595886,
+      "learning_rate": 0.0001,
+      "loss": 1.5176,
+      "step": 2017
+    },
+    {
+      "epoch": 0.23180747803112975,
+      "grad_norm": 0.37755805253982544,
+      "learning_rate": 0.0001,
+      "loss": 1.7714,
+      "step": 2018
+    },
+    {
+      "epoch": 0.23192234794095687,
+      "grad_norm": 0.36677590012550354,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 2019
+    },
+    {
+      "epoch": 0.232037217850784,
+      "grad_norm": 0.40023931860923767,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2321520877606111,
+      "grad_norm": 0.36305657029151917,
+      "learning_rate": 0.0001,
+      "loss": 1.726,
+      "step": 2021
+    },
+    {
+      "epoch": 0.23226695767043823,
+      "grad_norm": 0.3816761076450348,
+      "learning_rate": 0.0001,
+      "loss": 1.8716,
+      "step": 2022
+    },
+    {
+      "epoch": 0.23238182758026535,
+      "grad_norm": 0.3882976770401001,
+      "learning_rate": 0.0001,
+      "loss": 1.8066,
+      "step": 2023
+    },
+    {
+      "epoch": 0.23249669749009247,
+      "grad_norm": 0.3663192093372345,
+      "learning_rate": 0.0001,
+      "loss": 1.5011,
+      "step": 2024
+    },
+    {
+      "epoch": 0.2326115673999196,
+      "grad_norm": 0.3655019998550415,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 2025
+    },
+    {
+      "epoch": 0.2327264373097467,
+      "grad_norm": 0.3717341423034668,
+      "learning_rate": 0.0001,
+      "loss": 1.766,
+      "step": 2026
+    },
+    {
+      "epoch": 0.23284130721957383,
+      "grad_norm": 0.38333335518836975,
+      "learning_rate": 0.0001,
+      "loss": 1.8287,
+      "step": 2027
+    },
+    {
+      "epoch": 0.23295617712940095,
+      "grad_norm": 0.3660372197628021,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 2028
+    },
+    {
+      "epoch": 0.23307104703922807,
+      "grad_norm": 0.4071906507015228,
+      "learning_rate": 0.0001,
+      "loss": 1.6666,
+      "step": 2029
+    },
+    {
+      "epoch": 0.2331859169490552,
+      "grad_norm": 0.33285489678382874,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 2030
+    },
+    {
+      "epoch": 0.2333007868588823,
+      "grad_norm": 0.36184608936309814,
+      "learning_rate": 0.0001,
+      "loss": 1.7848,
+      "step": 2031
+    },
+    {
+      "epoch": 0.23341565676870943,
+      "grad_norm": 0.3524230718612671,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 2032
+    },
+    {
+      "epoch": 0.23353052667853655,
+      "grad_norm": 0.37483248114585876,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 2033
+    },
+    {
+      "epoch": 0.23364539658836367,
+      "grad_norm": 0.3705480992794037,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 2034
+    },
+    {
+      "epoch": 0.2337602664981908,
+      "grad_norm": 0.39686012268066406,
+      "learning_rate": 0.0001,
+      "loss": 1.7466,
+      "step": 2035
+    },
+    {
+      "epoch": 0.2338751364080179,
+      "grad_norm": 0.3723219633102417,
+      "learning_rate": 0.0001,
+      "loss": 1.6358,
+      "step": 2036
+    },
+    {
+      "epoch": 0.23399000631784503,
+      "grad_norm": 0.3748587965965271,
+      "learning_rate": 0.0001,
+      "loss": 1.7883,
+      "step": 2037
+    },
+    {
+      "epoch": 0.23410487622767215,
+      "grad_norm": 0.3536939024925232,
+      "learning_rate": 0.0001,
+      "loss": 1.8048,
+      "step": 2038
+    },
+    {
+      "epoch": 0.23421974613749927,
+      "grad_norm": 0.36391976475715637,
+      "learning_rate": 0.0001,
+      "loss": 1.7326,
+      "step": 2039
+    },
+    {
+      "epoch": 0.23433461604732642,
+      "grad_norm": 0.3579852879047394,
+      "learning_rate": 0.0001,
+      "loss": 1.7311,
+      "step": 2040
+    },
+    {
+      "epoch": 0.23444948595715354,
+      "grad_norm": 0.3592579960823059,
+      "learning_rate": 0.0001,
+      "loss": 1.7675,
+      "step": 2041
+    },
+    {
+      "epoch": 0.23456435586698066,
+      "grad_norm": 0.34196656942367554,
+      "learning_rate": 0.0001,
+      "loss": 1.4591,
+      "step": 2042
+    },
+    {
+      "epoch": 0.23467922577680778,
+      "grad_norm": 0.40393322706222534,
+      "learning_rate": 0.0001,
+      "loss": 1.8106,
+      "step": 2043
+    },
+    {
+      "epoch": 0.2347940956866349,
+      "grad_norm": 0.39818137884140015,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 2044
+    },
+    {
+      "epoch": 0.23490896559646202,
+      "grad_norm": 0.3436848521232605,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 2045
+    },
+    {
+      "epoch": 0.23502383550628914,
+      "grad_norm": 0.35754308104515076,
+      "learning_rate": 0.0001,
+      "loss": 1.7914,
+      "step": 2046
+    },
+    {
+      "epoch": 0.23513870541611626,
+      "grad_norm": 0.37163278460502625,
+      "learning_rate": 0.0001,
+      "loss": 1.7085,
+      "step": 2047
+    },
+    {
+      "epoch": 0.23525357532594338,
+      "grad_norm": 0.3725832402706146,
+      "learning_rate": 0.0001,
+      "loss": 1.7469,
+      "step": 2048
+    },
+    {
+      "epoch": 0.2353684452357705,
+      "grad_norm": 0.3440713584423065,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 2049
+    },
+    {
+      "epoch": 0.23548331514559762,
+      "grad_norm": 0.3776625096797943,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 2050
+    },
+    {
+      "epoch": 0.23559818505542474,
+      "grad_norm": 0.37252920866012573,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 2051
+    },
+    {
+      "epoch": 0.23571305496525186,
+      "grad_norm": 0.36468106508255005,
+      "learning_rate": 0.0001,
+      "loss": 1.7879,
+      "step": 2052
+    },
+    {
+      "epoch": 0.23582792487507898,
+      "grad_norm": 0.3846648037433624,
+      "learning_rate": 0.0001,
+      "loss": 1.7629,
+      "step": 2053
+    },
+    {
+      "epoch": 0.2359427947849061,
+      "grad_norm": 0.3676551580429077,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 2054
+    },
+    {
+      "epoch": 0.23605766469473322,
+      "grad_norm": 0.38455480337142944,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 2055
+    },
+    {
+      "epoch": 0.23617253460456034,
+      "grad_norm": 0.39479124546051025,
+      "learning_rate": 0.0001,
+      "loss": 1.8082,
+      "step": 2056
+    },
+    {
+      "epoch": 0.23628740451438746,
+      "grad_norm": 0.37550801038742065,
+      "learning_rate": 0.0001,
+      "loss": 1.7738,
+      "step": 2057
+    },
+    {
+      "epoch": 0.23640227442421458,
+      "grad_norm": 0.3539324104785919,
+      "learning_rate": 0.0001,
+      "loss": 1.4211,
+      "step": 2058
+    },
+    {
+      "epoch": 0.2365171443340417,
+      "grad_norm": 0.38804805278778076,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 2059
+    },
+    {
+      "epoch": 0.23663201424386882,
+      "grad_norm": 0.34223702549934387,
+      "learning_rate": 0.0001,
+      "loss": 1.6774,
+      "step": 2060
+    },
+    {
+      "epoch": 0.23674688415369594,
+      "grad_norm": 0.3723870515823364,
+      "learning_rate": 0.0001,
+      "loss": 1.7951,
+      "step": 2061
+    },
+    {
+      "epoch": 0.23686175406352306,
+      "grad_norm": 0.3813320994377136,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 2062
+    },
+    {
+      "epoch": 0.23697662397335018,
+      "grad_norm": 0.38882893323898315,
+      "learning_rate": 0.0001,
+      "loss": 1.7914,
+      "step": 2063
+    },
+    {
+      "epoch": 0.2370914938831773,
+      "grad_norm": 0.3598068058490753,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 2064
+    },
+    {
+      "epoch": 0.23720636379300442,
+      "grad_norm": 0.35870295763015747,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 2065
+    },
+    {
+      "epoch": 0.23732123370283154,
+      "grad_norm": 0.38279473781585693,
+      "learning_rate": 0.0001,
+      "loss": 1.7661,
+      "step": 2066
+    },
+    {
+      "epoch": 0.23743610361265866,
+      "grad_norm": 0.3647814393043518,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 2067
+    },
+    {
+      "epoch": 0.23755097352248578,
+      "grad_norm": 0.38286200165748596,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 2068
+    },
+    {
+      "epoch": 0.2376658434323129,
+      "grad_norm": 0.3622128367424011,
+      "learning_rate": 0.0001,
+      "loss": 1.8133,
+      "step": 2069
+    },
+    {
+      "epoch": 0.23778071334214002,
+      "grad_norm": 0.38336479663848877,
+      "learning_rate": 0.0001,
+      "loss": 1.8398,
+      "step": 2070
+    },
+    {
+      "epoch": 0.23789558325196714,
+      "grad_norm": 0.34429532289505005,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 2071
+    },
+    {
+      "epoch": 0.23801045316179426,
+      "grad_norm": 0.3539319932460785,
+      "learning_rate": 0.0001,
+      "loss": 1.8332,
+      "step": 2072
+    },
+    {
+      "epoch": 0.23812532307162138,
+      "grad_norm": 0.41278359293937683,
+      "learning_rate": 0.0001,
+      "loss": 1.7155,
+      "step": 2073
+    },
+    {
+      "epoch": 0.23824019298144852,
+      "grad_norm": 0.3446906507015228,
+      "learning_rate": 0.0001,
+      "loss": 1.6454,
+      "step": 2074
+    },
+    {
+      "epoch": 0.23835506289127564,
+      "grad_norm": 0.3742326498031616,
+      "learning_rate": 0.0001,
+      "loss": 1.7497,
+      "step": 2075
+    },
+    {
+      "epoch": 0.23846993280110276,
+      "grad_norm": 0.3751114308834076,
+      "learning_rate": 0.0001,
+      "loss": 1.766,
+      "step": 2076
+    },
+    {
+      "epoch": 0.23858480271092988,
+      "grad_norm": 0.358395516872406,
+      "learning_rate": 0.0001,
+      "loss": 1.8033,
+      "step": 2077
+    },
+    {
+      "epoch": 0.238699672620757,
+      "grad_norm": 0.3309650123119354,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 2078
+    },
+    {
+      "epoch": 0.23881454253058412,
+      "grad_norm": 0.3873356878757477,
+      "learning_rate": 0.0001,
+      "loss": 1.7555,
+      "step": 2079
+    },
+    {
+      "epoch": 0.23892941244041124,
+      "grad_norm": 0.3576614558696747,
+      "learning_rate": 0.0001,
+      "loss": 1.7634,
+      "step": 2080
+    },
+    {
+      "epoch": 0.23904428235023836,
+      "grad_norm": 0.4060707688331604,
+      "learning_rate": 0.0001,
+      "loss": 1.8486,
+      "step": 2081
+    },
+    {
+      "epoch": 0.23915915226006548,
+      "grad_norm": 0.3609447181224823,
+      "learning_rate": 0.0001,
+      "loss": 1.7257,
+      "step": 2082
+    },
+    {
+      "epoch": 0.2392740221698926,
+      "grad_norm": 0.39137205481529236,
+      "learning_rate": 0.0001,
+      "loss": 1.788,
+      "step": 2083
+    },
+    {
+      "epoch": 0.23938889207971972,
+      "grad_norm": 0.4108488857746124,
+      "learning_rate": 0.0001,
+      "loss": 1.5738,
+      "step": 2084
+    },
+    {
+      "epoch": 0.23950376198954684,
+      "grad_norm": 0.3932496905326843,
+      "learning_rate": 0.0001,
+      "loss": 1.9864,
+      "step": 2085
+    },
+    {
+      "epoch": 0.23961863189937396,
+      "grad_norm": 0.3582679331302643,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 2086
+    },
+    {
+      "epoch": 0.23973350180920108,
+      "grad_norm": 0.382107138633728,
+      "learning_rate": 0.0001,
+      "loss": 1.8809,
+      "step": 2087
+    },
+    {
+      "epoch": 0.2398483717190282,
+      "grad_norm": 0.37730252742767334,
+      "learning_rate": 0.0001,
+      "loss": 1.5719,
+      "step": 2088
+    },
+    {
+      "epoch": 0.23996324162885532,
+      "grad_norm": 0.39350810647010803,
+      "learning_rate": 0.0001,
+      "loss": 1.7602,
+      "step": 2089
+    },
+    {
+      "epoch": 0.24007811153868244,
+      "grad_norm": 0.3490624725818634,
+      "learning_rate": 0.0001,
+      "loss": 1.7111,
+      "step": 2090
+    },
+    {
+      "epoch": 0.24019298144850956,
+      "grad_norm": 0.39576220512390137,
+      "learning_rate": 0.0001,
+      "loss": 1.7265,
+      "step": 2091
+    },
+    {
+      "epoch": 0.24030785135833668,
+      "grad_norm": 0.34063565731048584,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 2092
+    },
+    {
+      "epoch": 0.2404227212681638,
+      "grad_norm": 0.36094650626182556,
+      "learning_rate": 0.0001,
+      "loss": 1.7337,
+      "step": 2093
+    },
+    {
+      "epoch": 0.24053759117799092,
+      "grad_norm": 0.39108943939208984,
+      "learning_rate": 0.0001,
+      "loss": 1.8195,
+      "step": 2094
+    },
+    {
+      "epoch": 0.24065246108781804,
+      "grad_norm": 0.3705921471118927,
+      "learning_rate": 0.0001,
+      "loss": 1.5288,
+      "step": 2095
+    },
+    {
+      "epoch": 0.24076733099764516,
+      "grad_norm": 0.38726896047592163,
+      "learning_rate": 0.0001,
+      "loss": 1.8124,
+      "step": 2096
+    },
+    {
+      "epoch": 0.24088220090747228,
+      "grad_norm": 0.3693424165248871,
+      "learning_rate": 0.0001,
+      "loss": 1.4665,
+      "step": 2097
+    },
+    {
+      "epoch": 0.2409970708172994,
+      "grad_norm": 0.37922948598861694,
+      "learning_rate": 0.0001,
+      "loss": 1.6475,
+      "step": 2098
+    },
+    {
+      "epoch": 0.24111194072712652,
+      "grad_norm": 0.3428456485271454,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 2099
+    },
+    {
+      "epoch": 0.24122681063695364,
+      "grad_norm": 0.38854971528053284,
+      "learning_rate": 0.0001,
+      "loss": 1.8648,
+      "step": 2100
+    },
+    {
+      "epoch": 0.24134168054678076,
+      "grad_norm": 0.3516068756580353,
+      "learning_rate": 0.0001,
+      "loss": 1.6517,
+      "step": 2101
+    },
+    {
+      "epoch": 0.24145655045660788,
+      "grad_norm": 0.3651220500469208,
+      "learning_rate": 0.0001,
+      "loss": 1.7118,
+      "step": 2102
+    },
+    {
+      "epoch": 0.241571420366435,
+      "grad_norm": 0.355566143989563,
+      "learning_rate": 0.0001,
+      "loss": 1.473,
+      "step": 2103
+    },
+    {
+      "epoch": 0.24168629027626212,
+      "grad_norm": 0.3593463599681854,
+      "learning_rate": 0.0001,
+      "loss": 1.6904,
+      "step": 2104
+    },
+    {
+      "epoch": 0.24180116018608924,
+      "grad_norm": 0.35307595133781433,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 2105
+    },
+    {
+      "epoch": 0.24191603009591636,
+      "grad_norm": 0.37137705087661743,
+      "learning_rate": 0.0001,
+      "loss": 1.6786,
+      "step": 2106
+    },
+    {
+      "epoch": 0.24203090000574348,
+      "grad_norm": 0.3728501498699188,
+      "learning_rate": 0.0001,
+      "loss": 1.8597,
+      "step": 2107
+    },
+    {
+      "epoch": 0.24214576991557063,
+      "grad_norm": 0.3600460886955261,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 2108
+    },
+    {
+      "epoch": 0.24226063982539775,
+      "grad_norm": 0.3472753167152405,
+      "learning_rate": 0.0001,
+      "loss": 1.5751,
+      "step": 2109
+    },
+    {
+      "epoch": 0.24237550973522487,
+      "grad_norm": 0.36193597316741943,
+      "learning_rate": 0.0001,
+      "loss": 1.6062,
+      "step": 2110
+    },
+    {
+      "epoch": 0.242490379645052,
+      "grad_norm": 0.3618480861186981,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 2111
+    },
+    {
+      "epoch": 0.2426052495548791,
+      "grad_norm": 0.358378529548645,
+      "learning_rate": 0.0001,
+      "loss": 1.381,
+      "step": 2112
+    },
+    {
+      "epoch": 0.24272011946470623,
+      "grad_norm": 0.386943519115448,
+      "learning_rate": 0.0001,
+      "loss": 1.896,
+      "step": 2113
+    },
+    {
+      "epoch": 0.24283498937453335,
+      "grad_norm": 0.38075029850006104,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 2114
+    },
+    {
+      "epoch": 0.24294985928436047,
+      "grad_norm": 0.3709038197994232,
+      "learning_rate": 0.0001,
+      "loss": 1.7108,
+      "step": 2115
+    },
+    {
+      "epoch": 0.2430647291941876,
+      "grad_norm": 0.3460537791252136,
+      "learning_rate": 0.0001,
+      "loss": 1.5338,
+      "step": 2116
+    },
+    {
+      "epoch": 0.2431795991040147,
+      "grad_norm": 0.3539460599422455,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 2117
+    },
+    {
+      "epoch": 0.24329446901384183,
+      "grad_norm": 0.3654404580593109,
+      "learning_rate": 0.0001,
+      "loss": 1.7205,
+      "step": 2118
+    },
+    {
+      "epoch": 0.24340933892366895,
+      "grad_norm": 0.37309154868125916,
+      "learning_rate": 0.0001,
+      "loss": 1.7026,
+      "step": 2119
+    },
+    {
+      "epoch": 0.24352420883349607,
+      "grad_norm": 0.3884303569793701,
+      "learning_rate": 0.0001,
+      "loss": 1.8393,
+      "step": 2120
+    },
+    {
+      "epoch": 0.2436390787433232,
+      "grad_norm": 0.41834697127342224,
+      "learning_rate": 0.0001,
+      "loss": 1.6144,
+      "step": 2121
+    },
+    {
+      "epoch": 0.2437539486531503,
+      "grad_norm": 0.48797813057899475,
+      "learning_rate": 0.0001,
+      "loss": 1.7264,
+      "step": 2122
+    },
+    {
+      "epoch": 0.24386881856297743,
+      "grad_norm": 0.3766729533672333,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 2123
+    },
+    {
+      "epoch": 0.24398368847280455,
+      "grad_norm": 0.3885788023471832,
+      "learning_rate": 0.0001,
+      "loss": 1.9709,
+      "step": 2124
+    },
+    {
+      "epoch": 0.24409855838263167,
+      "grad_norm": 0.36236581206321716,
+      "learning_rate": 0.0001,
+      "loss": 1.5214,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2442134282924588,
+      "grad_norm": 0.368348628282547,
+      "learning_rate": 0.0001,
+      "loss": 1.6698,
+      "step": 2126
+    },
+    {
+      "epoch": 0.2443282982022859,
+      "grad_norm": 0.3759993612766266,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 2127
+    },
+    {
+      "epoch": 0.24444316811211303,
+      "grad_norm": 0.3532876670360565,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 2128
+    },
+    {
+      "epoch": 0.24455803802194015,
+      "grad_norm": 0.3924901783466339,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 2129
+    },
+    {
+      "epoch": 0.24467290793176727,
+      "grad_norm": 0.36675912141799927,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2447877778415944,
+      "grad_norm": 0.3623403310775757,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 2131
+    },
+    {
+      "epoch": 0.2449026477514215,
+      "grad_norm": 0.4029271900653839,
+      "learning_rate": 0.0001,
+      "loss": 1.8573,
+      "step": 2132
+    },
+    {
+      "epoch": 0.24501751766124863,
+      "grad_norm": 0.38149070739746094,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 2133
+    },
+    {
+      "epoch": 0.24513238757107575,
+      "grad_norm": 0.36011314392089844,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 2134
+    },
+    {
+      "epoch": 0.24524725748090287,
+      "grad_norm": 0.35948145389556885,
+      "learning_rate": 0.0001,
+      "loss": 1.4656,
+      "step": 2135
+    },
+    {
+      "epoch": 0.24536212739073,
+      "grad_norm": 0.39184871315956116,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 2136
+    },
+    {
+      "epoch": 0.2454769973005571,
+      "grad_norm": 0.3509827256202698,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 2137
+    },
+    {
+      "epoch": 0.24559186721038423,
+      "grad_norm": 0.37394797801971436,
+      "learning_rate": 0.0001,
+      "loss": 1.7774,
+      "step": 2138
+    },
+    {
+      "epoch": 0.24570673712021135,
+      "grad_norm": 0.37039676308631897,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 2139
+    },
+    {
+      "epoch": 0.24582160703003847,
+      "grad_norm": 0.34239184856414795,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2459364769398656,
+      "grad_norm": 0.4217103123664856,
+      "learning_rate": 0.0001,
+      "loss": 2.0345,
+      "step": 2141
+    },
+    {
+      "epoch": 0.2460513468496927,
+      "grad_norm": 0.352780282497406,
+      "learning_rate": 0.0001,
+      "loss": 1.6952,
+      "step": 2142
+    },
+    {
+      "epoch": 0.24616621675951986,
+      "grad_norm": 0.34040459990501404,
+      "learning_rate": 0.0001,
+      "loss": 1.3958,
+      "step": 2143
+    },
+    {
+      "epoch": 0.24628108666934698,
+      "grad_norm": 0.3590191900730133,
+      "learning_rate": 0.0001,
+      "loss": 1.5738,
+      "step": 2144
+    },
+    {
+      "epoch": 0.2463959565791741,
+      "grad_norm": 0.3775222599506378,
+      "learning_rate": 0.0001,
+      "loss": 1.7631,
+      "step": 2145
+    },
+    {
+      "epoch": 0.24651082648900122,
+      "grad_norm": 0.3881835639476776,
+      "learning_rate": 0.0001,
+      "loss": 1.8367,
+      "step": 2146
+    },
+    {
+      "epoch": 0.24662569639882834,
+      "grad_norm": 0.35426759719848633,
+      "learning_rate": 0.0001,
+      "loss": 1.5748,
+      "step": 2147
+    },
+    {
+      "epoch": 0.24674056630865546,
+      "grad_norm": 0.43328168988227844,
+      "learning_rate": 0.0001,
+      "loss": 1.7385,
+      "step": 2148
+    },
+    {
+      "epoch": 0.24685543621848258,
+      "grad_norm": 0.38424152135849,
+      "learning_rate": 0.0001,
+      "loss": 1.8426,
+      "step": 2149
+    },
+    {
+      "epoch": 0.2469703061283097,
+      "grad_norm": 0.3645937144756317,
+      "learning_rate": 0.0001,
+      "loss": 1.7276,
+      "step": 2150
+    },
+    {
+      "epoch": 0.24708517603813682,
+      "grad_norm": 0.413036584854126,
+      "learning_rate": 0.0001,
+      "loss": 1.8982,
+      "step": 2151
+    },
+    {
+      "epoch": 0.24720004594796394,
+      "grad_norm": 0.35927438735961914,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 2152
+    },
+    {
+      "epoch": 0.24731491585779106,
+      "grad_norm": 0.3814374804496765,
+      "learning_rate": 0.0001,
+      "loss": 1.76,
+      "step": 2153
+    },
+    {
+      "epoch": 0.24742978576761818,
+      "grad_norm": 0.3744584321975708,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 2154
+    },
+    {
+      "epoch": 0.2475446556774453,
+      "grad_norm": 0.3815367817878723,
+      "learning_rate": 0.0001,
+      "loss": 1.8374,
+      "step": 2155
+    },
+    {
+      "epoch": 0.24765952558727242,
+      "grad_norm": 0.37428149580955505,
+      "learning_rate": 0.0001,
+      "loss": 1.7515,
+      "step": 2156
+    },
+    {
+      "epoch": 0.24777439549709954,
+      "grad_norm": 0.3383006751537323,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 2157
+    },
+    {
+      "epoch": 0.24788926540692666,
+      "grad_norm": 0.3787260055541992,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 2158
+    },
+    {
+      "epoch": 0.24800413531675378,
+      "grad_norm": 0.34863799810409546,
+      "learning_rate": 0.0001,
+      "loss": 1.7608,
+      "step": 2159
+    },
+    {
+      "epoch": 0.2481190052265809,
+      "grad_norm": 0.34521782398223877,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 2160
+    },
+    {
+      "epoch": 0.24823387513640802,
+      "grad_norm": 0.4506453275680542,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 2161
+    },
+    {
+      "epoch": 0.24834874504623514,
+      "grad_norm": 0.388887494802475,
+      "learning_rate": 0.0001,
+      "loss": 1.9017,
+      "step": 2162
+    },
+    {
+      "epoch": 0.24846361495606226,
+      "grad_norm": 0.3828551173210144,
+      "learning_rate": 0.0001,
+      "loss": 1.755,
+      "step": 2163
+    },
+    {
+      "epoch": 0.24857848486588938,
+      "grad_norm": 0.35569700598716736,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 2164
+    },
+    {
+      "epoch": 0.2486933547757165,
+      "grad_norm": 0.3651540279388428,
+      "learning_rate": 0.0001,
+      "loss": 1.7724,
+      "step": 2165
+    },
+    {
+      "epoch": 0.24880822468554362,
+      "grad_norm": 0.362402081489563,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 2166
+    },
+    {
+      "epoch": 0.24892309459537074,
+      "grad_norm": 0.36203300952911377,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 2167
+    },
+    {
+      "epoch": 0.24903796450519786,
+      "grad_norm": 0.37122076749801636,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 2168
+    },
+    {
+      "epoch": 0.24915283441502498,
+      "grad_norm": 0.357860803604126,
+      "learning_rate": 0.0001,
+      "loss": 1.6995,
+      "step": 2169
+    },
+    {
+      "epoch": 0.2492677043248521,
+      "grad_norm": 0.36631032824516296,
+      "learning_rate": 0.0001,
+      "loss": 1.7249,
+      "step": 2170
+    },
+    {
+      "epoch": 0.24938257423467922,
+      "grad_norm": 0.37046635150909424,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 2171
+    },
+    {
+      "epoch": 0.24949744414450634,
+      "grad_norm": 0.39707261323928833,
+      "learning_rate": 0.0001,
+      "loss": 1.9089,
+      "step": 2172
+    },
+    {
+      "epoch": 0.24961231405433346,
+      "grad_norm": 0.3699803650379181,
+      "learning_rate": 0.0001,
+      "loss": 1.6095,
+      "step": 2173
+    },
+    {
+      "epoch": 0.24972718396416058,
+      "grad_norm": 0.3902735710144043,
+      "learning_rate": 0.0001,
+      "loss": 1.7023,
+      "step": 2174
+    },
+    {
+      "epoch": 0.2498420538739877,
+      "grad_norm": 0.3696674406528473,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 2175
+    },
+    {
+      "epoch": 0.24995692378381482,
+      "grad_norm": 0.3827657699584961,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 2176
+    },
+    {
+      "epoch": 0.25007179369364196,
+      "grad_norm": 0.38356491923332214,
+      "learning_rate": 0.0001,
+      "loss": 1.7748,
+      "step": 2177
+    },
+    {
+      "epoch": 0.2501866636034691,
+      "grad_norm": 0.3400535583496094,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 2178
+    },
+    {
+      "epoch": 0.2503015335132962,
+      "grad_norm": 0.37260451912879944,
+      "learning_rate": 0.0001,
+      "loss": 1.7625,
+      "step": 2179
+    },
+    {
+      "epoch": 0.2504164034231233,
+      "grad_norm": 0.4127359688282013,
+      "learning_rate": 0.0001,
+      "loss": 1.4756,
+      "step": 2180
+    },
+    {
+      "epoch": 0.25053127333295044,
+      "grad_norm": 0.3528091609477997,
+      "learning_rate": 0.0001,
+      "loss": 1.6681,
+      "step": 2181
+    },
+    {
+      "epoch": 0.25064614324277756,
+      "grad_norm": 0.39472371339797974,
+      "learning_rate": 0.0001,
+      "loss": 1.8162,
+      "step": 2182
+    },
+    {
+      "epoch": 0.2507610131526047,
+      "grad_norm": 0.3909297585487366,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 2183
+    },
+    {
+      "epoch": 0.2508758830624318,
+      "grad_norm": 0.42550724744796753,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 2184
+    },
+    {
+      "epoch": 0.2509907529722589,
+      "grad_norm": 0.3795909881591797,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 2185
+    },
+    {
+      "epoch": 0.25110562288208604,
+      "grad_norm": 0.37554746866226196,
+      "learning_rate": 0.0001,
+      "loss": 1.8221,
+      "step": 2186
+    },
+    {
+      "epoch": 0.25122049279191316,
+      "grad_norm": 0.362114280462265,
+      "learning_rate": 0.0001,
+      "loss": 1.6699,
+      "step": 2187
+    },
+    {
+      "epoch": 0.2513353627017403,
+      "grad_norm": 0.36386096477508545,
+      "learning_rate": 0.0001,
+      "loss": 1.4788,
+      "step": 2188
+    },
+    {
+      "epoch": 0.2514502326115674,
+      "grad_norm": 0.3716920018196106,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 2189
+    },
+    {
+      "epoch": 0.2515651025213945,
+      "grad_norm": 0.3535931706428528,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 2190
+    },
+    {
+      "epoch": 0.25167997243122164,
+      "grad_norm": 0.3724220097064972,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 2191
+    },
+    {
+      "epoch": 0.25179484234104876,
+      "grad_norm": 0.40217527747154236,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 2192
+    },
+    {
+      "epoch": 0.2519097122508759,
+      "grad_norm": 0.383705198764801,
+      "learning_rate": 0.0001,
+      "loss": 1.6798,
+      "step": 2193
+    },
+    {
+      "epoch": 0.252024582160703,
+      "grad_norm": 0.35728463530540466,
+      "learning_rate": 0.0001,
+      "loss": 1.7787,
+      "step": 2194
+    },
+    {
+      "epoch": 0.2521394520705301,
+      "grad_norm": 0.34664642810821533,
+      "learning_rate": 0.0001,
+      "loss": 1.4345,
+      "step": 2195
+    },
+    {
+      "epoch": 0.25225432198035724,
+      "grad_norm": 0.3613569438457489,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 2196
+    },
+    {
+      "epoch": 0.25236919189018436,
+      "grad_norm": 0.36570632457733154,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 2197
+    },
+    {
+      "epoch": 0.2524840618000115,
+      "grad_norm": 0.36983218789100647,
+      "learning_rate": 0.0001,
+      "loss": 1.7606,
+      "step": 2198
+    },
+    {
+      "epoch": 0.2525989317098386,
+      "grad_norm": 0.35930711030960083,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 2199
+    },
+    {
+      "epoch": 0.2527138016196657,
+      "grad_norm": 0.39522892236709595,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 2200
+    },
+    {
+      "epoch": 0.25282867152949284,
+      "grad_norm": 0.3675948679447174,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 2201
+    },
+    {
+      "epoch": 0.25294354143931996,
+      "grad_norm": 0.3466283679008484,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 2202
+    },
+    {
+      "epoch": 0.2530584113491471,
+      "grad_norm": 0.3820589482784271,
+      "learning_rate": 0.0001,
+      "loss": 1.8305,
+      "step": 2203
+    },
+    {
+      "epoch": 0.2531732812589742,
+      "grad_norm": 0.45334169268608093,
+      "learning_rate": 0.0001,
+      "loss": 1.8272,
+      "step": 2204
+    },
+    {
+      "epoch": 0.2532881511688013,
+      "grad_norm": 0.36322474479675293,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 2205
+    },
+    {
+      "epoch": 0.25340302107862844,
+      "grad_norm": 0.3699181377887726,
+      "learning_rate": 0.0001,
+      "loss": 1.7263,
+      "step": 2206
+    },
+    {
+      "epoch": 0.25351789098845556,
+      "grad_norm": 0.3594902455806732,
+      "learning_rate": 0.0001,
+      "loss": 1.6536,
+      "step": 2207
+    },
+    {
+      "epoch": 0.2536327608982827,
+      "grad_norm": 0.3868749141693115,
+      "learning_rate": 0.0001,
+      "loss": 1.7946,
+      "step": 2208
+    },
+    {
+      "epoch": 0.2537476308081098,
+      "grad_norm": 0.39426884055137634,
+      "learning_rate": 0.0001,
+      "loss": 1.898,
+      "step": 2209
+    },
+    {
+      "epoch": 0.2538625007179369,
+      "grad_norm": 0.41421404480934143,
+      "learning_rate": 0.0001,
+      "loss": 1.8456,
+      "step": 2210
+    },
+    {
+      "epoch": 0.25397737062776404,
+      "grad_norm": 0.35613730549812317,
+      "learning_rate": 0.0001,
+      "loss": 1.706,
+      "step": 2211
+    },
+    {
+      "epoch": 0.25409224053759116,
+      "grad_norm": 0.38455823063850403,
+      "learning_rate": 0.0001,
+      "loss": 1.6824,
+      "step": 2212
+    },
+    {
+      "epoch": 0.2542071104474183,
+      "grad_norm": 0.3799576759338379,
+      "learning_rate": 0.0001,
+      "loss": 1.714,
+      "step": 2213
+    },
+    {
+      "epoch": 0.2543219803572454,
+      "grad_norm": 0.43304580450057983,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 2214
+    },
+    {
+      "epoch": 0.2544368502670725,
+      "grad_norm": 0.39498084783554077,
+      "learning_rate": 0.0001,
+      "loss": 1.8003,
+      "step": 2215
+    },
+    {
+      "epoch": 0.25455172017689964,
+      "grad_norm": 0.3838403522968292,
+      "learning_rate": 0.0001,
+      "loss": 1.6831,
+      "step": 2216
+    },
+    {
+      "epoch": 0.25466659008672676,
+      "grad_norm": 0.38586917519569397,
+      "learning_rate": 0.0001,
+      "loss": 1.8446,
+      "step": 2217
+    },
+    {
+      "epoch": 0.2547814599965539,
+      "grad_norm": 0.37291255593299866,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 2218
+    },
+    {
+      "epoch": 0.254896329906381,
+      "grad_norm": 0.37416213750839233,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 2219
+    },
+    {
+      "epoch": 0.2550111998162081,
+      "grad_norm": 0.37467607855796814,
+      "learning_rate": 0.0001,
+      "loss": 1.6916,
+      "step": 2220
+    },
+    {
+      "epoch": 0.25512606972603524,
+      "grad_norm": 0.3643462061882019,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 2221
+    },
+    {
+      "epoch": 0.25524093963586236,
+      "grad_norm": 0.3733225464820862,
+      "learning_rate": 0.0001,
+      "loss": 1.7953,
+      "step": 2222
+    },
+    {
+      "epoch": 0.2553558095456895,
+      "grad_norm": 0.3754335045814514,
+      "learning_rate": 0.0001,
+      "loss": 1.8093,
+      "step": 2223
+    },
+    {
+      "epoch": 0.2554706794555166,
+      "grad_norm": 0.3865663707256317,
+      "learning_rate": 0.0001,
+      "loss": 1.7088,
+      "step": 2224
+    },
+    {
+      "epoch": 0.2555855493653437,
+      "grad_norm": 0.38617023825645447,
+      "learning_rate": 0.0001,
+      "loss": 1.8311,
+      "step": 2225
+    },
+    {
+      "epoch": 0.25570041927517084,
+      "grad_norm": 0.37515413761138916,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 2226
+    },
+    {
+      "epoch": 0.25581528918499796,
+      "grad_norm": 0.38885021209716797,
+      "learning_rate": 0.0001,
+      "loss": 1.8163,
+      "step": 2227
+    },
+    {
+      "epoch": 0.25593015909482514,
+      "grad_norm": 0.37118270993232727,
+      "learning_rate": 0.0001,
+      "loss": 1.8075,
+      "step": 2228
+    },
+    {
+      "epoch": 0.25604502900465226,
+      "grad_norm": 0.389565110206604,
+      "learning_rate": 0.0001,
+      "loss": 1.6765,
+      "step": 2229
+    },
+    {
+      "epoch": 0.2561598989144794,
+      "grad_norm": 0.3385695517063141,
+      "learning_rate": 0.0001,
+      "loss": 1.3302,
+      "step": 2230
+    },
+    {
+      "epoch": 0.2562747688243065,
+      "grad_norm": 0.37382227182388306,
+      "learning_rate": 0.0001,
+      "loss": 1.7651,
+      "step": 2231
+    },
+    {
+      "epoch": 0.2563896387341336,
+      "grad_norm": 0.39900514483451843,
+      "learning_rate": 0.0001,
+      "loss": 1.7818,
+      "step": 2232
+    },
+    {
+      "epoch": 0.25650450864396074,
+      "grad_norm": 0.3673021197319031,
+      "learning_rate": 0.0001,
+      "loss": 1.814,
+      "step": 2233
+    },
+    {
+      "epoch": 0.25661937855378786,
+      "grad_norm": 0.3759724199771881,
+      "learning_rate": 0.0001,
+      "loss": 1.7816,
+      "step": 2234
+    },
+    {
+      "epoch": 0.256734248463615,
+      "grad_norm": 0.37381884455680847,
+      "learning_rate": 0.0001,
+      "loss": 1.7092,
+      "step": 2235
+    },
+    {
+      "epoch": 0.2568491183734421,
+      "grad_norm": 0.3776073157787323,
+      "learning_rate": 0.0001,
+      "loss": 1.6729,
+      "step": 2236
+    },
+    {
+      "epoch": 0.2569639882832692,
+      "grad_norm": 0.4105524718761444,
+      "learning_rate": 0.0001,
+      "loss": 1.5777,
+      "step": 2237
+    },
+    {
+      "epoch": 0.25707885819309634,
+      "grad_norm": 0.41020524501800537,
+      "learning_rate": 0.0001,
+      "loss": 1.9763,
+      "step": 2238
+    },
+    {
+      "epoch": 0.25719372810292346,
+      "grad_norm": 0.4077359139919281,
+      "learning_rate": 0.0001,
+      "loss": 2.0014,
+      "step": 2239
+    },
+    {
+      "epoch": 0.2573085980127506,
+      "grad_norm": 0.41229790449142456,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 2240
+    },
+    {
+      "epoch": 0.2574234679225777,
+      "grad_norm": 0.380489319562912,
+      "learning_rate": 0.0001,
+      "loss": 1.8079,
+      "step": 2241
+    },
+    {
+      "epoch": 0.2575383378324048,
+      "grad_norm": 0.3971484899520874,
+      "learning_rate": 0.0001,
+      "loss": 1.8367,
+      "step": 2242
+    },
+    {
+      "epoch": 0.25765320774223194,
+      "grad_norm": 0.37627938389778137,
+      "learning_rate": 0.0001,
+      "loss": 1.6502,
+      "step": 2243
+    },
+    {
+      "epoch": 0.25776807765205906,
+      "grad_norm": 0.3823041319847107,
+      "learning_rate": 0.0001,
+      "loss": 1.7642,
+      "step": 2244
+    },
+    {
+      "epoch": 0.2578829475618862,
+      "grad_norm": 0.3671165704727173,
+      "learning_rate": 0.0001,
+      "loss": 1.7102,
+      "step": 2245
+    },
+    {
+      "epoch": 0.2579978174717133,
+      "grad_norm": 0.39813950657844543,
+      "learning_rate": 0.0001,
+      "loss": 1.8005,
+      "step": 2246
+    },
+    {
+      "epoch": 0.2581126873815404,
+      "grad_norm": 0.3677361309528351,
+      "learning_rate": 0.0001,
+      "loss": 1.658,
+      "step": 2247
+    },
+    {
+      "epoch": 0.25822755729136754,
+      "grad_norm": 0.3524981439113617,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 2248
+    },
+    {
+      "epoch": 0.25834242720119466,
+      "grad_norm": 0.37544408440589905,
+      "learning_rate": 0.0001,
+      "loss": 1.6862,
+      "step": 2249
+    },
+    {
+      "epoch": 0.2584572971110218,
+      "grad_norm": 0.35381603240966797,
+      "learning_rate": 0.0001,
+      "loss": 1.5272,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2585721670208489,
+      "grad_norm": 0.39469513297080994,
+      "learning_rate": 0.0001,
+      "loss": 1.7863,
+      "step": 2251
+    },
+    {
+      "epoch": 0.258687036930676,
+      "grad_norm": 0.3802177906036377,
+      "learning_rate": 0.0001,
+      "loss": 1.7908,
+      "step": 2252
+    },
+    {
+      "epoch": 0.25880190684050314,
+      "grad_norm": 0.3482286036014557,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 2253
+    },
+    {
+      "epoch": 0.25891677675033026,
+      "grad_norm": 0.3742547035217285,
+      "learning_rate": 0.0001,
+      "loss": 1.6636,
+      "step": 2254
+    },
+    {
+      "epoch": 0.2590316466601574,
+      "grad_norm": 0.386465847492218,
+      "learning_rate": 0.0001,
+      "loss": 1.6659,
+      "step": 2255
+    },
+    {
+      "epoch": 0.2591465165699845,
+      "grad_norm": 0.365347295999527,
+      "learning_rate": 0.0001,
+      "loss": 1.7536,
+      "step": 2256
+    },
+    {
+      "epoch": 0.2592613864798116,
+      "grad_norm": 0.3640124797821045,
+      "learning_rate": 0.0001,
+      "loss": 1.7142,
+      "step": 2257
+    },
+    {
+      "epoch": 0.25937625638963874,
+      "grad_norm": 0.36901116371154785,
+      "learning_rate": 0.0001,
+      "loss": 1.477,
+      "step": 2258
+    },
+    {
+      "epoch": 0.25949112629946586,
+      "grad_norm": 0.36894455552101135,
+      "learning_rate": 0.0001,
+      "loss": 1.6933,
+      "step": 2259
+    },
+    {
+      "epoch": 0.259605996209293,
+      "grad_norm": 0.37590548396110535,
+      "learning_rate": 0.0001,
+      "loss": 1.7388,
+      "step": 2260
+    },
+    {
+      "epoch": 0.2597208661191201,
+      "grad_norm": 0.4689255654811859,
+      "learning_rate": 0.0001,
+      "loss": 2.1016,
+      "step": 2261
+    },
+    {
+      "epoch": 0.2598357360289472,
+      "grad_norm": 0.3811010420322418,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 2262
+    },
+    {
+      "epoch": 0.25995060593877434,
+      "grad_norm": 0.3560413420200348,
+      "learning_rate": 0.0001,
+      "loss": 1.6288,
+      "step": 2263
+    },
+    {
+      "epoch": 0.26006547584860146,
+      "grad_norm": 0.4040233790874481,
+      "learning_rate": 0.0001,
+      "loss": 1.8888,
+      "step": 2264
+    },
+    {
+      "epoch": 0.2601803457584286,
+      "grad_norm": 0.3741479516029358,
+      "learning_rate": 0.0001,
+      "loss": 1.6749,
+      "step": 2265
+    },
+    {
+      "epoch": 0.2602952156682557,
+      "grad_norm": 0.3752221465110779,
+      "learning_rate": 0.0001,
+      "loss": 1.8194,
+      "step": 2266
+    },
+    {
+      "epoch": 0.2604100855780828,
+      "grad_norm": 0.3529088497161865,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 2267
+    },
+    {
+      "epoch": 0.26052495548790994,
+      "grad_norm": 0.37346789240837097,
+      "learning_rate": 0.0001,
+      "loss": 1.694,
+      "step": 2268
+    },
+    {
+      "epoch": 0.26063982539773706,
+      "grad_norm": 0.38784459233283997,
+      "learning_rate": 0.0001,
+      "loss": 1.7968,
+      "step": 2269
+    },
+    {
+      "epoch": 0.2607546953075642,
+      "grad_norm": 0.3864971697330475,
+      "learning_rate": 0.0001,
+      "loss": 1.7894,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 0.3784656226634979,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 2271
+    },
+    {
+      "epoch": 0.2609844351272184,
+      "grad_norm": 0.3623911738395691,
+      "learning_rate": 0.0001,
+      "loss": 1.7436,
+      "step": 2272
+    },
+    {
+      "epoch": 0.26109930503704554,
+      "grad_norm": 0.3621329069137573,
+      "learning_rate": 0.0001,
+      "loss": 1.4809,
+      "step": 2273
+    },
+    {
+      "epoch": 0.26121417494687266,
+      "grad_norm": 0.40668338537216187,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 2274
+    },
+    {
+      "epoch": 0.2613290448566998,
+      "grad_norm": 0.3952760398387909,
+      "learning_rate": 0.0001,
+      "loss": 1.9019,
+      "step": 2275
+    },
+    {
+      "epoch": 0.2614439147665269,
+      "grad_norm": 0.39390838146209717,
+      "learning_rate": 0.0001,
+      "loss": 1.8203,
+      "step": 2276
+    },
+    {
+      "epoch": 0.261558784676354,
+      "grad_norm": 0.3634113073348999,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 2277
+    },
+    {
+      "epoch": 0.26167365458618114,
+      "grad_norm": 0.36678680777549744,
+      "learning_rate": 0.0001,
+      "loss": 1.7097,
+      "step": 2278
+    },
+    {
+      "epoch": 0.26178852449600826,
+      "grad_norm": 0.4072323143482208,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 2279
+    },
+    {
+      "epoch": 0.2619033944058354,
+      "grad_norm": 0.39810293912887573,
+      "learning_rate": 0.0001,
+      "loss": 1.8374,
+      "step": 2280
+    },
+    {
+      "epoch": 0.2620182643156625,
+      "grad_norm": 0.35911354422569275,
+      "learning_rate": 0.0001,
+      "loss": 1.7054,
+      "step": 2281
+    },
+    {
+      "epoch": 0.2621331342254896,
+      "grad_norm": 0.3710712492465973,
+      "learning_rate": 0.0001,
+      "loss": 1.7595,
+      "step": 2282
+    },
+    {
+      "epoch": 0.26224800413531674,
+      "grad_norm": 0.3721124231815338,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 2283
+    },
+    {
+      "epoch": 0.26236287404514386,
+      "grad_norm": 0.35447004437446594,
+      "learning_rate": 0.0001,
+      "loss": 1.6689,
+      "step": 2284
+    },
+    {
+      "epoch": 0.262477743954971,
+      "grad_norm": 0.37349700927734375,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 2285
+    },
+    {
+      "epoch": 0.2625926138647981,
+      "grad_norm": 0.3656753599643707,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 2286
+    },
+    {
+      "epoch": 0.2627074837746252,
+      "grad_norm": 0.35918521881103516,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 2287
+    },
+    {
+      "epoch": 0.26282235368445234,
+      "grad_norm": 0.36438190937042236,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 2288
+    },
+    {
+      "epoch": 0.26293722359427946,
+      "grad_norm": 0.365792453289032,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 2289
+    },
+    {
+      "epoch": 0.2630520935041066,
+      "grad_norm": 0.3362356424331665,
+      "learning_rate": 0.0001,
+      "loss": 1.3965,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2631669634139337,
+      "grad_norm": 0.3698473274707794,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 2291
+    },
+    {
+      "epoch": 0.2632818333237608,
+      "grad_norm": 0.447664737701416,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 2292
+    },
+    {
+      "epoch": 0.26339670323358794,
+      "grad_norm": 0.4149281978607178,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 2293
+    },
+    {
+      "epoch": 0.26351157314341506,
+      "grad_norm": 0.37246763706207275,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 2294
+    },
+    {
+      "epoch": 0.2636264430532422,
+      "grad_norm": 0.37586966156959534,
+      "learning_rate": 0.0001,
+      "loss": 1.8224,
+      "step": 2295
+    },
+    {
+      "epoch": 0.2637413129630693,
+      "grad_norm": 0.36681029200553894,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 2296
+    },
+    {
+      "epoch": 0.26385618287289647,
+      "grad_norm": 0.3731790781021118,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 2297
+    },
+    {
+      "epoch": 0.2639710527827236,
+      "grad_norm": 0.4063315987586975,
+      "learning_rate": 0.0001,
+      "loss": 1.8513,
+      "step": 2298
+    },
+    {
+      "epoch": 0.2640859226925507,
+      "grad_norm": 0.36280304193496704,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 2299
+    },
+    {
+      "epoch": 0.26420079260237783,
+      "grad_norm": 0.41365888714790344,
+      "learning_rate": 0.0001,
+      "loss": 1.7712,
+      "step": 2300
+    },
+    {
+      "epoch": 0.26431566251220495,
+      "grad_norm": 0.4301823079586029,
+      "learning_rate": 0.0001,
+      "loss": 2.0047,
+      "step": 2301
+    },
+    {
+      "epoch": 0.26443053242203207,
+      "grad_norm": 0.3758213222026825,
+      "learning_rate": 0.0001,
+      "loss": 1.6871,
+      "step": 2302
+    },
+    {
+      "epoch": 0.2645454023318592,
+      "grad_norm": 0.34419310092926025,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 2303
+    },
+    {
+      "epoch": 0.2646602722416863,
+      "grad_norm": 0.38945549726486206,
+      "learning_rate": 0.0001,
+      "loss": 1.7238,
+      "step": 2304
+    },
+    {
+      "epoch": 0.26477514215151343,
+      "grad_norm": 0.3707123398780823,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 2305
+    },
+    {
+      "epoch": 0.26489001206134055,
+      "grad_norm": 0.35695189237594604,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 2306
+    },
+    {
+      "epoch": 0.26500488197116767,
+      "grad_norm": 0.36474674940109253,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 2307
+    },
+    {
+      "epoch": 0.2651197518809948,
+      "grad_norm": 0.35917285084724426,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 2308
+    },
+    {
+      "epoch": 0.2652346217908219,
+      "grad_norm": 0.38368695974349976,
+      "learning_rate": 0.0001,
+      "loss": 1.7404,
+      "step": 2309
+    },
+    {
+      "epoch": 0.26534949170064903,
+      "grad_norm": 0.4073870778083801,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 2310
+    },
+    {
+      "epoch": 0.26546436161047615,
+      "grad_norm": 0.35575923323631287,
+      "learning_rate": 0.0001,
+      "loss": 1.6298,
+      "step": 2311
+    },
+    {
+      "epoch": 0.26557923152030327,
+      "grad_norm": 0.3653978109359741,
+      "learning_rate": 0.0001,
+      "loss": 1.7346,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2656941014301304,
+      "grad_norm": 0.35288769006729126,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 2313
+    },
+    {
+      "epoch": 0.2658089713399575,
+      "grad_norm": 0.36546453833580017,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 2314
+    },
+    {
+      "epoch": 0.26592384124978463,
+      "grad_norm": 0.3479709029197693,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 2315
+    },
+    {
+      "epoch": 0.26603871115961175,
+      "grad_norm": 0.3809703290462494,
+      "learning_rate": 0.0001,
+      "loss": 1.7787,
+      "step": 2316
+    },
+    {
+      "epoch": 0.26615358106943887,
+      "grad_norm": 0.3938155472278595,
+      "learning_rate": 0.0001,
+      "loss": 1.7848,
+      "step": 2317
+    },
+    {
+      "epoch": 0.266268450979266,
+      "grad_norm": 0.3979399502277374,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 2318
+    },
+    {
+      "epoch": 0.2663833208890931,
+      "grad_norm": 0.34814873337745667,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 2319
+    },
+    {
+      "epoch": 0.26649819079892023,
+      "grad_norm": 0.3895953595638275,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 2320
+    },
+    {
+      "epoch": 0.26661306070874735,
+      "grad_norm": 0.3680303692817688,
+      "learning_rate": 0.0001,
+      "loss": 1.6297,
+      "step": 2321
+    },
+    {
+      "epoch": 0.26672793061857447,
+      "grad_norm": 0.3696908950805664,
+      "learning_rate": 0.0001,
+      "loss": 1.7016,
+      "step": 2322
+    },
+    {
+      "epoch": 0.2668428005284016,
+      "grad_norm": 0.3959087133407593,
+      "learning_rate": 0.0001,
+      "loss": 1.6687,
+      "step": 2323
+    },
+    {
+      "epoch": 0.2669576704382287,
+      "grad_norm": 0.41594138741493225,
+      "learning_rate": 0.0001,
+      "loss": 1.7522,
+      "step": 2324
+    },
+    {
+      "epoch": 0.26707254034805583,
+      "grad_norm": 0.3591321110725403,
+      "learning_rate": 0.0001,
+      "loss": 1.7126,
+      "step": 2325
+    },
+    {
+      "epoch": 0.26718741025788295,
+      "grad_norm": 0.3633732795715332,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 2326
+    },
+    {
+      "epoch": 0.26730228016771007,
+      "grad_norm": 0.3702927529811859,
+      "learning_rate": 0.0001,
+      "loss": 1.7004,
+      "step": 2327
+    },
+    {
+      "epoch": 0.2674171500775372,
+      "grad_norm": 0.38838890194892883,
+      "learning_rate": 0.0001,
+      "loss": 1.7151,
+      "step": 2328
+    },
+    {
+      "epoch": 0.2675320199873643,
+      "grad_norm": 0.3630795180797577,
+      "learning_rate": 0.0001,
+      "loss": 1.7975,
+      "step": 2329
+    },
+    {
+      "epoch": 0.26764688989719143,
+      "grad_norm": 0.3541039228439331,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 2330
+    },
+    {
+      "epoch": 0.26776175980701855,
+      "grad_norm": 0.35986846685409546,
+      "learning_rate": 0.0001,
+      "loss": 1.8925,
+      "step": 2331
+    },
+    {
+      "epoch": 0.26787662971684567,
+      "grad_norm": 0.3914209008216858,
+      "learning_rate": 0.0001,
+      "loss": 1.9096,
+      "step": 2332
+    },
+    {
+      "epoch": 0.2679914996266728,
+      "grad_norm": 0.3626435399055481,
+      "learning_rate": 0.0001,
+      "loss": 1.8039,
+      "step": 2333
+    },
+    {
+      "epoch": 0.2681063695364999,
+      "grad_norm": 0.33067116141319275,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 2334
+    },
+    {
+      "epoch": 0.26822123944632703,
+      "grad_norm": 0.38686808943748474,
+      "learning_rate": 0.0001,
+      "loss": 1.5786,
+      "step": 2335
+    },
+    {
+      "epoch": 0.26833610935615415,
+      "grad_norm": 0.3926886022090912,
+      "learning_rate": 0.0001,
+      "loss": 1.8575,
+      "step": 2336
+    },
+    {
+      "epoch": 0.26845097926598127,
+      "grad_norm": 0.3721444606781006,
+      "learning_rate": 0.0001,
+      "loss": 1.6829,
+      "step": 2337
+    },
+    {
+      "epoch": 0.2685658491758084,
+      "grad_norm": 0.39007246494293213,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2686807190856355,
+      "grad_norm": 0.3614482283592224,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 2339
+    },
+    {
+      "epoch": 0.26879558899546263,
+      "grad_norm": 0.39546316862106323,
+      "learning_rate": 0.0001,
+      "loss": 1.4712,
+      "step": 2340
+    },
+    {
+      "epoch": 0.26891045890528975,
+      "grad_norm": 0.36849477887153625,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 2341
+    },
+    {
+      "epoch": 0.26902532881511687,
+      "grad_norm": 0.3788404166698456,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 2342
+    },
+    {
+      "epoch": 0.269140198724944,
+      "grad_norm": 0.35742640495300293,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 2343
+    },
+    {
+      "epoch": 0.2692550686347711,
+      "grad_norm": 0.37428852915763855,
+      "learning_rate": 0.0001,
+      "loss": 1.7477,
+      "step": 2344
+    },
+    {
+      "epoch": 0.26936993854459823,
+      "grad_norm": 0.42374229431152344,
+      "learning_rate": 0.0001,
+      "loss": 1.8088,
+      "step": 2345
+    },
+    {
+      "epoch": 0.26948480845442535,
+      "grad_norm": 0.3708469867706299,
+      "learning_rate": 0.0001,
+      "loss": 1.8116,
+      "step": 2346
+    },
+    {
+      "epoch": 0.26959967836425247,
+      "grad_norm": 0.37047696113586426,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 2347
+    },
+    {
+      "epoch": 0.2697145482740796,
+      "grad_norm": 0.4440860450267792,
+      "learning_rate": 0.0001,
+      "loss": 1.7974,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2698294181839067,
+      "grad_norm": 0.38337430357933044,
+      "learning_rate": 0.0001,
+      "loss": 1.7039,
+      "step": 2349
+    },
+    {
+      "epoch": 0.26994428809373383,
+      "grad_norm": 0.3824228048324585,
+      "learning_rate": 0.0001,
+      "loss": 1.7418,
+      "step": 2350
+    },
+    {
+      "epoch": 0.27005915800356095,
+      "grad_norm": 0.38111045956611633,
+      "learning_rate": 0.0001,
+      "loss": 1.7184,
+      "step": 2351
+    },
+    {
+      "epoch": 0.27017402791338807,
+      "grad_norm": 0.3806154727935791,
+      "learning_rate": 0.0001,
+      "loss": 1.8688,
+      "step": 2352
+    },
+    {
+      "epoch": 0.2702888978232152,
+      "grad_norm": 0.3737587332725525,
+      "learning_rate": 0.0001,
+      "loss": 1.7697,
+      "step": 2353
+    },
+    {
+      "epoch": 0.2704037677330423,
+      "grad_norm": 0.37875616550445557,
+      "learning_rate": 0.0001,
+      "loss": 1.7901,
+      "step": 2354
+    },
+    {
+      "epoch": 0.27051863764286943,
+      "grad_norm": 0.3805428445339203,
+      "learning_rate": 0.0001,
+      "loss": 1.8359,
+      "step": 2355
+    },
+    {
+      "epoch": 0.27063350755269655,
+      "grad_norm": 0.3564281761646271,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 2356
+    },
+    {
+      "epoch": 0.27074837746252367,
+      "grad_norm": 0.36830076575279236,
+      "learning_rate": 0.0001,
+      "loss": 1.7583,
+      "step": 2357
+    },
+    {
+      "epoch": 0.2708632473723508,
+      "grad_norm": 0.38345348834991455,
+      "learning_rate": 0.0001,
+      "loss": 1.8286,
+      "step": 2358
+    },
+    {
+      "epoch": 0.2709781172821779,
+      "grad_norm": 0.3683982789516449,
+      "learning_rate": 0.0001,
+      "loss": 1.6467,
+      "step": 2359
+    },
+    {
+      "epoch": 0.27109298719200503,
+      "grad_norm": 0.36231061816215515,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 2360
+    },
+    {
+      "epoch": 0.27120785710183215,
+      "grad_norm": 0.40081697702407837,
+      "learning_rate": 0.0001,
+      "loss": 1.7706,
+      "step": 2361
+    },
+    {
+      "epoch": 0.27132272701165927,
+      "grad_norm": 0.3795923590660095,
+      "learning_rate": 0.0001,
+      "loss": 1.8317,
+      "step": 2362
+    },
+    {
+      "epoch": 0.2714375969214864,
+      "grad_norm": 0.41488227248191833,
+      "learning_rate": 0.0001,
+      "loss": 1.8353,
+      "step": 2363
+    },
+    {
+      "epoch": 0.2715524668313135,
+      "grad_norm": 0.36446574330329895,
+      "learning_rate": 0.0001,
+      "loss": 1.7574,
+      "step": 2364
+    },
+    {
+      "epoch": 0.2716673367411407,
+      "grad_norm": 0.4068087935447693,
+      "learning_rate": 0.0001,
+      "loss": 1.6117,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2717822066509678,
+      "grad_norm": 0.3724612295627594,
+      "learning_rate": 0.0001,
+      "loss": 1.8087,
+      "step": 2366
+    },
+    {
+      "epoch": 0.2718970765607949,
+      "grad_norm": 0.3523292541503906,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 2367
+    },
+    {
+      "epoch": 0.27201194647062205,
+      "grad_norm": 0.3559991717338562,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 2368
+    },
+    {
+      "epoch": 0.27212681638044917,
+      "grad_norm": 0.370811402797699,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 2369
+    },
+    {
+      "epoch": 0.2722416862902763,
+      "grad_norm": 0.350815087556839,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2723565562001034,
+      "grad_norm": 0.3826122283935547,
+      "learning_rate": 0.0001,
+      "loss": 1.7605,
+      "step": 2371
+    },
+    {
+      "epoch": 0.2724714261099305,
+      "grad_norm": 0.3690153956413269,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 2372
+    },
+    {
+      "epoch": 0.27258629601975765,
+      "grad_norm": 0.3891347050666809,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 2373
+    },
+    {
+      "epoch": 0.27270116592958477,
+      "grad_norm": 0.3778075873851776,
+      "learning_rate": 0.0001,
+      "loss": 1.7759,
+      "step": 2374
+    },
+    {
+      "epoch": 0.2728160358394119,
+      "grad_norm": 0.35634365677833557,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 2375
+    },
+    {
+      "epoch": 0.272930905749239,
+      "grad_norm": 0.3891456127166748,
+      "learning_rate": 0.0001,
+      "loss": 1.7749,
+      "step": 2376
+    },
+    {
+      "epoch": 0.2730457756590661,
+      "grad_norm": 0.3792819380760193,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 2377
+    },
+    {
+      "epoch": 0.27316064556889325,
+      "grad_norm": 0.37712422013282776,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 2378
+    },
+    {
+      "epoch": 0.27327551547872037,
+      "grad_norm": 0.38600102066993713,
+      "learning_rate": 0.0001,
+      "loss": 1.8743,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2733903853885475,
+      "grad_norm": 0.3560759723186493,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2735052552983746,
+      "grad_norm": 0.3798205852508545,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 2381
+    },
+    {
+      "epoch": 0.2736201252082017,
+      "grad_norm": 0.40677618980407715,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 2382
+    },
+    {
+      "epoch": 0.27373499511802885,
+      "grad_norm": 0.3787870705127716,
+      "learning_rate": 0.0001,
+      "loss": 1.7416,
+      "step": 2383
+    },
+    {
+      "epoch": 0.27384986502785597,
+      "grad_norm": 0.39786311984062195,
+      "learning_rate": 0.0001,
+      "loss": 1.8658,
+      "step": 2384
+    },
+    {
+      "epoch": 0.2739647349376831,
+      "grad_norm": 0.39431723952293396,
+      "learning_rate": 0.0001,
+      "loss": 1.8427,
+      "step": 2385
+    },
+    {
+      "epoch": 0.2740796048475102,
+      "grad_norm": 0.40065842866897583,
+      "learning_rate": 0.0001,
+      "loss": 1.8135,
+      "step": 2386
+    },
+    {
+      "epoch": 0.2741944747573373,
+      "grad_norm": 0.3779159188270569,
+      "learning_rate": 0.0001,
+      "loss": 1.6793,
+      "step": 2387
+    },
+    {
+      "epoch": 0.27430934466716445,
+      "grad_norm": 0.35582754015922546,
+      "learning_rate": 0.0001,
+      "loss": 1.6984,
+      "step": 2388
+    },
+    {
+      "epoch": 0.27442421457699157,
+      "grad_norm": 0.3920300304889679,
+      "learning_rate": 0.0001,
+      "loss": 1.7694,
+      "step": 2389
+    },
+    {
+      "epoch": 0.2745390844868187,
+      "grad_norm": 0.394414484500885,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2746539543966458,
+      "grad_norm": 0.3930966258049011,
+      "learning_rate": 0.0001,
+      "loss": 1.5893,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2747688243064729,
+      "grad_norm": 0.3484657406806946,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 2392
+    },
+    {
+      "epoch": 0.27488369421630005,
+      "grad_norm": 0.41878804564476013,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 2393
+    },
+    {
+      "epoch": 0.27499856412612717,
+      "grad_norm": 0.36988186836242676,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 2394
+    },
+    {
+      "epoch": 0.2751134340359543,
+      "grad_norm": 0.3710579574108124,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 2395
+    },
+    {
+      "epoch": 0.2752283039457814,
+      "grad_norm": 0.3970875144004822,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2753431738556085,
+      "grad_norm": 0.3790924847126007,
+      "learning_rate": 0.0001,
+      "loss": 1.7784,
+      "step": 2397
+    },
+    {
+      "epoch": 0.27545804376543565,
+      "grad_norm": 0.35328909754753113,
+      "learning_rate": 0.0001,
+      "loss": 1.8243,
+      "step": 2398
+    },
+    {
+      "epoch": 0.27557291367526277,
+      "grad_norm": 0.360989511013031,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2756877835850899,
+      "grad_norm": 0.4168824553489685,
+      "learning_rate": 0.0001,
+      "loss": 1.7192,
+      "step": 2400
+    },
+    {
+      "epoch": 0.275802653494917,
+      "grad_norm": 0.3885016143321991,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 2401
+    },
+    {
+      "epoch": 0.2759175234047441,
+      "grad_norm": 0.3799929916858673,
+      "learning_rate": 0.0001,
+      "loss": 1.7446,
+      "step": 2402
+    },
+    {
+      "epoch": 0.27603239331457124,
+      "grad_norm": 0.42033493518829346,
+      "learning_rate": 0.0001,
+      "loss": 1.8286,
+      "step": 2403
+    },
+    {
+      "epoch": 0.27614726322439836,
+      "grad_norm": 0.39111387729644775,
+      "learning_rate": 0.0001,
+      "loss": 1.7342,
+      "step": 2404
+    },
+    {
+      "epoch": 0.2762621331342255,
+      "grad_norm": 0.34956350922584534,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 2405
+    },
+    {
+      "epoch": 0.2763770030440526,
+      "grad_norm": 0.3685983419418335,
+      "learning_rate": 0.0001,
+      "loss": 1.8086,
+      "step": 2406
+    },
+    {
+      "epoch": 0.2764918729538797,
+      "grad_norm": 0.35792601108551025,
+      "learning_rate": 0.0001,
+      "loss": 1.7419,
+      "step": 2407
+    },
+    {
+      "epoch": 0.27660674286370684,
+      "grad_norm": 0.39823225140571594,
+      "learning_rate": 0.0001,
+      "loss": 1.7881,
+      "step": 2408
+    },
+    {
+      "epoch": 0.27672161277353396,
+      "grad_norm": 0.39737337827682495,
+      "learning_rate": 0.0001,
+      "loss": 1.7751,
+      "step": 2409
+    },
+    {
+      "epoch": 0.2768364826833611,
+      "grad_norm": 0.35919415950775146,
+      "learning_rate": 0.0001,
+      "loss": 1.6735,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2769513525931882,
+      "grad_norm": 0.33741164207458496,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2770662225030153,
+      "grad_norm": 0.359222948551178,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 2412
+    },
+    {
+      "epoch": 0.27718109241284244,
+      "grad_norm": 0.3950967490673065,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 2413
+    },
+    {
+      "epoch": 0.27729596232266956,
+      "grad_norm": 0.3793290853500366,
+      "learning_rate": 0.0001,
+      "loss": 1.732,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2774108322324967,
+      "grad_norm": 0.38844072818756104,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2775257021423238,
+      "grad_norm": 0.3837685286998749,
+      "learning_rate": 0.0001,
+      "loss": 1.7407,
+      "step": 2416
+    },
+    {
+      "epoch": 0.2776405720521509,
+      "grad_norm": 0.4028517007827759,
+      "learning_rate": 0.0001,
+      "loss": 1.8119,
+      "step": 2417
+    },
+    {
+      "epoch": 0.27775544196197804,
+      "grad_norm": 0.38334351778030396,
+      "learning_rate": 0.0001,
+      "loss": 1.7055,
+      "step": 2418
+    },
+    {
+      "epoch": 0.27787031187180516,
+      "grad_norm": 0.3632005453109741,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 2419
+    },
+    {
+      "epoch": 0.2779851817816323,
+      "grad_norm": 0.36728063225746155,
+      "learning_rate": 0.0001,
+      "loss": 1.7756,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2781000516914594,
+      "grad_norm": 0.41414251923561096,
+      "learning_rate": 0.0001,
+      "loss": 1.9967,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2782149216012865,
+      "grad_norm": 0.3955191969871521,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 2422
+    },
+    {
+      "epoch": 0.27832979151111364,
+      "grad_norm": 0.37816205620765686,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 2423
+    },
+    {
+      "epoch": 0.27844466142094076,
+      "grad_norm": 0.4215168058872223,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2785595313307679,
+      "grad_norm": 0.3857564926147461,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 2425
+    },
+    {
+      "epoch": 0.278674401240595,
+      "grad_norm": 0.3551070988178253,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2787892711504221,
+      "grad_norm": 0.3506256937980652,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 2427
+    },
+    {
+      "epoch": 0.27890414106024924,
+      "grad_norm": 0.3387966752052307,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 2428
+    },
+    {
+      "epoch": 0.27901901097007636,
+      "grad_norm": 0.3767859637737274,
+      "learning_rate": 0.0001,
+      "loss": 1.7688,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2791338808799035,
+      "grad_norm": 0.38322755694389343,
+      "learning_rate": 0.0001,
+      "loss": 1.867,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2792487507897306,
+      "grad_norm": 0.3900243043899536,
+      "learning_rate": 0.0001,
+      "loss": 1.9258,
+      "step": 2431
+    },
+    {
+      "epoch": 0.2793636206995577,
+      "grad_norm": 0.3808649182319641,
+      "learning_rate": 0.0001,
+      "loss": 1.53,
+      "step": 2432
+    },
+    {
+      "epoch": 0.27947849060938484,
+      "grad_norm": 0.36853742599487305,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 2433
+    },
+    {
+      "epoch": 0.279593360519212,
+      "grad_norm": 0.36365145444869995,
+      "learning_rate": 0.0001,
+      "loss": 1.7215,
+      "step": 2434
+    },
+    {
+      "epoch": 0.27970823042903914,
+      "grad_norm": 0.361780047416687,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 2435
+    },
+    {
+      "epoch": 0.27982310033886626,
+      "grad_norm": 0.39141660928726196,
+      "learning_rate": 0.0001,
+      "loss": 1.8127,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2799379702486934,
+      "grad_norm": 0.3739947974681854,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2800528401585205,
+      "grad_norm": 0.3897460103034973,
+      "learning_rate": 0.0001,
+      "loss": 1.869,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2801677100683476,
+      "grad_norm": 0.3739219903945923,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 2439
+    },
+    {
+      "epoch": 0.28028257997817474,
+      "grad_norm": 0.3786073923110962,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 2440
+    },
+    {
+      "epoch": 0.28039744988800186,
+      "grad_norm": 0.37929031252861023,
+      "learning_rate": 0.0001,
+      "loss": 1.7422,
+      "step": 2441
+    },
+    {
+      "epoch": 0.280512319797829,
+      "grad_norm": 0.35074925422668457,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 2442
+    },
+    {
+      "epoch": 0.2806271897076561,
+      "grad_norm": 0.3912547528743744,
+      "learning_rate": 0.0001,
+      "loss": 1.8629,
+      "step": 2443
+    },
+    {
+      "epoch": 0.2807420596174832,
+      "grad_norm": 0.36757802963256836,
+      "learning_rate": 0.0001,
+      "loss": 1.557,
+      "step": 2444
+    },
+    {
+      "epoch": 0.28085692952731034,
+      "grad_norm": 0.3755451738834381,
+      "learning_rate": 0.0001,
+      "loss": 1.8249,
+      "step": 2445
+    },
+    {
+      "epoch": 0.28097179943713746,
+      "grad_norm": 0.3435642719268799,
+      "learning_rate": 0.0001,
+      "loss": 1.4253,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2810866693469646,
+      "grad_norm": 0.3567301332950592,
+      "learning_rate": 0.0001,
+      "loss": 1.4934,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2812015392567917,
+      "grad_norm": 0.3861426115036011,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2813164091666188,
+      "grad_norm": 0.38505983352661133,
+      "learning_rate": 0.0001,
+      "loss": 1.7413,
+      "step": 2449
+    },
+    {
+      "epoch": 0.28143127907644594,
+      "grad_norm": 0.3739560544490814,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 2450
+    },
+    {
+      "epoch": 0.28154614898627306,
+      "grad_norm": 0.359266996383667,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2816610188961002,
+      "grad_norm": 0.34818223118782043,
+      "learning_rate": 0.0001,
+      "loss": 1.4532,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2817758888059273,
+      "grad_norm": 0.40676432847976685,
+      "learning_rate": 0.0001,
+      "loss": 1.8238,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2818907587157544,
+      "grad_norm": 0.37692880630493164,
+      "learning_rate": 0.0001,
+      "loss": 1.7096,
+      "step": 2454
+    },
+    {
+      "epoch": 0.28200562862558154,
+      "grad_norm": 0.3589748442173004,
+      "learning_rate": 0.0001,
+      "loss": 1.5383,
+      "step": 2455
+    },
+    {
+      "epoch": 0.28212049853540866,
+      "grad_norm": 0.3927457928657532,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2822353684452358,
+      "grad_norm": 0.37036100029945374,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2823502383550629,
+      "grad_norm": 0.3649539649486542,
+      "learning_rate": 0.0001,
+      "loss": 1.4174,
+      "step": 2458
+    },
+    {
+      "epoch": 0.28246510826489,
+      "grad_norm": 0.3977210521697998,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 2459
+    },
+    {
+      "epoch": 0.28257997817471714,
+      "grad_norm": 0.37352606654167175,
+      "learning_rate": 0.0001,
+      "loss": 1.7877,
+      "step": 2460
+    },
+    {
+      "epoch": 0.28269484808454426,
+      "grad_norm": 0.37431058287620544,
+      "learning_rate": 0.0001,
+      "loss": 1.7231,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2828097179943714,
+      "grad_norm": 0.3667674660682678,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2829245879041985,
+      "grad_norm": 0.4098829925060272,
+      "learning_rate": 0.0001,
+      "loss": 1.7942,
+      "step": 2463
+    },
+    {
+      "epoch": 0.2830394578140256,
+      "grad_norm": 0.35849472880363464,
+      "learning_rate": 0.0001,
+      "loss": 1.7135,
+      "step": 2464
+    },
+    {
+      "epoch": 0.28315432772385274,
+      "grad_norm": 0.40055009722709656,
+      "learning_rate": 0.0001,
+      "loss": 1.7541,
+      "step": 2465
+    },
+    {
+      "epoch": 0.28326919763367986,
+      "grad_norm": 0.3810610771179199,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 2466
+    },
+    {
+      "epoch": 0.283384067543507,
+      "grad_norm": 0.3662252724170685,
+      "learning_rate": 0.0001,
+      "loss": 1.5483,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2834989374533341,
+      "grad_norm": 0.36314326524734497,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2836138073631612,
+      "grad_norm": 0.3729235529899597,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 2469
+    },
+    {
+      "epoch": 0.28372867727298834,
+      "grad_norm": 0.40618225932121277,
+      "learning_rate": 0.0001,
+      "loss": 1.7267,
+      "step": 2470
+    },
+    {
+      "epoch": 0.28384354718281546,
+      "grad_norm": 0.3927193284034729,
+      "learning_rate": 0.0001,
+      "loss": 1.8012,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2839584170926426,
+      "grad_norm": 0.3536290228366852,
+      "learning_rate": 0.0001,
+      "loss": 1.7102,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2840732870024697,
+      "grad_norm": 0.36351481080055237,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2841881569122968,
+      "grad_norm": 0.3685545027256012,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 2474
+    },
+    {
+      "epoch": 0.28430302682212394,
+      "grad_norm": 0.37130284309387207,
+      "learning_rate": 0.0001,
+      "loss": 1.7307,
+      "step": 2475
+    },
+    {
+      "epoch": 0.28441789673195106,
+      "grad_norm": 0.3823324143886566,
+      "learning_rate": 0.0001,
+      "loss": 1.7253,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2845327666417782,
+      "grad_norm": 0.3819986879825592,
+      "learning_rate": 0.0001,
+      "loss": 1.5769,
+      "step": 2477
+    },
+    {
+      "epoch": 0.2846476365516053,
+      "grad_norm": 0.4364182651042938,
+      "learning_rate": 0.0001,
+      "loss": 1.7868,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2847625064614324,
+      "grad_norm": 0.3844752907752991,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 2479
+    },
+    {
+      "epoch": 0.28487737637125954,
+      "grad_norm": 0.38411590456962585,
+      "learning_rate": 0.0001,
+      "loss": 1.5398,
+      "step": 2480
+    },
+    {
+      "epoch": 0.28499224628108666,
+      "grad_norm": 0.40151742100715637,
+      "learning_rate": 0.0001,
+      "loss": 1.8084,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2851071161909138,
+      "grad_norm": 0.41466403007507324,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2852219861007409,
+      "grad_norm": 0.37958747148513794,
+      "learning_rate": 0.0001,
+      "loss": 1.516,
+      "step": 2483
+    },
+    {
+      "epoch": 0.285336856010568,
+      "grad_norm": 0.3950232267379761,
+      "learning_rate": 0.0001,
+      "loss": 1.955,
+      "step": 2484
+    },
+    {
+      "epoch": 0.28545172592039514,
+      "grad_norm": 0.3635809123516083,
+      "learning_rate": 0.0001,
+      "loss": 1.7445,
+      "step": 2485
+    },
+    {
+      "epoch": 0.28556659583022226,
+      "grad_norm": 0.38769015669822693,
+      "learning_rate": 0.0001,
+      "loss": 1.6845,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2856814657400494,
+      "grad_norm": 0.37288132309913635,
+      "learning_rate": 0.0001,
+      "loss": 1.5452,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2857963356498765,
+      "grad_norm": 0.4224679172039032,
+      "learning_rate": 0.0001,
+      "loss": 1.8745,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2859112055597036,
+      "grad_norm": 0.4068784713745117,
+      "learning_rate": 0.0001,
+      "loss": 1.9496,
+      "step": 2489
+    },
+    {
+      "epoch": 0.28602607546953074,
+      "grad_norm": 0.39766570925712585,
+      "learning_rate": 0.0001,
+      "loss": 1.7433,
+      "step": 2490
+    },
+    {
+      "epoch": 0.28614094537935786,
+      "grad_norm": 0.4218822717666626,
+      "learning_rate": 0.0001,
+      "loss": 1.7239,
+      "step": 2491
+    },
+    {
+      "epoch": 0.286255815289185,
+      "grad_norm": 0.3917092978954315,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2863706851990121,
+      "grad_norm": 0.44831544160842896,
+      "learning_rate": 0.0001,
+      "loss": 1.7862,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2864855551088392,
+      "grad_norm": 0.39573273062705994,
+      "learning_rate": 0.0001,
+      "loss": 1.8192,
+      "step": 2494
+    },
+    {
+      "epoch": 0.28660042501866634,
+      "grad_norm": 0.37129390239715576,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 2495
+    },
+    {
+      "epoch": 0.28671529492849346,
+      "grad_norm": 0.3685044050216675,
+      "learning_rate": 0.0001,
+      "loss": 1.7772,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2868301648383206,
+      "grad_norm": 0.41524022817611694,
+      "learning_rate": 0.0001,
+      "loss": 1.825,
+      "step": 2497
+    },
+    {
+      "epoch": 0.2869450347481477,
+      "grad_norm": 0.39948517084121704,
+      "learning_rate": 0.0001,
+      "loss": 1.6756,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2870599046579748,
+      "grad_norm": 0.3788050413131714,
+      "learning_rate": 0.0001,
+      "loss": 1.8223,
+      "step": 2499
+    },
+    {
+      "epoch": 0.28717477456780194,
+      "grad_norm": 0.3545449674129486,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 2500
+    },
+    {
+      "epoch": 0.28728964447762906,
+      "grad_norm": 0.37108996510505676,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 2501
+    },
+    {
+      "epoch": 0.28740451438745623,
+      "grad_norm": 0.37973251938819885,
+      "learning_rate": 0.0001,
+      "loss": 1.9102,
+      "step": 2502
+    },
+    {
+      "epoch": 0.28751938429728335,
+      "grad_norm": 0.3596639931201935,
+      "learning_rate": 0.0001,
+      "loss": 1.6966,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2876342542071105,
+      "grad_norm": 0.39931946992874146,
+      "learning_rate": 0.0001,
+      "loss": 1.664,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2877491241169376,
+      "grad_norm": 0.3801995515823364,
+      "learning_rate": 0.0001,
+      "loss": 1.6762,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2878639940267647,
+      "grad_norm": 0.3975699245929718,
+      "learning_rate": 0.0001,
+      "loss": 1.8263,
+      "step": 2506
+    },
+    {
+      "epoch": 0.28797886393659183,
+      "grad_norm": 0.41587021946907043,
+      "learning_rate": 0.0001,
+      "loss": 1.8763,
+      "step": 2507
+    },
+    {
+      "epoch": 0.28809373384641895,
+      "grad_norm": 0.3884534239768982,
+      "learning_rate": 0.0001,
+      "loss": 1.8323,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2882086037562461,
+      "grad_norm": 0.376265287399292,
+      "learning_rate": 0.0001,
+      "loss": 1.735,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2883234736660732,
+      "grad_norm": 0.349988728761673,
+      "learning_rate": 0.0001,
+      "loss": 1.5764,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2884383435759003,
+      "grad_norm": 0.3495781123638153,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 2511
+    },
+    {
+      "epoch": 0.28855321348572743,
+      "grad_norm": 0.40810123085975647,
+      "learning_rate": 0.0001,
+      "loss": 1.7881,
+      "step": 2512
+    },
+    {
+      "epoch": 0.28866808339555455,
+      "grad_norm": 0.37921836972236633,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2887829533053817,
+      "grad_norm": 0.3909505009651184,
+      "learning_rate": 0.0001,
+      "loss": 1.7327,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2888978232152088,
+      "grad_norm": 0.3837600648403168,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2890126931250359,
+      "grad_norm": 0.4128781855106354,
+      "learning_rate": 0.0001,
+      "loss": 1.8047,
+      "step": 2516
+    },
+    {
+      "epoch": 0.28912756303486303,
+      "grad_norm": 0.3738429546356201,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 2517
+    },
+    {
+      "epoch": 0.28924243294469015,
+      "grad_norm": 0.40288591384887695,
+      "learning_rate": 0.0001,
+      "loss": 1.8381,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2893573028545173,
+      "grad_norm": 0.37133219838142395,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2894721727643444,
+      "grad_norm": 0.3895220160484314,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2895870426741715,
+      "grad_norm": 0.42748066782951355,
+      "learning_rate": 0.0001,
+      "loss": 1.7745,
+      "step": 2521
+    },
+    {
+      "epoch": 0.28970191258399863,
+      "grad_norm": 0.3681427836418152,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 2522
+    },
+    {
+      "epoch": 0.28981678249382575,
+      "grad_norm": 0.3755578100681305,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2899316524036529,
+      "grad_norm": 0.3791626989841461,
+      "learning_rate": 0.0001,
+      "loss": 1.6502,
+      "step": 2524
+    },
+    {
+      "epoch": 0.29004652231348,
+      "grad_norm": 0.4189550578594208,
+      "learning_rate": 0.0001,
+      "loss": 1.9463,
+      "step": 2525
+    },
+    {
+      "epoch": 0.2901613922233071,
+      "grad_norm": 0.4706687033176422,
+      "learning_rate": 0.0001,
+      "loss": 1.8105,
+      "step": 2526
+    },
+    {
+      "epoch": 0.29027626213313423,
+      "grad_norm": 0.38283270597457886,
+      "learning_rate": 0.0001,
+      "loss": 1.7055,
+      "step": 2527
+    },
+    {
+      "epoch": 0.29039113204296135,
+      "grad_norm": 0.38680872321128845,
+      "learning_rate": 0.0001,
+      "loss": 1.7457,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2905060019527885,
+      "grad_norm": 0.34589383006095886,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2906208718626156,
+      "grad_norm": 0.33116042613983154,
+      "learning_rate": 0.0001,
+      "loss": 1.3595,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2907357417724427,
+      "grad_norm": 0.37206852436065674,
+      "learning_rate": 0.0001,
+      "loss": 1.8634,
+      "step": 2531
+    },
+    {
+      "epoch": 0.29085061168226983,
+      "grad_norm": 0.39483073353767395,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 2532
+    },
+    {
+      "epoch": 0.29096548159209695,
+      "grad_norm": 0.3861776888370514,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2910803515019241,
+      "grad_norm": 0.3621794283390045,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2911952214117512,
+      "grad_norm": 0.40973809361457825,
+      "learning_rate": 0.0001,
+      "loss": 1.7886,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2913100913215783,
+      "grad_norm": 0.36391711235046387,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 2536
+    },
+    {
+      "epoch": 0.29142496123140543,
+      "grad_norm": 0.38137203454971313,
+      "learning_rate": 0.0001,
+      "loss": 1.4499,
+      "step": 2537
+    },
+    {
+      "epoch": 0.29153983114123255,
+      "grad_norm": 0.3694712817668915,
+      "learning_rate": 0.0001,
+      "loss": 1.5863,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2916547010510597,
+      "grad_norm": 0.38381004333496094,
+      "learning_rate": 0.0001,
+      "loss": 1.8341,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2917695709608868,
+      "grad_norm": 0.39511749148368835,
+      "learning_rate": 0.0001,
+      "loss": 1.8543,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2918844408707139,
+      "grad_norm": 0.41761019825935364,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 2541
+    },
+    {
+      "epoch": 0.29199931078054103,
+      "grad_norm": 0.4057486355304718,
+      "learning_rate": 0.0001,
+      "loss": 1.7036,
+      "step": 2542
+    },
+    {
+      "epoch": 0.29211418069036815,
+      "grad_norm": 0.35340383648872375,
+      "learning_rate": 0.0001,
+      "loss": 1.4367,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2922290506001953,
+      "grad_norm": 0.3808727264404297,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2923439205100224,
+      "grad_norm": 0.38112786412239075,
+      "learning_rate": 0.0001,
+      "loss": 1.6942,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2924587904198495,
+      "grad_norm": 0.41311514377593994,
+      "learning_rate": 0.0001,
+      "loss": 1.8936,
+      "step": 2546
+    },
+    {
+      "epoch": 0.29257366032967663,
+      "grad_norm": 0.37865912914276123,
+      "learning_rate": 0.0001,
+      "loss": 1.7468,
+      "step": 2547
+    },
+    {
+      "epoch": 0.29268853023950375,
+      "grad_norm": 0.379802942276001,
+      "learning_rate": 0.0001,
+      "loss": 1.6966,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2928034001493309,
+      "grad_norm": 0.4265817105770111,
+      "learning_rate": 0.0001,
+      "loss": 1.6835,
+      "step": 2549
+    },
+    {
+      "epoch": 0.292918270059158,
+      "grad_norm": 0.38082998991012573,
+      "learning_rate": 0.0001,
+      "loss": 1.6804,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2930331399689851,
+      "grad_norm": 0.38684189319610596,
+      "learning_rate": 0.0001,
+      "loss": 1.9011,
+      "step": 2551
+    },
+    {
+      "epoch": 0.29314800987881223,
+      "grad_norm": 0.38656875491142273,
+      "learning_rate": 0.0001,
+      "loss": 1.7264,
+      "step": 2552
+    },
+    {
+      "epoch": 0.29326287978863935,
+      "grad_norm": 0.365608274936676,
+      "learning_rate": 0.0001,
+      "loss": 1.4035,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2933777496984665,
+      "grad_norm": 0.3454169034957886,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2934926196082936,
+      "grad_norm": 0.3723081052303314,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2936074895181207,
+      "grad_norm": 0.35232463479042053,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 2556
+    },
+    {
+      "epoch": 0.29372235942794783,
+      "grad_norm": 0.36890724301338196,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 2557
+    },
+    {
+      "epoch": 0.29383722933777495,
+      "grad_norm": 0.39997896552085876,
+      "learning_rate": 0.0001,
+      "loss": 1.9178,
+      "step": 2558
+    },
+    {
+      "epoch": 0.29395209924760207,
+      "grad_norm": 0.3915958106517792,
+      "learning_rate": 0.0001,
+      "loss": 1.6637,
+      "step": 2559
+    },
+    {
+      "epoch": 0.2940669691574292,
+      "grad_norm": 0.3597055971622467,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2941818390672563,
+      "grad_norm": 0.38754889369010925,
+      "learning_rate": 0.0001,
+      "loss": 1.68,
+      "step": 2561
+    },
+    {
+      "epoch": 0.29429670897708343,
+      "grad_norm": 0.37158629298210144,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 2562
+    },
+    {
+      "epoch": 0.29441157888691055,
+      "grad_norm": 0.393255352973938,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 2563
+    },
+    {
+      "epoch": 0.29452644879673767,
+      "grad_norm": 0.4069103002548218,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2946413187065648,
+      "grad_norm": 0.3824077844619751,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2947561886163919,
+      "grad_norm": 0.3937399387359619,
+      "learning_rate": 0.0001,
+      "loss": 1.7788,
+      "step": 2566
+    },
+    {
+      "epoch": 0.29487105852621903,
+      "grad_norm": 0.4155014753341675,
+      "learning_rate": 0.0001,
+      "loss": 1.8913,
+      "step": 2567
+    },
+    {
+      "epoch": 0.29498592843604615,
+      "grad_norm": 0.3631512224674225,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 2568
+    },
+    {
+      "epoch": 0.29510079834587327,
+      "grad_norm": 0.37146568298339844,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2952156682557004,
+      "grad_norm": 0.3793732523918152,
+      "learning_rate": 0.0001,
+      "loss": 1.77,
+      "step": 2570
+    },
+    {
+      "epoch": 0.29533053816552757,
+      "grad_norm": 0.39151185750961304,
+      "learning_rate": 0.0001,
+      "loss": 1.7726,
+      "step": 2571
+    },
+    {
+      "epoch": 0.2954454080753547,
+      "grad_norm": 0.38116058707237244,
+      "learning_rate": 0.0001,
+      "loss": 1.8428,
+      "step": 2572
+    },
+    {
+      "epoch": 0.2955602779851818,
+      "grad_norm": 0.3952963650226593,
+      "learning_rate": 0.0001,
+      "loss": 1.6293,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2956751478950089,
+      "grad_norm": 0.3993338942527771,
+      "learning_rate": 0.0001,
+      "loss": 1.9019,
+      "step": 2574
+    },
+    {
+      "epoch": 0.29579001780483605,
+      "grad_norm": 0.37218764424324036,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 2575
+    },
+    {
+      "epoch": 0.29590488771466317,
+      "grad_norm": 0.3602159023284912,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 2576
+    },
+    {
+      "epoch": 0.2960197576244903,
+      "grad_norm": 0.4018074572086334,
+      "learning_rate": 0.0001,
+      "loss": 1.7806,
+      "step": 2577
+    },
+    {
+      "epoch": 0.2961346275343174,
+      "grad_norm": 0.38347718119621277,
+      "learning_rate": 0.0001,
+      "loss": 1.6941,
+      "step": 2578
+    },
+    {
+      "epoch": 0.2962494974441445,
+      "grad_norm": 0.3954737186431885,
+      "learning_rate": 0.0001,
+      "loss": 1.8642,
+      "step": 2579
+    },
+    {
+      "epoch": 0.29636436735397165,
+      "grad_norm": 0.4196760952472687,
+      "learning_rate": 0.0001,
+      "loss": 1.6487,
+      "step": 2580
+    },
+    {
+      "epoch": 0.29647923726379877,
+      "grad_norm": 0.3532737195491791,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 2581
+    },
+    {
+      "epoch": 0.2965941071736259,
+      "grad_norm": 0.36641091108322144,
+      "learning_rate": 0.0001,
+      "loss": 1.4252,
+      "step": 2582
+    },
+    {
+      "epoch": 0.296708977083453,
+      "grad_norm": 0.37007638812065125,
+      "learning_rate": 0.0001,
+      "loss": 1.6963,
+      "step": 2583
+    },
+    {
+      "epoch": 0.2968238469932801,
+      "grad_norm": 0.3760312795639038,
+      "learning_rate": 0.0001,
+      "loss": 1.613,
+      "step": 2584
+    },
+    {
+      "epoch": 0.29693871690310725,
+      "grad_norm": 0.392160564661026,
+      "learning_rate": 0.0001,
+      "loss": 1.759,
+      "step": 2585
+    },
+    {
+      "epoch": 0.29705358681293437,
+      "grad_norm": 0.3935618996620178,
+      "learning_rate": 0.0001,
+      "loss": 1.787,
+      "step": 2586
+    },
+    {
+      "epoch": 0.2971684567227615,
+      "grad_norm": 0.3662956655025482,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 2587
+    },
+    {
+      "epoch": 0.2972833266325886,
+      "grad_norm": 0.4194296598434448,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 2588
+    },
+    {
+      "epoch": 0.2973981965424157,
+      "grad_norm": 0.37482190132141113,
+      "learning_rate": 0.0001,
+      "loss": 1.8223,
+      "step": 2589
+    },
+    {
+      "epoch": 0.29751306645224285,
+      "grad_norm": 0.38104763627052307,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 2590
+    },
+    {
+      "epoch": 0.29762793636206997,
+      "grad_norm": 0.37571361660957336,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 2591
+    },
+    {
+      "epoch": 0.2977428062718971,
+      "grad_norm": 0.4095185697078705,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 2592
+    },
+    {
+      "epoch": 0.2978576761817242,
+      "grad_norm": 0.4440751075744629,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 2593
+    },
+    {
+      "epoch": 0.2979725460915513,
+      "grad_norm": 0.3924051821231842,
+      "learning_rate": 0.0001,
+      "loss": 1.6179,
+      "step": 2594
+    },
+    {
+      "epoch": 0.29808741600137845,
+      "grad_norm": 0.3902948796749115,
+      "learning_rate": 0.0001,
+      "loss": 1.7332,
+      "step": 2595
+    },
+    {
+      "epoch": 0.29820228591120557,
+      "grad_norm": 0.3790493607521057,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 2596
+    },
+    {
+      "epoch": 0.2983171558210327,
+      "grad_norm": 0.38171616196632385,
+      "learning_rate": 0.0001,
+      "loss": 1.6307,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2984320257308598,
+      "grad_norm": 0.40960413217544556,
+      "learning_rate": 0.0001,
+      "loss": 1.9166,
+      "step": 2598
+    },
+    {
+      "epoch": 0.2985468956406869,
+      "grad_norm": 0.3882502019405365,
+      "learning_rate": 0.0001,
+      "loss": 1.7118,
+      "step": 2599
+    },
+    {
+      "epoch": 0.29866176555051405,
+      "grad_norm": 0.3626171350479126,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 2600
+    },
+    {
+      "epoch": 0.29877663546034117,
+      "grad_norm": 0.37243038415908813,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 2601
+    },
+    {
+      "epoch": 0.2988915053701683,
+      "grad_norm": 0.40221109986305237,
+      "learning_rate": 0.0001,
+      "loss": 1.6998,
+      "step": 2602
+    },
+    {
+      "epoch": 0.2990063752799954,
+      "grad_norm": 0.3771398663520813,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2991212451898225,
+      "grad_norm": 0.3790495991706848,
+      "learning_rate": 0.0001,
+      "loss": 1.6981,
+      "step": 2604
+    },
+    {
+      "epoch": 0.29923611509964965,
+      "grad_norm": 0.3962880074977875,
+      "learning_rate": 0.0001,
+      "loss": 1.7409,
+      "step": 2605
+    },
+    {
+      "epoch": 0.29935098500947677,
+      "grad_norm": 0.3856731057167053,
+      "learning_rate": 0.0001,
+      "loss": 1.8138,
+      "step": 2606
+    },
+    {
+      "epoch": 0.2994658549193039,
+      "grad_norm": 0.37686559557914734,
+      "learning_rate": 0.0001,
+      "loss": 1.7306,
+      "step": 2607
+    },
+    {
+      "epoch": 0.299580724829131,
+      "grad_norm": 0.4329466223716736,
+      "learning_rate": 0.0001,
+      "loss": 2.0436,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2996955947389581,
+      "grad_norm": 0.3704317808151245,
+      "learning_rate": 0.0001,
+      "loss": 1.6667,
+      "step": 2609
+    },
+    {
+      "epoch": 0.29981046464878525,
+      "grad_norm": 0.42832422256469727,
+      "learning_rate": 0.0001,
+      "loss": 1.7882,
+      "step": 2610
+    },
+    {
+      "epoch": 0.29992533455861237,
+      "grad_norm": 0.416471928358078,
+      "learning_rate": 0.0001,
+      "loss": 1.8709,
+      "step": 2611
+    },
+    {
+      "epoch": 0.3000402044684395,
+      "grad_norm": 0.35787421464920044,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 2612
+    },
+    {
+      "epoch": 0.3001550743782666,
+      "grad_norm": 0.37828290462493896,
+      "learning_rate": 0.0001,
+      "loss": 1.5254,
+      "step": 2613
+    },
+    {
+      "epoch": 0.3002699442880937,
+      "grad_norm": 0.3505316972732544,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 2614
+    },
+    {
+      "epoch": 0.30038481419792085,
+      "grad_norm": 0.3572443127632141,
+      "learning_rate": 0.0001,
+      "loss": 1.6333,
+      "step": 2615
+    },
+    {
+      "epoch": 0.30049968410774797,
+      "grad_norm": 0.3744872808456421,
+      "learning_rate": 0.0001,
+      "loss": 1.7276,
+      "step": 2616
+    },
+    {
+      "epoch": 0.3006145540175751,
+      "grad_norm": 0.38314029574394226,
+      "learning_rate": 0.0001,
+      "loss": 1.6795,
+      "step": 2617
+    },
+    {
+      "epoch": 0.3007294239274022,
+      "grad_norm": 0.38417288661003113,
+      "learning_rate": 0.0001,
+      "loss": 1.7798,
+      "step": 2618
+    },
+    {
+      "epoch": 0.3008442938372293,
+      "grad_norm": 0.39451834559440613,
+      "learning_rate": 0.0001,
+      "loss": 1.7956,
+      "step": 2619
+    },
+    {
+      "epoch": 0.30095916374705645,
+      "grad_norm": 0.3972351849079132,
+      "learning_rate": 0.0001,
+      "loss": 1.654,
+      "step": 2620
+    },
+    {
+      "epoch": 0.30107403365688357,
+      "grad_norm": 0.4091535806655884,
+      "learning_rate": 0.0001,
+      "loss": 1.9206,
+      "step": 2621
+    },
+    {
+      "epoch": 0.3011889035667107,
+      "grad_norm": 0.3716078996658325,
+      "learning_rate": 0.0001,
+      "loss": 1.432,
+      "step": 2622
+    },
+    {
+      "epoch": 0.3013037734765378,
+      "grad_norm": 0.34581923484802246,
+      "learning_rate": 0.0001,
+      "loss": 1.3475,
+      "step": 2623
+    },
+    {
+      "epoch": 0.3014186433863649,
+      "grad_norm": 0.3731369376182556,
+      "learning_rate": 0.0001,
+      "loss": 1.613,
+      "step": 2624
+    },
+    {
+      "epoch": 0.30153351329619205,
+      "grad_norm": 0.36073705554008484,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 2625
+    },
+    {
+      "epoch": 0.30164838320601917,
+      "grad_norm": 0.3785097002983093,
+      "learning_rate": 0.0001,
+      "loss": 1.7506,
+      "step": 2626
+    },
+    {
+      "epoch": 0.3017632531158463,
+      "grad_norm": 0.346171498298645,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 2627
+    },
+    {
+      "epoch": 0.3018781230256734,
+      "grad_norm": 0.353345662355423,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 2628
+    },
+    {
+      "epoch": 0.3019929929355005,
+      "grad_norm": 0.3921557366847992,
+      "learning_rate": 0.0001,
+      "loss": 1.8869,
+      "step": 2629
+    },
+    {
+      "epoch": 0.30210786284532765,
+      "grad_norm": 0.377298504114151,
+      "learning_rate": 0.0001,
+      "loss": 1.8358,
+      "step": 2630
+    },
+    {
+      "epoch": 0.30222273275515477,
+      "grad_norm": 0.3824778199195862,
+      "learning_rate": 0.0001,
+      "loss": 1.705,
+      "step": 2631
+    },
+    {
+      "epoch": 0.3023376026649819,
+      "grad_norm": 0.371186763048172,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 2632
+    },
+    {
+      "epoch": 0.302452472574809,
+      "grad_norm": 0.3520771563053131,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 2633
+    },
+    {
+      "epoch": 0.3025673424846361,
+      "grad_norm": 0.3969862461090088,
+      "learning_rate": 0.0001,
+      "loss": 1.7946,
+      "step": 2634
+    },
+    {
+      "epoch": 0.30268221239446325,
+      "grad_norm": 0.39209333062171936,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 2635
+    },
+    {
+      "epoch": 0.30279708230429037,
+      "grad_norm": 0.4001356065273285,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 2636
+    },
+    {
+      "epoch": 0.3029119522141175,
+      "grad_norm": 0.3858399987220764,
+      "learning_rate": 0.0001,
+      "loss": 1.7268,
+      "step": 2637
+    },
+    {
+      "epoch": 0.3030268221239446,
+      "grad_norm": 0.3735487163066864,
+      "learning_rate": 0.0001,
+      "loss": 1.5653,
+      "step": 2638
+    },
+    {
+      "epoch": 0.3031416920337717,
+      "grad_norm": 0.39118990302085876,
+      "learning_rate": 0.0001,
+      "loss": 1.6705,
+      "step": 2639
+    },
+    {
+      "epoch": 0.3032565619435989,
+      "grad_norm": 0.42595022916793823,
+      "learning_rate": 0.0001,
+      "loss": 1.7263,
+      "step": 2640
+    },
+    {
+      "epoch": 0.303371431853426,
+      "grad_norm": 0.41437122225761414,
+      "learning_rate": 0.0001,
+      "loss": 1.8724,
+      "step": 2641
+    },
+    {
+      "epoch": 0.30348630176325314,
+      "grad_norm": 0.3900952637195587,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 2642
+    },
+    {
+      "epoch": 0.30360117167308026,
+      "grad_norm": 0.38442274928092957,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 2643
+    },
+    {
+      "epoch": 0.3037160415829074,
+      "grad_norm": 0.39064112305641174,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 2644
+    },
+    {
+      "epoch": 0.3038309114927345,
+      "grad_norm": 0.3769366443157196,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3039457814025616,
+      "grad_norm": 0.39268919825553894,
+      "learning_rate": 0.0001,
+      "loss": 1.6863,
+      "step": 2646
+    },
+    {
+      "epoch": 0.30406065131238874,
+      "grad_norm": 0.3635852038860321,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 2647
+    },
+    {
+      "epoch": 0.30417552122221586,
+      "grad_norm": 0.36260101199150085,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 2648
+    },
+    {
+      "epoch": 0.304290391132043,
+      "grad_norm": 0.37813135981559753,
+      "learning_rate": 0.0001,
+      "loss": 1.7067,
+      "step": 2649
+    },
+    {
+      "epoch": 0.3044052610418701,
+      "grad_norm": 0.39040499925613403,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3045201309516972,
+      "grad_norm": 0.36264294385910034,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 2651
+    },
+    {
+      "epoch": 0.30463500086152434,
+      "grad_norm": 0.3571374714374542,
+      "learning_rate": 0.0001,
+      "loss": 1.5062,
+      "step": 2652
+    },
+    {
+      "epoch": 0.30474987077135146,
+      "grad_norm": 0.3698153793811798,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 2653
+    },
+    {
+      "epoch": 0.3048647406811786,
+      "grad_norm": 0.37908482551574707,
+      "learning_rate": 0.0001,
+      "loss": 1.7519,
+      "step": 2654
+    },
+    {
+      "epoch": 0.3049796105910057,
+      "grad_norm": 0.4145006835460663,
+      "learning_rate": 0.0001,
+      "loss": 2.0428,
+      "step": 2655
+    },
+    {
+      "epoch": 0.3050944805008328,
+      "grad_norm": 0.37752121686935425,
+      "learning_rate": 0.0001,
+      "loss": 1.7289,
+      "step": 2656
+    },
+    {
+      "epoch": 0.30520935041065994,
+      "grad_norm": 0.41512149572372437,
+      "learning_rate": 0.0001,
+      "loss": 2.0011,
+      "step": 2657
+    },
+    {
+      "epoch": 0.30532422032048706,
+      "grad_norm": 0.3637721836566925,
+      "learning_rate": 0.0001,
+      "loss": 1.6781,
+      "step": 2658
+    },
+    {
+      "epoch": 0.3054390902303142,
+      "grad_norm": 0.3608452081680298,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 2659
+    },
+    {
+      "epoch": 0.3055539601401413,
+      "grad_norm": 0.4033581018447876,
+      "learning_rate": 0.0001,
+      "loss": 1.8266,
+      "step": 2660
+    },
+    {
+      "epoch": 0.3056688300499684,
+      "grad_norm": 0.371520459651947,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 2661
+    },
+    {
+      "epoch": 0.30578369995979554,
+      "grad_norm": 0.36730000376701355,
+      "learning_rate": 0.0001,
+      "loss": 1.5817,
+      "step": 2662
+    },
+    {
+      "epoch": 0.30589856986962266,
+      "grad_norm": 0.376396119594574,
+      "learning_rate": 0.0001,
+      "loss": 1.7396,
+      "step": 2663
+    },
+    {
+      "epoch": 0.3060134397794498,
+      "grad_norm": 0.35428953170776367,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 2664
+    },
+    {
+      "epoch": 0.3061283096892769,
+      "grad_norm": 0.4117322266101837,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 2665
+    },
+    {
+      "epoch": 0.306243179599104,
+      "grad_norm": 0.39247551560401917,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 2666
+    },
+    {
+      "epoch": 0.30635804950893114,
+      "grad_norm": 0.39449337124824524,
+      "learning_rate": 0.0001,
+      "loss": 1.8517,
+      "step": 2667
+    },
+    {
+      "epoch": 0.30647291941875826,
+      "grad_norm": 0.3695959150791168,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 2668
+    },
+    {
+      "epoch": 0.3065877893285854,
+      "grad_norm": 0.3867664635181427,
+      "learning_rate": 0.0001,
+      "loss": 1.7339,
+      "step": 2669
+    },
+    {
+      "epoch": 0.3067026592384125,
+      "grad_norm": 0.4185912311077118,
+      "learning_rate": 0.0001,
+      "loss": 1.9492,
+      "step": 2670
+    },
+    {
+      "epoch": 0.3068175291482396,
+      "grad_norm": 0.365018755197525,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 2671
+    },
+    {
+      "epoch": 0.30693239905806674,
+      "grad_norm": 0.4161297380924225,
+      "learning_rate": 0.0001,
+      "loss": 1.7588,
+      "step": 2672
+    },
+    {
+      "epoch": 0.30704726896789386,
+      "grad_norm": 0.4104420840740204,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 2673
+    },
+    {
+      "epoch": 0.307162138877721,
+      "grad_norm": 0.393228143453598,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 2674
+    },
+    {
+      "epoch": 0.3072770087875481,
+      "grad_norm": 0.37595561146736145,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 2675
+    },
+    {
+      "epoch": 0.3073918786973752,
+      "grad_norm": 0.3587210178375244,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 2676
+    },
+    {
+      "epoch": 0.30750674860720234,
+      "grad_norm": 0.37775328755378723,
+      "learning_rate": 0.0001,
+      "loss": 1.6793,
+      "step": 2677
+    },
+    {
+      "epoch": 0.30762161851702946,
+      "grad_norm": 0.3870543837547302,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 2678
+    },
+    {
+      "epoch": 0.3077364884268566,
+      "grad_norm": 0.3819582760334015,
+      "learning_rate": 0.0001,
+      "loss": 1.6968,
+      "step": 2679
+    },
+    {
+      "epoch": 0.3078513583366837,
+      "grad_norm": 0.3661853075027466,
+      "learning_rate": 0.0001,
+      "loss": 1.6436,
+      "step": 2680
+    },
+    {
+      "epoch": 0.3079662282465108,
+      "grad_norm": 0.356086790561676,
+      "learning_rate": 0.0001,
+      "loss": 1.7138,
+      "step": 2681
+    },
+    {
+      "epoch": 0.30808109815633794,
+      "grad_norm": 0.34564852714538574,
+      "learning_rate": 0.0001,
+      "loss": 1.4758,
+      "step": 2682
+    },
+    {
+      "epoch": 0.30819596806616506,
+      "grad_norm": 0.37625306844711304,
+      "learning_rate": 0.0001,
+      "loss": 1.8159,
+      "step": 2683
+    },
+    {
+      "epoch": 0.3083108379759922,
+      "grad_norm": 0.4367530345916748,
+      "learning_rate": 0.0001,
+      "loss": 1.9309,
+      "step": 2684
+    },
+    {
+      "epoch": 0.3084257078858193,
+      "grad_norm": 0.41043978929519653,
+      "learning_rate": 0.0001,
+      "loss": 1.9249,
+      "step": 2685
+    },
+    {
+      "epoch": 0.3085405777956464,
+      "grad_norm": 0.36539000272750854,
+      "learning_rate": 0.0001,
+      "loss": 1.689,
+      "step": 2686
+    },
+    {
+      "epoch": 0.30865544770547354,
+      "grad_norm": 0.3621140718460083,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 2687
+    },
+    {
+      "epoch": 0.30877031761530066,
+      "grad_norm": 0.34792810678482056,
+      "learning_rate": 0.0001,
+      "loss": 1.5316,
+      "step": 2688
+    },
+    {
+      "epoch": 0.3088851875251278,
+      "grad_norm": 0.37253010272979736,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 2689
+    },
+    {
+      "epoch": 0.3090000574349549,
+      "grad_norm": 0.3895919919013977,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 2690
+    },
+    {
+      "epoch": 0.309114927344782,
+      "grad_norm": 0.3458951711654663,
+      "learning_rate": 0.0001,
+      "loss": 1.5419,
+      "step": 2691
+    },
+    {
+      "epoch": 0.30922979725460914,
+      "grad_norm": 0.34888923168182373,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 2692
+    },
+    {
+      "epoch": 0.30934466716443626,
+      "grad_norm": 0.3860279321670532,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 2693
+    },
+    {
+      "epoch": 0.3094595370742634,
+      "grad_norm": 0.3899478018283844,
+      "learning_rate": 0.0001,
+      "loss": 1.8091,
+      "step": 2694
+    },
+    {
+      "epoch": 0.3095744069840905,
+      "grad_norm": 0.3502478003501892,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 2695
+    },
+    {
+      "epoch": 0.3096892768939176,
+      "grad_norm": 0.3814723789691925,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 2696
+    },
+    {
+      "epoch": 0.30980414680374474,
+      "grad_norm": 0.4145774245262146,
+      "learning_rate": 0.0001,
+      "loss": 1.6666,
+      "step": 2697
+    },
+    {
+      "epoch": 0.30991901671357186,
+      "grad_norm": 0.3777678906917572,
+      "learning_rate": 0.0001,
+      "loss": 1.7871,
+      "step": 2698
+    },
+    {
+      "epoch": 0.310033886623399,
+      "grad_norm": 0.39043325185775757,
+      "learning_rate": 0.0001,
+      "loss": 1.8022,
+      "step": 2699
+    },
+    {
+      "epoch": 0.3101487565332261,
+      "grad_norm": 0.39468830823898315,
+      "learning_rate": 0.0001,
+      "loss": 1.6472,
+      "step": 2700
+    },
+    {
+      "epoch": 0.3102636264430532,
+      "grad_norm": 0.41475868225097656,
+      "learning_rate": 0.0001,
+      "loss": 1.8264,
+      "step": 2701
+    },
+    {
+      "epoch": 0.31037849635288034,
+      "grad_norm": 0.387824147939682,
+      "learning_rate": 0.0001,
+      "loss": 1.7159,
+      "step": 2702
+    },
+    {
+      "epoch": 0.31049336626270746,
+      "grad_norm": 0.38441115617752075,
+      "learning_rate": 0.0001,
+      "loss": 1.6584,
+      "step": 2703
+    },
+    {
+      "epoch": 0.3106082361725346,
+      "grad_norm": 0.374197781085968,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 2704
+    },
+    {
+      "epoch": 0.3107231060823617,
+      "grad_norm": 0.3562909960746765,
+      "learning_rate": 0.0001,
+      "loss": 1.4979,
+      "step": 2705
+    },
+    {
+      "epoch": 0.3108379759921888,
+      "grad_norm": 0.38204044103622437,
+      "learning_rate": 0.0001,
+      "loss": 1.8704,
+      "step": 2706
+    },
+    {
+      "epoch": 0.31095284590201594,
+      "grad_norm": 0.3814204931259155,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 2707
+    },
+    {
+      "epoch": 0.3110677158118431,
+      "grad_norm": 0.4282824993133545,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 2708
+    },
+    {
+      "epoch": 0.31118258572167024,
+      "grad_norm": 0.3564637005329132,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 2709
+    },
+    {
+      "epoch": 0.31129745563149736,
+      "grad_norm": 0.4151432514190674,
+      "learning_rate": 0.0001,
+      "loss": 1.9175,
+      "step": 2710
+    },
+    {
+      "epoch": 0.3114123255413245,
+      "grad_norm": 0.37644943594932556,
+      "learning_rate": 0.0001,
+      "loss": 1.5751,
+      "step": 2711
+    },
+    {
+      "epoch": 0.3115271954511516,
+      "grad_norm": 0.39474377036094666,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 2712
+    },
+    {
+      "epoch": 0.3116420653609787,
+      "grad_norm": 0.3828750550746918,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 2713
+    },
+    {
+      "epoch": 0.31175693527080584,
+      "grad_norm": 0.38936948776245117,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 2714
+    },
+    {
+      "epoch": 0.31187180518063295,
+      "grad_norm": 0.40668943524360657,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 2715
+    },
+    {
+      "epoch": 0.3119866750904601,
+      "grad_norm": 0.36367443203926086,
+      "learning_rate": 0.0001,
+      "loss": 1.6902,
+      "step": 2716
+    },
+    {
+      "epoch": 0.3121015450002872,
+      "grad_norm": 0.4468287229537964,
+      "learning_rate": 0.0001,
+      "loss": 1.5658,
+      "step": 2717
+    },
+    {
+      "epoch": 0.3122164149101143,
+      "grad_norm": 0.3429298400878906,
+      "learning_rate": 0.0001,
+      "loss": 1.4857,
+      "step": 2718
+    },
+    {
+      "epoch": 0.31233128481994143,
+      "grad_norm": 0.4072478711605072,
+      "learning_rate": 0.0001,
+      "loss": 1.7749,
+      "step": 2719
+    },
+    {
+      "epoch": 0.31244615472976855,
+      "grad_norm": 0.37284785509109497,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 2720
+    },
+    {
+      "epoch": 0.3125610246395957,
+      "grad_norm": 0.41400986909866333,
+      "learning_rate": 0.0001,
+      "loss": 1.936,
+      "step": 2721
+    },
+    {
+      "epoch": 0.3126758945494228,
+      "grad_norm": 0.3585307002067566,
+      "learning_rate": 0.0001,
+      "loss": 1.3443,
+      "step": 2722
+    },
+    {
+      "epoch": 0.3127907644592499,
+      "grad_norm": 0.41940388083457947,
+      "learning_rate": 0.0001,
+      "loss": 1.8787,
+      "step": 2723
+    },
+    {
+      "epoch": 0.31290563436907703,
+      "grad_norm": 0.3974437713623047,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 2724
+    },
+    {
+      "epoch": 0.31302050427890415,
+      "grad_norm": 0.39561134576797485,
+      "learning_rate": 0.0001,
+      "loss": 1.7365,
+      "step": 2725
+    },
+    {
+      "epoch": 0.3131353741887313,
+      "grad_norm": 0.36751341819763184,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 2726
+    },
+    {
+      "epoch": 0.3132502440985584,
+      "grad_norm": 0.41230806708335876,
+      "learning_rate": 0.0001,
+      "loss": 1.5398,
+      "step": 2727
+    },
+    {
+      "epoch": 0.3133651140083855,
+      "grad_norm": 0.3846902847290039,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 2728
+    },
+    {
+      "epoch": 0.31347998391821263,
+      "grad_norm": 0.40803879499435425,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 2729
+    },
+    {
+      "epoch": 0.31359485382803975,
+      "grad_norm": 0.35604923963546753,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 2730
+    },
+    {
+      "epoch": 0.3137097237378669,
+      "grad_norm": 0.38761159777641296,
+      "learning_rate": 0.0001,
+      "loss": 1.7401,
+      "step": 2731
+    },
+    {
+      "epoch": 0.313824593647694,
+      "grad_norm": 0.42147189378738403,
+      "learning_rate": 0.0001,
+      "loss": 1.7959,
+      "step": 2732
+    },
+    {
+      "epoch": 0.3139394635575211,
+      "grad_norm": 0.3850533366203308,
+      "learning_rate": 0.0001,
+      "loss": 1.8302,
+      "step": 2733
+    },
+    {
+      "epoch": 0.31405433346734823,
+      "grad_norm": 0.3670084476470947,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 2734
+    },
+    {
+      "epoch": 0.31416920337717535,
+      "grad_norm": 0.3647415041923523,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 2735
+    },
+    {
+      "epoch": 0.3142840732870025,
+      "grad_norm": 0.38492029905319214,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 2736
+    },
+    {
+      "epoch": 0.3143989431968296,
+      "grad_norm": 0.4054207503795624,
+      "learning_rate": 0.0001,
+      "loss": 1.8419,
+      "step": 2737
+    },
+    {
+      "epoch": 0.3145138131066567,
+      "grad_norm": 0.36467689275741577,
+      "learning_rate": 0.0001,
+      "loss": 1.5701,
+      "step": 2738
+    },
+    {
+      "epoch": 0.31462868301648383,
+      "grad_norm": 0.3815039098262787,
+      "learning_rate": 0.0001,
+      "loss": 1.6893,
+      "step": 2739
+    },
+    {
+      "epoch": 0.31474355292631095,
+      "grad_norm": 0.3768649697303772,
+      "learning_rate": 0.0001,
+      "loss": 1.5021,
+      "step": 2740
+    },
+    {
+      "epoch": 0.3148584228361381,
+      "grad_norm": 0.36210617423057556,
+      "learning_rate": 0.0001,
+      "loss": 1.7335,
+      "step": 2741
+    },
+    {
+      "epoch": 0.3149732927459652,
+      "grad_norm": 0.41380202770233154,
+      "learning_rate": 0.0001,
+      "loss": 1.799,
+      "step": 2742
+    },
+    {
+      "epoch": 0.3150881626557923,
+      "grad_norm": 0.38201257586479187,
+      "learning_rate": 0.0001,
+      "loss": 1.669,
+      "step": 2743
+    },
+    {
+      "epoch": 0.31520303256561943,
+      "grad_norm": 0.383025586605072,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 2744
+    },
+    {
+      "epoch": 0.31531790247544655,
+      "grad_norm": 0.3883838653564453,
+      "learning_rate": 0.0001,
+      "loss": 1.6941,
+      "step": 2745
+    },
+    {
+      "epoch": 0.3154327723852737,
+      "grad_norm": 0.35281357169151306,
+      "learning_rate": 0.0001,
+      "loss": 1.4192,
+      "step": 2746
+    },
+    {
+      "epoch": 0.3155476422951008,
+      "grad_norm": 0.40777984261512756,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 2747
+    },
+    {
+      "epoch": 0.3156625122049279,
+      "grad_norm": 0.36390420794487,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 2748
+    },
+    {
+      "epoch": 0.31577738211475503,
+      "grad_norm": 0.38784828782081604,
+      "learning_rate": 0.0001,
+      "loss": 1.7435,
+      "step": 2749
+    },
+    {
+      "epoch": 0.31589225202458215,
+      "grad_norm": 0.4247525632381439,
+      "learning_rate": 0.0001,
+      "loss": 1.7565,
+      "step": 2750
+    },
+    {
+      "epoch": 0.3160071219344093,
+      "grad_norm": 0.38927143812179565,
+      "learning_rate": 0.0001,
+      "loss": 1.6956,
+      "step": 2751
+    },
+    {
+      "epoch": 0.3161219918442364,
+      "grad_norm": 0.34700581431388855,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 2752
+    },
+    {
+      "epoch": 0.3162368617540635,
+      "grad_norm": 0.3716479241847992,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 2753
+    },
+    {
+      "epoch": 0.31635173166389063,
+      "grad_norm": 0.4048490822315216,
+      "learning_rate": 0.0001,
+      "loss": 1.7723,
+      "step": 2754
+    },
+    {
+      "epoch": 0.31646660157371775,
+      "grad_norm": 0.36082956194877625,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3165814714835449,
+      "grad_norm": 0.40957361459732056,
+      "learning_rate": 0.0001,
+      "loss": 1.845,
+      "step": 2756
+    },
+    {
+      "epoch": 0.316696341393372,
+      "grad_norm": 0.3455057442188263,
+      "learning_rate": 0.0001,
+      "loss": 1.5038,
+      "step": 2757
+    },
+    {
+      "epoch": 0.3168112113031991,
+      "grad_norm": 0.3513345718383789,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 2758
+    },
+    {
+      "epoch": 0.31692608121302623,
+      "grad_norm": 0.3828970491886139,
+      "learning_rate": 0.0001,
+      "loss": 1.6297,
+      "step": 2759
+    },
+    {
+      "epoch": 0.31704095112285335,
+      "grad_norm": 0.370225191116333,
+      "learning_rate": 0.0001,
+      "loss": 1.6051,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3171558210326805,
+      "grad_norm": 0.4304163157939911,
+      "learning_rate": 0.0001,
+      "loss": 1.9163,
+      "step": 2761
+    },
+    {
+      "epoch": 0.3172706909425076,
+      "grad_norm": 0.3837917745113373,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 2762
+    },
+    {
+      "epoch": 0.3173855608523347,
+      "grad_norm": 0.3771938383579254,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 2763
+    },
+    {
+      "epoch": 0.31750043076216183,
+      "grad_norm": 0.3861342668533325,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 2764
+    },
+    {
+      "epoch": 0.31761530067198895,
+      "grad_norm": 0.36387091875076294,
+      "learning_rate": 0.0001,
+      "loss": 1.6573,
+      "step": 2765
+    },
+    {
+      "epoch": 0.3177301705818161,
+      "grad_norm": 0.3894106447696686,
+      "learning_rate": 0.0001,
+      "loss": 1.4844,
+      "step": 2766
+    },
+    {
+      "epoch": 0.3178450404916432,
+      "grad_norm": 0.37301284074783325,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 2767
+    },
+    {
+      "epoch": 0.3179599104014703,
+      "grad_norm": 0.38504043221473694,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 2768
+    },
+    {
+      "epoch": 0.31807478031129743,
+      "grad_norm": 0.37516283988952637,
+      "learning_rate": 0.0001,
+      "loss": 1.6915,
+      "step": 2769
+    },
+    {
+      "epoch": 0.31818965022112455,
+      "grad_norm": 0.3629774749279022,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 2770
+    },
+    {
+      "epoch": 0.3183045201309517,
+      "grad_norm": 0.3853417634963989,
+      "learning_rate": 0.0001,
+      "loss": 1.7261,
+      "step": 2771
+    },
+    {
+      "epoch": 0.3184193900407788,
+      "grad_norm": 0.36345425248146057,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 2772
+    },
+    {
+      "epoch": 0.3185342599506059,
+      "grad_norm": 0.3774106204509735,
+      "learning_rate": 0.0001,
+      "loss": 1.6075,
+      "step": 2773
+    },
+    {
+      "epoch": 0.31864912986043303,
+      "grad_norm": 0.40592819452285767,
+      "learning_rate": 0.0001,
+      "loss": 1.8993,
+      "step": 2774
+    },
+    {
+      "epoch": 0.31876399977026015,
+      "grad_norm": 0.38778918981552124,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 2775
+    },
+    {
+      "epoch": 0.3188788696800873,
+      "grad_norm": 0.39623749256134033,
+      "learning_rate": 0.0001,
+      "loss": 1.7249,
+      "step": 2776
+    },
+    {
+      "epoch": 0.31899373958991445,
+      "grad_norm": 0.3900299072265625,
+      "learning_rate": 0.0001,
+      "loss": 1.7291,
+      "step": 2777
+    },
+    {
+      "epoch": 0.31910860949974157,
+      "grad_norm": 0.3717004358768463,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 2778
+    },
+    {
+      "epoch": 0.3192234794095687,
+      "grad_norm": 0.38834843039512634,
+      "learning_rate": 0.0001,
+      "loss": 1.7519,
+      "step": 2779
+    },
+    {
+      "epoch": 0.3193383493193958,
+      "grad_norm": 0.3893420994281769,
+      "learning_rate": 0.0001,
+      "loss": 1.7821,
+      "step": 2780
+    },
+    {
+      "epoch": 0.31945321922922293,
+      "grad_norm": 0.4312572479248047,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 2781
+    },
+    {
+      "epoch": 0.31956808913905005,
+      "grad_norm": 0.3759611248970032,
+      "learning_rate": 0.0001,
+      "loss": 1.5577,
+      "step": 2782
+    },
+    {
+      "epoch": 0.31968295904887717,
+      "grad_norm": 0.37230929732322693,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 2783
+    },
+    {
+      "epoch": 0.3197978289587043,
+      "grad_norm": 0.3799343407154083,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 2784
+    },
+    {
+      "epoch": 0.3199126988685314,
+      "grad_norm": 0.38527607917785645,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 2785
+    },
+    {
+      "epoch": 0.32002756877835853,
+      "grad_norm": 0.3624141812324524,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 2786
+    },
+    {
+      "epoch": 0.32014243868818565,
+      "grad_norm": 0.36637428402900696,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 2787
+    },
+    {
+      "epoch": 0.32025730859801277,
+      "grad_norm": 0.4166329503059387,
+      "learning_rate": 0.0001,
+      "loss": 1.8372,
+      "step": 2788
+    },
+    {
+      "epoch": 0.3203721785078399,
+      "grad_norm": 0.38629505038261414,
+      "learning_rate": 0.0001,
+      "loss": 1.726,
+      "step": 2789
+    },
+    {
+      "epoch": 0.320487048417667,
+      "grad_norm": 0.3948490023612976,
+      "learning_rate": 0.0001,
+      "loss": 1.7453,
+      "step": 2790
+    },
+    {
+      "epoch": 0.32060191832749413,
+      "grad_norm": 0.3907056450843811,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 2791
+    },
+    {
+      "epoch": 0.32071678823732125,
+      "grad_norm": 0.36693593859672546,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 2792
+    },
+    {
+      "epoch": 0.32083165814714837,
+      "grad_norm": 0.36896297335624695,
+      "learning_rate": 0.0001,
+      "loss": 1.4488,
+      "step": 2793
+    },
+    {
+      "epoch": 0.3209465280569755,
+      "grad_norm": 0.38584834337234497,
+      "learning_rate": 0.0001,
+      "loss": 1.8108,
+      "step": 2794
+    },
+    {
+      "epoch": 0.3210613979668026,
+      "grad_norm": 0.3919477164745331,
+      "learning_rate": 0.0001,
+      "loss": 1.4851,
+      "step": 2795
+    },
+    {
+      "epoch": 0.32117626787662973,
+      "grad_norm": 0.41220781207084656,
+      "learning_rate": 0.0001,
+      "loss": 1.6354,
+      "step": 2796
+    },
+    {
+      "epoch": 0.32129113778645685,
+      "grad_norm": 0.3902750015258789,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 2797
+    },
+    {
+      "epoch": 0.32140600769628397,
+      "grad_norm": 0.3956315219402313,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 2798
+    },
+    {
+      "epoch": 0.3215208776061111,
+      "grad_norm": 0.35562974214553833,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 2799
+    },
+    {
+      "epoch": 0.3216357475159382,
+      "grad_norm": 0.39182206988334656,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 2800
+    },
+    {
+      "epoch": 0.32175061742576533,
+      "grad_norm": 0.41521987318992615,
+      "learning_rate": 0.0001,
+      "loss": 1.7274,
+      "step": 2801
+    },
+    {
+      "epoch": 0.32186548733559245,
+      "grad_norm": 0.38131776452064514,
+      "learning_rate": 0.0001,
+      "loss": 1.5591,
+      "step": 2802
+    },
+    {
+      "epoch": 0.32198035724541957,
+      "grad_norm": 0.4228755533695221,
+      "learning_rate": 0.0001,
+      "loss": 1.8683,
+      "step": 2803
+    },
+    {
+      "epoch": 0.3220952271552467,
+      "grad_norm": 0.3989662230014801,
+      "learning_rate": 0.0001,
+      "loss": 1.8067,
+      "step": 2804
+    },
+    {
+      "epoch": 0.3222100970650738,
+      "grad_norm": 0.3961024284362793,
+      "learning_rate": 0.0001,
+      "loss": 1.8281,
+      "step": 2805
+    },
+    {
+      "epoch": 0.32232496697490093,
+      "grad_norm": 0.4035508930683136,
+      "learning_rate": 0.0001,
+      "loss": 1.7393,
+      "step": 2806
+    },
+    {
+      "epoch": 0.32243983688472805,
+      "grad_norm": 0.38078513741493225,
+      "learning_rate": 0.0001,
+      "loss": 1.7108,
+      "step": 2807
+    },
+    {
+      "epoch": 0.32255470679455517,
+      "grad_norm": 0.367631196975708,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 2808
+    },
+    {
+      "epoch": 0.3226695767043823,
+      "grad_norm": 0.3607901632785797,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 2809
+    },
+    {
+      "epoch": 0.3227844466142094,
+      "grad_norm": 0.3930343687534332,
+      "learning_rate": 0.0001,
+      "loss": 1.4403,
+      "step": 2810
+    },
+    {
+      "epoch": 0.32289931652403653,
+      "grad_norm": 0.37640708684921265,
+      "learning_rate": 0.0001,
+      "loss": 1.798,
+      "step": 2811
+    },
+    {
+      "epoch": 0.32301418643386365,
+      "grad_norm": 0.36390334367752075,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 2812
+    },
+    {
+      "epoch": 0.32312905634369077,
+      "grad_norm": 0.3854324519634247,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 2813
+    },
+    {
+      "epoch": 0.3232439262535179,
+      "grad_norm": 0.37264391779899597,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 2814
+    },
+    {
+      "epoch": 0.323358796163345,
+      "grad_norm": 0.40931811928749084,
+      "learning_rate": 0.0001,
+      "loss": 1.6462,
+      "step": 2815
+    },
+    {
+      "epoch": 0.32347366607317213,
+      "grad_norm": 0.36656174063682556,
+      "learning_rate": 0.0001,
+      "loss": 1.618,
+      "step": 2816
+    },
+    {
+      "epoch": 0.32358853598299925,
+      "grad_norm": 0.3858596682548523,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 2817
+    },
+    {
+      "epoch": 0.32370340589282637,
+      "grad_norm": 0.3784213066101074,
+      "learning_rate": 0.0001,
+      "loss": 1.8733,
+      "step": 2818
+    },
+    {
+      "epoch": 0.3238182758026535,
+      "grad_norm": 0.3836335241794586,
+      "learning_rate": 0.0001,
+      "loss": 1.6904,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3239331457124806,
+      "grad_norm": 0.3633041977882385,
+      "learning_rate": 0.0001,
+      "loss": 1.7444,
+      "step": 2820
+    },
+    {
+      "epoch": 0.32404801562230773,
+      "grad_norm": 0.3903411626815796,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 2821
+    },
+    {
+      "epoch": 0.32416288553213485,
+      "grad_norm": 0.4233011305332184,
+      "learning_rate": 0.0001,
+      "loss": 1.6829,
+      "step": 2822
+    },
+    {
+      "epoch": 0.32427775544196197,
+      "grad_norm": 0.4304129481315613,
+      "learning_rate": 0.0001,
+      "loss": 1.859,
+      "step": 2823
+    },
+    {
+      "epoch": 0.3243926253517891,
+      "grad_norm": 0.37114959955215454,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 2824
+    },
+    {
+      "epoch": 0.3245074952616162,
+      "grad_norm": 0.3852083683013916,
+      "learning_rate": 0.0001,
+      "loss": 1.6829,
+      "step": 2825
+    },
+    {
+      "epoch": 0.32462236517144333,
+      "grad_norm": 0.3652872145175934,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 2826
+    },
+    {
+      "epoch": 0.32473723508127045,
+      "grad_norm": 0.3811475932598114,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 2827
+    },
+    {
+      "epoch": 0.32485210499109757,
+      "grad_norm": 0.3886179029941559,
+      "learning_rate": 0.0001,
+      "loss": 1.7562,
+      "step": 2828
+    },
+    {
+      "epoch": 0.3249669749009247,
+      "grad_norm": 0.39535319805145264,
+      "learning_rate": 0.0001,
+      "loss": 1.5934,
+      "step": 2829
+    },
+    {
+      "epoch": 0.3250818448107518,
+      "grad_norm": 0.40873974561691284,
+      "learning_rate": 0.0001,
+      "loss": 1.8607,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3251967147205789,
+      "grad_norm": 0.38622164726257324,
+      "learning_rate": 0.0001,
+      "loss": 1.6962,
+      "step": 2831
+    },
+    {
+      "epoch": 0.32531158463040605,
+      "grad_norm": 0.42052188515663147,
+      "learning_rate": 0.0001,
+      "loss": 1.7905,
+      "step": 2832
+    },
+    {
+      "epoch": 0.32542645454023317,
+      "grad_norm": 0.3537195026874542,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 2833
+    },
+    {
+      "epoch": 0.3255413244500603,
+      "grad_norm": 0.3691607415676117,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 2834
+    },
+    {
+      "epoch": 0.3256561943598874,
+      "grad_norm": 0.40789857506752014,
+      "learning_rate": 0.0001,
+      "loss": 1.9685,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3257710642697145,
+      "grad_norm": 0.3981241285800934,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 2836
+    },
+    {
+      "epoch": 0.32588593417954165,
+      "grad_norm": 0.36344999074935913,
+      "learning_rate": 0.0001,
+      "loss": 1.6628,
+      "step": 2837
+    },
+    {
+      "epoch": 0.32600080408936877,
+      "grad_norm": 0.3493889272212982,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 2838
+    },
+    {
+      "epoch": 0.3261156739991959,
+      "grad_norm": 0.39113399386405945,
+      "learning_rate": 0.0001,
+      "loss": 1.7783,
+      "step": 2839
+    },
+    {
+      "epoch": 0.326230543909023,
+      "grad_norm": 0.3922522962093353,
+      "learning_rate": 0.0001,
+      "loss": 1.8986,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3263454138188501,
+      "grad_norm": 0.35591554641723633,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 2841
+    },
+    {
+      "epoch": 0.32646028372867725,
+      "grad_norm": 0.38669490814208984,
+      "learning_rate": 0.0001,
+      "loss": 1.8257,
+      "step": 2842
+    },
+    {
+      "epoch": 0.32657515363850437,
+      "grad_norm": 0.37732958793640137,
+      "learning_rate": 0.0001,
+      "loss": 1.7277,
+      "step": 2843
+    },
+    {
+      "epoch": 0.3266900235483315,
+      "grad_norm": 0.3958292603492737,
+      "learning_rate": 0.0001,
+      "loss": 1.8254,
+      "step": 2844
+    },
+    {
+      "epoch": 0.32680489345815866,
+      "grad_norm": 0.368966668844223,
+      "learning_rate": 0.0001,
+      "loss": 1.728,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3269197633679858,
+      "grad_norm": 0.3690721094608307,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 2846
+    },
+    {
+      "epoch": 0.3270346332778129,
+      "grad_norm": 0.37405163049697876,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 2847
+    },
+    {
+      "epoch": 0.32714950318764,
+      "grad_norm": 0.43797191977500916,
+      "learning_rate": 0.0001,
+      "loss": 1.9247,
+      "step": 2848
+    },
+    {
+      "epoch": 0.32726437309746714,
+      "grad_norm": 0.38513773679733276,
+      "learning_rate": 0.0001,
+      "loss": 1.7059,
+      "step": 2849
+    },
+    {
+      "epoch": 0.32737924300729426,
+      "grad_norm": 0.36473625898361206,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3274941129171214,
+      "grad_norm": 0.38011401891708374,
+      "learning_rate": 0.0001,
+      "loss": 1.6667,
+      "step": 2851
+    },
+    {
+      "epoch": 0.3276089828269485,
+      "grad_norm": 0.3875674605369568,
+      "learning_rate": 0.0001,
+      "loss": 1.8539,
+      "step": 2852
+    },
+    {
+      "epoch": 0.3277238527367756,
+      "grad_norm": 0.4060609042644501,
+      "learning_rate": 0.0001,
+      "loss": 1.6604,
+      "step": 2853
+    },
+    {
+      "epoch": 0.32783872264660274,
+      "grad_norm": 0.38478556275367737,
+      "learning_rate": 0.0001,
+      "loss": 1.8449,
+      "step": 2854
+    },
+    {
+      "epoch": 0.32795359255642986,
+      "grad_norm": 0.414813756942749,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 2855
+    },
+    {
+      "epoch": 0.328068462466257,
+      "grad_norm": 0.38957443833351135,
+      "learning_rate": 0.0001,
+      "loss": 1.8987,
+      "step": 2856
+    },
+    {
+      "epoch": 0.3281833323760841,
+      "grad_norm": 0.3751903772354126,
+      "learning_rate": 0.0001,
+      "loss": 1.7977,
+      "step": 2857
+    },
+    {
+      "epoch": 0.3282982022859112,
+      "grad_norm": 0.3669251799583435,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 2858
+    },
+    {
+      "epoch": 0.32841307219573834,
+      "grad_norm": 0.36348956823349,
+      "learning_rate": 0.0001,
+      "loss": 1.7146,
+      "step": 2859
+    },
+    {
+      "epoch": 0.32852794210556546,
+      "grad_norm": 0.35992637276649475,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 2860
+    },
+    {
+      "epoch": 0.3286428120153926,
+      "grad_norm": 0.3643839955329895,
+      "learning_rate": 0.0001,
+      "loss": 1.7187,
+      "step": 2861
+    },
+    {
+      "epoch": 0.3287576819252197,
+      "grad_norm": 0.40732714533805847,
+      "learning_rate": 0.0001,
+      "loss": 1.8751,
+      "step": 2862
+    },
+    {
+      "epoch": 0.3288725518350468,
+      "grad_norm": 0.38161808252334595,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 2863
+    },
+    {
+      "epoch": 0.32898742174487394,
+      "grad_norm": 0.40559515357017517,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 2864
+    },
+    {
+      "epoch": 0.32910229165470106,
+      "grad_norm": 0.3841257393360138,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 2865
+    },
+    {
+      "epoch": 0.3292171615645282,
+      "grad_norm": 0.36284148693084717,
+      "learning_rate": 0.0001,
+      "loss": 1.4342,
+      "step": 2866
+    },
+    {
+      "epoch": 0.3293320314743553,
+      "grad_norm": 0.3603561818599701,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 2867
+    },
+    {
+      "epoch": 0.3294469013841824,
+      "grad_norm": 0.39351412653923035,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 2868
+    },
+    {
+      "epoch": 0.32956177129400954,
+      "grad_norm": 0.408401757478714,
+      "learning_rate": 0.0001,
+      "loss": 1.7423,
+      "step": 2869
+    },
+    {
+      "epoch": 0.32967664120383666,
+      "grad_norm": 0.37314414978027344,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3297915111136638,
+      "grad_norm": 0.369933545589447,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 2871
+    },
+    {
+      "epoch": 0.3299063810234909,
+      "grad_norm": 0.4098500907421112,
+      "learning_rate": 0.0001,
+      "loss": 1.8109,
+      "step": 2872
+    },
+    {
+      "epoch": 0.330021250933318,
+      "grad_norm": 0.40561625361442566,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 2873
+    },
+    {
+      "epoch": 0.33013612084314514,
+      "grad_norm": 0.3818763792514801,
+      "learning_rate": 0.0001,
+      "loss": 1.6462,
+      "step": 2874
+    },
+    {
+      "epoch": 0.33025099075297226,
+      "grad_norm": 0.4518624544143677,
+      "learning_rate": 0.0001,
+      "loss": 1.6921,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3303658606627994,
+      "grad_norm": 0.3883214294910431,
+      "learning_rate": 0.0001,
+      "loss": 1.7215,
+      "step": 2876
+    },
+    {
+      "epoch": 0.3304807305726265,
+      "grad_norm": 0.4087526500225067,
+      "learning_rate": 0.0001,
+      "loss": 1.919,
+      "step": 2877
+    },
+    {
+      "epoch": 0.3305956004824536,
+      "grad_norm": 0.37639445066452026,
+      "learning_rate": 0.0001,
+      "loss": 1.7867,
+      "step": 2878
+    },
+    {
+      "epoch": 0.33071047039228074,
+      "grad_norm": 0.42621272802352905,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 2879
+    },
+    {
+      "epoch": 0.33082534030210786,
+      "grad_norm": 0.36311256885528564,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 2880
+    },
+    {
+      "epoch": 0.330940210211935,
+      "grad_norm": 0.35729482769966125,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 2881
+    },
+    {
+      "epoch": 0.3310550801217621,
+      "grad_norm": 0.3519277274608612,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 2882
+    },
+    {
+      "epoch": 0.3311699500315892,
+      "grad_norm": 0.40030479431152344,
+      "learning_rate": 0.0001,
+      "loss": 1.8654,
+      "step": 2883
+    },
+    {
+      "epoch": 0.33128481994141634,
+      "grad_norm": 0.3615962862968445,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 2884
+    },
+    {
+      "epoch": 0.33139968985124346,
+      "grad_norm": 0.38331496715545654,
+      "learning_rate": 0.0001,
+      "loss": 1.7308,
+      "step": 2885
+    },
+    {
+      "epoch": 0.3315145597610706,
+      "grad_norm": 0.3735135793685913,
+      "learning_rate": 0.0001,
+      "loss": 1.4669,
+      "step": 2886
+    },
+    {
+      "epoch": 0.3316294296708977,
+      "grad_norm": 0.39380258321762085,
+      "learning_rate": 0.0001,
+      "loss": 1.5611,
+      "step": 2887
+    },
+    {
+      "epoch": 0.3317442995807248,
+      "grad_norm": 0.4285888373851776,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 2888
+    },
+    {
+      "epoch": 0.33185916949055194,
+      "grad_norm": 0.36404624581336975,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 2889
+    },
+    {
+      "epoch": 0.33197403940037906,
+      "grad_norm": 0.42275553941726685,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 2890
+    },
+    {
+      "epoch": 0.3320889093102062,
+      "grad_norm": 0.37147074937820435,
+      "learning_rate": 0.0001,
+      "loss": 1.5029,
+      "step": 2891
+    },
+    {
+      "epoch": 0.3322037792200333,
+      "grad_norm": 0.3962612450122833,
+      "learning_rate": 0.0001,
+      "loss": 1.8306,
+      "step": 2892
+    },
+    {
+      "epoch": 0.3323186491298604,
+      "grad_norm": 0.36973974108695984,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 2893
+    },
+    {
+      "epoch": 0.33243351903968754,
+      "grad_norm": 0.3744032680988312,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 2894
+    },
+    {
+      "epoch": 0.33254838894951466,
+      "grad_norm": 0.3863425552845001,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 2895
+    },
+    {
+      "epoch": 0.3326632588593418,
+      "grad_norm": 0.3814421594142914,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 2896
+    },
+    {
+      "epoch": 0.3327781287691689,
+      "grad_norm": 0.3841194808483124,
+      "learning_rate": 0.0001,
+      "loss": 1.7178,
+      "step": 2897
+    },
+    {
+      "epoch": 0.332892998678996,
+      "grad_norm": 0.3847556412220001,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 2898
+    },
+    {
+      "epoch": 0.33300786858882314,
+      "grad_norm": 0.38841572403907776,
+      "learning_rate": 0.0001,
+      "loss": 1.8005,
+      "step": 2899
+    },
+    {
+      "epoch": 0.33312273849865026,
+      "grad_norm": 0.41473743319511414,
+      "learning_rate": 0.0001,
+      "loss": 1.6702,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3332376084084774,
+      "grad_norm": 0.37773972749710083,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 2901
+    },
+    {
+      "epoch": 0.3333524783183045,
+      "grad_norm": 0.38093826174736023,
+      "learning_rate": 0.0001,
+      "loss": 1.7681,
+      "step": 2902
+    },
+    {
+      "epoch": 0.3334673482281316,
+      "grad_norm": 0.39234668016433716,
+      "learning_rate": 0.0001,
+      "loss": 1.7385,
+      "step": 2903
+    },
+    {
+      "epoch": 0.33358221813795874,
+      "grad_norm": 0.4004804790019989,
+      "learning_rate": 0.0001,
+      "loss": 1.7519,
+      "step": 2904
+    },
+    {
+      "epoch": 0.33369708804778586,
+      "grad_norm": 0.35611221194267273,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 2905
+    },
+    {
+      "epoch": 0.333811957957613,
+      "grad_norm": 0.383696585893631,
+      "learning_rate": 0.0001,
+      "loss": 1.7001,
+      "step": 2906
+    },
+    {
+      "epoch": 0.3339268278674401,
+      "grad_norm": 0.3913770020008087,
+      "learning_rate": 0.0001,
+      "loss": 1.7143,
+      "step": 2907
+    },
+    {
+      "epoch": 0.3340416977772672,
+      "grad_norm": 0.38301557302474976,
+      "learning_rate": 0.0001,
+      "loss": 1.649,
+      "step": 2908
+    },
+    {
+      "epoch": 0.33415656768709434,
+      "grad_norm": 0.3856869041919708,
+      "learning_rate": 0.0001,
+      "loss": 1.7601,
+      "step": 2909
+    },
+    {
+      "epoch": 0.33427143759692146,
+      "grad_norm": 0.3839951753616333,
+      "learning_rate": 0.0001,
+      "loss": 1.8057,
+      "step": 2910
+    },
+    {
+      "epoch": 0.3343863075067486,
+      "grad_norm": 0.38269492983818054,
+      "learning_rate": 0.0001,
+      "loss": 1.5389,
+      "step": 2911
+    },
+    {
+      "epoch": 0.3345011774165757,
+      "grad_norm": 0.3791959583759308,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 2912
+    },
+    {
+      "epoch": 0.3346160473264028,
+      "grad_norm": 0.36112257838249207,
+      "learning_rate": 0.0001,
+      "loss": 1.6216,
+      "step": 2913
+    },
+    {
+      "epoch": 0.33473091723623,
+      "grad_norm": 0.3655812740325928,
+      "learning_rate": 0.0001,
+      "loss": 1.4784,
+      "step": 2914
+    },
+    {
+      "epoch": 0.3348457871460571,
+      "grad_norm": 0.3751130998134613,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 2915
+    },
+    {
+      "epoch": 0.33496065705588424,
+      "grad_norm": 0.3733077049255371,
+      "learning_rate": 0.0001,
+      "loss": 1.8166,
+      "step": 2916
+    },
+    {
+      "epoch": 0.33507552696571136,
+      "grad_norm": 0.3870159983634949,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 2917
+    },
+    {
+      "epoch": 0.3351903968755385,
+      "grad_norm": 0.3635254204273224,
+      "learning_rate": 0.0001,
+      "loss": 1.4773,
+      "step": 2918
+    },
+    {
+      "epoch": 0.3353052667853656,
+      "grad_norm": 0.3966655135154724,
+      "learning_rate": 0.0001,
+      "loss": 1.4219,
+      "step": 2919
+    },
+    {
+      "epoch": 0.3354201366951927,
+      "grad_norm": 0.3748622536659241,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 2920
+    },
+    {
+      "epoch": 0.33553500660501984,
+      "grad_norm": 0.38931792974472046,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 2921
+    },
+    {
+      "epoch": 0.33564987651484696,
+      "grad_norm": 0.4402804970741272,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 2922
+    },
+    {
+      "epoch": 0.3357647464246741,
+      "grad_norm": 0.3441646099090576,
+      "learning_rate": 0.0001,
+      "loss": 1.5717,
+      "step": 2923
+    },
+    {
+      "epoch": 0.3358796163345012,
+      "grad_norm": 0.39570891857147217,
+      "learning_rate": 0.0001,
+      "loss": 1.714,
+      "step": 2924
+    },
+    {
+      "epoch": 0.3359944862443283,
+      "grad_norm": 0.39201679825782776,
+      "learning_rate": 0.0001,
+      "loss": 1.7896,
+      "step": 2925
+    },
+    {
+      "epoch": 0.33610935615415544,
+      "grad_norm": 0.3926868140697479,
+      "learning_rate": 0.0001,
+      "loss": 1.6262,
+      "step": 2926
+    },
+    {
+      "epoch": 0.33622422606398256,
+      "grad_norm": 0.3830588757991791,
+      "learning_rate": 0.0001,
+      "loss": 1.6365,
+      "step": 2927
+    },
+    {
+      "epoch": 0.3363390959738097,
+      "grad_norm": 0.3714669346809387,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 2928
+    },
+    {
+      "epoch": 0.3364539658836368,
+      "grad_norm": 0.4032626748085022,
+      "learning_rate": 0.0001,
+      "loss": 1.7987,
+      "step": 2929
+    },
+    {
+      "epoch": 0.3365688357934639,
+      "grad_norm": 0.40316057205200195,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 2930
+    },
+    {
+      "epoch": 0.33668370570329104,
+      "grad_norm": 0.4364853799343109,
+      "learning_rate": 0.0001,
+      "loss": 1.5832,
+      "step": 2931
+    },
+    {
+      "epoch": 0.33679857561311816,
+      "grad_norm": 0.3844012916088104,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 2932
+    },
+    {
+      "epoch": 0.3369134455229453,
+      "grad_norm": 0.3774738311767578,
+      "learning_rate": 0.0001,
+      "loss": 1.7877,
+      "step": 2933
+    },
+    {
+      "epoch": 0.3370283154327724,
+      "grad_norm": 0.4184546172618866,
+      "learning_rate": 0.0001,
+      "loss": 1.8513,
+      "step": 2934
+    },
+    {
+      "epoch": 0.3371431853425995,
+      "grad_norm": 0.3983631432056427,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 2935
+    },
+    {
+      "epoch": 0.33725805525242664,
+      "grad_norm": 0.356240451335907,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 2936
+    },
+    {
+      "epoch": 0.33737292516225376,
+      "grad_norm": 0.3877936601638794,
+      "learning_rate": 0.0001,
+      "loss": 1.6252,
+      "step": 2937
+    },
+    {
+      "epoch": 0.3374877950720809,
+      "grad_norm": 0.3945756256580353,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 2938
+    },
+    {
+      "epoch": 0.337602664981908,
+      "grad_norm": 0.3544231355190277,
+      "learning_rate": 0.0001,
+      "loss": 1.3307,
+      "step": 2939
+    },
+    {
+      "epoch": 0.3377175348917351,
+      "grad_norm": 0.3833335041999817,
+      "learning_rate": 0.0001,
+      "loss": 1.7065,
+      "step": 2940
+    },
+    {
+      "epoch": 0.33783240480156224,
+      "grad_norm": 0.3731600046157837,
+      "learning_rate": 0.0001,
+      "loss": 1.7736,
+      "step": 2941
+    },
+    {
+      "epoch": 0.33794727471138936,
+      "grad_norm": 0.4063700735569,
+      "learning_rate": 0.0001,
+      "loss": 1.6827,
+      "step": 2942
+    },
+    {
+      "epoch": 0.3380621446212165,
+      "grad_norm": 0.4960021674633026,
+      "learning_rate": 0.0001,
+      "loss": 1.9989,
+      "step": 2943
+    },
+    {
+      "epoch": 0.3381770145310436,
+      "grad_norm": 0.4238811433315277,
+      "learning_rate": 0.0001,
+      "loss": 1.7855,
+      "step": 2944
+    },
+    {
+      "epoch": 0.3382918844408707,
+      "grad_norm": 0.4114185571670532,
+      "learning_rate": 0.0001,
+      "loss": 1.8296,
+      "step": 2945
+    },
+    {
+      "epoch": 0.33840675435069784,
+      "grad_norm": 0.40994930267333984,
+      "learning_rate": 0.0001,
+      "loss": 1.8185,
+      "step": 2946
+    },
+    {
+      "epoch": 0.33852162426052496,
+      "grad_norm": 0.39755189418792725,
+      "learning_rate": 0.0001,
+      "loss": 1.7911,
+      "step": 2947
+    },
+    {
+      "epoch": 0.3386364941703521,
+      "grad_norm": 0.39836958050727844,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 2948
+    },
+    {
+      "epoch": 0.3387513640801792,
+      "grad_norm": 0.3699915409088135,
+      "learning_rate": 0.0001,
+      "loss": 1.603,
+      "step": 2949
+    },
+    {
+      "epoch": 0.3388662339900063,
+      "grad_norm": 0.42995521426200867,
+      "learning_rate": 0.0001,
+      "loss": 1.852,
+      "step": 2950
+    },
+    {
+      "epoch": 0.33898110389983344,
+      "grad_norm": 0.398151695728302,
+      "learning_rate": 0.0001,
+      "loss": 1.8414,
+      "step": 2951
+    },
+    {
+      "epoch": 0.33909597380966056,
+      "grad_norm": 0.36557191610336304,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 2952
+    },
+    {
+      "epoch": 0.3392108437194877,
+      "grad_norm": 0.3784855008125305,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 2953
+    },
+    {
+      "epoch": 0.3393257136293148,
+      "grad_norm": 0.40273427963256836,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 2954
+    },
+    {
+      "epoch": 0.3394405835391419,
+      "grad_norm": 0.37005481123924255,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 2955
+    },
+    {
+      "epoch": 0.33955545344896904,
+      "grad_norm": 0.3850763142108917,
+      "learning_rate": 0.0001,
+      "loss": 1.7957,
+      "step": 2956
+    },
+    {
+      "epoch": 0.33967032335879616,
+      "grad_norm": 0.3557315170764923,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 2957
+    },
+    {
+      "epoch": 0.3397851932686233,
+      "grad_norm": 0.44396260380744934,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 2958
+    },
+    {
+      "epoch": 0.3399000631784504,
+      "grad_norm": 0.3347325325012207,
+      "learning_rate": 0.0001,
+      "loss": 1.3493,
+      "step": 2959
+    },
+    {
+      "epoch": 0.3400149330882775,
+      "grad_norm": 0.3799315094947815,
+      "learning_rate": 0.0001,
+      "loss": 1.698,
+      "step": 2960
+    },
+    {
+      "epoch": 0.34012980299810464,
+      "grad_norm": 0.3979965150356293,
+      "learning_rate": 0.0001,
+      "loss": 1.6106,
+      "step": 2961
+    },
+    {
+      "epoch": 0.34024467290793176,
+      "grad_norm": 0.3687105178833008,
+      "learning_rate": 0.0001,
+      "loss": 1.6298,
+      "step": 2962
+    },
+    {
+      "epoch": 0.3403595428177589,
+      "grad_norm": 0.3896116018295288,
+      "learning_rate": 0.0001,
+      "loss": 1.6547,
+      "step": 2963
+    },
+    {
+      "epoch": 0.340474412727586,
+      "grad_norm": 0.38803455233573914,
+      "learning_rate": 0.0001,
+      "loss": 1.7099,
+      "step": 2964
+    },
+    {
+      "epoch": 0.3405892826374131,
+      "grad_norm": 0.37791207432746887,
+      "learning_rate": 0.0001,
+      "loss": 1.4235,
+      "step": 2965
+    },
+    {
+      "epoch": 0.34070415254724024,
+      "grad_norm": 0.3701097071170807,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 2966
+    },
+    {
+      "epoch": 0.34081902245706736,
+      "grad_norm": 0.3952276408672333,
+      "learning_rate": 0.0001,
+      "loss": 1.7784,
+      "step": 2967
+    },
+    {
+      "epoch": 0.3409338923668945,
+      "grad_norm": 0.6198942065238953,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 2968
+    },
+    {
+      "epoch": 0.3410487622767216,
+      "grad_norm": 0.39042800664901733,
+      "learning_rate": 0.0001,
+      "loss": 1.715,
+      "step": 2969
+    },
+    {
+      "epoch": 0.3411636321865487,
+      "grad_norm": 0.39784181118011475,
+      "learning_rate": 0.0001,
+      "loss": 1.7825,
+      "step": 2970
+    },
+    {
+      "epoch": 0.34127850209637584,
+      "grad_norm": 0.42527255415916443,
+      "learning_rate": 0.0001,
+      "loss": 1.6846,
+      "step": 2971
+    },
+    {
+      "epoch": 0.34139337200620296,
+      "grad_norm": 0.35908040404319763,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 2972
+    },
+    {
+      "epoch": 0.3415082419160301,
+      "grad_norm": 0.3641200661659241,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 2973
+    },
+    {
+      "epoch": 0.3416231118258572,
+      "grad_norm": 0.38720008730888367,
+      "learning_rate": 0.0001,
+      "loss": 1.6965,
+      "step": 2974
+    },
+    {
+      "epoch": 0.3417379817356843,
+      "grad_norm": 0.40444648265838623,
+      "learning_rate": 0.0001,
+      "loss": 1.9605,
+      "step": 2975
+    },
+    {
+      "epoch": 0.34185285164551144,
+      "grad_norm": 0.3607354462146759,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 2976
+    },
+    {
+      "epoch": 0.34196772155533856,
+      "grad_norm": 0.3962652087211609,
+      "learning_rate": 0.0001,
+      "loss": 1.85,
+      "step": 2977
+    },
+    {
+      "epoch": 0.3420825914651657,
+      "grad_norm": 0.36956319212913513,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 2978
+    },
+    {
+      "epoch": 0.3421974613749928,
+      "grad_norm": 0.39066120982170105,
+      "learning_rate": 0.0001,
+      "loss": 1.7077,
+      "step": 2979
+    },
+    {
+      "epoch": 0.3423123312848199,
+      "grad_norm": 0.3526730537414551,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 2980
+    },
+    {
+      "epoch": 0.34242720119464704,
+      "grad_norm": 0.39420273900032043,
+      "learning_rate": 0.0001,
+      "loss": 1.7418,
+      "step": 2981
+    },
+    {
+      "epoch": 0.3425420711044742,
+      "grad_norm": 0.3418049216270447,
+      "learning_rate": 0.0001,
+      "loss": 1.3557,
+      "step": 2982
+    },
+    {
+      "epoch": 0.34265694101430133,
+      "grad_norm": 0.4102267920970917,
+      "learning_rate": 0.0001,
+      "loss": 1.8232,
+      "step": 2983
+    },
+    {
+      "epoch": 0.34277181092412845,
+      "grad_norm": 0.3934805691242218,
+      "learning_rate": 0.0001,
+      "loss": 1.7044,
+      "step": 2984
+    },
+    {
+      "epoch": 0.34288668083395557,
+      "grad_norm": 0.3770129978656769,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3430015507437827,
+      "grad_norm": 0.39099806547164917,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 2986
+    },
+    {
+      "epoch": 0.3431164206536098,
+      "grad_norm": 0.36751532554626465,
+      "learning_rate": 0.0001,
+      "loss": 1.7394,
+      "step": 2987
+    },
+    {
+      "epoch": 0.34323129056343693,
+      "grad_norm": 0.3834240734577179,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 2988
+    },
+    {
+      "epoch": 0.34334616047326405,
+      "grad_norm": 0.3922926187515259,
+      "learning_rate": 0.0001,
+      "loss": 1.6312,
+      "step": 2989
+    },
+    {
+      "epoch": 0.34346103038309117,
+      "grad_norm": 0.38694506883621216,
+      "learning_rate": 0.0001,
+      "loss": 1.81,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3435759002929183,
+      "grad_norm": 0.3988105356693268,
+      "learning_rate": 0.0001,
+      "loss": 1.6714,
+      "step": 2991
+    },
+    {
+      "epoch": 0.3436907702027454,
+      "grad_norm": 0.42653176188468933,
+      "learning_rate": 0.0001,
+      "loss": 1.8591,
+      "step": 2992
+    },
+    {
+      "epoch": 0.34380564011257253,
+      "grad_norm": 0.4091017246246338,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 2993
+    },
+    {
+      "epoch": 0.34392051002239965,
+      "grad_norm": 0.4520750641822815,
+      "learning_rate": 0.0001,
+      "loss": 1.8694,
+      "step": 2994
+    },
+    {
+      "epoch": 0.34403537993222677,
+      "grad_norm": 0.38830214738845825,
+      "learning_rate": 0.0001,
+      "loss": 1.7041,
+      "step": 2995
+    },
+    {
+      "epoch": 0.3441502498420539,
+      "grad_norm": 0.40523776412010193,
+      "learning_rate": 0.0001,
+      "loss": 1.7478,
+      "step": 2996
+    },
+    {
+      "epoch": 0.344265119751881,
+      "grad_norm": 0.3667933940887451,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 2997
+    },
+    {
+      "epoch": 0.34437998966170813,
+      "grad_norm": 0.43613407015800476,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 2998
+    },
+    {
+      "epoch": 0.34449485957153525,
+      "grad_norm": 0.38262253999710083,
+      "learning_rate": 0.0001,
+      "loss": 1.5915,
+      "step": 2999
+    },
+    {
+      "epoch": 0.34460972948136237,
+      "grad_norm": 0.39579108357429504,
+      "learning_rate": 0.0001,
+      "loss": 1.658,
+      "step": 3000
+    },
+    {
+      "epoch": 0.3447245993911895,
+      "grad_norm": 0.38654452562332153,
+      "learning_rate": 0.0001,
+      "loss": 1.3892,
+      "step": 3001
+    },
+    {
+      "epoch": 0.3448394693010166,
+      "grad_norm": 0.35805824398994446,
+      "learning_rate": 0.0001,
+      "loss": 1.4611,
+      "step": 3002
+    },
+    {
+      "epoch": 0.34495433921084373,
+      "grad_norm": 0.3665701448917389,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 3003
+    },
+    {
+      "epoch": 0.34506920912067085,
+      "grad_norm": 0.40592774748802185,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 3004
+    },
+    {
+      "epoch": 0.34518407903049797,
+      "grad_norm": 0.3932124376296997,
+      "learning_rate": 0.0001,
+      "loss": 1.6414,
+      "step": 3005
+    },
+    {
+      "epoch": 0.3452989489403251,
+      "grad_norm": 0.42479029297828674,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 3006
+    },
+    {
+      "epoch": 0.3454138188501522,
+      "grad_norm": 0.43539106845855713,
+      "learning_rate": 0.0001,
+      "loss": 1.7239,
+      "step": 3007
+    },
+    {
+      "epoch": 0.34552868875997933,
+      "grad_norm": 0.40625861287117004,
+      "learning_rate": 0.0001,
+      "loss": 1.9526,
+      "step": 3008
+    },
+    {
+      "epoch": 0.34564355866980645,
+      "grad_norm": 0.3962743282318115,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 3009
+    },
+    {
+      "epoch": 0.34575842857963357,
+      "grad_norm": 0.37623900175094604,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 3010
+    },
+    {
+      "epoch": 0.3458732984894607,
+      "grad_norm": 0.3827407658100128,
+      "learning_rate": 0.0001,
+      "loss": 1.6511,
+      "step": 3011
+    },
+    {
+      "epoch": 0.3459881683992878,
+      "grad_norm": 0.3819064795970917,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 3012
+    },
+    {
+      "epoch": 0.34610303830911493,
+      "grad_norm": 0.38156652450561523,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 3013
+    },
+    {
+      "epoch": 0.34621790821894205,
+      "grad_norm": 0.42006298899650574,
+      "learning_rate": 0.0001,
+      "loss": 1.8345,
+      "step": 3014
+    },
+    {
+      "epoch": 0.34633277812876917,
+      "grad_norm": 0.3623389005661011,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 3015
+    },
+    {
+      "epoch": 0.3464476480385963,
+      "grad_norm": 0.39460188150405884,
+      "learning_rate": 0.0001,
+      "loss": 1.8264,
+      "step": 3016
+    },
+    {
+      "epoch": 0.3465625179484234,
+      "grad_norm": 0.40314847230911255,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 3017
+    },
+    {
+      "epoch": 0.34667738785825053,
+      "grad_norm": 0.37284019589424133,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 3018
+    },
+    {
+      "epoch": 0.34679225776807765,
+      "grad_norm": 0.4431118071079254,
+      "learning_rate": 0.0001,
+      "loss": 1.9377,
+      "step": 3019
+    },
+    {
+      "epoch": 0.34690712767790477,
+      "grad_norm": 0.3849484622478485,
+      "learning_rate": 0.0001,
+      "loss": 1.6559,
+      "step": 3020
+    },
+    {
+      "epoch": 0.3470219975877319,
+      "grad_norm": 0.3909108340740204,
+      "learning_rate": 0.0001,
+      "loss": 1.8541,
+      "step": 3021
+    },
+    {
+      "epoch": 0.347136867497559,
+      "grad_norm": 0.38009119033813477,
+      "learning_rate": 0.0001,
+      "loss": 1.7868,
+      "step": 3022
+    },
+    {
+      "epoch": 0.34725173740738613,
+      "grad_norm": 0.3676866292953491,
+      "learning_rate": 0.0001,
+      "loss": 1.5971,
+      "step": 3023
+    },
+    {
+      "epoch": 0.34736660731721325,
+      "grad_norm": 0.3783824145793915,
+      "learning_rate": 0.0001,
+      "loss": 1.6714,
+      "step": 3024
+    },
+    {
+      "epoch": 0.34748147722704037,
+      "grad_norm": 0.3574279546737671,
+      "learning_rate": 0.0001,
+      "loss": 1.4742,
+      "step": 3025
+    },
+    {
+      "epoch": 0.3475963471368675,
+      "grad_norm": 0.3630661070346832,
+      "learning_rate": 0.0001,
+      "loss": 1.3814,
+      "step": 3026
+    },
+    {
+      "epoch": 0.3477112170466946,
+      "grad_norm": 0.37314149737358093,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 3027
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.38733503222465515,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 3028
+    },
+    {
+      "epoch": 0.34794095686634885,
+      "grad_norm": 0.42135870456695557,
+      "learning_rate": 0.0001,
+      "loss": 1.9024,
+      "step": 3029
+    },
+    {
+      "epoch": 0.34805582677617597,
+      "grad_norm": 0.40700820088386536,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3481706966860031,
+      "grad_norm": 0.389607310295105,
+      "learning_rate": 0.0001,
+      "loss": 1.751,
+      "step": 3031
+    },
+    {
+      "epoch": 0.3482855665958302,
+      "grad_norm": 0.39443865418434143,
+      "learning_rate": 0.0001,
+      "loss": 1.4554,
+      "step": 3032
+    },
+    {
+      "epoch": 0.34840043650565733,
+      "grad_norm": 0.4073096513748169,
+      "learning_rate": 0.0001,
+      "loss": 1.7826,
+      "step": 3033
+    },
+    {
+      "epoch": 0.34851530641548445,
+      "grad_norm": 0.38107818365097046,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 3034
+    },
+    {
+      "epoch": 0.34863017632531157,
+      "grad_norm": 0.39980536699295044,
+      "learning_rate": 0.0001,
+      "loss": 1.7422,
+      "step": 3035
+    },
+    {
+      "epoch": 0.3487450462351387,
+      "grad_norm": 0.3910979628562927,
+      "learning_rate": 0.0001,
+      "loss": 1.7604,
+      "step": 3036
+    },
+    {
+      "epoch": 0.3488599161449658,
+      "grad_norm": 0.3955102562904358,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 3037
+    },
+    {
+      "epoch": 0.34897478605479293,
+      "grad_norm": 0.3812708258628845,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 3038
+    },
+    {
+      "epoch": 0.34908965596462005,
+      "grad_norm": 0.402920126914978,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 3039
+    },
+    {
+      "epoch": 0.34920452587444717,
+      "grad_norm": 0.3769501745700836,
+      "learning_rate": 0.0001,
+      "loss": 1.6784,
+      "step": 3040
+    },
+    {
+      "epoch": 0.3493193957842743,
+      "grad_norm": 0.3954136371612549,
+      "learning_rate": 0.0001,
+      "loss": 1.7905,
+      "step": 3041
+    },
+    {
+      "epoch": 0.3494342656941014,
+      "grad_norm": 0.36639055609703064,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 3042
+    },
+    {
+      "epoch": 0.34954913560392853,
+      "grad_norm": 0.42278578877449036,
+      "learning_rate": 0.0001,
+      "loss": 1.9002,
+      "step": 3043
+    },
+    {
+      "epoch": 0.34966400551375565,
+      "grad_norm": 0.37817153334617615,
+      "learning_rate": 0.0001,
+      "loss": 1.4365,
+      "step": 3044
+    },
+    {
+      "epoch": 0.34977887542358277,
+      "grad_norm": 0.3958953320980072,
+      "learning_rate": 0.0001,
+      "loss": 1.8305,
+      "step": 3045
+    },
+    {
+      "epoch": 0.3498937453334099,
+      "grad_norm": 0.3798516094684601,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 3046
+    },
+    {
+      "epoch": 0.350008615243237,
+      "grad_norm": 0.37097060680389404,
+      "learning_rate": 0.0001,
+      "loss": 1.4808,
+      "step": 3047
+    },
+    {
+      "epoch": 0.35012348515306413,
+      "grad_norm": 0.3970964848995209,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 3048
+    },
+    {
+      "epoch": 0.35023835506289125,
+      "grad_norm": 0.3985350728034973,
+      "learning_rate": 0.0001,
+      "loss": 1.9372,
+      "step": 3049
+    },
+    {
+      "epoch": 0.35035322497271837,
+      "grad_norm": 0.3665081262588501,
+      "learning_rate": 0.0001,
+      "loss": 1.4447,
+      "step": 3050
+    },
+    {
+      "epoch": 0.35046809488254554,
+      "grad_norm": 0.3866146504878998,
+      "learning_rate": 0.0001,
+      "loss": 1.6509,
+      "step": 3051
+    },
+    {
+      "epoch": 0.35058296479237266,
+      "grad_norm": 0.37819328904151917,
+      "learning_rate": 0.0001,
+      "loss": 1.6386,
+      "step": 3052
+    },
+    {
+      "epoch": 0.3506978347021998,
+      "grad_norm": 0.3822932243347168,
+      "learning_rate": 0.0001,
+      "loss": 1.6757,
+      "step": 3053
+    },
+    {
+      "epoch": 0.3508127046120269,
+      "grad_norm": 0.3891104757785797,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 3054
+    },
+    {
+      "epoch": 0.350927574521854,
+      "grad_norm": 0.40191301703453064,
+      "learning_rate": 0.0001,
+      "loss": 1.8016,
+      "step": 3055
+    },
+    {
+      "epoch": 0.35104244443168114,
+      "grad_norm": 0.40188851952552795,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 3056
+    },
+    {
+      "epoch": 0.35115731434150826,
+      "grad_norm": 0.3913547694683075,
+      "learning_rate": 0.0001,
+      "loss": 1.5879,
+      "step": 3057
+    },
+    {
+      "epoch": 0.3512721842513354,
+      "grad_norm": 0.410559743642807,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 3058
+    },
+    {
+      "epoch": 0.3513870541611625,
+      "grad_norm": 0.3729799687862396,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 3059
+    },
+    {
+      "epoch": 0.3515019240709896,
+      "grad_norm": 0.39681947231292725,
+      "learning_rate": 0.0001,
+      "loss": 1.7177,
+      "step": 3060
+    },
+    {
+      "epoch": 0.35161679398081674,
+      "grad_norm": 0.38647258281707764,
+      "learning_rate": 0.0001,
+      "loss": 1.8458,
+      "step": 3061
+    },
+    {
+      "epoch": 0.35173166389064386,
+      "grad_norm": 0.3754447102546692,
+      "learning_rate": 0.0001,
+      "loss": 1.78,
+      "step": 3062
+    },
+    {
+      "epoch": 0.351846533800471,
+      "grad_norm": 0.35854676365852356,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 3063
+    },
+    {
+      "epoch": 0.3519614037102981,
+      "grad_norm": 0.3889663815498352,
+      "learning_rate": 0.0001,
+      "loss": 1.8046,
+      "step": 3064
+    },
+    {
+      "epoch": 0.3520762736201252,
+      "grad_norm": 0.36157429218292236,
+      "learning_rate": 0.0001,
+      "loss": 1.3026,
+      "step": 3065
+    },
+    {
+      "epoch": 0.35219114352995234,
+      "grad_norm": 0.44542935490608215,
+      "learning_rate": 0.0001,
+      "loss": 1.6889,
+      "step": 3066
+    },
+    {
+      "epoch": 0.35230601343977946,
+      "grad_norm": 0.3814290463924408,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 3067
+    },
+    {
+      "epoch": 0.3524208833496066,
+      "grad_norm": 0.36910781264305115,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 3068
+    },
+    {
+      "epoch": 0.3525357532594337,
+      "grad_norm": 0.4063052535057068,
+      "learning_rate": 0.0001,
+      "loss": 1.6937,
+      "step": 3069
+    },
+    {
+      "epoch": 0.3526506231692608,
+      "grad_norm": 0.3729820251464844,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 3070
+    },
+    {
+      "epoch": 0.35276549307908794,
+      "grad_norm": 0.3980967700481415,
+      "learning_rate": 0.0001,
+      "loss": 1.6192,
+      "step": 3071
+    },
+    {
+      "epoch": 0.35288036298891506,
+      "grad_norm": 0.3677471876144409,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 3072
+    },
+    {
+      "epoch": 0.3529952328987422,
+      "grad_norm": 0.38200077414512634,
+      "learning_rate": 0.0001,
+      "loss": 1.6235,
+      "step": 3073
+    },
+    {
+      "epoch": 0.3531101028085693,
+      "grad_norm": 0.37423157691955566,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 3074
+    },
+    {
+      "epoch": 0.3532249727183964,
+      "grad_norm": 0.37253043055534363,
+      "learning_rate": 0.0001,
+      "loss": 1.6977,
+      "step": 3075
+    },
+    {
+      "epoch": 0.35333984262822354,
+      "grad_norm": 0.3927091360092163,
+      "learning_rate": 0.0001,
+      "loss": 1.7231,
+      "step": 3076
+    },
+    {
+      "epoch": 0.35345471253805066,
+      "grad_norm": 0.36988648772239685,
+      "learning_rate": 0.0001,
+      "loss": 1.506,
+      "step": 3077
+    },
+    {
+      "epoch": 0.3535695824478778,
+      "grad_norm": 0.38558465242385864,
+      "learning_rate": 0.0001,
+      "loss": 1.7394,
+      "step": 3078
+    },
+    {
+      "epoch": 0.3536844523577049,
+      "grad_norm": 0.44665199518203735,
+      "learning_rate": 0.0001,
+      "loss": 1.6499,
+      "step": 3079
+    },
+    {
+      "epoch": 0.353799322267532,
+      "grad_norm": 0.3623279929161072,
+      "learning_rate": 0.0001,
+      "loss": 1.4641,
+      "step": 3080
+    },
+    {
+      "epoch": 0.35391419217735914,
+      "grad_norm": 0.3726842701435089,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 3081
+    },
+    {
+      "epoch": 0.35402906208718626,
+      "grad_norm": 0.3852822184562683,
+      "learning_rate": 0.0001,
+      "loss": 1.7239,
+      "step": 3082
+    },
+    {
+      "epoch": 0.3541439319970134,
+      "grad_norm": 0.3527339994907379,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 3083
+    },
+    {
+      "epoch": 0.3542588019068405,
+      "grad_norm": 0.3964180052280426,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 3084
+    },
+    {
+      "epoch": 0.3543736718166676,
+      "grad_norm": 0.37189149856567383,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 3085
+    },
+    {
+      "epoch": 0.35448854172649474,
+      "grad_norm": 0.3664330840110779,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 3086
+    },
+    {
+      "epoch": 0.35460341163632186,
+      "grad_norm": 0.43963682651519775,
+      "learning_rate": 0.0001,
+      "loss": 1.8784,
+      "step": 3087
+    },
+    {
+      "epoch": 0.354718281546149,
+      "grad_norm": 0.41229376196861267,
+      "learning_rate": 0.0001,
+      "loss": 2.0028,
+      "step": 3088
+    },
+    {
+      "epoch": 0.3548331514559761,
+      "grad_norm": 0.37985897064208984,
+      "learning_rate": 0.0001,
+      "loss": 1.522,
+      "step": 3089
+    },
+    {
+      "epoch": 0.3549480213658032,
+      "grad_norm": 0.4061013162136078,
+      "learning_rate": 0.0001,
+      "loss": 1.767,
+      "step": 3090
+    },
+    {
+      "epoch": 0.35506289127563034,
+      "grad_norm": 0.4098246395587921,
+      "learning_rate": 0.0001,
+      "loss": 1.6884,
+      "step": 3091
+    },
+    {
+      "epoch": 0.35517776118545746,
+      "grad_norm": 0.4013693630695343,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 3092
+    },
+    {
+      "epoch": 0.3552926310952846,
+      "grad_norm": 0.3519565463066101,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 3093
+    },
+    {
+      "epoch": 0.3554075010051117,
+      "grad_norm": 0.3677949011325836,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 3094
+    },
+    {
+      "epoch": 0.3555223709149388,
+      "grad_norm": 0.38344481587409973,
+      "learning_rate": 0.0001,
+      "loss": 1.7043,
+      "step": 3095
+    },
+    {
+      "epoch": 0.35563724082476594,
+      "grad_norm": 0.3987760543823242,
+      "learning_rate": 0.0001,
+      "loss": 1.5873,
+      "step": 3096
+    },
+    {
+      "epoch": 0.35575211073459306,
+      "grad_norm": 0.37182289361953735,
+      "learning_rate": 0.0001,
+      "loss": 1.5454,
+      "step": 3097
+    },
+    {
+      "epoch": 0.3558669806444202,
+      "grad_norm": 0.36989521980285645,
+      "learning_rate": 0.0001,
+      "loss": 1.6257,
+      "step": 3098
+    },
+    {
+      "epoch": 0.3559818505542473,
+      "grad_norm": 0.38265496492385864,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 3099
+    },
+    {
+      "epoch": 0.3560967204640744,
+      "grad_norm": 0.4006288945674896,
+      "learning_rate": 0.0001,
+      "loss": 1.787,
+      "step": 3100
+    },
+    {
+      "epoch": 0.35621159037390154,
+      "grad_norm": 0.38467901945114136,
+      "learning_rate": 0.0001,
+      "loss": 1.639,
+      "step": 3101
+    },
+    {
+      "epoch": 0.35632646028372866,
+      "grad_norm": 0.3911599814891815,
+      "learning_rate": 0.0001,
+      "loss": 1.6307,
+      "step": 3102
+    },
+    {
+      "epoch": 0.3564413301935558,
+      "grad_norm": 0.36751917004585266,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 3103
+    },
+    {
+      "epoch": 0.3565562001033829,
+      "grad_norm": 0.3881228566169739,
+      "learning_rate": 0.0001,
+      "loss": 1.7045,
+      "step": 3104
+    },
+    {
+      "epoch": 0.35667107001321,
+      "grad_norm": 0.36318087577819824,
+      "learning_rate": 0.0001,
+      "loss": 1.4927,
+      "step": 3105
+    },
+    {
+      "epoch": 0.35678593992303714,
+      "grad_norm": 0.3672310709953308,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 3106
+    },
+    {
+      "epoch": 0.35690080983286426,
+      "grad_norm": 0.35635697841644287,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 3107
+    },
+    {
+      "epoch": 0.3570156797426914,
+      "grad_norm": 0.38991066813468933,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 3108
+    },
+    {
+      "epoch": 0.3571305496525185,
+      "grad_norm": 0.3975917100906372,
+      "learning_rate": 0.0001,
+      "loss": 1.6502,
+      "step": 3109
+    },
+    {
+      "epoch": 0.3572454195623456,
+      "grad_norm": 0.39171603322029114,
+      "learning_rate": 0.0001,
+      "loss": 1.7254,
+      "step": 3110
+    },
+    {
+      "epoch": 0.35736028947217274,
+      "grad_norm": 0.39555254578590393,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 3111
+    },
+    {
+      "epoch": 0.35747515938199986,
+      "grad_norm": 0.36104774475097656,
+      "learning_rate": 0.0001,
+      "loss": 1.4724,
+      "step": 3112
+    },
+    {
+      "epoch": 0.357590029291827,
+      "grad_norm": 0.38759103417396545,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 3113
+    },
+    {
+      "epoch": 0.3577048992016541,
+      "grad_norm": 0.4347061216831207,
+      "learning_rate": 0.0001,
+      "loss": 1.8388,
+      "step": 3114
+    },
+    {
+      "epoch": 0.3578197691114812,
+      "grad_norm": 0.36423662304878235,
+      "learning_rate": 0.0001,
+      "loss": 1.7236,
+      "step": 3115
+    },
+    {
+      "epoch": 0.35793463902130834,
+      "grad_norm": 0.4022355377674103,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 3116
+    },
+    {
+      "epoch": 0.35804950893113546,
+      "grad_norm": 0.3823007047176361,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 3117
+    },
+    {
+      "epoch": 0.3581643788409626,
+      "grad_norm": 0.3590918481349945,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 3118
+    },
+    {
+      "epoch": 0.35827924875078976,
+      "grad_norm": 0.3881797790527344,
+      "learning_rate": 0.0001,
+      "loss": 1.7983,
+      "step": 3119
+    },
+    {
+      "epoch": 0.3583941186606169,
+      "grad_norm": 0.41121959686279297,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 3120
+    },
+    {
+      "epoch": 0.358508988570444,
+      "grad_norm": 0.37650343775749207,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 3121
+    },
+    {
+      "epoch": 0.3586238584802711,
+      "grad_norm": 0.4044298827648163,
+      "learning_rate": 0.0001,
+      "loss": 1.715,
+      "step": 3122
+    },
+    {
+      "epoch": 0.35873872839009824,
+      "grad_norm": 0.3791009485721588,
+      "learning_rate": 0.0001,
+      "loss": 1.6326,
+      "step": 3123
+    },
+    {
+      "epoch": 0.35885359829992536,
+      "grad_norm": 0.425060898065567,
+      "learning_rate": 0.0001,
+      "loss": 1.8154,
+      "step": 3124
+    },
+    {
+      "epoch": 0.3589684682097525,
+      "grad_norm": 0.3991539776325226,
+      "learning_rate": 0.0001,
+      "loss": 1.7491,
+      "step": 3125
+    },
+    {
+      "epoch": 0.3590833381195796,
+      "grad_norm": 0.40489476919174194,
+      "learning_rate": 0.0001,
+      "loss": 1.8376,
+      "step": 3126
+    },
+    {
+      "epoch": 0.3591982080294067,
+      "grad_norm": 0.40158843994140625,
+      "learning_rate": 0.0001,
+      "loss": 1.818,
+      "step": 3127
+    },
+    {
+      "epoch": 0.35931307793923384,
+      "grad_norm": 0.4292317032814026,
+      "learning_rate": 0.0001,
+      "loss": 1.8346,
+      "step": 3128
+    },
+    {
+      "epoch": 0.35942794784906096,
+      "grad_norm": 0.4303164780139923,
+      "learning_rate": 0.0001,
+      "loss": 1.9145,
+      "step": 3129
+    },
+    {
+      "epoch": 0.3595428177588881,
+      "grad_norm": 0.406921923160553,
+      "learning_rate": 0.0001,
+      "loss": 1.7952,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3596576876687152,
+      "grad_norm": 0.39321935176849365,
+      "learning_rate": 0.0001,
+      "loss": 1.8008,
+      "step": 3131
+    },
+    {
+      "epoch": 0.3597725575785423,
+      "grad_norm": 0.36769285798072815,
+      "learning_rate": 0.0001,
+      "loss": 1.7165,
+      "step": 3132
+    },
+    {
+      "epoch": 0.35988742748836944,
+      "grad_norm": 0.37966418266296387,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 3133
+    },
+    {
+      "epoch": 0.36000229739819656,
+      "grad_norm": 0.4449335038661957,
+      "learning_rate": 0.0001,
+      "loss": 1.8674,
+      "step": 3134
+    },
+    {
+      "epoch": 0.3601171673080237,
+      "grad_norm": 0.37845826148986816,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 3135
+    },
+    {
+      "epoch": 0.3602320372178508,
+      "grad_norm": 0.43711939454078674,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 3136
+    },
+    {
+      "epoch": 0.3603469071276779,
+      "grad_norm": 0.3823258876800537,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 3137
+    },
+    {
+      "epoch": 0.36046177703750504,
+      "grad_norm": 0.38525834679603577,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 3138
+    },
+    {
+      "epoch": 0.36057664694733216,
+      "grad_norm": 0.38904234766960144,
+      "learning_rate": 0.0001,
+      "loss": 1.5454,
+      "step": 3139
+    },
+    {
+      "epoch": 0.3606915168571593,
+      "grad_norm": 0.39216116070747375,
+      "learning_rate": 0.0001,
+      "loss": 1.7976,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3608063867669864,
+      "grad_norm": 0.3699231743812561,
+      "learning_rate": 0.0001,
+      "loss": 1.5785,
+      "step": 3141
+    },
+    {
+      "epoch": 0.3609212566768135,
+      "grad_norm": 0.3960736393928528,
+      "learning_rate": 0.0001,
+      "loss": 1.6743,
+      "step": 3142
+    },
+    {
+      "epoch": 0.36103612658664064,
+      "grad_norm": 0.37303784489631653,
+      "learning_rate": 0.0001,
+      "loss": 1.415,
+      "step": 3143
+    },
+    {
+      "epoch": 0.36115099649646776,
+      "grad_norm": 0.4160480797290802,
+      "learning_rate": 0.0001,
+      "loss": 1.7466,
+      "step": 3144
+    },
+    {
+      "epoch": 0.3612658664062949,
+      "grad_norm": 0.4043230712413788,
+      "learning_rate": 0.0001,
+      "loss": 1.7383,
+      "step": 3145
+    },
+    {
+      "epoch": 0.361380736316122,
+      "grad_norm": 0.3847825825214386,
+      "learning_rate": 0.0001,
+      "loss": 1.6484,
+      "step": 3146
+    },
+    {
+      "epoch": 0.3614956062259491,
+      "grad_norm": 0.40463775396347046,
+      "learning_rate": 0.0001,
+      "loss": 1.6482,
+      "step": 3147
+    },
+    {
+      "epoch": 0.36161047613577624,
+      "grad_norm": 0.3750319480895996,
+      "learning_rate": 0.0001,
+      "loss": 1.6921,
+      "step": 3148
+    },
+    {
+      "epoch": 0.36172534604560336,
+      "grad_norm": 0.37009164690971375,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 3149
+    },
+    {
+      "epoch": 0.3618402159554305,
+      "grad_norm": 0.34993571043014526,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 3150
+    },
+    {
+      "epoch": 0.3619550858652576,
+      "grad_norm": 0.3800502419471741,
+      "learning_rate": 0.0001,
+      "loss": 1.6974,
+      "step": 3151
+    },
+    {
+      "epoch": 0.3620699557750847,
+      "grad_norm": 0.3825697600841522,
+      "learning_rate": 0.0001,
+      "loss": 1.4917,
+      "step": 3152
+    },
+    {
+      "epoch": 0.36218482568491184,
+      "grad_norm": 0.384375661611557,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 3153
+    },
+    {
+      "epoch": 0.36229969559473896,
+      "grad_norm": 0.4005104899406433,
+      "learning_rate": 0.0001,
+      "loss": 1.7287,
+      "step": 3154
+    },
+    {
+      "epoch": 0.3624145655045661,
+      "grad_norm": 0.3781236708164215,
+      "learning_rate": 0.0001,
+      "loss": 1.4272,
+      "step": 3155
+    },
+    {
+      "epoch": 0.3625294354143932,
+      "grad_norm": 0.37591099739074707,
+      "learning_rate": 0.0001,
+      "loss": 1.5723,
+      "step": 3156
+    },
+    {
+      "epoch": 0.3626443053242203,
+      "grad_norm": 0.3829016089439392,
+      "learning_rate": 0.0001,
+      "loss": 1.4127,
+      "step": 3157
+    },
+    {
+      "epoch": 0.36275917523404744,
+      "grad_norm": 0.4181736707687378,
+      "learning_rate": 0.0001,
+      "loss": 1.8398,
+      "step": 3158
+    },
+    {
+      "epoch": 0.36287404514387456,
+      "grad_norm": 0.3961413502693176,
+      "learning_rate": 0.0001,
+      "loss": 1.651,
+      "step": 3159
+    },
+    {
+      "epoch": 0.3629889150537017,
+      "grad_norm": 0.3490610420703888,
+      "learning_rate": 0.0001,
+      "loss": 1.6495,
+      "step": 3160
+    },
+    {
+      "epoch": 0.3631037849635288,
+      "grad_norm": 0.45399048924446106,
+      "learning_rate": 0.0001,
+      "loss": 1.7701,
+      "step": 3161
+    },
+    {
+      "epoch": 0.3632186548733559,
+      "grad_norm": 0.3794902563095093,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 3162
+    },
+    {
+      "epoch": 0.36333352478318304,
+      "grad_norm": 0.41195976734161377,
+      "learning_rate": 0.0001,
+      "loss": 1.7688,
+      "step": 3163
+    },
+    {
+      "epoch": 0.36344839469301016,
+      "grad_norm": 0.4826469123363495,
+      "learning_rate": 0.0001,
+      "loss": 1.5523,
+      "step": 3164
+    },
+    {
+      "epoch": 0.3635632646028373,
+      "grad_norm": 0.39594000577926636,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 3165
+    },
+    {
+      "epoch": 0.3636781345126644,
+      "grad_norm": 0.41155776381492615,
+      "learning_rate": 0.0001,
+      "loss": 1.7261,
+      "step": 3166
+    },
+    {
+      "epoch": 0.3637930044224915,
+      "grad_norm": 0.413484126329422,
+      "learning_rate": 0.0001,
+      "loss": 1.7269,
+      "step": 3167
+    },
+    {
+      "epoch": 0.36390787433231864,
+      "grad_norm": 0.41094526648521423,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 3168
+    },
+    {
+      "epoch": 0.36402274424214576,
+      "grad_norm": 0.4104525148868561,
+      "learning_rate": 0.0001,
+      "loss": 1.7229,
+      "step": 3169
+    },
+    {
+      "epoch": 0.3641376141519729,
+      "grad_norm": 0.40125396847724915,
+      "learning_rate": 0.0001,
+      "loss": 1.7518,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3642524840618,
+      "grad_norm": 0.37167468667030334,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 3171
+    },
+    {
+      "epoch": 0.3643673539716271,
+      "grad_norm": 0.4034722149372101,
+      "learning_rate": 0.0001,
+      "loss": 1.6484,
+      "step": 3172
+    },
+    {
+      "epoch": 0.36448222388145424,
+      "grad_norm": 0.3839128613471985,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 3173
+    },
+    {
+      "epoch": 0.36459709379128136,
+      "grad_norm": 0.36040592193603516,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 3174
+    },
+    {
+      "epoch": 0.3647119637011085,
+      "grad_norm": 0.3928232192993164,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 3175
+    },
+    {
+      "epoch": 0.3648268336109356,
+      "grad_norm": 0.3935447931289673,
+      "learning_rate": 0.0001,
+      "loss": 1.6686,
+      "step": 3176
+    },
+    {
+      "epoch": 0.3649417035207627,
+      "grad_norm": 0.37371543049812317,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 3177
+    },
+    {
+      "epoch": 0.36505657343058984,
+      "grad_norm": 0.40319526195526123,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 3178
+    },
+    {
+      "epoch": 0.36517144334041696,
+      "grad_norm": 0.42261865735054016,
+      "learning_rate": 0.0001,
+      "loss": 1.5549,
+      "step": 3179
+    },
+    {
+      "epoch": 0.3652863132502441,
+      "grad_norm": 0.4181409180164337,
+      "learning_rate": 0.0001,
+      "loss": 1.6968,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3654011831600712,
+      "grad_norm": 0.420463889837265,
+      "learning_rate": 0.0001,
+      "loss": 1.8145,
+      "step": 3181
+    },
+    {
+      "epoch": 0.3655160530698983,
+      "grad_norm": 0.3919375538825989,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 3182
+    },
+    {
+      "epoch": 0.36563092297972544,
+      "grad_norm": 0.6190922260284424,
+      "learning_rate": 0.0001,
+      "loss": 1.6232,
+      "step": 3183
+    },
+    {
+      "epoch": 0.36574579288955256,
+      "grad_norm": 0.3965461254119873,
+      "learning_rate": 0.0001,
+      "loss": 1.6968,
+      "step": 3184
+    },
+    {
+      "epoch": 0.3658606627993797,
+      "grad_norm": 0.3818904757499695,
+      "learning_rate": 0.0001,
+      "loss": 1.7041,
+      "step": 3185
+    },
+    {
+      "epoch": 0.3659755327092068,
+      "grad_norm": 0.38142475485801697,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 3186
+    },
+    {
+      "epoch": 0.3660904026190339,
+      "grad_norm": 0.4176258444786072,
+      "learning_rate": 0.0001,
+      "loss": 1.678,
+      "step": 3187
+    },
+    {
+      "epoch": 0.3662052725288611,
+      "grad_norm": 0.3882141709327698,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 3188
+    },
+    {
+      "epoch": 0.3663201424386882,
+      "grad_norm": 0.3725062608718872,
+      "learning_rate": 0.0001,
+      "loss": 1.5094,
+      "step": 3189
+    },
+    {
+      "epoch": 0.36643501234851533,
+      "grad_norm": 0.4642198979854584,
+      "learning_rate": 0.0001,
+      "loss": 2.0719,
+      "step": 3190
+    },
+    {
+      "epoch": 0.36654988225834245,
+      "grad_norm": 0.4120834171772003,
+      "learning_rate": 0.0001,
+      "loss": 1.475,
+      "step": 3191
+    },
+    {
+      "epoch": 0.36666475216816957,
+      "grad_norm": 0.44245001673698425,
+      "learning_rate": 0.0001,
+      "loss": 1.818,
+      "step": 3192
+    },
+    {
+      "epoch": 0.3667796220779967,
+      "grad_norm": 0.3757948875427246,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 3193
+    },
+    {
+      "epoch": 0.3668944919878238,
+      "grad_norm": 0.386857271194458,
+      "learning_rate": 0.0001,
+      "loss": 1.7447,
+      "step": 3194
+    },
+    {
+      "epoch": 0.36700936189765093,
+      "grad_norm": 0.37118127942085266,
+      "learning_rate": 0.0001,
+      "loss": 1.7593,
+      "step": 3195
+    },
+    {
+      "epoch": 0.36712423180747805,
+      "grad_norm": 0.39029768109321594,
+      "learning_rate": 0.0001,
+      "loss": 1.6951,
+      "step": 3196
+    },
+    {
+      "epoch": 0.36723910171730517,
+      "grad_norm": 0.4252242147922516,
+      "learning_rate": 0.0001,
+      "loss": 1.817,
+      "step": 3197
+    },
+    {
+      "epoch": 0.3673539716271323,
+      "grad_norm": 0.39361050724983215,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 3198
+    },
+    {
+      "epoch": 0.3674688415369594,
+      "grad_norm": 0.4176224172115326,
+      "learning_rate": 0.0001,
+      "loss": 1.8173,
+      "step": 3199
+    },
+    {
+      "epoch": 0.36758371144678653,
+      "grad_norm": 0.42337673902511597,
+      "learning_rate": 0.0001,
+      "loss": 1.6874,
+      "step": 3200
+    },
+    {
+      "epoch": 0.36769858135661365,
+      "grad_norm": 0.3670978546142578,
+      "learning_rate": 0.0001,
+      "loss": 1.6705,
+      "step": 3201
+    },
+    {
+      "epoch": 0.36781345126644077,
+      "grad_norm": 0.3871794044971466,
+      "learning_rate": 0.0001,
+      "loss": 1.7955,
+      "step": 3202
+    },
+    {
+      "epoch": 0.3679283211762679,
+      "grad_norm": 0.39479079842567444,
+      "learning_rate": 0.0001,
+      "loss": 1.8078,
+      "step": 3203
+    },
+    {
+      "epoch": 0.368043191086095,
+      "grad_norm": 0.405738890171051,
+      "learning_rate": 0.0001,
+      "loss": 1.8309,
+      "step": 3204
+    },
+    {
+      "epoch": 0.36815806099592213,
+      "grad_norm": 0.37141597270965576,
+      "learning_rate": 0.0001,
+      "loss": 1.5382,
+      "step": 3205
+    },
+    {
+      "epoch": 0.36827293090574925,
+      "grad_norm": 0.4640069007873535,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 3206
+    },
+    {
+      "epoch": 0.36838780081557637,
+      "grad_norm": 0.39117664098739624,
+      "learning_rate": 0.0001,
+      "loss": 1.8055,
+      "step": 3207
+    },
+    {
+      "epoch": 0.3685026707254035,
+      "grad_norm": 0.3931315839290619,
+      "learning_rate": 0.0001,
+      "loss": 1.6636,
+      "step": 3208
+    },
+    {
+      "epoch": 0.3686175406352306,
+      "grad_norm": 0.406044065952301,
+      "learning_rate": 0.0001,
+      "loss": 1.7652,
+      "step": 3209
+    },
+    {
+      "epoch": 0.36873241054505773,
+      "grad_norm": 0.37992948293685913,
+      "learning_rate": 0.0001,
+      "loss": 1.5028,
+      "step": 3210
+    },
+    {
+      "epoch": 0.36884728045488485,
+      "grad_norm": 0.4231666326522827,
+      "learning_rate": 0.0001,
+      "loss": 1.7238,
+      "step": 3211
+    },
+    {
+      "epoch": 0.36896215036471197,
+      "grad_norm": 0.4150424599647522,
+      "learning_rate": 0.0001,
+      "loss": 1.8108,
+      "step": 3212
+    },
+    {
+      "epoch": 0.3690770202745391,
+      "grad_norm": 0.3594525158405304,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 3213
+    },
+    {
+      "epoch": 0.3691918901843662,
+      "grad_norm": 0.381798654794693,
+      "learning_rate": 0.0001,
+      "loss": 1.6687,
+      "step": 3214
+    },
+    {
+      "epoch": 0.36930676009419333,
+      "grad_norm": 0.38499388098716736,
+      "learning_rate": 0.0001,
+      "loss": 1.7469,
+      "step": 3215
+    },
+    {
+      "epoch": 0.36942163000402045,
+      "grad_norm": 0.4006686210632324,
+      "learning_rate": 0.0001,
+      "loss": 1.7326,
+      "step": 3216
+    },
+    {
+      "epoch": 0.36953649991384757,
+      "grad_norm": 0.396440327167511,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 3217
+    },
+    {
+      "epoch": 0.3696513698236747,
+      "grad_norm": 0.3961047828197479,
+      "learning_rate": 0.0001,
+      "loss": 1.762,
+      "step": 3218
+    },
+    {
+      "epoch": 0.3697662397335018,
+      "grad_norm": 0.4057855010032654,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 3219
+    },
+    {
+      "epoch": 0.36988110964332893,
+      "grad_norm": 0.4407478868961334,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 3220
+    },
+    {
+      "epoch": 0.36999597955315605,
+      "grad_norm": 0.37952542304992676,
+      "learning_rate": 0.0001,
+      "loss": 1.7687,
+      "step": 3221
+    },
+    {
+      "epoch": 0.37011084946298317,
+      "grad_norm": 0.3981403708457947,
+      "learning_rate": 0.0001,
+      "loss": 1.6591,
+      "step": 3222
+    },
+    {
+      "epoch": 0.3702257193728103,
+      "grad_norm": 0.3874553143978119,
+      "learning_rate": 0.0001,
+      "loss": 1.6144,
+      "step": 3223
+    },
+    {
+      "epoch": 0.3703405892826374,
+      "grad_norm": 0.40902000665664673,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 3224
+    },
+    {
+      "epoch": 0.37045545919246453,
+      "grad_norm": 0.42551088333129883,
+      "learning_rate": 0.0001,
+      "loss": 1.7355,
+      "step": 3225
+    },
+    {
+      "epoch": 0.37057032910229165,
+      "grad_norm": 0.3739902079105377,
+      "learning_rate": 0.0001,
+      "loss": 1.4641,
+      "step": 3226
+    },
+    {
+      "epoch": 0.37068519901211877,
+      "grad_norm": 0.41387036442756653,
+      "learning_rate": 0.0001,
+      "loss": 1.7716,
+      "step": 3227
+    },
+    {
+      "epoch": 0.3708000689219459,
+      "grad_norm": 0.393655002117157,
+      "learning_rate": 0.0001,
+      "loss": 1.5306,
+      "step": 3228
+    },
+    {
+      "epoch": 0.370914938831773,
+      "grad_norm": 0.39244723320007324,
+      "learning_rate": 0.0001,
+      "loss": 1.7408,
+      "step": 3229
+    },
+    {
+      "epoch": 0.37102980874160013,
+      "grad_norm": 0.3959055542945862,
+      "learning_rate": 0.0001,
+      "loss": 1.8051,
+      "step": 3230
+    },
+    {
+      "epoch": 0.37114467865142725,
+      "grad_norm": 0.39641687273979187,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 3231
+    },
+    {
+      "epoch": 0.37125954856125437,
+      "grad_norm": 0.37749791145324707,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 3232
+    },
+    {
+      "epoch": 0.3713744184710815,
+      "grad_norm": 0.40520602464675903,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 3233
+    },
+    {
+      "epoch": 0.3714892883809086,
+      "grad_norm": 0.37299278378486633,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 3234
+    },
+    {
+      "epoch": 0.37160415829073573,
+      "grad_norm": 0.41653332114219666,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 3235
+    },
+    {
+      "epoch": 0.37171902820056285,
+      "grad_norm": 0.4918358623981476,
+      "learning_rate": 0.0001,
+      "loss": 1.894,
+      "step": 3236
+    },
+    {
+      "epoch": 0.37183389811038997,
+      "grad_norm": 0.40961939096450806,
+      "learning_rate": 0.0001,
+      "loss": 1.6471,
+      "step": 3237
+    },
+    {
+      "epoch": 0.3719487680202171,
+      "grad_norm": 0.39822736382484436,
+      "learning_rate": 0.0001,
+      "loss": 1.6876,
+      "step": 3238
+    },
+    {
+      "epoch": 0.3720636379300442,
+      "grad_norm": 0.42709147930145264,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 3239
+    },
+    {
+      "epoch": 0.37217850783987133,
+      "grad_norm": 0.3828563690185547,
+      "learning_rate": 0.0001,
+      "loss": 1.5153,
+      "step": 3240
+    },
+    {
+      "epoch": 0.37229337774969845,
+      "grad_norm": 0.3979793190956116,
+      "learning_rate": 0.0001,
+      "loss": 1.7632,
+      "step": 3241
+    },
+    {
+      "epoch": 0.37240824765952557,
+      "grad_norm": 0.4186950623989105,
+      "learning_rate": 0.0001,
+      "loss": 1.6498,
+      "step": 3242
+    },
+    {
+      "epoch": 0.3725231175693527,
+      "grad_norm": 0.3847867548465729,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 3243
+    },
+    {
+      "epoch": 0.3726379874791798,
+      "grad_norm": 0.4237745404243469,
+      "learning_rate": 0.0001,
+      "loss": 1.6776,
+      "step": 3244
+    },
+    {
+      "epoch": 0.37275285738900693,
+      "grad_norm": 0.41460272669792175,
+      "learning_rate": 0.0001,
+      "loss": 1.8457,
+      "step": 3245
+    },
+    {
+      "epoch": 0.37286772729883405,
+      "grad_norm": 0.37694281339645386,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 3246
+    },
+    {
+      "epoch": 0.37298259720866117,
+      "grad_norm": 0.3860875368118286,
+      "learning_rate": 0.0001,
+      "loss": 1.7283,
+      "step": 3247
+    },
+    {
+      "epoch": 0.3730974671184883,
+      "grad_norm": 0.4228290319442749,
+      "learning_rate": 0.0001,
+      "loss": 1.9065,
+      "step": 3248
+    },
+    {
+      "epoch": 0.3732123370283154,
+      "grad_norm": 0.4396858811378479,
+      "learning_rate": 0.0001,
+      "loss": 1.8495,
+      "step": 3249
+    },
+    {
+      "epoch": 0.37332720693814253,
+      "grad_norm": 0.44928887486457825,
+      "learning_rate": 0.0001,
+      "loss": 1.9191,
+      "step": 3250
+    },
+    {
+      "epoch": 0.37344207684796965,
+      "grad_norm": 0.4096646010875702,
+      "learning_rate": 0.0001,
+      "loss": 1.7892,
+      "step": 3251
+    },
+    {
+      "epoch": 0.37355694675779677,
+      "grad_norm": 0.36828523874282837,
+      "learning_rate": 0.0001,
+      "loss": 1.396,
+      "step": 3252
+    },
+    {
+      "epoch": 0.3736718166676239,
+      "grad_norm": 0.3781738877296448,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 3253
+    },
+    {
+      "epoch": 0.373786686577451,
+      "grad_norm": 0.3941371738910675,
+      "learning_rate": 0.0001,
+      "loss": 1.7005,
+      "step": 3254
+    },
+    {
+      "epoch": 0.37390155648727813,
+      "grad_norm": 0.3932330310344696,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 3255
+    },
+    {
+      "epoch": 0.3740164263971053,
+      "grad_norm": 0.40510842204093933,
+      "learning_rate": 0.0001,
+      "loss": 1.7448,
+      "step": 3256
+    },
+    {
+      "epoch": 0.3741312963069324,
+      "grad_norm": 0.40375015139579773,
+      "learning_rate": 0.0001,
+      "loss": 1.7276,
+      "step": 3257
+    },
+    {
+      "epoch": 0.37424616621675955,
+      "grad_norm": 0.4124782085418701,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 3258
+    },
+    {
+      "epoch": 0.37436103612658667,
+      "grad_norm": 0.3871046006679535,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 3259
+    },
+    {
+      "epoch": 0.3744759060364138,
+      "grad_norm": 0.40944960713386536,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3745907759462409,
+      "grad_norm": 0.39792561531066895,
+      "learning_rate": 0.0001,
+      "loss": 1.6789,
+      "step": 3261
+    },
+    {
+      "epoch": 0.374705645856068,
+      "grad_norm": 0.38910964131355286,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 3262
+    },
+    {
+      "epoch": 0.37482051576589515,
+      "grad_norm": 0.37065139412879944,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 3263
+    },
+    {
+      "epoch": 0.37493538567572227,
+      "grad_norm": 0.3843998908996582,
+      "learning_rate": 0.0001,
+      "loss": 1.7721,
+      "step": 3264
+    },
+    {
+      "epoch": 0.3750502555855494,
+      "grad_norm": 0.4312693178653717,
+      "learning_rate": 0.0001,
+      "loss": 1.9565,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3751651254953765,
+      "grad_norm": 0.3939458727836609,
+      "learning_rate": 0.0001,
+      "loss": 1.7894,
+      "step": 3266
+    },
+    {
+      "epoch": 0.3752799954052036,
+      "grad_norm": 0.3727447986602783,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 3267
+    },
+    {
+      "epoch": 0.37539486531503075,
+      "grad_norm": 0.3730446994304657,
+      "learning_rate": 0.0001,
+      "loss": 1.5062,
+      "step": 3268
+    },
+    {
+      "epoch": 0.37550973522485787,
+      "grad_norm": 0.4053596258163452,
+      "learning_rate": 0.0001,
+      "loss": 1.7005,
+      "step": 3269
+    },
+    {
+      "epoch": 0.375624605134685,
+      "grad_norm": 0.4294143617153168,
+      "learning_rate": 0.0001,
+      "loss": 1.5863,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3757394750445121,
+      "grad_norm": 0.3998027443885803,
+      "learning_rate": 0.0001,
+      "loss": 1.6598,
+      "step": 3271
+    },
+    {
+      "epoch": 0.3758543449543392,
+      "grad_norm": 0.39624300599098206,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 3272
+    },
+    {
+      "epoch": 0.37596921486416635,
+      "grad_norm": 0.38765525817871094,
+      "learning_rate": 0.0001,
+      "loss": 1.7249,
+      "step": 3273
+    },
+    {
+      "epoch": 0.37608408477399347,
+      "grad_norm": 0.38132765889167786,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 3274
+    },
+    {
+      "epoch": 0.3761989546838206,
+      "grad_norm": 0.386642187833786,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 3275
+    },
+    {
+      "epoch": 0.3763138245936477,
+      "grad_norm": 0.3864053189754486,
+      "learning_rate": 0.0001,
+      "loss": 1.5038,
+      "step": 3276
+    },
+    {
+      "epoch": 0.3764286945034748,
+      "grad_norm": 0.40680447220802307,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 3277
+    },
+    {
+      "epoch": 0.37654356441330195,
+      "grad_norm": 0.38861724734306335,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 3278
+    },
+    {
+      "epoch": 0.37665843432312907,
+      "grad_norm": 0.3894346058368683,
+      "learning_rate": 0.0001,
+      "loss": 1.7633,
+      "step": 3279
+    },
+    {
+      "epoch": 0.3767733042329562,
+      "grad_norm": 0.4006745517253876,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 3280
+    },
+    {
+      "epoch": 0.3768881741427833,
+      "grad_norm": 0.4053511917591095,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 3281
+    },
+    {
+      "epoch": 0.3770030440526104,
+      "grad_norm": 0.44916823506355286,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 3282
+    },
+    {
+      "epoch": 0.37711791396243755,
+      "grad_norm": 0.4060499668121338,
+      "learning_rate": 0.0001,
+      "loss": 1.6009,
+      "step": 3283
+    },
+    {
+      "epoch": 0.37723278387226467,
+      "grad_norm": 0.43672630190849304,
+      "learning_rate": 0.0001,
+      "loss": 1.7468,
+      "step": 3284
+    },
+    {
+      "epoch": 0.3773476537820918,
+      "grad_norm": 0.4144275486469269,
+      "learning_rate": 0.0001,
+      "loss": 1.7347,
+      "step": 3285
+    },
+    {
+      "epoch": 0.3774625236919189,
+      "grad_norm": 0.3954853415489197,
+      "learning_rate": 0.0001,
+      "loss": 1.7747,
+      "step": 3286
+    },
+    {
+      "epoch": 0.377577393601746,
+      "grad_norm": 0.4159530699253082,
+      "learning_rate": 0.0001,
+      "loss": 1.868,
+      "step": 3287
+    },
+    {
+      "epoch": 0.37769226351157315,
+      "grad_norm": 0.4597531855106354,
+      "learning_rate": 0.0001,
+      "loss": 1.6728,
+      "step": 3288
+    },
+    {
+      "epoch": 0.37780713342140027,
+      "grad_norm": 0.3839264214038849,
+      "learning_rate": 0.0001,
+      "loss": 1.4713,
+      "step": 3289
+    },
+    {
+      "epoch": 0.3779220033312274,
+      "grad_norm": 0.40157079696655273,
+      "learning_rate": 0.0001,
+      "loss": 1.7359,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3780368732410545,
+      "grad_norm": 0.38555535674095154,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 3291
+    },
+    {
+      "epoch": 0.3781517431508816,
+      "grad_norm": 0.41350314021110535,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 3292
+    },
+    {
+      "epoch": 0.37826661306070875,
+      "grad_norm": 0.4091131389141083,
+      "learning_rate": 0.0001,
+      "loss": 1.7817,
+      "step": 3293
+    },
+    {
+      "epoch": 0.37838148297053587,
+      "grad_norm": 0.39197850227355957,
+      "learning_rate": 0.0001,
+      "loss": 1.7373,
+      "step": 3294
+    },
+    {
+      "epoch": 0.378496352880363,
+      "grad_norm": 0.3582981824874878,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 3295
+    },
+    {
+      "epoch": 0.3786112227901901,
+      "grad_norm": 0.4097733497619629,
+      "learning_rate": 0.0001,
+      "loss": 1.815,
+      "step": 3296
+    },
+    {
+      "epoch": 0.3787260927000172,
+      "grad_norm": 0.405089408159256,
+      "learning_rate": 0.0001,
+      "loss": 1.7199,
+      "step": 3297
+    },
+    {
+      "epoch": 0.37884096260984435,
+      "grad_norm": 0.3839695155620575,
+      "learning_rate": 0.0001,
+      "loss": 1.7156,
+      "step": 3298
+    },
+    {
+      "epoch": 0.37895583251967146,
+      "grad_norm": 0.4171733260154724,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 3299
+    },
+    {
+      "epoch": 0.3790707024294986,
+      "grad_norm": 0.39771202206611633,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 3300
+    },
+    {
+      "epoch": 0.3791855723393257,
+      "grad_norm": 0.4176054894924164,
+      "learning_rate": 0.0001,
+      "loss": 1.6608,
+      "step": 3301
+    },
+    {
+      "epoch": 0.3793004422491528,
+      "grad_norm": 0.3709241449832916,
+      "learning_rate": 0.0001,
+      "loss": 1.4327,
+      "step": 3302
+    },
+    {
+      "epoch": 0.37941531215897994,
+      "grad_norm": 0.4202117919921875,
+      "learning_rate": 0.0001,
+      "loss": 1.7664,
+      "step": 3303
+    },
+    {
+      "epoch": 0.37953018206880706,
+      "grad_norm": 0.4221265912055969,
+      "learning_rate": 0.0001,
+      "loss": 1.8107,
+      "step": 3304
+    },
+    {
+      "epoch": 0.3796450519786342,
+      "grad_norm": 0.3954116404056549,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 3305
+    },
+    {
+      "epoch": 0.3797599218884613,
+      "grad_norm": 0.3891046643257141,
+      "learning_rate": 0.0001,
+      "loss": 1.4281,
+      "step": 3306
+    },
+    {
+      "epoch": 0.3798747917982884,
+      "grad_norm": 0.4024218022823334,
+      "learning_rate": 0.0001,
+      "loss": 1.7119,
+      "step": 3307
+    },
+    {
+      "epoch": 0.37998966170811554,
+      "grad_norm": 0.45850181579589844,
+      "learning_rate": 0.0001,
+      "loss": 1.7876,
+      "step": 3308
+    },
+    {
+      "epoch": 0.38010453161794266,
+      "grad_norm": 0.4152509868144989,
+      "learning_rate": 0.0001,
+      "loss": 1.847,
+      "step": 3309
+    },
+    {
+      "epoch": 0.3802194015277698,
+      "grad_norm": 0.4059930443763733,
+      "learning_rate": 0.0001,
+      "loss": 1.6807,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3803342714375969,
+      "grad_norm": 0.4039838910102844,
+      "learning_rate": 0.0001,
+      "loss": 1.7222,
+      "step": 3311
+    },
+    {
+      "epoch": 0.380449141347424,
+      "grad_norm": 0.40002793073654175,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 3312
+    },
+    {
+      "epoch": 0.38056401125725114,
+      "grad_norm": 0.3877595067024231,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 3313
+    },
+    {
+      "epoch": 0.38067888116707826,
+      "grad_norm": 0.41167452931404114,
+      "learning_rate": 0.0001,
+      "loss": 1.7559,
+      "step": 3314
+    },
+    {
+      "epoch": 0.3807937510769054,
+      "grad_norm": 0.3951496481895447,
+      "learning_rate": 0.0001,
+      "loss": 1.6941,
+      "step": 3315
+    },
+    {
+      "epoch": 0.3809086209867325,
+      "grad_norm": 0.4057624936103821,
+      "learning_rate": 0.0001,
+      "loss": 1.7384,
+      "step": 3316
+    },
+    {
+      "epoch": 0.3810234908965596,
+      "grad_norm": 0.41710686683654785,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 3317
+    },
+    {
+      "epoch": 0.38113836080638674,
+      "grad_norm": 0.38492968678474426,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 3318
+    },
+    {
+      "epoch": 0.38125323071621386,
+      "grad_norm": 0.3713145852088928,
+      "learning_rate": 0.0001,
+      "loss": 1.4979,
+      "step": 3319
+    },
+    {
+      "epoch": 0.381368100626041,
+      "grad_norm": 0.3968343734741211,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 3320
+    },
+    {
+      "epoch": 0.3814829705358681,
+      "grad_norm": 0.3992241621017456,
+      "learning_rate": 0.0001,
+      "loss": 1.6667,
+      "step": 3321
+    },
+    {
+      "epoch": 0.3815978404456952,
+      "grad_norm": 0.3984096050262451,
+      "learning_rate": 0.0001,
+      "loss": 1.6411,
+      "step": 3322
+    },
+    {
+      "epoch": 0.38171271035552234,
+      "grad_norm": 0.4242939352989197,
+      "learning_rate": 0.0001,
+      "loss": 1.8618,
+      "step": 3323
+    },
+    {
+      "epoch": 0.38182758026534946,
+      "grad_norm": 0.4698657691478729,
+      "learning_rate": 0.0001,
+      "loss": 1.7804,
+      "step": 3324
+    },
+    {
+      "epoch": 0.38194245017517664,
+      "grad_norm": 0.3952822685241699,
+      "learning_rate": 0.0001,
+      "loss": 1.762,
+      "step": 3325
+    },
+    {
+      "epoch": 0.38205732008500376,
+      "grad_norm": 0.3968496322631836,
+      "learning_rate": 0.0001,
+      "loss": 1.7494,
+      "step": 3326
+    },
+    {
+      "epoch": 0.3821721899948309,
+      "grad_norm": 0.35102754831314087,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 3327
+    },
+    {
+      "epoch": 0.382287059904658,
+      "grad_norm": 0.3999139070510864,
+      "learning_rate": 0.0001,
+      "loss": 1.422,
+      "step": 3328
+    },
+    {
+      "epoch": 0.3824019298144851,
+      "grad_norm": 0.3803260922431946,
+      "learning_rate": 0.0001,
+      "loss": 1.5255,
+      "step": 3329
+    },
+    {
+      "epoch": 0.38251679972431224,
+      "grad_norm": 0.3881869614124298,
+      "learning_rate": 0.0001,
+      "loss": 1.6423,
+      "step": 3330
+    },
+    {
+      "epoch": 0.38263166963413936,
+      "grad_norm": 0.3832140266895294,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 3331
+    },
+    {
+      "epoch": 0.3827465395439665,
+      "grad_norm": 0.4469684064388275,
+      "learning_rate": 0.0001,
+      "loss": 1.7267,
+      "step": 3332
+    },
+    {
+      "epoch": 0.3828614094537936,
+      "grad_norm": 0.3962031602859497,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 3333
+    },
+    {
+      "epoch": 0.3829762793636207,
+      "grad_norm": 0.408916175365448,
+      "learning_rate": 0.0001,
+      "loss": 1.6637,
+      "step": 3334
+    },
+    {
+      "epoch": 0.38309114927344784,
+      "grad_norm": 0.4158702790737152,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 3335
+    },
+    {
+      "epoch": 0.38320601918327496,
+      "grad_norm": 0.3798253536224365,
+      "learning_rate": 0.0001,
+      "loss": 1.5571,
+      "step": 3336
+    },
+    {
+      "epoch": 0.3833208890931021,
+      "grad_norm": 0.39568817615509033,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 3337
+    },
+    {
+      "epoch": 0.3834357590029292,
+      "grad_norm": 0.38767796754837036,
+      "learning_rate": 0.0001,
+      "loss": 1.6802,
+      "step": 3338
+    },
+    {
+      "epoch": 0.3835506289127563,
+      "grad_norm": 0.3866712749004364,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 3339
+    },
+    {
+      "epoch": 0.38366549882258344,
+      "grad_norm": 0.41397809982299805,
+      "learning_rate": 0.0001,
+      "loss": 1.8039,
+      "step": 3340
+    },
+    {
+      "epoch": 0.38378036873241056,
+      "grad_norm": 0.39398789405822754,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 3341
+    },
+    {
+      "epoch": 0.3838952386422377,
+      "grad_norm": 0.4054300785064697,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 3342
+    },
+    {
+      "epoch": 0.3840101085520648,
+      "grad_norm": 0.4255220890045166,
+      "learning_rate": 0.0001,
+      "loss": 1.9429,
+      "step": 3343
+    },
+    {
+      "epoch": 0.3841249784618919,
+      "grad_norm": 0.4655556082725525,
+      "learning_rate": 0.0001,
+      "loss": 1.7391,
+      "step": 3344
+    },
+    {
+      "epoch": 0.38423984837171904,
+      "grad_norm": 0.43720874190330505,
+      "learning_rate": 0.0001,
+      "loss": 1.7068,
+      "step": 3345
+    },
+    {
+      "epoch": 0.38435471828154616,
+      "grad_norm": 0.443059504032135,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 3346
+    },
+    {
+      "epoch": 0.3844695881913733,
+      "grad_norm": 0.4180503189563751,
+      "learning_rate": 0.0001,
+      "loss": 1.7699,
+      "step": 3347
+    },
+    {
+      "epoch": 0.3845844581012004,
+      "grad_norm": 0.3981085419654846,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 3348
+    },
+    {
+      "epoch": 0.3846993280110275,
+      "grad_norm": 0.38549384474754333,
+      "learning_rate": 0.0001,
+      "loss": 1.5092,
+      "step": 3349
+    },
+    {
+      "epoch": 0.38481419792085464,
+      "grad_norm": 0.4354805052280426,
+      "learning_rate": 0.0001,
+      "loss": 1.7426,
+      "step": 3350
+    },
+    {
+      "epoch": 0.38492906783068176,
+      "grad_norm": 0.39129069447517395,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 3351
+    },
+    {
+      "epoch": 0.3850439377405089,
+      "grad_norm": 0.40617823600769043,
+      "learning_rate": 0.0001,
+      "loss": 1.7721,
+      "step": 3352
+    },
+    {
+      "epoch": 0.385158807650336,
+      "grad_norm": 0.3931029438972473,
+      "learning_rate": 0.0001,
+      "loss": 1.4904,
+      "step": 3353
+    },
+    {
+      "epoch": 0.3852736775601631,
+      "grad_norm": 0.37746840715408325,
+      "learning_rate": 0.0001,
+      "loss": 1.6286,
+      "step": 3354
+    },
+    {
+      "epoch": 0.38538854746999024,
+      "grad_norm": 0.4153432250022888,
+      "learning_rate": 0.0001,
+      "loss": 1.8866,
+      "step": 3355
+    },
+    {
+      "epoch": 0.38550341737981736,
+      "grad_norm": 0.3631424605846405,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 3356
+    },
+    {
+      "epoch": 0.3856182872896445,
+      "grad_norm": 0.42953309416770935,
+      "learning_rate": 0.0001,
+      "loss": 1.9191,
+      "step": 3357
+    },
+    {
+      "epoch": 0.3857331571994716,
+      "grad_norm": 0.4635213613510132,
+      "learning_rate": 0.0001,
+      "loss": 1.9103,
+      "step": 3358
+    },
+    {
+      "epoch": 0.3858480271092987,
+      "grad_norm": 0.40578657388687134,
+      "learning_rate": 0.0001,
+      "loss": 1.7202,
+      "step": 3359
+    },
+    {
+      "epoch": 0.38596289701912584,
+      "grad_norm": 0.3969693183898926,
+      "learning_rate": 0.0001,
+      "loss": 1.7364,
+      "step": 3360
+    },
+    {
+      "epoch": 0.38607776692895296,
+      "grad_norm": 0.39727887511253357,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 3361
+    },
+    {
+      "epoch": 0.3861926368387801,
+      "grad_norm": 0.3983653485774994,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 3362
+    },
+    {
+      "epoch": 0.3863075067486072,
+      "grad_norm": 0.3816327452659607,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 3363
+    },
+    {
+      "epoch": 0.3864223766584343,
+      "grad_norm": 0.38533225655555725,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 3364
+    },
+    {
+      "epoch": 0.38653724656826144,
+      "grad_norm": 0.40804487466812134,
+      "learning_rate": 0.0001,
+      "loss": 1.7621,
+      "step": 3365
+    },
+    {
+      "epoch": 0.38665211647808856,
+      "grad_norm": 0.3823128044605255,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 3366
+    },
+    {
+      "epoch": 0.3867669863879157,
+      "grad_norm": 0.39397284388542175,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 3367
+    },
+    {
+      "epoch": 0.3868818562977428,
+      "grad_norm": 0.38799893856048584,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 3368
+    },
+    {
+      "epoch": 0.3869967262075699,
+      "grad_norm": 0.3756542205810547,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 3369
+    },
+    {
+      "epoch": 0.38711159611739704,
+      "grad_norm": 0.40256842970848083,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 3370
+    },
+    {
+      "epoch": 0.38722646602722416,
+      "grad_norm": 0.43613606691360474,
+      "learning_rate": 0.0001,
+      "loss": 1.8034,
+      "step": 3371
+    },
+    {
+      "epoch": 0.3873413359370513,
+      "grad_norm": 0.38475316762924194,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 3372
+    },
+    {
+      "epoch": 0.3874562058468784,
+      "grad_norm": 0.40157008171081543,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 3373
+    },
+    {
+      "epoch": 0.3875710757567055,
+      "grad_norm": 0.39951568841934204,
+      "learning_rate": 0.0001,
+      "loss": 1.5993,
+      "step": 3374
+    },
+    {
+      "epoch": 0.38768594566653264,
+      "grad_norm": 0.42573562264442444,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 3375
+    },
+    {
+      "epoch": 0.38780081557635976,
+      "grad_norm": 0.3692642152309418,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 3376
+    },
+    {
+      "epoch": 0.3879156854861869,
+      "grad_norm": 0.3984464108943939,
+      "learning_rate": 0.0001,
+      "loss": 1.8419,
+      "step": 3377
+    },
+    {
+      "epoch": 0.388030555396014,
+      "grad_norm": 0.40837594866752625,
+      "learning_rate": 0.0001,
+      "loss": 1.8327,
+      "step": 3378
+    },
+    {
+      "epoch": 0.3881454253058411,
+      "grad_norm": 0.4153634011745453,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 3379
+    },
+    {
+      "epoch": 0.38826029521566824,
+      "grad_norm": 0.40997132658958435,
+      "learning_rate": 0.0001,
+      "loss": 1.7067,
+      "step": 3380
+    },
+    {
+      "epoch": 0.38837516512549536,
+      "grad_norm": 0.43204736709594727,
+      "learning_rate": 0.0001,
+      "loss": 1.7734,
+      "step": 3381
+    },
+    {
+      "epoch": 0.3884900350353225,
+      "grad_norm": 0.4231989085674286,
+      "learning_rate": 0.0001,
+      "loss": 1.7414,
+      "step": 3382
+    },
+    {
+      "epoch": 0.3886049049451496,
+      "grad_norm": 0.3816749155521393,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 3383
+    },
+    {
+      "epoch": 0.3887197748549767,
+      "grad_norm": 0.4512353539466858,
+      "learning_rate": 0.0001,
+      "loss": 1.9125,
+      "step": 3384
+    },
+    {
+      "epoch": 0.38883464476480384,
+      "grad_norm": 0.3650393486022949,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 3385
+    },
+    {
+      "epoch": 0.38894951467463096,
+      "grad_norm": 0.3751547038555145,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 3386
+    },
+    {
+      "epoch": 0.3890643845844581,
+      "grad_norm": 0.36982572078704834,
+      "learning_rate": 0.0001,
+      "loss": 1.6857,
+      "step": 3387
+    },
+    {
+      "epoch": 0.3891792544942852,
+      "grad_norm": 0.43625834584236145,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 3388
+    },
+    {
+      "epoch": 0.3892941244041123,
+      "grad_norm": 0.40629592537879944,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 3389
+    },
+    {
+      "epoch": 0.38940899431393944,
+      "grad_norm": 0.41246914863586426,
+      "learning_rate": 0.0001,
+      "loss": 1.8241,
+      "step": 3390
+    },
+    {
+      "epoch": 0.38952386422376656,
+      "grad_norm": 0.40135088562965393,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 3391
+    },
+    {
+      "epoch": 0.3896387341335937,
+      "grad_norm": 0.4104592502117157,
+      "learning_rate": 0.0001,
+      "loss": 1.8265,
+      "step": 3392
+    },
+    {
+      "epoch": 0.3897536040434208,
+      "grad_norm": 0.35968706011772156,
+      "learning_rate": 0.0001,
+      "loss": 1.5164,
+      "step": 3393
+    },
+    {
+      "epoch": 0.389868473953248,
+      "grad_norm": 0.4108087420463562,
+      "learning_rate": 0.0001,
+      "loss": 1.6904,
+      "step": 3394
+    },
+    {
+      "epoch": 0.3899833438630751,
+      "grad_norm": 0.39649486541748047,
+      "learning_rate": 0.0001,
+      "loss": 1.6845,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3900982137729022,
+      "grad_norm": 0.448559045791626,
+      "learning_rate": 0.0001,
+      "loss": 1.8202,
+      "step": 3396
+    },
+    {
+      "epoch": 0.39021308368272933,
+      "grad_norm": 0.40180444717407227,
+      "learning_rate": 0.0001,
+      "loss": 1.7422,
+      "step": 3397
+    },
+    {
+      "epoch": 0.39032795359255645,
+      "grad_norm": 0.4011480510234833,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 3398
+    },
+    {
+      "epoch": 0.3904428235023836,
+      "grad_norm": 0.41441354155540466,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 3399
+    },
+    {
+      "epoch": 0.3905576934122107,
+      "grad_norm": 0.41887158155441284,
+      "learning_rate": 0.0001,
+      "loss": 1.7279,
+      "step": 3400
+    },
+    {
+      "epoch": 0.3906725633220378,
+      "grad_norm": 0.4123765826225281,
+      "learning_rate": 0.0001,
+      "loss": 1.7621,
+      "step": 3401
+    },
+    {
+      "epoch": 0.39078743323186493,
+      "grad_norm": 0.3798231780529022,
+      "learning_rate": 0.0001,
+      "loss": 1.4537,
+      "step": 3402
+    },
+    {
+      "epoch": 0.39090230314169205,
+      "grad_norm": 0.40583983063697815,
+      "learning_rate": 0.0001,
+      "loss": 1.609,
+      "step": 3403
+    },
+    {
+      "epoch": 0.3910171730515192,
+      "grad_norm": 0.41971534490585327,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 3404
+    },
+    {
+      "epoch": 0.3911320429613463,
+      "grad_norm": 0.4053299129009247,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 3405
+    },
+    {
+      "epoch": 0.3912469128711734,
+      "grad_norm": 0.4222469925880432,
+      "learning_rate": 0.0001,
+      "loss": 1.9102,
+      "step": 3406
+    },
+    {
+      "epoch": 0.39136178278100053,
+      "grad_norm": 0.3722541630268097,
+      "learning_rate": 0.0001,
+      "loss": 1.6368,
+      "step": 3407
+    },
+    {
+      "epoch": 0.39147665269082765,
+      "grad_norm": 0.3922461271286011,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 3408
+    },
+    {
+      "epoch": 0.3915915226006548,
+      "grad_norm": 0.43620502948760986,
+      "learning_rate": 0.0001,
+      "loss": 1.9101,
+      "step": 3409
+    },
+    {
+      "epoch": 0.3917063925104819,
+      "grad_norm": 0.38486000895500183,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 3410
+    },
+    {
+      "epoch": 0.391821262420309,
+      "grad_norm": 0.417163223028183,
+      "learning_rate": 0.0001,
+      "loss": 1.4222,
+      "step": 3411
+    },
+    {
+      "epoch": 0.39193613233013613,
+      "grad_norm": 0.37510520219802856,
+      "learning_rate": 0.0001,
+      "loss": 1.2839,
+      "step": 3412
+    },
+    {
+      "epoch": 0.39205100223996325,
+      "grad_norm": 0.4312410056591034,
+      "learning_rate": 0.0001,
+      "loss": 1.7017,
+      "step": 3413
+    },
+    {
+      "epoch": 0.3921658721497904,
+      "grad_norm": 0.4008169174194336,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 3414
+    },
+    {
+      "epoch": 0.3922807420596175,
+      "grad_norm": 0.3868180811405182,
+      "learning_rate": 0.0001,
+      "loss": 1.6447,
+      "step": 3415
+    },
+    {
+      "epoch": 0.3923956119694446,
+      "grad_norm": 0.38877058029174805,
+      "learning_rate": 0.0001,
+      "loss": 1.8274,
+      "step": 3416
+    },
+    {
+      "epoch": 0.39251048187927173,
+      "grad_norm": 0.39784759283065796,
+      "learning_rate": 0.0001,
+      "loss": 1.7979,
+      "step": 3417
+    },
+    {
+      "epoch": 0.39262535178909885,
+      "grad_norm": 0.3924922049045563,
+      "learning_rate": 0.0001,
+      "loss": 1.713,
+      "step": 3418
+    },
+    {
+      "epoch": 0.392740221698926,
+      "grad_norm": 0.40661346912384033,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 3419
+    },
+    {
+      "epoch": 0.3928550916087531,
+      "grad_norm": 0.4197205901145935,
+      "learning_rate": 0.0001,
+      "loss": 1.7735,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3929699615185802,
+      "grad_norm": 0.3987165689468384,
+      "learning_rate": 0.0001,
+      "loss": 1.8069,
+      "step": 3421
+    },
+    {
+      "epoch": 0.39308483142840733,
+      "grad_norm": 0.38276979327201843,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 3422
+    },
+    {
+      "epoch": 0.39319970133823445,
+      "grad_norm": 0.3716558814048767,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 3423
+    },
+    {
+      "epoch": 0.3933145712480616,
+      "grad_norm": 0.3954230844974518,
+      "learning_rate": 0.0001,
+      "loss": 1.7698,
+      "step": 3424
+    },
+    {
+      "epoch": 0.3934294411578887,
+      "grad_norm": 0.39034202694892883,
+      "learning_rate": 0.0001,
+      "loss": 1.783,
+      "step": 3425
+    },
+    {
+      "epoch": 0.3935443110677158,
+      "grad_norm": 0.36752232909202576,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 3426
+    },
+    {
+      "epoch": 0.39365918097754293,
+      "grad_norm": 0.3931824564933777,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 3427
+    },
+    {
+      "epoch": 0.39377405088737005,
+      "grad_norm": 0.3895040452480316,
+      "learning_rate": 0.0001,
+      "loss": 1.717,
+      "step": 3428
+    },
+    {
+      "epoch": 0.3938889207971972,
+      "grad_norm": 0.4126931428909302,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3940037907070243,
+      "grad_norm": 0.39863380789756775,
+      "learning_rate": 0.0001,
+      "loss": 1.8072,
+      "step": 3430
+    },
+    {
+      "epoch": 0.3941186606168514,
+      "grad_norm": 0.409013032913208,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 3431
+    },
+    {
+      "epoch": 0.39423353052667853,
+      "grad_norm": 0.39997661113739014,
+      "learning_rate": 0.0001,
+      "loss": 1.4622,
+      "step": 3432
+    },
+    {
+      "epoch": 0.39434840043650565,
+      "grad_norm": 0.38587263226509094,
+      "learning_rate": 0.0001,
+      "loss": 1.7621,
+      "step": 3433
+    },
+    {
+      "epoch": 0.3944632703463328,
+      "grad_norm": 0.41442370414733887,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 3434
+    },
+    {
+      "epoch": 0.3945781402561599,
+      "grad_norm": 0.387526273727417,
+      "learning_rate": 0.0001,
+      "loss": 1.6732,
+      "step": 3435
+    },
+    {
+      "epoch": 0.394693010165987,
+      "grad_norm": 0.4018903970718384,
+      "learning_rate": 0.0001,
+      "loss": 1.7994,
+      "step": 3436
+    },
+    {
+      "epoch": 0.39480788007581413,
+      "grad_norm": 0.38275036215782166,
+      "learning_rate": 0.0001,
+      "loss": 1.6699,
+      "step": 3437
+    },
+    {
+      "epoch": 0.39492274998564125,
+      "grad_norm": 0.392865926027298,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 3438
+    },
+    {
+      "epoch": 0.3950376198954684,
+      "grad_norm": 0.3883517384529114,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 3439
+    },
+    {
+      "epoch": 0.3951524898052955,
+      "grad_norm": 0.43552684783935547,
+      "learning_rate": 0.0001,
+      "loss": 1.835,
+      "step": 3440
+    },
+    {
+      "epoch": 0.3952673597151226,
+      "grad_norm": 0.38498955965042114,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 3441
+    },
+    {
+      "epoch": 0.39538222962494973,
+      "grad_norm": 0.4317042827606201,
+      "learning_rate": 0.0001,
+      "loss": 1.9773,
+      "step": 3442
+    },
+    {
+      "epoch": 0.39549709953477685,
+      "grad_norm": 0.4095926880836487,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 3443
+    },
+    {
+      "epoch": 0.395611969444604,
+      "grad_norm": 0.4393211007118225,
+      "learning_rate": 0.0001,
+      "loss": 1.8934,
+      "step": 3444
+    },
+    {
+      "epoch": 0.3957268393544311,
+      "grad_norm": 0.3967788517475128,
+      "learning_rate": 0.0001,
+      "loss": 1.6129,
+      "step": 3445
+    },
+    {
+      "epoch": 0.3958417092642582,
+      "grad_norm": 0.3683384656906128,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 3446
+    },
+    {
+      "epoch": 0.39595657917408533,
+      "grad_norm": 0.37010353803634644,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 3447
+    },
+    {
+      "epoch": 0.39607144908391245,
+      "grad_norm": 0.4637109637260437,
+      "learning_rate": 0.0001,
+      "loss": 1.7736,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3961863189937396,
+      "grad_norm": 0.3716694116592407,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 3449
+    },
+    {
+      "epoch": 0.3963011889035667,
+      "grad_norm": 0.3940329849720001,
+      "learning_rate": 0.0001,
+      "loss": 1.8883,
+      "step": 3450
+    },
+    {
+      "epoch": 0.3964160588133938,
+      "grad_norm": 0.3837946653366089,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 3451
+    },
+    {
+      "epoch": 0.39653092872322093,
+      "grad_norm": 0.4419655203819275,
+      "learning_rate": 0.0001,
+      "loss": 1.7845,
+      "step": 3452
+    },
+    {
+      "epoch": 0.39664579863304805,
+      "grad_norm": 0.39042928814888,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 3453
+    },
+    {
+      "epoch": 0.3967606685428752,
+      "grad_norm": 0.43041858077049255,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 3454
+    },
+    {
+      "epoch": 0.3968755384527023,
+      "grad_norm": 0.36053743958473206,
+      "learning_rate": 0.0001,
+      "loss": 1.4182,
+      "step": 3455
+    },
+    {
+      "epoch": 0.3969904083625294,
+      "grad_norm": 0.3963378965854645,
+      "learning_rate": 0.0001,
+      "loss": 1.8062,
+      "step": 3456
+    },
+    {
+      "epoch": 0.39710527827235653,
+      "grad_norm": 0.391775906085968,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 3457
+    },
+    {
+      "epoch": 0.39722014818218365,
+      "grad_norm": 0.39680591225624084,
+      "learning_rate": 0.0001,
+      "loss": 1.7523,
+      "step": 3458
+    },
+    {
+      "epoch": 0.39733501809201077,
+      "grad_norm": 0.4133915901184082,
+      "learning_rate": 0.0001,
+      "loss": 1.6781,
+      "step": 3459
+    },
+    {
+      "epoch": 0.3974498880018379,
+      "grad_norm": 0.41504111886024475,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 3460
+    },
+    {
+      "epoch": 0.397564757911665,
+      "grad_norm": 0.39782634377479553,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3976796278214922,
+      "grad_norm": 0.4147406220436096,
+      "learning_rate": 0.0001,
+      "loss": 1.6262,
+      "step": 3462
+    },
+    {
+      "epoch": 0.3977944977313193,
+      "grad_norm": 0.3900047838687897,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 3463
+    },
+    {
+      "epoch": 0.3979093676411464,
+      "grad_norm": 0.3884030282497406,
+      "learning_rate": 0.0001,
+      "loss": 1.695,
+      "step": 3464
+    },
+    {
+      "epoch": 0.39802423755097355,
+      "grad_norm": 0.422282874584198,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 3465
+    },
+    {
+      "epoch": 0.39813910746080067,
+      "grad_norm": 0.39325493574142456,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 3466
+    },
+    {
+      "epoch": 0.3982539773706278,
+      "grad_norm": 0.3940114974975586,
+      "learning_rate": 0.0001,
+      "loss": 1.6235,
+      "step": 3467
+    },
+    {
+      "epoch": 0.3983688472804549,
+      "grad_norm": 0.4041246175765991,
+      "learning_rate": 0.0001,
+      "loss": 1.6107,
+      "step": 3468
+    },
+    {
+      "epoch": 0.398483717190282,
+      "grad_norm": 0.37534579634666443,
+      "learning_rate": 0.0001,
+      "loss": 1.737,
+      "step": 3469
+    },
+    {
+      "epoch": 0.39859858710010915,
+      "grad_norm": 0.40846285223960876,
+      "learning_rate": 0.0001,
+      "loss": 1.7085,
+      "step": 3470
+    },
+    {
+      "epoch": 0.39871345700993627,
+      "grad_norm": 0.43047085404396057,
+      "learning_rate": 0.0001,
+      "loss": 1.7536,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3988283269197634,
+      "grad_norm": 0.3942980170249939,
+      "learning_rate": 0.0001,
+      "loss": 1.7882,
+      "step": 3472
+    },
+    {
+      "epoch": 0.3989431968295905,
+      "grad_norm": 0.408866822719574,
+      "learning_rate": 0.0001,
+      "loss": 1.6989,
+      "step": 3473
+    },
+    {
+      "epoch": 0.3990580667394176,
+      "grad_norm": 0.3862482011318207,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 3474
+    },
+    {
+      "epoch": 0.39917293664924475,
+      "grad_norm": 0.3940208852291107,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 3475
+    },
+    {
+      "epoch": 0.39928780655907187,
+      "grad_norm": 0.42412546277046204,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 3476
+    },
+    {
+      "epoch": 0.399402676468899,
+      "grad_norm": 0.4334600567817688,
+      "learning_rate": 0.0001,
+      "loss": 1.5058,
+      "step": 3477
+    },
+    {
+      "epoch": 0.3995175463787261,
+      "grad_norm": 0.3741202652454376,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 3478
+    },
+    {
+      "epoch": 0.3996324162885532,
+      "grad_norm": 0.37549999356269836,
+      "learning_rate": 0.0001,
+      "loss": 1.4891,
+      "step": 3479
+    },
+    {
+      "epoch": 0.39974728619838035,
+      "grad_norm": 0.41318750381469727,
+      "learning_rate": 0.0001,
+      "loss": 1.7634,
+      "step": 3480
+    },
+    {
+      "epoch": 0.39986215610820747,
+      "grad_norm": 0.39875760674476624,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 3481
+    },
+    {
+      "epoch": 0.3999770260180346,
+      "grad_norm": 0.4000686705112457,
+      "learning_rate": 0.0001,
+      "loss": 1.5764,
+      "step": 3482
+    },
+    {
+      "epoch": 0.4000918959278617,
+      "grad_norm": 0.4229472577571869,
+      "learning_rate": 0.0001,
+      "loss": 1.5092,
+      "step": 3483
+    },
+    {
+      "epoch": 0.4002067658376888,
+      "grad_norm": 0.4480687975883484,
+      "learning_rate": 0.0001,
+      "loss": 1.8381,
+      "step": 3484
+    },
+    {
+      "epoch": 0.40032163574751595,
+      "grad_norm": 0.3803386688232422,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 3485
+    },
+    {
+      "epoch": 0.40043650565734307,
+      "grad_norm": 0.40477797389030457,
+      "learning_rate": 0.0001,
+      "loss": 1.7031,
+      "step": 3486
+    },
+    {
+      "epoch": 0.4005513755671702,
+      "grad_norm": 0.4002171456813812,
+      "learning_rate": 0.0001,
+      "loss": 1.7292,
+      "step": 3487
+    },
+    {
+      "epoch": 0.4006662454769973,
+      "grad_norm": 0.4127141833305359,
+      "learning_rate": 0.0001,
+      "loss": 1.8455,
+      "step": 3488
+    },
+    {
+      "epoch": 0.4007811153868244,
+      "grad_norm": 0.4116250276565552,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 3489
+    },
+    {
+      "epoch": 0.40089598529665155,
+      "grad_norm": 0.3903186619281769,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 3490
+    },
+    {
+      "epoch": 0.40101085520647867,
+      "grad_norm": 0.3886655867099762,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 3491
+    },
+    {
+      "epoch": 0.4011257251163058,
+      "grad_norm": 0.40146976709365845,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 3492
+    },
+    {
+      "epoch": 0.4012405950261329,
+      "grad_norm": 0.4304850697517395,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 3493
+    },
+    {
+      "epoch": 0.40135546493596,
+      "grad_norm": 0.41273269057273865,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 3494
+    },
+    {
+      "epoch": 0.40147033484578715,
+      "grad_norm": 0.42954984307289124,
+      "learning_rate": 0.0001,
+      "loss": 1.7387,
+      "step": 3495
+    },
+    {
+      "epoch": 0.40158520475561427,
+      "grad_norm": 0.3892943859100342,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 3496
+    },
+    {
+      "epoch": 0.4017000746654414,
+      "grad_norm": 0.43084821105003357,
+      "learning_rate": 0.0001,
+      "loss": 1.8832,
+      "step": 3497
+    },
+    {
+      "epoch": 0.4018149445752685,
+      "grad_norm": 0.3853071331977844,
+      "learning_rate": 0.0001,
+      "loss": 1.4343,
+      "step": 3498
+    },
+    {
+      "epoch": 0.4019298144850956,
+      "grad_norm": 0.3769509196281433,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 3499
+    },
+    {
+      "epoch": 0.40204468439492275,
+      "grad_norm": 0.41047728061676025,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 3500
+    },
+    {
+      "epoch": 0.40215955430474987,
+      "grad_norm": 0.3837498128414154,
+      "learning_rate": 0.0001,
+      "loss": 1.6573,
+      "step": 3501
+    },
+    {
+      "epoch": 0.402274424214577,
+      "grad_norm": 0.4407394826412201,
+      "learning_rate": 0.0001,
+      "loss": 1.6908,
+      "step": 3502
+    },
+    {
+      "epoch": 0.4023892941244041,
+      "grad_norm": 0.3970791697502136,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 3503
+    },
+    {
+      "epoch": 0.4025041640342312,
+      "grad_norm": 0.402908593416214,
+      "learning_rate": 0.0001,
+      "loss": 1.7421,
+      "step": 3504
+    },
+    {
+      "epoch": 0.40261903394405835,
+      "grad_norm": 0.3898850977420807,
+      "learning_rate": 0.0001,
+      "loss": 1.5148,
+      "step": 3505
+    },
+    {
+      "epoch": 0.40273390385388547,
+      "grad_norm": 0.4609782099723816,
+      "learning_rate": 0.0001,
+      "loss": 1.9721,
+      "step": 3506
+    },
+    {
+      "epoch": 0.4028487737637126,
+      "grad_norm": 0.4256317615509033,
+      "learning_rate": 0.0001,
+      "loss": 1.8592,
+      "step": 3507
+    },
+    {
+      "epoch": 0.4029636436735397,
+      "grad_norm": 0.38240641355514526,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 3508
+    },
+    {
+      "epoch": 0.4030785135833668,
+      "grad_norm": 0.40074291825294495,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 3509
+    },
+    {
+      "epoch": 0.40319338349319395,
+      "grad_norm": 0.38915419578552246,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 3510
+    },
+    {
+      "epoch": 0.40330825340302107,
+      "grad_norm": 0.39490392804145813,
+      "learning_rate": 0.0001,
+      "loss": 1.7036,
+      "step": 3511
+    },
+    {
+      "epoch": 0.4034231233128482,
+      "grad_norm": 0.37011227011680603,
+      "learning_rate": 0.0001,
+      "loss": 1.6374,
+      "step": 3512
+    },
+    {
+      "epoch": 0.4035379932226753,
+      "grad_norm": 0.41320136189460754,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 3513
+    },
+    {
+      "epoch": 0.4036528631325024,
+      "grad_norm": 0.43924593925476074,
+      "learning_rate": 0.0001,
+      "loss": 1.7871,
+      "step": 3514
+    },
+    {
+      "epoch": 0.40376773304232955,
+      "grad_norm": 0.3919188380241394,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 3515
+    },
+    {
+      "epoch": 0.40388260295215667,
+      "grad_norm": 0.40619519352912903,
+      "learning_rate": 0.0001,
+      "loss": 1.5834,
+      "step": 3516
+    },
+    {
+      "epoch": 0.4039974728619838,
+      "grad_norm": 0.4137546420097351,
+      "learning_rate": 0.0001,
+      "loss": 1.7776,
+      "step": 3517
+    },
+    {
+      "epoch": 0.4041123427718109,
+      "grad_norm": 0.4008006453514099,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 3518
+    },
+    {
+      "epoch": 0.404227212681638,
+      "grad_norm": 0.43822675943374634,
+      "learning_rate": 0.0001,
+      "loss": 1.8259,
+      "step": 3519
+    },
+    {
+      "epoch": 0.40434208259146515,
+      "grad_norm": 0.41821324825286865,
+      "learning_rate": 0.0001,
+      "loss": 1.7496,
+      "step": 3520
+    },
+    {
+      "epoch": 0.40445695250129227,
+      "grad_norm": 0.45367196202278137,
+      "learning_rate": 0.0001,
+      "loss": 1.9165,
+      "step": 3521
+    },
+    {
+      "epoch": 0.4045718224111194,
+      "grad_norm": 0.38345637917518616,
+      "learning_rate": 0.0001,
+      "loss": 1.7299,
+      "step": 3522
+    },
+    {
+      "epoch": 0.4046866923209465,
+      "grad_norm": 0.3878767490386963,
+      "learning_rate": 0.0001,
+      "loss": 1.683,
+      "step": 3523
+    },
+    {
+      "epoch": 0.4048015622307736,
+      "grad_norm": 0.43922069668769836,
+      "learning_rate": 0.0001,
+      "loss": 1.8087,
+      "step": 3524
+    },
+    {
+      "epoch": 0.40491643214060075,
+      "grad_norm": 0.38794466853141785,
+      "learning_rate": 0.0001,
+      "loss": 1.6888,
+      "step": 3525
+    },
+    {
+      "epoch": 0.40503130205042787,
+      "grad_norm": 0.39647185802459717,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 3526
+    },
+    {
+      "epoch": 0.405146171960255,
+      "grad_norm": 0.38950473070144653,
+      "learning_rate": 0.0001,
+      "loss": 1.7733,
+      "step": 3527
+    },
+    {
+      "epoch": 0.4052610418700821,
+      "grad_norm": 0.43412071466445923,
+      "learning_rate": 0.0001,
+      "loss": 1.8643,
+      "step": 3528
+    },
+    {
+      "epoch": 0.4053759117799092,
+      "grad_norm": 0.4561465382575989,
+      "learning_rate": 0.0001,
+      "loss": 1.9004,
+      "step": 3529
+    },
+    {
+      "epoch": 0.40549078168973635,
+      "grad_norm": 0.3911570906639099,
+      "learning_rate": 0.0001,
+      "loss": 1.7185,
+      "step": 3530
+    },
+    {
+      "epoch": 0.4056056515995635,
+      "grad_norm": 0.3808756172657013,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 3531
+    },
+    {
+      "epoch": 0.40572052150939064,
+      "grad_norm": 0.41304758191108704,
+      "learning_rate": 0.0001,
+      "loss": 1.704,
+      "step": 3532
+    },
+    {
+      "epoch": 0.40583539141921776,
+      "grad_norm": 0.3868066072463989,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 3533
+    },
+    {
+      "epoch": 0.4059502613290449,
+      "grad_norm": 0.47195518016815186,
+      "learning_rate": 0.0001,
+      "loss": 1.85,
+      "step": 3534
+    },
+    {
+      "epoch": 0.406065131238872,
+      "grad_norm": 0.40682026743888855,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 3535
+    },
+    {
+      "epoch": 0.4061800011486991,
+      "grad_norm": 0.4269334375858307,
+      "learning_rate": 0.0001,
+      "loss": 1.743,
+      "step": 3536
+    },
+    {
+      "epoch": 0.40629487105852624,
+      "grad_norm": 0.40638071298599243,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 3537
+    },
+    {
+      "epoch": 0.40640974096835336,
+      "grad_norm": 0.3894198536872864,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 3538
+    },
+    {
+      "epoch": 0.4065246108781805,
+      "grad_norm": 0.4187344014644623,
+      "learning_rate": 0.0001,
+      "loss": 1.647,
+      "step": 3539
+    },
+    {
+      "epoch": 0.4066394807880076,
+      "grad_norm": 0.3857315182685852,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 3540
+    },
+    {
+      "epoch": 0.4067543506978347,
+      "grad_norm": 0.3764125108718872,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 3541
+    },
+    {
+      "epoch": 0.40686922060766184,
+      "grad_norm": 0.38731610774993896,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 3542
+    },
+    {
+      "epoch": 0.40698409051748896,
+      "grad_norm": 0.38628220558166504,
+      "learning_rate": 0.0001,
+      "loss": 1.599,
+      "step": 3543
+    },
+    {
+      "epoch": 0.4070989604273161,
+      "grad_norm": 0.39660200476646423,
+      "learning_rate": 0.0001,
+      "loss": 1.6402,
+      "step": 3544
+    },
+    {
+      "epoch": 0.4072138303371432,
+      "grad_norm": 0.4349413812160492,
+      "learning_rate": 0.0001,
+      "loss": 1.9247,
+      "step": 3545
+    },
+    {
+      "epoch": 0.4073287002469703,
+      "grad_norm": 0.41799604892730713,
+      "learning_rate": 0.0001,
+      "loss": 1.7725,
+      "step": 3546
+    },
+    {
+      "epoch": 0.40744357015679744,
+      "grad_norm": 0.4117201566696167,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 3547
+    },
+    {
+      "epoch": 0.40755844006662456,
+      "grad_norm": 0.3982154130935669,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 3548
+    },
+    {
+      "epoch": 0.4076733099764517,
+      "grad_norm": 0.386783242225647,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 3549
+    },
+    {
+      "epoch": 0.4077881798862788,
+      "grad_norm": 0.3949221670627594,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 3550
+    },
+    {
+      "epoch": 0.4079030497961059,
+      "grad_norm": 0.38319775462150574,
+      "learning_rate": 0.0001,
+      "loss": 1.6634,
+      "step": 3551
+    },
+    {
+      "epoch": 0.40801791970593304,
+      "grad_norm": 0.4165116250514984,
+      "learning_rate": 0.0001,
+      "loss": 1.7474,
+      "step": 3552
+    },
+    {
+      "epoch": 0.40813278961576016,
+      "grad_norm": 0.4256540536880493,
+      "learning_rate": 0.0001,
+      "loss": 1.8328,
+      "step": 3553
+    },
+    {
+      "epoch": 0.4082476595255873,
+      "grad_norm": 0.4176916182041168,
+      "learning_rate": 0.0001,
+      "loss": 1.7,
+      "step": 3554
+    },
+    {
+      "epoch": 0.4083625294354144,
+      "grad_norm": 0.40990862250328064,
+      "learning_rate": 0.0001,
+      "loss": 1.3584,
+      "step": 3555
+    },
+    {
+      "epoch": 0.4084773993452415,
+      "grad_norm": 0.39748623967170715,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 3556
+    },
+    {
+      "epoch": 0.40859226925506864,
+      "grad_norm": 0.3950015604496002,
+      "learning_rate": 0.0001,
+      "loss": 1.5983,
+      "step": 3557
+    },
+    {
+      "epoch": 0.40870713916489576,
+      "grad_norm": 0.3774421215057373,
+      "learning_rate": 0.0001,
+      "loss": 1.4646,
+      "step": 3558
+    },
+    {
+      "epoch": 0.4088220090747229,
+      "grad_norm": 0.3629385530948639,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 3559
+    },
+    {
+      "epoch": 0.40893687898455,
+      "grad_norm": 0.3893386125564575,
+      "learning_rate": 0.0001,
+      "loss": 1.6006,
+      "step": 3560
+    },
+    {
+      "epoch": 0.4090517488943771,
+      "grad_norm": 0.3888409435749054,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 3561
+    },
+    {
+      "epoch": 0.40916661880420424,
+      "grad_norm": 0.415077805519104,
+      "learning_rate": 0.0001,
+      "loss": 1.6517,
+      "step": 3562
+    },
+    {
+      "epoch": 0.40928148871403136,
+      "grad_norm": 0.420685350894928,
+      "learning_rate": 0.0001,
+      "loss": 1.7946,
+      "step": 3563
+    },
+    {
+      "epoch": 0.4093963586238585,
+      "grad_norm": 0.4362063407897949,
+      "learning_rate": 0.0001,
+      "loss": 1.7735,
+      "step": 3564
+    },
+    {
+      "epoch": 0.4095112285336856,
+      "grad_norm": 0.42826390266418457,
+      "learning_rate": 0.0001,
+      "loss": 1.8189,
+      "step": 3565
+    },
+    {
+      "epoch": 0.4096260984435127,
+      "grad_norm": 0.3822219669818878,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 3566
+    },
+    {
+      "epoch": 0.40974096835333984,
+      "grad_norm": 0.4273683428764343,
+      "learning_rate": 0.0001,
+      "loss": 1.8192,
+      "step": 3567
+    },
+    {
+      "epoch": 0.40985583826316696,
+      "grad_norm": 0.401048481464386,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 3568
+    },
+    {
+      "epoch": 0.4099707081729941,
+      "grad_norm": 0.4532688558101654,
+      "learning_rate": 0.0001,
+      "loss": 1.6273,
+      "step": 3569
+    },
+    {
+      "epoch": 0.4100855780828212,
+      "grad_norm": 0.40526339411735535,
+      "learning_rate": 0.0001,
+      "loss": 1.8208,
+      "step": 3570
+    },
+    {
+      "epoch": 0.4102004479926483,
+      "grad_norm": 0.3863803446292877,
+      "learning_rate": 0.0001,
+      "loss": 1.6683,
+      "step": 3571
+    },
+    {
+      "epoch": 0.41031531790247544,
+      "grad_norm": 0.38212668895721436,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 3572
+    },
+    {
+      "epoch": 0.41043018781230256,
+      "grad_norm": 0.38503432273864746,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 3573
+    },
+    {
+      "epoch": 0.4105450577221297,
+      "grad_norm": 0.4347050189971924,
+      "learning_rate": 0.0001,
+      "loss": 1.7337,
+      "step": 3574
+    },
+    {
+      "epoch": 0.4106599276319568,
+      "grad_norm": 0.4009620249271393,
+      "learning_rate": 0.0001,
+      "loss": 1.4972,
+      "step": 3575
+    },
+    {
+      "epoch": 0.4107747975417839,
+      "grad_norm": 0.40209561586380005,
+      "learning_rate": 0.0001,
+      "loss": 1.716,
+      "step": 3576
+    },
+    {
+      "epoch": 0.41088966745161104,
+      "grad_norm": 0.42424091696739197,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 3577
+    },
+    {
+      "epoch": 0.41100453736143816,
+      "grad_norm": 0.405749648809433,
+      "learning_rate": 0.0001,
+      "loss": 1.3783,
+      "step": 3578
+    },
+    {
+      "epoch": 0.4111194072712653,
+      "grad_norm": 0.4051080346107483,
+      "learning_rate": 0.0001,
+      "loss": 1.7914,
+      "step": 3579
+    },
+    {
+      "epoch": 0.4112342771810924,
+      "grad_norm": 0.3940957188606262,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 3580
+    },
+    {
+      "epoch": 0.4113491470909195,
+      "grad_norm": 0.4148561358451843,
+      "learning_rate": 0.0001,
+      "loss": 1.815,
+      "step": 3581
+    },
+    {
+      "epoch": 0.41146401700074664,
+      "grad_norm": 0.37967970967292786,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 3582
+    },
+    {
+      "epoch": 0.41157888691057376,
+      "grad_norm": 0.4085034728050232,
+      "learning_rate": 0.0001,
+      "loss": 1.7586,
+      "step": 3583
+    },
+    {
+      "epoch": 0.4116937568204009,
+      "grad_norm": 0.43780767917633057,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 3584
+    },
+    {
+      "epoch": 0.411808626730228,
+      "grad_norm": 0.4053542912006378,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 3585
+    },
+    {
+      "epoch": 0.4119234966400551,
+      "grad_norm": 0.42248353362083435,
+      "learning_rate": 0.0001,
+      "loss": 1.7921,
+      "step": 3586
+    },
+    {
+      "epoch": 0.41203836654988224,
+      "grad_norm": 0.38272780179977417,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 3587
+    },
+    {
+      "epoch": 0.41215323645970936,
+      "grad_norm": 0.3969421088695526,
+      "learning_rate": 0.0001,
+      "loss": 1.7657,
+      "step": 3588
+    },
+    {
+      "epoch": 0.4122681063695365,
+      "grad_norm": 0.3919724225997925,
+      "learning_rate": 0.0001,
+      "loss": 1.5506,
+      "step": 3589
+    },
+    {
+      "epoch": 0.4123829762793636,
+      "grad_norm": 0.37912601232528687,
+      "learning_rate": 0.0001,
+      "loss": 1.6657,
+      "step": 3590
+    },
+    {
+      "epoch": 0.4124978461891907,
+      "grad_norm": 0.4223651885986328,
+      "learning_rate": 0.0001,
+      "loss": 1.7099,
+      "step": 3591
+    },
+    {
+      "epoch": 0.41261271609901784,
+      "grad_norm": 0.39578935503959656,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 3592
+    },
+    {
+      "epoch": 0.41272758600884496,
+      "grad_norm": 0.40267515182495117,
+      "learning_rate": 0.0001,
+      "loss": 1.5518,
+      "step": 3593
+    },
+    {
+      "epoch": 0.4128424559186721,
+      "grad_norm": 0.4238963723182678,
+      "learning_rate": 0.0001,
+      "loss": 1.9176,
+      "step": 3594
+    },
+    {
+      "epoch": 0.4129573258284992,
+      "grad_norm": 0.43114468455314636,
+      "learning_rate": 0.0001,
+      "loss": 1.7819,
+      "step": 3595
+    },
+    {
+      "epoch": 0.4130721957383263,
+      "grad_norm": 0.36643150448799133,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 3596
+    },
+    {
+      "epoch": 0.41318706564815344,
+      "grad_norm": 0.37914469838142395,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 3597
+    },
+    {
+      "epoch": 0.41330193555798056,
+      "grad_norm": 0.4015408158302307,
+      "learning_rate": 0.0001,
+      "loss": 1.6765,
+      "step": 3598
+    },
+    {
+      "epoch": 0.41341680546780774,
+      "grad_norm": 0.40977799892425537,
+      "learning_rate": 0.0001,
+      "loss": 1.445,
+      "step": 3599
+    },
+    {
+      "epoch": 0.41353167537763486,
+      "grad_norm": 0.4055366814136505,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 3600
+    },
+    {
+      "epoch": 0.413646545287462,
+      "grad_norm": 0.4310015141963959,
+      "learning_rate": 0.0001,
+      "loss": 1.6015,
+      "step": 3601
+    },
+    {
+      "epoch": 0.4137614151972891,
+      "grad_norm": 0.4342171847820282,
+      "learning_rate": 0.0001,
+      "loss": 1.7804,
+      "step": 3602
+    },
+    {
+      "epoch": 0.4138762851071162,
+      "grad_norm": 0.3718089759349823,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 3603
+    },
+    {
+      "epoch": 0.41399115501694334,
+      "grad_norm": 0.38151004910469055,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 3604
+    },
+    {
+      "epoch": 0.41410602492677046,
+      "grad_norm": 0.4037642776966095,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 3605
+    },
+    {
+      "epoch": 0.4142208948365976,
+      "grad_norm": 0.40361806750297546,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 3606
+    },
+    {
+      "epoch": 0.4143357647464247,
+      "grad_norm": 0.4126417934894562,
+      "learning_rate": 0.0001,
+      "loss": 1.703,
+      "step": 3607
+    },
+    {
+      "epoch": 0.4144506346562518,
+      "grad_norm": 0.409773051738739,
+      "learning_rate": 0.0001,
+      "loss": 1.6464,
+      "step": 3608
+    },
+    {
+      "epoch": 0.41456550456607894,
+      "grad_norm": 0.44128933548927307,
+      "learning_rate": 0.0001,
+      "loss": 1.7266,
+      "step": 3609
+    },
+    {
+      "epoch": 0.41468037447590606,
+      "grad_norm": 0.3855801820755005,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 3610
+    },
+    {
+      "epoch": 0.4147952443857332,
+      "grad_norm": 0.3980732262134552,
+      "learning_rate": 0.0001,
+      "loss": 1.7958,
+      "step": 3611
+    },
+    {
+      "epoch": 0.4149101142955603,
+      "grad_norm": 0.4035704433917999,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 3612
+    },
+    {
+      "epoch": 0.4150249842053874,
+      "grad_norm": 0.3865931034088135,
+      "learning_rate": 0.0001,
+      "loss": 1.778,
+      "step": 3613
+    },
+    {
+      "epoch": 0.41513985411521453,
+      "grad_norm": 0.4053201377391815,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 3614
+    },
+    {
+      "epoch": 0.41525472402504165,
+      "grad_norm": 0.3935595452785492,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 3615
+    },
+    {
+      "epoch": 0.4153695939348688,
+      "grad_norm": 0.3803655803203583,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 3616
+    },
+    {
+      "epoch": 0.4154844638446959,
+      "grad_norm": 0.4078107476234436,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 3617
+    },
+    {
+      "epoch": 0.415599333754523,
+      "grad_norm": 0.40983328223228455,
+      "learning_rate": 0.0001,
+      "loss": 1.8803,
+      "step": 3618
+    },
+    {
+      "epoch": 0.41571420366435013,
+      "grad_norm": 0.3894909918308258,
+      "learning_rate": 0.0001,
+      "loss": 1.6771,
+      "step": 3619
+    },
+    {
+      "epoch": 0.41582907357417725,
+      "grad_norm": 0.41159242391586304,
+      "learning_rate": 0.0001,
+      "loss": 1.7309,
+      "step": 3620
+    },
+    {
+      "epoch": 0.4159439434840044,
+      "grad_norm": 0.4065288305282593,
+      "learning_rate": 0.0001,
+      "loss": 1.7908,
+      "step": 3621
+    },
+    {
+      "epoch": 0.4160588133938315,
+      "grad_norm": 0.4101758599281311,
+      "learning_rate": 0.0001,
+      "loss": 1.7292,
+      "step": 3622
+    },
+    {
+      "epoch": 0.4161736833036586,
+      "grad_norm": 0.41453102231025696,
+      "learning_rate": 0.0001,
+      "loss": 1.8015,
+      "step": 3623
+    },
+    {
+      "epoch": 0.41628855321348573,
+      "grad_norm": 0.39582541584968567,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 3624
+    },
+    {
+      "epoch": 0.41640342312331285,
+      "grad_norm": 0.4312124252319336,
+      "learning_rate": 0.0001,
+      "loss": 1.7743,
+      "step": 3625
+    },
+    {
+      "epoch": 0.41651829303314,
+      "grad_norm": 0.40501224994659424,
+      "learning_rate": 0.0001,
+      "loss": 1.8335,
+      "step": 3626
+    },
+    {
+      "epoch": 0.4166331629429671,
+      "grad_norm": 0.39315342903137207,
+      "learning_rate": 0.0001,
+      "loss": 1.5686,
+      "step": 3627
+    },
+    {
+      "epoch": 0.4167480328527942,
+      "grad_norm": 0.4335770606994629,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 3628
+    },
+    {
+      "epoch": 0.41686290276262133,
+      "grad_norm": 0.3941115736961365,
+      "learning_rate": 0.0001,
+      "loss": 1.6904,
+      "step": 3629
+    },
+    {
+      "epoch": 0.41697777267244845,
+      "grad_norm": 0.40810754895210266,
+      "learning_rate": 0.0001,
+      "loss": 1.6955,
+      "step": 3630
+    },
+    {
+      "epoch": 0.4170926425822756,
+      "grad_norm": 0.3764353394508362,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 3631
+    },
+    {
+      "epoch": 0.4172075124921027,
+      "grad_norm": 0.41539716720581055,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 3632
+    },
+    {
+      "epoch": 0.4173223824019298,
+      "grad_norm": 0.3807416260242462,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 3633
+    },
+    {
+      "epoch": 0.41743725231175693,
+      "grad_norm": 0.38666248321533203,
+      "learning_rate": 0.0001,
+      "loss": 1.5018,
+      "step": 3634
+    },
+    {
+      "epoch": 0.41755212222158405,
+      "grad_norm": 0.40492692589759827,
+      "learning_rate": 0.0001,
+      "loss": 1.5823,
+      "step": 3635
+    },
+    {
+      "epoch": 0.4176669921314112,
+      "grad_norm": 0.38428235054016113,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 3636
+    },
+    {
+      "epoch": 0.4177818620412383,
+      "grad_norm": 0.3912535607814789,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 3637
+    },
+    {
+      "epoch": 0.4178967319510654,
+      "grad_norm": 0.3759515881538391,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 3638
+    },
+    {
+      "epoch": 0.41801160186089253,
+      "grad_norm": 0.38995474576950073,
+      "learning_rate": 0.0001,
+      "loss": 1.6871,
+      "step": 3639
+    },
+    {
+      "epoch": 0.41812647177071965,
+      "grad_norm": 0.38808655738830566,
+      "learning_rate": 0.0001,
+      "loss": 1.3153,
+      "step": 3640
+    },
+    {
+      "epoch": 0.4182413416805468,
+      "grad_norm": 0.3956000804901123,
+      "learning_rate": 0.0001,
+      "loss": 1.575,
+      "step": 3641
+    },
+    {
+      "epoch": 0.4183562115903739,
+      "grad_norm": 0.43580135703086853,
+      "learning_rate": 0.0001,
+      "loss": 1.6672,
+      "step": 3642
+    },
+    {
+      "epoch": 0.418471081500201,
+      "grad_norm": 0.4119028151035309,
+      "learning_rate": 0.0001,
+      "loss": 1.7918,
+      "step": 3643
+    },
+    {
+      "epoch": 0.41858595141002813,
+      "grad_norm": 0.44993263483047485,
+      "learning_rate": 0.0001,
+      "loss": 1.88,
+      "step": 3644
+    },
+    {
+      "epoch": 0.41870082131985525,
+      "grad_norm": 0.4024428129196167,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 3645
+    },
+    {
+      "epoch": 0.4188156912296824,
+      "grad_norm": 0.4061081111431122,
+      "learning_rate": 0.0001,
+      "loss": 1.7381,
+      "step": 3646
+    },
+    {
+      "epoch": 0.4189305611395095,
+      "grad_norm": 0.3835539221763611,
+      "learning_rate": 0.0001,
+      "loss": 1.4161,
+      "step": 3647
+    },
+    {
+      "epoch": 0.4190454310493366,
+      "grad_norm": 0.41590166091918945,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 3648
+    },
+    {
+      "epoch": 0.41916030095916373,
+      "grad_norm": 0.40571367740631104,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 3649
+    },
+    {
+      "epoch": 0.41927517086899085,
+      "grad_norm": 0.43440738320350647,
+      "learning_rate": 0.0001,
+      "loss": 1.7795,
+      "step": 3650
+    },
+    {
+      "epoch": 0.419390040778818,
+      "grad_norm": 0.40848132967948914,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 3651
+    },
+    {
+      "epoch": 0.4195049106886451,
+      "grad_norm": 0.3971754014492035,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 3652
+    },
+    {
+      "epoch": 0.4196197805984722,
+      "grad_norm": 0.3942772448062897,
+      "learning_rate": 0.0001,
+      "loss": 1.59,
+      "step": 3653
+    },
+    {
+      "epoch": 0.41973465050829933,
+      "grad_norm": 0.4283560812473297,
+      "learning_rate": 0.0001,
+      "loss": 1.7446,
+      "step": 3654
+    },
+    {
+      "epoch": 0.41984952041812645,
+      "grad_norm": 0.39290323853492737,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 3655
+    },
+    {
+      "epoch": 0.4199643903279536,
+      "grad_norm": 0.3974771797657013,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 3656
+    },
+    {
+      "epoch": 0.4200792602377807,
+      "grad_norm": 0.40750113129615784,
+      "learning_rate": 0.0001,
+      "loss": 1.8154,
+      "step": 3657
+    },
+    {
+      "epoch": 0.4201941301476078,
+      "grad_norm": 0.4657525420188904,
+      "learning_rate": 0.0001,
+      "loss": 1.8862,
+      "step": 3658
+    },
+    {
+      "epoch": 0.42030900005743493,
+      "grad_norm": 0.40977463126182556,
+      "learning_rate": 0.0001,
+      "loss": 1.7574,
+      "step": 3659
+    },
+    {
+      "epoch": 0.42042386996726205,
+      "grad_norm": 0.4081574082374573,
+      "learning_rate": 0.0001,
+      "loss": 1.7681,
+      "step": 3660
+    },
+    {
+      "epoch": 0.4205387398770892,
+      "grad_norm": 0.46353742480278015,
+      "learning_rate": 0.0001,
+      "loss": 1.9854,
+      "step": 3661
+    },
+    {
+      "epoch": 0.4206536097869163,
+      "grad_norm": 0.3851860463619232,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 3662
+    },
+    {
+      "epoch": 0.4207684796967434,
+      "grad_norm": 0.38905036449432373,
+      "learning_rate": 0.0001,
+      "loss": 1.5134,
+      "step": 3663
+    },
+    {
+      "epoch": 0.42088334960657053,
+      "grad_norm": 0.3887808620929718,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 3664
+    },
+    {
+      "epoch": 0.42099821951639765,
+      "grad_norm": 0.39309823513031006,
+      "learning_rate": 0.0001,
+      "loss": 1.7099,
+      "step": 3665
+    },
+    {
+      "epoch": 0.4211130894262248,
+      "grad_norm": 0.389574259519577,
+      "learning_rate": 0.0001,
+      "loss": 1.7133,
+      "step": 3666
+    },
+    {
+      "epoch": 0.4212279593360519,
+      "grad_norm": 0.3817690312862396,
+      "learning_rate": 0.0001,
+      "loss": 1.4487,
+      "step": 3667
+    },
+    {
+      "epoch": 0.42134282924587907,
+      "grad_norm": 0.41336965560913086,
+      "learning_rate": 0.0001,
+      "loss": 1.7784,
+      "step": 3668
+    },
+    {
+      "epoch": 0.4214576991557062,
+      "grad_norm": 0.42329856753349304,
+      "learning_rate": 0.0001,
+      "loss": 1.7214,
+      "step": 3669
+    },
+    {
+      "epoch": 0.4215725690655333,
+      "grad_norm": 0.3827388882637024,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 3670
+    },
+    {
+      "epoch": 0.42168743897536043,
+      "grad_norm": 0.37781426310539246,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 3671
+    },
+    {
+      "epoch": 0.42180230888518755,
+      "grad_norm": 0.40145260095596313,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 3672
+    },
+    {
+      "epoch": 0.42191717879501467,
+      "grad_norm": 0.37764087319374084,
+      "learning_rate": 0.0001,
+      "loss": 1.5918,
+      "step": 3673
+    },
+    {
+      "epoch": 0.4220320487048418,
+      "grad_norm": 0.40427038073539734,
+      "learning_rate": 0.0001,
+      "loss": 1.8588,
+      "step": 3674
+    },
+    {
+      "epoch": 0.4221469186146689,
+      "grad_norm": 0.3810051679611206,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 3675
+    },
+    {
+      "epoch": 0.42226178852449603,
+      "grad_norm": 0.4347747266292572,
+      "learning_rate": 0.0001,
+      "loss": 1.81,
+      "step": 3676
+    },
+    {
+      "epoch": 0.42237665843432315,
+      "grad_norm": 0.4019508361816406,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 3677
+    },
+    {
+      "epoch": 0.42249152834415027,
+      "grad_norm": 0.40587884187698364,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 3678
+    },
+    {
+      "epoch": 0.4226063982539774,
+      "grad_norm": 0.3974643051624298,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 3679
+    },
+    {
+      "epoch": 0.4227212681638045,
+      "grad_norm": 0.443687379360199,
+      "learning_rate": 0.0001,
+      "loss": 1.9096,
+      "step": 3680
+    },
+    {
+      "epoch": 0.42283613807363163,
+      "grad_norm": 0.4022139608860016,
+      "learning_rate": 0.0001,
+      "loss": 1.8164,
+      "step": 3681
+    },
+    {
+      "epoch": 0.42295100798345875,
+      "grad_norm": 0.46298283338546753,
+      "learning_rate": 0.0001,
+      "loss": 1.9288,
+      "step": 3682
+    },
+    {
+      "epoch": 0.42306587789328587,
+      "grad_norm": 0.3819207549095154,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 3683
+    },
+    {
+      "epoch": 0.423180747803113,
+      "grad_norm": 0.39071542024612427,
+      "learning_rate": 0.0001,
+      "loss": 1.4206,
+      "step": 3684
+    },
+    {
+      "epoch": 0.4232956177129401,
+      "grad_norm": 0.39119791984558105,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 3685
+    },
+    {
+      "epoch": 0.42341048762276723,
+      "grad_norm": 0.41179823875427246,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 3686
+    },
+    {
+      "epoch": 0.42352535753259435,
+      "grad_norm": 0.40467724204063416,
+      "learning_rate": 0.0001,
+      "loss": 1.6625,
+      "step": 3687
+    },
+    {
+      "epoch": 0.42364022744242147,
+      "grad_norm": 0.4044126570224762,
+      "learning_rate": 0.0001,
+      "loss": 1.7322,
+      "step": 3688
+    },
+    {
+      "epoch": 0.4237550973522486,
+      "grad_norm": 0.39904049038887024,
+      "learning_rate": 0.0001,
+      "loss": 1.4788,
+      "step": 3689
+    },
+    {
+      "epoch": 0.4238699672620757,
+      "grad_norm": 0.4178762137889862,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 3690
+    },
+    {
+      "epoch": 0.42398483717190283,
+      "grad_norm": 0.4076288640499115,
+      "learning_rate": 0.0001,
+      "loss": 1.6828,
+      "step": 3691
+    },
+    {
+      "epoch": 0.42409970708172995,
+      "grad_norm": 0.4018821716308594,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 3692
+    },
+    {
+      "epoch": 0.42421457699155707,
+      "grad_norm": 0.40769195556640625,
+      "learning_rate": 0.0001,
+      "loss": 1.6397,
+      "step": 3693
+    },
+    {
+      "epoch": 0.4243294469013842,
+      "grad_norm": 0.38684284687042236,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 3694
+    },
+    {
+      "epoch": 0.4244443168112113,
+      "grad_norm": 0.41944825649261475,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 3695
+    },
+    {
+      "epoch": 0.42455918672103843,
+      "grad_norm": 0.38351449370384216,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 3696
+    },
+    {
+      "epoch": 0.42467405663086555,
+      "grad_norm": 0.429304838180542,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 3697
+    },
+    {
+      "epoch": 0.42478892654069267,
+      "grad_norm": 0.42793336510658264,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 3698
+    },
+    {
+      "epoch": 0.4249037964505198,
+      "grad_norm": 0.41888338327407837,
+      "learning_rate": 0.0001,
+      "loss": 1.7443,
+      "step": 3699
+    },
+    {
+      "epoch": 0.4250186663603469,
+      "grad_norm": 0.3938450217247009,
+      "learning_rate": 0.0001,
+      "loss": 1.4583,
+      "step": 3700
+    },
+    {
+      "epoch": 0.42513353627017403,
+      "grad_norm": 0.3846457004547119,
+      "learning_rate": 0.0001,
+      "loss": 1.4163,
+      "step": 3701
+    },
+    {
+      "epoch": 0.42524840618000115,
+      "grad_norm": 0.37341800332069397,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 3702
+    },
+    {
+      "epoch": 0.42536327608982827,
+      "grad_norm": 0.39832285046577454,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 3703
+    },
+    {
+      "epoch": 0.4254781459996554,
+      "grad_norm": 0.374226838350296,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 3704
+    },
+    {
+      "epoch": 0.4255930159094825,
+      "grad_norm": 0.4002193212509155,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 3705
+    },
+    {
+      "epoch": 0.42570788581930963,
+      "grad_norm": 0.3768205940723419,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 3706
+    },
+    {
+      "epoch": 0.42582275572913675,
+      "grad_norm": 0.3951648771762848,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 3707
+    },
+    {
+      "epoch": 0.42593762563896387,
+      "grad_norm": 0.4215008616447449,
+      "learning_rate": 0.0001,
+      "loss": 1.8617,
+      "step": 3708
+    },
+    {
+      "epoch": 0.426052495548791,
+      "grad_norm": 0.4458306133747101,
+      "learning_rate": 0.0001,
+      "loss": 1.7462,
+      "step": 3709
+    },
+    {
+      "epoch": 0.4261673654586181,
+      "grad_norm": 0.40523743629455566,
+      "learning_rate": 0.0001,
+      "loss": 1.7303,
+      "step": 3710
+    },
+    {
+      "epoch": 0.42628223536844523,
+      "grad_norm": 0.412463515996933,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 3711
+    },
+    {
+      "epoch": 0.42639710527827235,
+      "grad_norm": 0.4172394871711731,
+      "learning_rate": 0.0001,
+      "loss": 1.7788,
+      "step": 3712
+    },
+    {
+      "epoch": 0.42651197518809947,
+      "grad_norm": 0.38772034645080566,
+      "learning_rate": 0.0001,
+      "loss": 1.6891,
+      "step": 3713
+    },
+    {
+      "epoch": 0.4266268450979266,
+      "grad_norm": 0.43306443095207214,
+      "learning_rate": 0.0001,
+      "loss": 1.7567,
+      "step": 3714
+    },
+    {
+      "epoch": 0.4267417150077537,
+      "grad_norm": 0.4059670865535736,
+      "learning_rate": 0.0001,
+      "loss": 1.8066,
+      "step": 3715
+    },
+    {
+      "epoch": 0.42685658491758083,
+      "grad_norm": 0.392241507768631,
+      "learning_rate": 0.0001,
+      "loss": 1.6895,
+      "step": 3716
+    },
+    {
+      "epoch": 0.42697145482740795,
+      "grad_norm": 0.42141708731651306,
+      "learning_rate": 0.0001,
+      "loss": 1.7378,
+      "step": 3717
+    },
+    {
+      "epoch": 0.42708632473723507,
+      "grad_norm": 0.42226916551589966,
+      "learning_rate": 0.0001,
+      "loss": 1.7086,
+      "step": 3718
+    },
+    {
+      "epoch": 0.4272011946470622,
+      "grad_norm": 0.4137056767940521,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 3719
+    },
+    {
+      "epoch": 0.4273160645568893,
+      "grad_norm": 0.4272863566875458,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 3720
+    },
+    {
+      "epoch": 0.42743093446671643,
+      "grad_norm": 0.4106510579586029,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 3721
+    },
+    {
+      "epoch": 0.42754580437654355,
+      "grad_norm": 0.4005192518234253,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 3722
+    },
+    {
+      "epoch": 0.42766067428637067,
+      "grad_norm": 0.4278923273086548,
+      "learning_rate": 0.0001,
+      "loss": 1.7448,
+      "step": 3723
+    },
+    {
+      "epoch": 0.4277755441961978,
+      "grad_norm": 0.4044369161128998,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 3724
+    },
+    {
+      "epoch": 0.4278904141060249,
+      "grad_norm": 0.40069693326950073,
+      "learning_rate": 0.0001,
+      "loss": 1.7594,
+      "step": 3725
+    },
+    {
+      "epoch": 0.42800528401585203,
+      "grad_norm": 0.4300965964794159,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 3726
+    },
+    {
+      "epoch": 0.42812015392567915,
+      "grad_norm": 0.39623257517814636,
+      "learning_rate": 0.0001,
+      "loss": 1.4162,
+      "step": 3727
+    },
+    {
+      "epoch": 0.42823502383550627,
+      "grad_norm": 0.39578771591186523,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 3728
+    },
+    {
+      "epoch": 0.4283498937453334,
+      "grad_norm": 0.4229432940483093,
+      "learning_rate": 0.0001,
+      "loss": 1.7019,
+      "step": 3729
+    },
+    {
+      "epoch": 0.4284647636551605,
+      "grad_norm": 0.40856388211250305,
+      "learning_rate": 0.0001,
+      "loss": 1.8467,
+      "step": 3730
+    },
+    {
+      "epoch": 0.4285796335649876,
+      "grad_norm": 0.4008654057979584,
+      "learning_rate": 0.0001,
+      "loss": 1.7282,
+      "step": 3731
+    },
+    {
+      "epoch": 0.42869450347481475,
+      "grad_norm": 0.4332294464111328,
+      "learning_rate": 0.0001,
+      "loss": 1.8846,
+      "step": 3732
+    },
+    {
+      "epoch": 0.42880937338464187,
+      "grad_norm": 0.39115771651268005,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 3733
+    },
+    {
+      "epoch": 0.428924243294469,
+      "grad_norm": 0.42474624514579773,
+      "learning_rate": 0.0001,
+      "loss": 1.7635,
+      "step": 3734
+    },
+    {
+      "epoch": 0.4290391132042961,
+      "grad_norm": 0.3834139406681061,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 3735
+    },
+    {
+      "epoch": 0.4291539831141233,
+      "grad_norm": 0.42314034700393677,
+      "learning_rate": 0.0001,
+      "loss": 1.7384,
+      "step": 3736
+    },
+    {
+      "epoch": 0.4292688530239504,
+      "grad_norm": 0.41689667105674744,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 3737
+    },
+    {
+      "epoch": 0.4293837229337775,
+      "grad_norm": 0.38845548033714294,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 3738
+    },
+    {
+      "epoch": 0.42949859284360464,
+      "grad_norm": 0.4228481352329254,
+      "learning_rate": 0.0001,
+      "loss": 1.7028,
+      "step": 3739
+    },
+    {
+      "epoch": 0.42961346275343176,
+      "grad_norm": 0.38769879937171936,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 3740
+    },
+    {
+      "epoch": 0.4297283326632589,
+      "grad_norm": 0.43911242485046387,
+      "learning_rate": 0.0001,
+      "loss": 1.7073,
+      "step": 3741
+    },
+    {
+      "epoch": 0.429843202573086,
+      "grad_norm": 0.40347546339035034,
+      "learning_rate": 0.0001,
+      "loss": 1.5931,
+      "step": 3742
+    },
+    {
+      "epoch": 0.4299580724829131,
+      "grad_norm": 0.43745940923690796,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 3743
+    },
+    {
+      "epoch": 0.43007294239274024,
+      "grad_norm": 0.41324344277381897,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 3744
+    },
+    {
+      "epoch": 0.43018781230256736,
+      "grad_norm": 0.43059617280960083,
+      "learning_rate": 0.0001,
+      "loss": 1.7581,
+      "step": 3745
+    },
+    {
+      "epoch": 0.4303026822123945,
+      "grad_norm": 0.40253952145576477,
+      "learning_rate": 0.0001,
+      "loss": 1.5059,
+      "step": 3746
+    },
+    {
+      "epoch": 0.4304175521222216,
+      "grad_norm": 0.3712945580482483,
+      "learning_rate": 0.0001,
+      "loss": 1.4402,
+      "step": 3747
+    },
+    {
+      "epoch": 0.4305324220320487,
+      "grad_norm": 0.41729727387428284,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 3748
+    },
+    {
+      "epoch": 0.43064729194187584,
+      "grad_norm": 0.40478816628456116,
+      "learning_rate": 0.0001,
+      "loss": 1.7401,
+      "step": 3749
+    },
+    {
+      "epoch": 0.43076216185170296,
+      "grad_norm": 0.37169450521469116,
+      "learning_rate": 0.0001,
+      "loss": 1.45,
+      "step": 3750
+    },
+    {
+      "epoch": 0.4308770317615301,
+      "grad_norm": 0.4352746903896332,
+      "learning_rate": 0.0001,
+      "loss": 1.711,
+      "step": 3751
+    },
+    {
+      "epoch": 0.4309919016713572,
+      "grad_norm": 0.41806161403656006,
+      "learning_rate": 0.0001,
+      "loss": 1.7793,
+      "step": 3752
+    },
+    {
+      "epoch": 0.4311067715811843,
+      "grad_norm": 0.41790369153022766,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 3753
+    },
+    {
+      "epoch": 0.43122164149101144,
+      "grad_norm": 0.40323588252067566,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 3754
+    },
+    {
+      "epoch": 0.43133651140083856,
+      "grad_norm": 0.4263021945953369,
+      "learning_rate": 0.0001,
+      "loss": 1.795,
+      "step": 3755
+    },
+    {
+      "epoch": 0.4314513813106657,
+      "grad_norm": 0.4722081124782562,
+      "learning_rate": 0.0001,
+      "loss": 1.8166,
+      "step": 3756
+    },
+    {
+      "epoch": 0.4315662512204928,
+      "grad_norm": 0.4170565605163574,
+      "learning_rate": 0.0001,
+      "loss": 1.7488,
+      "step": 3757
+    },
+    {
+      "epoch": 0.4316811211303199,
+      "grad_norm": 0.3915204703807831,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 3758
+    },
+    {
+      "epoch": 0.43179599104014704,
+      "grad_norm": 0.38846898078918457,
+      "learning_rate": 0.0001,
+      "loss": 1.6482,
+      "step": 3759
+    },
+    {
+      "epoch": 0.43191086094997416,
+      "grad_norm": 0.3822706639766693,
+      "learning_rate": 0.0001,
+      "loss": 1.7103,
+      "step": 3760
+    },
+    {
+      "epoch": 0.4320257308598013,
+      "grad_norm": 0.3950004279613495,
+      "learning_rate": 0.0001,
+      "loss": 1.5763,
+      "step": 3761
+    },
+    {
+      "epoch": 0.4321406007696284,
+      "grad_norm": 0.3973482549190521,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 3762
+    },
+    {
+      "epoch": 0.4322554706794555,
+      "grad_norm": 0.40038934350013733,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 3763
+    },
+    {
+      "epoch": 0.43237034058928264,
+      "grad_norm": 0.3878767490386963,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 3764
+    },
+    {
+      "epoch": 0.43248521049910976,
+      "grad_norm": 0.40222638845443726,
+      "learning_rate": 0.0001,
+      "loss": 1.7212,
+      "step": 3765
+    },
+    {
+      "epoch": 0.4326000804089369,
+      "grad_norm": 0.38121557235717773,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 3766
+    },
+    {
+      "epoch": 0.432714950318764,
+      "grad_norm": 0.4091407060623169,
+      "learning_rate": 0.0001,
+      "loss": 1.8233,
+      "step": 3767
+    },
+    {
+      "epoch": 0.4328298202285911,
+      "grad_norm": 0.39854979515075684,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 3768
+    },
+    {
+      "epoch": 0.43294469013841824,
+      "grad_norm": 0.41306614875793457,
+      "learning_rate": 0.0001,
+      "loss": 1.8204,
+      "step": 3769
+    },
+    {
+      "epoch": 0.43305956004824536,
+      "grad_norm": 0.4026402235031128,
+      "learning_rate": 0.0001,
+      "loss": 1.6817,
+      "step": 3770
+    },
+    {
+      "epoch": 0.4331744299580725,
+      "grad_norm": 0.3648960292339325,
+      "learning_rate": 0.0001,
+      "loss": 1.3872,
+      "step": 3771
+    },
+    {
+      "epoch": 0.4332892998678996,
+      "grad_norm": 0.3978753983974457,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 3772
+    },
+    {
+      "epoch": 0.4334041697777267,
+      "grad_norm": 0.3977111876010895,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 3773
+    },
+    {
+      "epoch": 0.43351903968755384,
+      "grad_norm": 0.4535907506942749,
+      "learning_rate": 0.0001,
+      "loss": 1.8803,
+      "step": 3774
+    },
+    {
+      "epoch": 0.43363390959738096,
+      "grad_norm": 0.4159374535083771,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 3775
+    },
+    {
+      "epoch": 0.4337487795072081,
+      "grad_norm": 0.40071192383766174,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 3776
+    },
+    {
+      "epoch": 0.4338636494170352,
+      "grad_norm": 0.40635019540786743,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 3777
+    },
+    {
+      "epoch": 0.4339785193268623,
+      "grad_norm": 0.4317454993724823,
+      "learning_rate": 0.0001,
+      "loss": 1.7709,
+      "step": 3778
+    },
+    {
+      "epoch": 0.43409338923668944,
+      "grad_norm": 0.3959444761276245,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 3779
+    },
+    {
+      "epoch": 0.43420825914651656,
+      "grad_norm": 0.3976995050907135,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 3780
+    },
+    {
+      "epoch": 0.4343231290563437,
+      "grad_norm": 0.4593839943408966,
+      "learning_rate": 0.0001,
+      "loss": 2.0386,
+      "step": 3781
+    },
+    {
+      "epoch": 0.4344379989661708,
+      "grad_norm": 0.4245525598526001,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 3782
+    },
+    {
+      "epoch": 0.4345528688759979,
+      "grad_norm": 0.3964807689189911,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 3783
+    },
+    {
+      "epoch": 0.43466773878582504,
+      "grad_norm": 0.3919978439807892,
+      "learning_rate": 0.0001,
+      "loss": 1.598,
+      "step": 3784
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 0.41812998056411743,
+      "learning_rate": 0.0001,
+      "loss": 1.7921,
+      "step": 3785
+    },
+    {
+      "epoch": 0.4348974786054793,
+      "grad_norm": 0.3976697325706482,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 3786
+    },
+    {
+      "epoch": 0.4350123485153064,
+      "grad_norm": 0.39994823932647705,
+      "learning_rate": 0.0001,
+      "loss": 1.7524,
+      "step": 3787
+    },
+    {
+      "epoch": 0.4351272184251335,
+      "grad_norm": 0.41444066166877747,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 3788
+    },
+    {
+      "epoch": 0.43524208833496064,
+      "grad_norm": 0.41670259833335876,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 3789
+    },
+    {
+      "epoch": 0.43535695824478776,
+      "grad_norm": 0.4110081195831299,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 3790
+    },
+    {
+      "epoch": 0.4354718281546149,
+      "grad_norm": 0.41741257905960083,
+      "learning_rate": 0.0001,
+      "loss": 1.7861,
+      "step": 3791
+    },
+    {
+      "epoch": 0.435586698064442,
+      "grad_norm": 0.40086445212364197,
+      "learning_rate": 0.0001,
+      "loss": 1.6162,
+      "step": 3792
+    },
+    {
+      "epoch": 0.4357015679742691,
+      "grad_norm": 0.38740214705467224,
+      "learning_rate": 0.0001,
+      "loss": 1.5713,
+      "step": 3793
+    },
+    {
+      "epoch": 0.43581643788409624,
+      "grad_norm": 0.43951183557510376,
+      "learning_rate": 0.0001,
+      "loss": 1.795,
+      "step": 3794
+    },
+    {
+      "epoch": 0.43593130779392336,
+      "grad_norm": 0.4288491904735565,
+      "learning_rate": 0.0001,
+      "loss": 1.7329,
+      "step": 3795
+    },
+    {
+      "epoch": 0.4360461777037505,
+      "grad_norm": 0.4051986038684845,
+      "learning_rate": 0.0001,
+      "loss": 1.6826,
+      "step": 3796
+    },
+    {
+      "epoch": 0.4361610476135776,
+      "grad_norm": 0.36968207359313965,
+      "learning_rate": 0.0001,
+      "loss": 1.4493,
+      "step": 3797
+    },
+    {
+      "epoch": 0.4362759175234047,
+      "grad_norm": 0.39007097482681274,
+      "learning_rate": 0.0001,
+      "loss": 1.4471,
+      "step": 3798
+    },
+    {
+      "epoch": 0.43639078743323184,
+      "grad_norm": 0.40956395864486694,
+      "learning_rate": 0.0001,
+      "loss": 1.5975,
+      "step": 3799
+    },
+    {
+      "epoch": 0.43650565734305896,
+      "grad_norm": 0.4181691110134125,
+      "learning_rate": 0.0001,
+      "loss": 1.7006,
+      "step": 3800
+    },
+    {
+      "epoch": 0.4366205272528861,
+      "grad_norm": 0.4084911644458771,
+      "learning_rate": 0.0001,
+      "loss": 1.7662,
+      "step": 3801
+    },
+    {
+      "epoch": 0.4367353971627132,
+      "grad_norm": 0.4404086470603943,
+      "learning_rate": 0.0001,
+      "loss": 1.7702,
+      "step": 3802
+    },
+    {
+      "epoch": 0.4368502670725403,
+      "grad_norm": 0.43644893169403076,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 3803
+    },
+    {
+      "epoch": 0.43696513698236744,
+      "grad_norm": 0.42488721013069153,
+      "learning_rate": 0.0001,
+      "loss": 1.795,
+      "step": 3804
+    },
+    {
+      "epoch": 0.4370800068921946,
+      "grad_norm": 0.4115133583545685,
+      "learning_rate": 0.0001,
+      "loss": 1.7059,
+      "step": 3805
+    },
+    {
+      "epoch": 0.43719487680202174,
+      "grad_norm": 0.4241204857826233,
+      "learning_rate": 0.0001,
+      "loss": 1.7201,
+      "step": 3806
+    },
+    {
+      "epoch": 0.43730974671184886,
+      "grad_norm": 0.3907410204410553,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 3807
+    },
+    {
+      "epoch": 0.437424616621676,
+      "grad_norm": 0.42004403471946716,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 3808
+    },
+    {
+      "epoch": 0.4375394865315031,
+      "grad_norm": 0.39130643010139465,
+      "learning_rate": 0.0001,
+      "loss": 1.5863,
+      "step": 3809
+    },
+    {
+      "epoch": 0.4376543564413302,
+      "grad_norm": 0.420646071434021,
+      "learning_rate": 0.0001,
+      "loss": 1.6687,
+      "step": 3810
+    },
+    {
+      "epoch": 0.43776922635115734,
+      "grad_norm": 0.43482112884521484,
+      "learning_rate": 0.0001,
+      "loss": 1.7814,
+      "step": 3811
+    },
+    {
+      "epoch": 0.43788409626098446,
+      "grad_norm": 0.4364151656627655,
+      "learning_rate": 0.0001,
+      "loss": 1.7378,
+      "step": 3812
+    },
+    {
+      "epoch": 0.4379989661708116,
+      "grad_norm": 0.44241318106651306,
+      "learning_rate": 0.0001,
+      "loss": 1.8158,
+      "step": 3813
+    },
+    {
+      "epoch": 0.4381138360806387,
+      "grad_norm": 0.3973522186279297,
+      "learning_rate": 0.0001,
+      "loss": 1.6807,
+      "step": 3814
+    },
+    {
+      "epoch": 0.4382287059904658,
+      "grad_norm": 0.42336922883987427,
+      "learning_rate": 0.0001,
+      "loss": 1.8086,
+      "step": 3815
+    },
+    {
+      "epoch": 0.43834357590029294,
+      "grad_norm": 0.40706026554107666,
+      "learning_rate": 0.0001,
+      "loss": 1.7805,
+      "step": 3816
+    },
+    {
+      "epoch": 0.43845844581012006,
+      "grad_norm": 0.4009002149105072,
+      "learning_rate": 0.0001,
+      "loss": 1.7439,
+      "step": 3817
+    },
+    {
+      "epoch": 0.4385733157199472,
+      "grad_norm": 0.38294774293899536,
+      "learning_rate": 0.0001,
+      "loss": 1.6179,
+      "step": 3818
+    },
+    {
+      "epoch": 0.4386881856297743,
+      "grad_norm": 0.40694519877433777,
+      "learning_rate": 0.0001,
+      "loss": 1.5189,
+      "step": 3819
+    },
+    {
+      "epoch": 0.4388030555396014,
+      "grad_norm": 0.3984982371330261,
+      "learning_rate": 0.0001,
+      "loss": 1.6288,
+      "step": 3820
+    },
+    {
+      "epoch": 0.43891792544942854,
+      "grad_norm": 0.44056373834609985,
+      "learning_rate": 0.0001,
+      "loss": 1.7716,
+      "step": 3821
+    },
+    {
+      "epoch": 0.43903279535925566,
+      "grad_norm": 0.41726037859916687,
+      "learning_rate": 0.0001,
+      "loss": 1.8127,
+      "step": 3822
+    },
+    {
+      "epoch": 0.4391476652690828,
+      "grad_norm": 0.4185958802700043,
+      "learning_rate": 0.0001,
+      "loss": 1.6932,
+      "step": 3823
+    },
+    {
+      "epoch": 0.4392625351789099,
+      "grad_norm": 0.3969833254814148,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 3824
+    },
+    {
+      "epoch": 0.439377405088737,
+      "grad_norm": 0.42625898122787476,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 3825
+    },
+    {
+      "epoch": 0.43949227499856414,
+      "grad_norm": 0.3792615532875061,
+      "learning_rate": 0.0001,
+      "loss": 1.5973,
+      "step": 3826
+    },
+    {
+      "epoch": 0.43960714490839126,
+      "grad_norm": 0.3736385107040405,
+      "learning_rate": 0.0001,
+      "loss": 1.3924,
+      "step": 3827
+    },
+    {
+      "epoch": 0.4397220148182184,
+      "grad_norm": 0.4105335772037506,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 3828
+    },
+    {
+      "epoch": 0.4398368847280455,
+      "grad_norm": 0.4096393585205078,
+      "learning_rate": 0.0001,
+      "loss": 1.722,
+      "step": 3829
+    },
+    {
+      "epoch": 0.4399517546378726,
+      "grad_norm": 0.3897015154361725,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 3830
+    },
+    {
+      "epoch": 0.44006662454769974,
+      "grad_norm": 0.4688592851161957,
+      "learning_rate": 0.0001,
+      "loss": 1.7616,
+      "step": 3831
+    },
+    {
+      "epoch": 0.44018149445752686,
+      "grad_norm": 0.4732673168182373,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 3832
+    },
+    {
+      "epoch": 0.440296364367354,
+      "grad_norm": 0.40647318959236145,
+      "learning_rate": 0.0001,
+      "loss": 1.745,
+      "step": 3833
+    },
+    {
+      "epoch": 0.4404112342771811,
+      "grad_norm": 0.38484612107276917,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 3834
+    },
+    {
+      "epoch": 0.4405261041870082,
+      "grad_norm": 0.39425164461135864,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 3835
+    },
+    {
+      "epoch": 0.44064097409683534,
+      "grad_norm": 0.38018524646759033,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 3836
+    },
+    {
+      "epoch": 0.44075584400666246,
+      "grad_norm": 0.40597987174987793,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 3837
+    },
+    {
+      "epoch": 0.4408707139164896,
+      "grad_norm": 0.4145064949989319,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 3838
+    },
+    {
+      "epoch": 0.4409855838263167,
+      "grad_norm": 0.41554126143455505,
+      "learning_rate": 0.0001,
+      "loss": 1.7004,
+      "step": 3839
+    },
+    {
+      "epoch": 0.4411004537361438,
+      "grad_norm": 0.3642127811908722,
+      "learning_rate": 0.0001,
+      "loss": 1.5098,
+      "step": 3840
+    },
+    {
+      "epoch": 0.44121532364597094,
+      "grad_norm": 0.40661707520484924,
+      "learning_rate": 0.0001,
+      "loss": 1.7808,
+      "step": 3841
+    },
+    {
+      "epoch": 0.44133019355579806,
+      "grad_norm": 0.42845696210861206,
+      "learning_rate": 0.0001,
+      "loss": 1.6614,
+      "step": 3842
+    },
+    {
+      "epoch": 0.4414450634656252,
+      "grad_norm": 0.43189069628715515,
+      "learning_rate": 0.0001,
+      "loss": 1.7699,
+      "step": 3843
+    },
+    {
+      "epoch": 0.4415599333754523,
+      "grad_norm": 0.42555564641952515,
+      "learning_rate": 0.0001,
+      "loss": 1.7836,
+      "step": 3844
+    },
+    {
+      "epoch": 0.4416748032852794,
+      "grad_norm": 0.4227694571018219,
+      "learning_rate": 0.0001,
+      "loss": 1.7315,
+      "step": 3845
+    },
+    {
+      "epoch": 0.44178967319510654,
+      "grad_norm": 0.4147578477859497,
+      "learning_rate": 0.0001,
+      "loss": 1.7124,
+      "step": 3846
+    },
+    {
+      "epoch": 0.44190454310493366,
+      "grad_norm": 0.39049941301345825,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 3847
+    },
+    {
+      "epoch": 0.4420194130147608,
+      "grad_norm": 0.4264180362224579,
+      "learning_rate": 0.0001,
+      "loss": 1.7393,
+      "step": 3848
+    },
+    {
+      "epoch": 0.4421342829245879,
+      "grad_norm": 0.416720449924469,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 3849
+    },
+    {
+      "epoch": 0.442249152834415,
+      "grad_norm": 0.4004881680011749,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 3850
+    },
+    {
+      "epoch": 0.44236402274424214,
+      "grad_norm": 0.3840608298778534,
+      "learning_rate": 0.0001,
+      "loss": 1.4561,
+      "step": 3851
+    },
+    {
+      "epoch": 0.44247889265406926,
+      "grad_norm": 0.398914098739624,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 3852
+    },
+    {
+      "epoch": 0.4425937625638964,
+      "grad_norm": 0.4018940329551697,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 3853
+    },
+    {
+      "epoch": 0.4427086324737235,
+      "grad_norm": 0.4263051450252533,
+      "learning_rate": 0.0001,
+      "loss": 1.6775,
+      "step": 3854
+    },
+    {
+      "epoch": 0.4428235023835506,
+      "grad_norm": 0.4021928608417511,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 3855
+    },
+    {
+      "epoch": 0.44293837229337774,
+      "grad_norm": 0.3968657851219177,
+      "learning_rate": 0.0001,
+      "loss": 1.4729,
+      "step": 3856
+    },
+    {
+      "epoch": 0.44305324220320486,
+      "grad_norm": 0.3775864541530609,
+      "learning_rate": 0.0001,
+      "loss": 1.6014,
+      "step": 3857
+    },
+    {
+      "epoch": 0.443168112113032,
+      "grad_norm": 0.3850625157356262,
+      "learning_rate": 0.0001,
+      "loss": 1.6046,
+      "step": 3858
+    },
+    {
+      "epoch": 0.4432829820228591,
+      "grad_norm": 0.4146713614463806,
+      "learning_rate": 0.0001,
+      "loss": 1.6827,
+      "step": 3859
+    },
+    {
+      "epoch": 0.4433978519326862,
+      "grad_norm": 0.3781043291091919,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 3860
+    },
+    {
+      "epoch": 0.44351272184251334,
+      "grad_norm": 0.38465172052383423,
+      "learning_rate": 0.0001,
+      "loss": 1.6866,
+      "step": 3861
+    },
+    {
+      "epoch": 0.44362759175234046,
+      "grad_norm": 0.41216936707496643,
+      "learning_rate": 0.0001,
+      "loss": 1.7803,
+      "step": 3862
+    },
+    {
+      "epoch": 0.4437424616621676,
+      "grad_norm": 0.4183763563632965,
+      "learning_rate": 0.0001,
+      "loss": 1.6628,
+      "step": 3863
+    },
+    {
+      "epoch": 0.4438573315719947,
+      "grad_norm": 0.39858147501945496,
+      "learning_rate": 0.0001,
+      "loss": 1.641,
+      "step": 3864
+    },
+    {
+      "epoch": 0.4439722014818218,
+      "grad_norm": 0.4076845347881317,
+      "learning_rate": 0.0001,
+      "loss": 1.7546,
+      "step": 3865
+    },
+    {
+      "epoch": 0.44408707139164894,
+      "grad_norm": 0.4265919625759125,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 3866
+    },
+    {
+      "epoch": 0.44420194130147606,
+      "grad_norm": 0.4252798855304718,
+      "learning_rate": 0.0001,
+      "loss": 1.698,
+      "step": 3867
+    },
+    {
+      "epoch": 0.4443168112113032,
+      "grad_norm": 0.3964008688926697,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 3868
+    },
+    {
+      "epoch": 0.4444316811211303,
+      "grad_norm": 0.3971204459667206,
+      "learning_rate": 0.0001,
+      "loss": 1.5393,
+      "step": 3869
+    },
+    {
+      "epoch": 0.4445465510309574,
+      "grad_norm": 0.39955177903175354,
+      "learning_rate": 0.0001,
+      "loss": 1.7192,
+      "step": 3870
+    },
+    {
+      "epoch": 0.44466142094078454,
+      "grad_norm": 0.40053296089172363,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 3871
+    },
+    {
+      "epoch": 0.44477629085061166,
+      "grad_norm": 0.3930230438709259,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 3872
+    },
+    {
+      "epoch": 0.44489116076043883,
+      "grad_norm": 0.41546955704689026,
+      "learning_rate": 0.0001,
+      "loss": 1.7292,
+      "step": 3873
+    },
+    {
+      "epoch": 0.44500603067026595,
+      "grad_norm": 0.40702396631240845,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 3874
+    },
+    {
+      "epoch": 0.44512090058009307,
+      "grad_norm": 0.451388955116272,
+      "learning_rate": 0.0001,
+      "loss": 1.8062,
+      "step": 3875
+    },
+    {
+      "epoch": 0.4452357704899202,
+      "grad_norm": 0.40768948197364807,
+      "learning_rate": 0.0001,
+      "loss": 1.6942,
+      "step": 3876
+    },
+    {
+      "epoch": 0.4453506403997473,
+      "grad_norm": 0.4409973919391632,
+      "learning_rate": 0.0001,
+      "loss": 1.7809,
+      "step": 3877
+    },
+    {
+      "epoch": 0.44546551030957443,
+      "grad_norm": 0.4067426323890686,
+      "learning_rate": 0.0001,
+      "loss": 1.6392,
+      "step": 3878
+    },
+    {
+      "epoch": 0.44558038021940155,
+      "grad_norm": 0.37034404277801514,
+      "learning_rate": 0.0001,
+      "loss": 1.6065,
+      "step": 3879
+    },
+    {
+      "epoch": 0.44569525012922867,
+      "grad_norm": 0.41960740089416504,
+      "learning_rate": 0.0001,
+      "loss": 1.9168,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4458101200390558,
+      "grad_norm": 0.40706804394721985,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 3881
+    },
+    {
+      "epoch": 0.4459249899488829,
+      "grad_norm": 0.38883423805236816,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 3882
+    },
+    {
+      "epoch": 0.44603985985871003,
+      "grad_norm": 0.3777417540550232,
+      "learning_rate": 0.0001,
+      "loss": 1.4643,
+      "step": 3883
+    },
+    {
+      "epoch": 0.44615472976853715,
+      "grad_norm": 0.3927326798439026,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 3884
+    },
+    {
+      "epoch": 0.44626959967836427,
+      "grad_norm": 0.39336463809013367,
+      "learning_rate": 0.0001,
+      "loss": 1.6599,
+      "step": 3885
+    },
+    {
+      "epoch": 0.4463844695881914,
+      "grad_norm": 0.38999754190444946,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 3886
+    },
+    {
+      "epoch": 0.4464993394980185,
+      "grad_norm": 0.37875038385391235,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 3887
+    },
+    {
+      "epoch": 0.44661420940784563,
+      "grad_norm": 0.4335520267486572,
+      "learning_rate": 0.0001,
+      "loss": 1.9779,
+      "step": 3888
+    },
+    {
+      "epoch": 0.44672907931767275,
+      "grad_norm": 0.461556613445282,
+      "learning_rate": 0.0001,
+      "loss": 1.9997,
+      "step": 3889
+    },
+    {
+      "epoch": 0.44684394922749987,
+      "grad_norm": 0.4445992410182953,
+      "learning_rate": 0.0001,
+      "loss": 2.0335,
+      "step": 3890
+    },
+    {
+      "epoch": 0.446958819137327,
+      "grad_norm": 0.39221543073654175,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 3891
+    },
+    {
+      "epoch": 0.4470736890471541,
+      "grad_norm": 0.3761284351348877,
+      "learning_rate": 0.0001,
+      "loss": 1.6046,
+      "step": 3892
+    },
+    {
+      "epoch": 0.44718855895698123,
+      "grad_norm": 0.473550409078598,
+      "learning_rate": 0.0001,
+      "loss": 1.9831,
+      "step": 3893
+    },
+    {
+      "epoch": 0.44730342886680835,
+      "grad_norm": 0.39917153120040894,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 3894
+    },
+    {
+      "epoch": 0.44741829877663547,
+      "grad_norm": 0.3896121680736542,
+      "learning_rate": 0.0001,
+      "loss": 1.5956,
+      "step": 3895
+    },
+    {
+      "epoch": 0.4475331686864626,
+      "grad_norm": 0.39092063903808594,
+      "learning_rate": 0.0001,
+      "loss": 1.5873,
+      "step": 3896
+    },
+    {
+      "epoch": 0.4476480385962897,
+      "grad_norm": 0.427628755569458,
+      "learning_rate": 0.0001,
+      "loss": 1.7593,
+      "step": 3897
+    },
+    {
+      "epoch": 0.44776290850611683,
+      "grad_norm": 0.4042544364929199,
+      "learning_rate": 0.0001,
+      "loss": 1.7025,
+      "step": 3898
+    },
+    {
+      "epoch": 0.44787777841594395,
+      "grad_norm": 0.3957765996456146,
+      "learning_rate": 0.0001,
+      "loss": 1.6169,
+      "step": 3899
+    },
+    {
+      "epoch": 0.44799264832577107,
+      "grad_norm": 0.41617143154144287,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 3900
+    },
+    {
+      "epoch": 0.4481075182355982,
+      "grad_norm": 0.4180884063243866,
+      "learning_rate": 0.0001,
+      "loss": 1.4391,
+      "step": 3901
+    },
+    {
+      "epoch": 0.4482223881454253,
+      "grad_norm": 0.38561975955963135,
+      "learning_rate": 0.0001,
+      "loss": 1.4084,
+      "step": 3902
+    },
+    {
+      "epoch": 0.44833725805525243,
+      "grad_norm": 0.4009731709957123,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 3903
+    },
+    {
+      "epoch": 0.44845212796507955,
+      "grad_norm": 0.4165489077568054,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 3904
+    },
+    {
+      "epoch": 0.44856699787490667,
+      "grad_norm": 0.41844233870506287,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 3905
+    },
+    {
+      "epoch": 0.4486818677847338,
+      "grad_norm": 0.4142603576183319,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 3906
+    },
+    {
+      "epoch": 0.4487967376945609,
+      "grad_norm": 0.41738462448120117,
+      "learning_rate": 0.0001,
+      "loss": 1.7553,
+      "step": 3907
+    },
+    {
+      "epoch": 0.44891160760438803,
+      "grad_norm": 0.40905338525772095,
+      "learning_rate": 0.0001,
+      "loss": 1.5517,
+      "step": 3908
+    },
+    {
+      "epoch": 0.44902647751421515,
+      "grad_norm": 0.4092770516872406,
+      "learning_rate": 0.0001,
+      "loss": 1.7373,
+      "step": 3909
+    },
+    {
+      "epoch": 0.44914134742404227,
+      "grad_norm": 0.3912902772426605,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4492562173338694,
+      "grad_norm": 0.42974889278411865,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 3911
+    },
+    {
+      "epoch": 0.4493710872436965,
+      "grad_norm": 0.4151856601238251,
+      "learning_rate": 0.0001,
+      "loss": 1.511,
+      "step": 3912
+    },
+    {
+      "epoch": 0.44948595715352363,
+      "grad_norm": 0.4048108458518982,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 3913
+    },
+    {
+      "epoch": 0.44960082706335075,
+      "grad_norm": 0.3995741307735443,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 3914
+    },
+    {
+      "epoch": 0.44971569697317787,
+      "grad_norm": 0.41176357865333557,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 3915
+    },
+    {
+      "epoch": 0.449830566883005,
+      "grad_norm": 0.4261864721775055,
+      "learning_rate": 0.0001,
+      "loss": 1.719,
+      "step": 3916
+    },
+    {
+      "epoch": 0.4499454367928321,
+      "grad_norm": 0.41965851187705994,
+      "learning_rate": 0.0001,
+      "loss": 1.8569,
+      "step": 3917
+    },
+    {
+      "epoch": 0.45006030670265923,
+      "grad_norm": 0.4024384617805481,
+      "learning_rate": 0.0001,
+      "loss": 1.682,
+      "step": 3918
+    },
+    {
+      "epoch": 0.45017517661248635,
+      "grad_norm": 0.42263585329055786,
+      "learning_rate": 0.0001,
+      "loss": 1.7975,
+      "step": 3919
+    },
+    {
+      "epoch": 0.45029004652231347,
+      "grad_norm": 0.37755391001701355,
+      "learning_rate": 0.0001,
+      "loss": 1.6863,
+      "step": 3920
+    },
+    {
+      "epoch": 0.4504049164321406,
+      "grad_norm": 0.4409322440624237,
+      "learning_rate": 0.0001,
+      "loss": 1.9805,
+      "step": 3921
+    },
+    {
+      "epoch": 0.4505197863419677,
+      "grad_norm": 0.4285867512226105,
+      "learning_rate": 0.0001,
+      "loss": 1.7172,
+      "step": 3922
+    },
+    {
+      "epoch": 0.45063465625179483,
+      "grad_norm": 0.4555914103984833,
+      "learning_rate": 0.0001,
+      "loss": 1.7151,
+      "step": 3923
+    },
+    {
+      "epoch": 0.45074952616162195,
+      "grad_norm": 0.399330198764801,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 3924
+    },
+    {
+      "epoch": 0.45086439607144907,
+      "grad_norm": 0.4097477197647095,
+      "learning_rate": 0.0001,
+      "loss": 1.7046,
+      "step": 3925
+    },
+    {
+      "epoch": 0.4509792659812762,
+      "grad_norm": 0.4111149311065674,
+      "learning_rate": 0.0001,
+      "loss": 1.6762,
+      "step": 3926
+    },
+    {
+      "epoch": 0.4510941358911033,
+      "grad_norm": 0.4010746479034424,
+      "learning_rate": 0.0001,
+      "loss": 1.2525,
+      "step": 3927
+    },
+    {
+      "epoch": 0.45120900580093043,
+      "grad_norm": 0.39972639083862305,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 3928
+    },
+    {
+      "epoch": 0.45132387571075755,
+      "grad_norm": 0.4085247218608856,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 3929
+    },
+    {
+      "epoch": 0.45143874562058467,
+      "grad_norm": 0.3858582377433777,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4515536155304118,
+      "grad_norm": 0.4039546251296997,
+      "learning_rate": 0.0001,
+      "loss": 1.4808,
+      "step": 3931
+    },
+    {
+      "epoch": 0.4516684854402389,
+      "grad_norm": 0.42480626702308655,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 3932
+    },
+    {
+      "epoch": 0.45178335535006603,
+      "grad_norm": 0.4410436451435089,
+      "learning_rate": 0.0001,
+      "loss": 1.7151,
+      "step": 3933
+    },
+    {
+      "epoch": 0.45189822525989315,
+      "grad_norm": 0.3930107057094574,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 3934
+    },
+    {
+      "epoch": 0.45201309516972027,
+      "grad_norm": 0.3937776982784271,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 3935
+    },
+    {
+      "epoch": 0.4521279650795474,
+      "grad_norm": 0.41633737087249756,
+      "learning_rate": 0.0001,
+      "loss": 1.9046,
+      "step": 3936
+    },
+    {
+      "epoch": 0.4522428349893745,
+      "grad_norm": 0.41131532192230225,
+      "learning_rate": 0.0001,
+      "loss": 1.7626,
+      "step": 3937
+    },
+    {
+      "epoch": 0.45235770489920163,
+      "grad_norm": 0.4197579026222229,
+      "learning_rate": 0.0001,
+      "loss": 1.686,
+      "step": 3938
+    },
+    {
+      "epoch": 0.45247257480902875,
+      "grad_norm": 0.4465826749801636,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 3939
+    },
+    {
+      "epoch": 0.45258744471885587,
+      "grad_norm": 0.3683403730392456,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 3940
+    },
+    {
+      "epoch": 0.452702314628683,
+      "grad_norm": 0.4257335066795349,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 3941
+    },
+    {
+      "epoch": 0.45281718453851016,
+      "grad_norm": 0.42209717631340027,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 3942
+    },
+    {
+      "epoch": 0.4529320544483373,
+      "grad_norm": 0.3728243112564087,
+      "learning_rate": 0.0001,
+      "loss": 1.3088,
+      "step": 3943
+    },
+    {
+      "epoch": 0.4530469243581644,
+      "grad_norm": 0.44413092732429504,
+      "learning_rate": 0.0001,
+      "loss": 1.7914,
+      "step": 3944
+    },
+    {
+      "epoch": 0.4531617942679915,
+      "grad_norm": 0.40472888946533203,
+      "learning_rate": 0.0001,
+      "loss": 1.4247,
+      "step": 3945
+    },
+    {
+      "epoch": 0.45327666417781864,
+      "grad_norm": 0.44987305998802185,
+      "learning_rate": 0.0001,
+      "loss": 1.8835,
+      "step": 3946
+    },
+    {
+      "epoch": 0.45339153408764576,
+      "grad_norm": 0.4118070602416992,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 3947
+    },
+    {
+      "epoch": 0.4535064039974729,
+      "grad_norm": 0.3929295241832733,
+      "learning_rate": 0.0001,
+      "loss": 1.6814,
+      "step": 3948
+    },
+    {
+      "epoch": 0.4536212739073,
+      "grad_norm": 0.43117576837539673,
+      "learning_rate": 0.0001,
+      "loss": 1.6748,
+      "step": 3949
+    },
+    {
+      "epoch": 0.4537361438171271,
+      "grad_norm": 0.434177041053772,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 3950
+    },
+    {
+      "epoch": 0.45385101372695424,
+      "grad_norm": 0.4338892698287964,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 3951
+    },
+    {
+      "epoch": 0.45396588363678136,
+      "grad_norm": 0.4370008409023285,
+      "learning_rate": 0.0001,
+      "loss": 1.6605,
+      "step": 3952
+    },
+    {
+      "epoch": 0.4540807535466085,
+      "grad_norm": 0.42377498745918274,
+      "learning_rate": 0.0001,
+      "loss": 1.6392,
+      "step": 3953
+    },
+    {
+      "epoch": 0.4541956234564356,
+      "grad_norm": 0.3817870318889618,
+      "learning_rate": 0.0001,
+      "loss": 1.5003,
+      "step": 3954
+    },
+    {
+      "epoch": 0.4543104933662627,
+      "grad_norm": 0.37805989384651184,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 3955
+    },
+    {
+      "epoch": 0.45442536327608984,
+      "grad_norm": 0.4458475708961487,
+      "learning_rate": 0.0001,
+      "loss": 1.8117,
+      "step": 3956
+    },
+    {
+      "epoch": 0.45454023318591696,
+      "grad_norm": 0.44833672046661377,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 3957
+    },
+    {
+      "epoch": 0.4546551030957441,
+      "grad_norm": 0.40606266260147095,
+      "learning_rate": 0.0001,
+      "loss": 1.7301,
+      "step": 3958
+    },
+    {
+      "epoch": 0.4547699730055712,
+      "grad_norm": 0.4011266529560089,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 3959
+    },
+    {
+      "epoch": 0.4548848429153983,
+      "grad_norm": 0.4372859299182892,
+      "learning_rate": 0.0001,
+      "loss": 1.7811,
+      "step": 3960
+    },
+    {
+      "epoch": 0.45499971282522544,
+      "grad_norm": 0.41698816418647766,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 3961
+    },
+    {
+      "epoch": 0.45511458273505256,
+      "grad_norm": 0.41979551315307617,
+      "learning_rate": 0.0001,
+      "loss": 1.8385,
+      "step": 3962
+    },
+    {
+      "epoch": 0.4552294526448797,
+      "grad_norm": 0.4077071249485016,
+      "learning_rate": 0.0001,
+      "loss": 1.6945,
+      "step": 3963
+    },
+    {
+      "epoch": 0.4553443225547068,
+      "grad_norm": 0.3627765476703644,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 3964
+    },
+    {
+      "epoch": 0.4554591924645339,
+      "grad_norm": 0.3952348828315735,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 3965
+    },
+    {
+      "epoch": 0.45557406237436104,
+      "grad_norm": 0.44601067900657654,
+      "learning_rate": 0.0001,
+      "loss": 1.8478,
+      "step": 3966
+    },
+    {
+      "epoch": 0.45568893228418816,
+      "grad_norm": 0.3901880085468292,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 3967
+    },
+    {
+      "epoch": 0.4558038021940153,
+      "grad_norm": 0.4016532599925995,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 3968
+    },
+    {
+      "epoch": 0.4559186721038424,
+      "grad_norm": 0.4067118763923645,
+      "learning_rate": 0.0001,
+      "loss": 1.4878,
+      "step": 3969
+    },
+    {
+      "epoch": 0.4560335420136695,
+      "grad_norm": 0.49588122963905334,
+      "learning_rate": 0.0001,
+      "loss": 1.8831,
+      "step": 3970
+    },
+    {
+      "epoch": 0.45614841192349664,
+      "grad_norm": 0.4378882348537445,
+      "learning_rate": 0.0001,
+      "loss": 1.7759,
+      "step": 3971
+    },
+    {
+      "epoch": 0.45626328183332376,
+      "grad_norm": 0.3988548517227173,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 3972
+    },
+    {
+      "epoch": 0.4563781517431509,
+      "grad_norm": 0.4102991819381714,
+      "learning_rate": 0.0001,
+      "loss": 1.6752,
+      "step": 3973
+    },
+    {
+      "epoch": 0.456493021652978,
+      "grad_norm": 0.41767358779907227,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 3974
+    },
+    {
+      "epoch": 0.4566078915628051,
+      "grad_norm": 0.40786105394363403,
+      "learning_rate": 0.0001,
+      "loss": 1.6357,
+      "step": 3975
+    },
+    {
+      "epoch": 0.45672276147263224,
+      "grad_norm": 0.4227571487426758,
+      "learning_rate": 0.0001,
+      "loss": 1.6638,
+      "step": 3976
+    },
+    {
+      "epoch": 0.45683763138245936,
+      "grad_norm": 0.4906644821166992,
+      "learning_rate": 0.0001,
+      "loss": 1.8612,
+      "step": 3977
+    },
+    {
+      "epoch": 0.4569525012922865,
+      "grad_norm": 0.4171657860279083,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 3978
+    },
+    {
+      "epoch": 0.4570673712021136,
+      "grad_norm": 0.5092725157737732,
+      "learning_rate": 0.0001,
+      "loss": 1.9038,
+      "step": 3979
+    },
+    {
+      "epoch": 0.4571822411119407,
+      "grad_norm": 0.3923867642879486,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 3980
+    },
+    {
+      "epoch": 0.45729711102176784,
+      "grad_norm": 0.41276079416275024,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 3981
+    },
+    {
+      "epoch": 0.45741198093159496,
+      "grad_norm": 0.41370689868927,
+      "learning_rate": 0.0001,
+      "loss": 1.6337,
+      "step": 3982
+    },
+    {
+      "epoch": 0.4575268508414221,
+      "grad_norm": 0.40904295444488525,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 3983
+    },
+    {
+      "epoch": 0.4576417207512492,
+      "grad_norm": 0.4055366516113281,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 3984
+    },
+    {
+      "epoch": 0.4577565906610763,
+      "grad_norm": 0.40704345703125,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 3985
+    },
+    {
+      "epoch": 0.45787146057090344,
+      "grad_norm": 0.395580530166626,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 3986
+    },
+    {
+      "epoch": 0.45798633048073056,
+      "grad_norm": 0.5622934103012085,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 3987
+    },
+    {
+      "epoch": 0.4581012003905577,
+      "grad_norm": 0.4147200584411621,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 3988
+    },
+    {
+      "epoch": 0.4582160703003848,
+      "grad_norm": 0.4014433026313782,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 3989
+    },
+    {
+      "epoch": 0.4583309402102119,
+      "grad_norm": 0.438319593667984,
+      "learning_rate": 0.0001,
+      "loss": 1.5005,
+      "step": 3990
+    },
+    {
+      "epoch": 0.45844581012003904,
+      "grad_norm": 0.4093727767467499,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 3991
+    },
+    {
+      "epoch": 0.45856068002986616,
+      "grad_norm": 0.40968188643455505,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 3992
+    },
+    {
+      "epoch": 0.4586755499396933,
+      "grad_norm": 0.43956807255744934,
+      "learning_rate": 0.0001,
+      "loss": 1.6531,
+      "step": 3993
+    },
+    {
+      "epoch": 0.4587904198495204,
+      "grad_norm": 0.40947967767715454,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 3994
+    },
+    {
+      "epoch": 0.4589052897593475,
+      "grad_norm": 0.40524375438690186,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 3995
+    },
+    {
+      "epoch": 0.45902015966917464,
+      "grad_norm": 0.4079066812992096,
+      "learning_rate": 0.0001,
+      "loss": 1.7688,
+      "step": 3996
+    },
+    {
+      "epoch": 0.45913502957900176,
+      "grad_norm": 0.40221965312957764,
+      "learning_rate": 0.0001,
+      "loss": 1.8436,
+      "step": 3997
+    },
+    {
+      "epoch": 0.4592498994888289,
+      "grad_norm": 0.4226916432380676,
+      "learning_rate": 0.0001,
+      "loss": 1.8167,
+      "step": 3998
+    },
+    {
+      "epoch": 0.459364769398656,
+      "grad_norm": 0.3903824985027313,
+      "learning_rate": 0.0001,
+      "loss": 1.7388,
+      "step": 3999
+    },
+    {
+      "epoch": 0.4594796393084831,
+      "grad_norm": 0.4148130416870117,
+      "learning_rate": 0.0001,
+      "loss": 1.8559,
+      "step": 4000
+    },
+    {
+      "epoch": 0.45959450921831024,
+      "grad_norm": 0.39708560705184937,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 4001
+    },
+    {
+      "epoch": 0.45970937912813736,
+      "grad_norm": 0.4335632920265198,
+      "learning_rate": 0.0001,
+      "loss": 1.7744,
+      "step": 4002
+    },
+    {
+      "epoch": 0.4598242490379645,
+      "grad_norm": 0.4105009436607361,
+      "learning_rate": 0.0001,
+      "loss": 1.7787,
+      "step": 4003
+    },
+    {
+      "epoch": 0.4599391189477916,
+      "grad_norm": 0.4119328558444977,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 4004
+    },
+    {
+      "epoch": 0.4600539888576187,
+      "grad_norm": 0.4249887466430664,
+      "learning_rate": 0.0001,
+      "loss": 1.6975,
+      "step": 4005
+    },
+    {
+      "epoch": 0.46016885876744584,
+      "grad_norm": 0.4140334129333496,
+      "learning_rate": 0.0001,
+      "loss": 1.6846,
+      "step": 4006
+    },
+    {
+      "epoch": 0.46028372867727296,
+      "grad_norm": 0.41865649819374084,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 4007
+    },
+    {
+      "epoch": 0.4603985985871001,
+      "grad_norm": 0.3861866295337677,
+      "learning_rate": 0.0001,
+      "loss": 1.3934,
+      "step": 4008
+    },
+    {
+      "epoch": 0.4605134684969272,
+      "grad_norm": 0.3830776512622833,
+      "learning_rate": 0.0001,
+      "loss": 1.6086,
+      "step": 4009
+    },
+    {
+      "epoch": 0.4606283384067543,
+      "grad_norm": 0.4259714186191559,
+      "learning_rate": 0.0001,
+      "loss": 1.7963,
+      "step": 4010
+    },
+    {
+      "epoch": 0.4607432083165815,
+      "grad_norm": 0.3694056570529938,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 4011
+    },
+    {
+      "epoch": 0.4608580782264086,
+      "grad_norm": 0.41248619556427,
+      "learning_rate": 0.0001,
+      "loss": 1.7967,
+      "step": 4012
+    },
+    {
+      "epoch": 0.46097294813623574,
+      "grad_norm": 0.40641582012176514,
+      "learning_rate": 0.0001,
+      "loss": 1.7708,
+      "step": 4013
+    },
+    {
+      "epoch": 0.46108781804606286,
+      "grad_norm": 0.43671518564224243,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 4014
+    },
+    {
+      "epoch": 0.46120268795589,
+      "grad_norm": 0.4056967496871948,
+      "learning_rate": 0.0001,
+      "loss": 1.7083,
+      "step": 4015
+    },
+    {
+      "epoch": 0.4613175578657171,
+      "grad_norm": 0.4197401702404022,
+      "learning_rate": 0.0001,
+      "loss": 1.763,
+      "step": 4016
+    },
+    {
+      "epoch": 0.4614324277755442,
+      "grad_norm": 0.38381877541542053,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 4017
+    },
+    {
+      "epoch": 0.46154729768537134,
+      "grad_norm": 0.39994266629219055,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 4018
+    },
+    {
+      "epoch": 0.46166216759519846,
+      "grad_norm": 0.40306952595710754,
+      "learning_rate": 0.0001,
+      "loss": 1.7447,
+      "step": 4019
+    },
+    {
+      "epoch": 0.4617770375050256,
+      "grad_norm": 0.4068452715873718,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 4020
+    },
+    {
+      "epoch": 0.4618919074148527,
+      "grad_norm": 0.3850820064544678,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 4021
+    },
+    {
+      "epoch": 0.4620067773246798,
+      "grad_norm": 0.420269250869751,
+      "learning_rate": 0.0001,
+      "loss": 1.814,
+      "step": 4022
+    },
+    {
+      "epoch": 0.46212164723450694,
+      "grad_norm": 0.404812216758728,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 4023
+    },
+    {
+      "epoch": 0.46223651714433406,
+      "grad_norm": 0.40154874324798584,
+      "learning_rate": 0.0001,
+      "loss": 1.7349,
+      "step": 4024
+    },
+    {
+      "epoch": 0.4623513870541612,
+      "grad_norm": 0.40221109986305237,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 4025
+    },
+    {
+      "epoch": 0.4624662569639883,
+      "grad_norm": 0.4642849564552307,
+      "learning_rate": 0.0001,
+      "loss": 1.6938,
+      "step": 4026
+    },
+    {
+      "epoch": 0.4625811268738154,
+      "grad_norm": 0.4198199212551117,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 4027
+    },
+    {
+      "epoch": 0.46269599678364254,
+      "grad_norm": 0.40827077627182007,
+      "learning_rate": 0.0001,
+      "loss": 1.7611,
+      "step": 4028
+    },
+    {
+      "epoch": 0.46281086669346966,
+      "grad_norm": 0.41351550817489624,
+      "learning_rate": 0.0001,
+      "loss": 1.6947,
+      "step": 4029
+    },
+    {
+      "epoch": 0.4629257366032968,
+      "grad_norm": 0.41555505990982056,
+      "learning_rate": 0.0001,
+      "loss": 1.7832,
+      "step": 4030
+    },
+    {
+      "epoch": 0.4630406065131239,
+      "grad_norm": 0.4276737570762634,
+      "learning_rate": 0.0001,
+      "loss": 1.7332,
+      "step": 4031
+    },
+    {
+      "epoch": 0.463155476422951,
+      "grad_norm": 0.39254918694496155,
+      "learning_rate": 0.0001,
+      "loss": 1.7283,
+      "step": 4032
+    },
+    {
+      "epoch": 0.46327034633277814,
+      "grad_norm": 0.40760576725006104,
+      "learning_rate": 0.0001,
+      "loss": 1.6464,
+      "step": 4033
+    },
+    {
+      "epoch": 0.46338521624260526,
+      "grad_norm": 0.3900972902774811,
+      "learning_rate": 0.0001,
+      "loss": 1.6067,
+      "step": 4034
+    },
+    {
+      "epoch": 0.4635000861524324,
+      "grad_norm": 0.42098426818847656,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 4035
+    },
+    {
+      "epoch": 0.4636149560622595,
+      "grad_norm": 0.41556769609451294,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 4036
+    },
+    {
+      "epoch": 0.4637298259720866,
+      "grad_norm": 0.4194001853466034,
+      "learning_rate": 0.0001,
+      "loss": 1.6771,
+      "step": 4037
+    },
+    {
+      "epoch": 0.46384469588191374,
+      "grad_norm": 0.4148116111755371,
+      "learning_rate": 0.0001,
+      "loss": 1.7292,
+      "step": 4038
+    },
+    {
+      "epoch": 0.46395956579174086,
+      "grad_norm": 0.39787450432777405,
+      "learning_rate": 0.0001,
+      "loss": 1.4533,
+      "step": 4039
+    },
+    {
+      "epoch": 0.464074435701568,
+      "grad_norm": 0.4328993856906891,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 4040
+    },
+    {
+      "epoch": 0.4641893056113951,
+      "grad_norm": 0.45193320512771606,
+      "learning_rate": 0.0001,
+      "loss": 1.9133,
+      "step": 4041
+    },
+    {
+      "epoch": 0.4643041755212222,
+      "grad_norm": 0.4238536059856415,
+      "learning_rate": 0.0001,
+      "loss": 1.7697,
+      "step": 4042
+    },
+    {
+      "epoch": 0.46441904543104934,
+      "grad_norm": 0.4080374836921692,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 4043
+    },
+    {
+      "epoch": 0.46453391534087646,
+      "grad_norm": 0.40985575318336487,
+      "learning_rate": 0.0001,
+      "loss": 1.742,
+      "step": 4044
+    },
+    {
+      "epoch": 0.4646487852507036,
+      "grad_norm": 0.40973904728889465,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 4045
+    },
+    {
+      "epoch": 0.4647636551605307,
+      "grad_norm": 0.4287373423576355,
+      "learning_rate": 0.0001,
+      "loss": 1.6476,
+      "step": 4046
+    },
+    {
+      "epoch": 0.4648785250703578,
+      "grad_norm": 0.41029754281044006,
+      "learning_rate": 0.0001,
+      "loss": 1.7405,
+      "step": 4047
+    },
+    {
+      "epoch": 0.46499339498018494,
+      "grad_norm": 0.4067857563495636,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 4048
+    },
+    {
+      "epoch": 0.46510826489001206,
+      "grad_norm": 0.42624571919441223,
+      "learning_rate": 0.0001,
+      "loss": 1.7049,
+      "step": 4049
+    },
+    {
+      "epoch": 0.4652231347998392,
+      "grad_norm": 0.3797012269496918,
+      "learning_rate": 0.0001,
+      "loss": 1.3009,
+      "step": 4050
+    },
+    {
+      "epoch": 0.4653380047096663,
+      "grad_norm": 0.37980806827545166,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 4051
+    },
+    {
+      "epoch": 0.4654528746194934,
+      "grad_norm": 0.42795923352241516,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 4052
+    },
+    {
+      "epoch": 0.46556774452932054,
+      "grad_norm": 0.4233475625514984,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 4053
+    },
+    {
+      "epoch": 0.46568261443914766,
+      "grad_norm": 0.3736623525619507,
+      "learning_rate": 0.0001,
+      "loss": 1.5268,
+      "step": 4054
+    },
+    {
+      "epoch": 0.4657974843489748,
+      "grad_norm": 0.43289387226104736,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 4055
+    },
+    {
+      "epoch": 0.4659123542588019,
+      "grad_norm": 0.39889997243881226,
+      "learning_rate": 0.0001,
+      "loss": 1.6901,
+      "step": 4056
+    },
+    {
+      "epoch": 0.466027224168629,
+      "grad_norm": 0.40544313192367554,
+      "learning_rate": 0.0001,
+      "loss": 1.7284,
+      "step": 4057
+    },
+    {
+      "epoch": 0.46614209407845614,
+      "grad_norm": 0.3909522294998169,
+      "learning_rate": 0.0001,
+      "loss": 1.5248,
+      "step": 4058
+    },
+    {
+      "epoch": 0.46625696398828326,
+      "grad_norm": 0.4280645549297333,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 4059
+    },
+    {
+      "epoch": 0.4663718338981104,
+      "grad_norm": 0.40522295236587524,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 4060
+    },
+    {
+      "epoch": 0.4664867038079375,
+      "grad_norm": 0.43785059452056885,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 4061
+    },
+    {
+      "epoch": 0.4666015737177646,
+      "grad_norm": 0.39573609828948975,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 4062
+    },
+    {
+      "epoch": 0.46671644362759174,
+      "grad_norm": 0.39389288425445557,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 4063
+    },
+    {
+      "epoch": 0.46683131353741886,
+      "grad_norm": 0.43171268701553345,
+      "learning_rate": 0.0001,
+      "loss": 1.8756,
+      "step": 4064
+    },
+    {
+      "epoch": 0.466946183447246,
+      "grad_norm": 0.4238601624965668,
+      "learning_rate": 0.0001,
+      "loss": 1.9024,
+      "step": 4065
+    },
+    {
+      "epoch": 0.4670610533570731,
+      "grad_norm": 0.4299545884132385,
+      "learning_rate": 0.0001,
+      "loss": 1.7863,
+      "step": 4066
+    },
+    {
+      "epoch": 0.4671759232669002,
+      "grad_norm": 0.38363876938819885,
+      "learning_rate": 0.0001,
+      "loss": 1.5237,
+      "step": 4067
+    },
+    {
+      "epoch": 0.46729079317672734,
+      "grad_norm": 0.39695143699645996,
+      "learning_rate": 0.0001,
+      "loss": 1.6357,
+      "step": 4068
+    },
+    {
+      "epoch": 0.46740566308655446,
+      "grad_norm": 0.4232119023799896,
+      "learning_rate": 0.0001,
+      "loss": 1.7294,
+      "step": 4069
+    },
+    {
+      "epoch": 0.4675205329963816,
+      "grad_norm": 0.39182764291763306,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 4070
+    },
+    {
+      "epoch": 0.4676354029062087,
+      "grad_norm": 0.38294535875320435,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 4071
+    },
+    {
+      "epoch": 0.4677502728160358,
+      "grad_norm": 0.39663416147232056,
+      "learning_rate": 0.0001,
+      "loss": 1.4622,
+      "step": 4072
+    },
+    {
+      "epoch": 0.46786514272586294,
+      "grad_norm": 0.4411957263946533,
+      "learning_rate": 0.0001,
+      "loss": 1.7604,
+      "step": 4073
+    },
+    {
+      "epoch": 0.46798001263569006,
+      "grad_norm": 0.43797844648361206,
+      "learning_rate": 0.0001,
+      "loss": 1.871,
+      "step": 4074
+    },
+    {
+      "epoch": 0.4680948825455172,
+      "grad_norm": 0.4069494605064392,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 4075
+    },
+    {
+      "epoch": 0.4682097524553443,
+      "grad_norm": 0.4068356454372406,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 4076
+    },
+    {
+      "epoch": 0.4683246223651714,
+      "grad_norm": 0.4451930820941925,
+      "learning_rate": 0.0001,
+      "loss": 1.6094,
+      "step": 4077
+    },
+    {
+      "epoch": 0.46843949227499854,
+      "grad_norm": 0.4141843616962433,
+      "learning_rate": 0.0001,
+      "loss": 1.6697,
+      "step": 4078
+    },
+    {
+      "epoch": 0.4685543621848257,
+      "grad_norm": 0.37937334179878235,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 4079
+    },
+    {
+      "epoch": 0.46866923209465283,
+      "grad_norm": 0.3883213996887207,
+      "learning_rate": 0.0001,
+      "loss": 1.7144,
+      "step": 4080
+    },
+    {
+      "epoch": 0.46878410200447995,
+      "grad_norm": 0.4085143506526947,
+      "learning_rate": 0.0001,
+      "loss": 1.6192,
+      "step": 4081
+    },
+    {
+      "epoch": 0.46889897191430707,
+      "grad_norm": 0.4008790850639343,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 4082
+    },
+    {
+      "epoch": 0.4690138418241342,
+      "grad_norm": 0.40579459071159363,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 4083
+    },
+    {
+      "epoch": 0.4691287117339613,
+      "grad_norm": 0.41397562623023987,
+      "learning_rate": 0.0001,
+      "loss": 1.7478,
+      "step": 4084
+    },
+    {
+      "epoch": 0.46924358164378843,
+      "grad_norm": 0.40979722142219543,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 4085
+    },
+    {
+      "epoch": 0.46935845155361555,
+      "grad_norm": 0.3959348201751709,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 4086
+    },
+    {
+      "epoch": 0.46947332146344267,
+      "grad_norm": 0.4095418453216553,
+      "learning_rate": 0.0001,
+      "loss": 1.6808,
+      "step": 4087
+    },
+    {
+      "epoch": 0.4695881913732698,
+      "grad_norm": 0.4053293466567993,
+      "learning_rate": 0.0001,
+      "loss": 1.6511,
+      "step": 4088
+    },
+    {
+      "epoch": 0.4697030612830969,
+      "grad_norm": 0.41647395491600037,
+      "learning_rate": 0.0001,
+      "loss": 1.7705,
+      "step": 4089
+    },
+    {
+      "epoch": 0.46981793119292403,
+      "grad_norm": 0.420941025018692,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 4090
+    },
+    {
+      "epoch": 0.46993280110275115,
+      "grad_norm": 0.4197264015674591,
+      "learning_rate": 0.0001,
+      "loss": 1.7921,
+      "step": 4091
+    },
+    {
+      "epoch": 0.47004767101257827,
+      "grad_norm": 0.40832772850990295,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 4092
+    },
+    {
+      "epoch": 0.4701625409224054,
+      "grad_norm": 0.4231039881706238,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 4093
+    },
+    {
+      "epoch": 0.4702774108322325,
+      "grad_norm": 0.4249723255634308,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 4094
+    },
+    {
+      "epoch": 0.47039228074205963,
+      "grad_norm": 0.41998720169067383,
+      "learning_rate": 0.0001,
+      "loss": 1.7629,
+      "step": 4095
+    },
+    {
+      "epoch": 0.47050715065188675,
+      "grad_norm": 0.38457614183425903,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 4096
+    },
+    {
+      "epoch": 0.47062202056171387,
+      "grad_norm": 0.3916681408882141,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 4097
+    },
+    {
+      "epoch": 0.470736890471541,
+      "grad_norm": 0.40213701128959656,
+      "learning_rate": 0.0001,
+      "loss": 1.6876,
+      "step": 4098
+    },
+    {
+      "epoch": 0.4708517603813681,
+      "grad_norm": 0.3882690668106079,
+      "learning_rate": 0.0001,
+      "loss": 1.6579,
+      "step": 4099
+    },
+    {
+      "epoch": 0.47096663029119523,
+      "grad_norm": 0.4170128107070923,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 4100
+    },
+    {
+      "epoch": 0.47108150020102235,
+      "grad_norm": 0.3979085087776184,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 4101
+    },
+    {
+      "epoch": 0.47119637011084947,
+      "grad_norm": 0.41884052753448486,
+      "learning_rate": 0.0001,
+      "loss": 1.7235,
+      "step": 4102
+    },
+    {
+      "epoch": 0.4713112400206766,
+      "grad_norm": 0.4105212688446045,
+      "learning_rate": 0.0001,
+      "loss": 1.6214,
+      "step": 4103
+    },
+    {
+      "epoch": 0.4714261099305037,
+      "grad_norm": 0.4247276186943054,
+      "learning_rate": 0.0001,
+      "loss": 1.7434,
+      "step": 4104
+    },
+    {
+      "epoch": 0.47154097984033083,
+      "grad_norm": 0.4294239580631256,
+      "learning_rate": 0.0001,
+      "loss": 1.7376,
+      "step": 4105
+    },
+    {
+      "epoch": 0.47165584975015795,
+      "grad_norm": 0.4285709857940674,
+      "learning_rate": 0.0001,
+      "loss": 1.7798,
+      "step": 4106
+    },
+    {
+      "epoch": 0.47177071965998507,
+      "grad_norm": 0.44262823462486267,
+      "learning_rate": 0.0001,
+      "loss": 1.8391,
+      "step": 4107
+    },
+    {
+      "epoch": 0.4718855895698122,
+      "grad_norm": 0.4055534601211548,
+      "learning_rate": 0.0001,
+      "loss": 1.6681,
+      "step": 4108
+    },
+    {
+      "epoch": 0.4720004594796393,
+      "grad_norm": 0.39059194922447205,
+      "learning_rate": 0.0001,
+      "loss": 1.7303,
+      "step": 4109
+    },
+    {
+      "epoch": 0.47211532938946643,
+      "grad_norm": 0.3832624554634094,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 4110
+    },
+    {
+      "epoch": 0.47223019929929355,
+      "grad_norm": 0.42001864314079285,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 4111
+    },
+    {
+      "epoch": 0.47234506920912067,
+      "grad_norm": 0.4122453033924103,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 4112
+    },
+    {
+      "epoch": 0.4724599391189478,
+      "grad_norm": 0.38872405886650085,
+      "learning_rate": 0.0001,
+      "loss": 1.5834,
+      "step": 4113
+    },
+    {
+      "epoch": 0.4725748090287749,
+      "grad_norm": 0.3932107388973236,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 4114
+    },
+    {
+      "epoch": 0.47268967893860203,
+      "grad_norm": 0.3980741798877716,
+      "learning_rate": 0.0001,
+      "loss": 1.6973,
+      "step": 4115
+    },
+    {
+      "epoch": 0.47280454884842915,
+      "grad_norm": 0.4428998827934265,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 4116
+    },
+    {
+      "epoch": 0.47291941875825627,
+      "grad_norm": 0.37861019372940063,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 4117
+    },
+    {
+      "epoch": 0.4730342886680834,
+      "grad_norm": 0.43132567405700684,
+      "learning_rate": 0.0001,
+      "loss": 1.8072,
+      "step": 4118
+    },
+    {
+      "epoch": 0.4731491585779105,
+      "grad_norm": 0.4315222501754761,
+      "learning_rate": 0.0001,
+      "loss": 1.7475,
+      "step": 4119
+    },
+    {
+      "epoch": 0.47326402848773763,
+      "grad_norm": 0.3742518424987793,
+      "learning_rate": 0.0001,
+      "loss": 1.4972,
+      "step": 4120
+    },
+    {
+      "epoch": 0.47337889839756475,
+      "grad_norm": 0.39365971088409424,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 4121
+    },
+    {
+      "epoch": 0.47349376830739187,
+      "grad_norm": 0.43931326270103455,
+      "learning_rate": 0.0001,
+      "loss": 1.7912,
+      "step": 4122
+    },
+    {
+      "epoch": 0.473608638217219,
+      "grad_norm": 0.4146181046962738,
+      "learning_rate": 0.0001,
+      "loss": 1.7775,
+      "step": 4123
+    },
+    {
+      "epoch": 0.4737235081270461,
+      "grad_norm": 0.4833734929561615,
+      "learning_rate": 0.0001,
+      "loss": 1.8526,
+      "step": 4124
+    },
+    {
+      "epoch": 0.47383837803687323,
+      "grad_norm": 0.38752907514572144,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 4125
+    },
+    {
+      "epoch": 0.47395324794670035,
+      "grad_norm": 0.38719457387924194,
+      "learning_rate": 0.0001,
+      "loss": 1.3979,
+      "step": 4126
+    },
+    {
+      "epoch": 0.47406811785652747,
+      "grad_norm": 0.4148174226284027,
+      "learning_rate": 0.0001,
+      "loss": 1.7373,
+      "step": 4127
+    },
+    {
+      "epoch": 0.4741829877663546,
+      "grad_norm": 0.41567835211753845,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 4128
+    },
+    {
+      "epoch": 0.4742978576761817,
+      "grad_norm": 0.45437824726104736,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 4129
+    },
+    {
+      "epoch": 0.47441272758600883,
+      "grad_norm": 0.41790398955345154,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 4130
+    },
+    {
+      "epoch": 0.47452759749583595,
+      "grad_norm": 0.40303486585617065,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 4131
+    },
+    {
+      "epoch": 0.47464246740566307,
+      "grad_norm": 0.391946405172348,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 4132
+    },
+    {
+      "epoch": 0.4747573373154902,
+      "grad_norm": 0.3910079896450043,
+      "learning_rate": 0.0001,
+      "loss": 1.6995,
+      "step": 4133
+    },
+    {
+      "epoch": 0.4748722072253173,
+      "grad_norm": 0.3853004276752472,
+      "learning_rate": 0.0001,
+      "loss": 1.6034,
+      "step": 4134
+    },
+    {
+      "epoch": 0.47498707713514443,
+      "grad_norm": 0.41160422563552856,
+      "learning_rate": 0.0001,
+      "loss": 1.7027,
+      "step": 4135
+    },
+    {
+      "epoch": 0.47510194704497155,
+      "grad_norm": 0.40281742811203003,
+      "learning_rate": 0.0001,
+      "loss": 1.7533,
+      "step": 4136
+    },
+    {
+      "epoch": 0.47521681695479867,
+      "grad_norm": 0.39282864332199097,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 4137
+    },
+    {
+      "epoch": 0.4753316868646258,
+      "grad_norm": 0.3944832384586334,
+      "learning_rate": 0.0001,
+      "loss": 1.5693,
+      "step": 4138
+    },
+    {
+      "epoch": 0.4754465567744529,
+      "grad_norm": 0.4138934910297394,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 4139
+    },
+    {
+      "epoch": 0.47556142668428003,
+      "grad_norm": 0.40532898902893066,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 4140
+    },
+    {
+      "epoch": 0.47567629659410715,
+      "grad_norm": 0.3844515383243561,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 4141
+    },
+    {
+      "epoch": 0.47579116650393427,
+      "grad_norm": 0.3857368230819702,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 4142
+    },
+    {
+      "epoch": 0.4759060364137614,
+      "grad_norm": 0.42560672760009766,
+      "learning_rate": 0.0001,
+      "loss": 1.8693,
+      "step": 4143
+    },
+    {
+      "epoch": 0.4760209063235885,
+      "grad_norm": 0.43299341201782227,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 4144
+    },
+    {
+      "epoch": 0.47613577623341563,
+      "grad_norm": 0.431902676820755,
+      "learning_rate": 0.0001,
+      "loss": 1.8618,
+      "step": 4145
+    },
+    {
+      "epoch": 0.47625064614324275,
+      "grad_norm": 0.43173423409461975,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 4146
+    },
+    {
+      "epoch": 0.47636551605306987,
+      "grad_norm": 0.4200470447540283,
+      "learning_rate": 0.0001,
+      "loss": 1.7881,
+      "step": 4147
+    },
+    {
+      "epoch": 0.47648038596289705,
+      "grad_norm": 0.3650665581226349,
+      "learning_rate": 0.0001,
+      "loss": 1.2449,
+      "step": 4148
+    },
+    {
+      "epoch": 0.47659525587272417,
+      "grad_norm": 0.40055355429649353,
+      "learning_rate": 0.0001,
+      "loss": 1.4084,
+      "step": 4149
+    },
+    {
+      "epoch": 0.4767101257825513,
+      "grad_norm": 0.41404348611831665,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4768249956923784,
+      "grad_norm": 0.42411214113235474,
+      "learning_rate": 0.0001,
+      "loss": 1.4006,
+      "step": 4151
+    },
+    {
+      "epoch": 0.4769398656022055,
+      "grad_norm": 0.4193376302719116,
+      "learning_rate": 0.0001,
+      "loss": 1.7433,
+      "step": 4152
+    },
+    {
+      "epoch": 0.47705473551203265,
+      "grad_norm": 0.41245442628860474,
+      "learning_rate": 0.0001,
+      "loss": 1.5723,
+      "step": 4153
+    },
+    {
+      "epoch": 0.47716960542185977,
+      "grad_norm": 0.4130132496356964,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 4154
+    },
+    {
+      "epoch": 0.4772844753316869,
+      "grad_norm": 0.3704391419887543,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 4155
+    },
+    {
+      "epoch": 0.477399345241514,
+      "grad_norm": 0.4236399531364441,
+      "learning_rate": 0.0001,
+      "loss": 1.7303,
+      "step": 4156
+    },
+    {
+      "epoch": 0.4775142151513411,
+      "grad_norm": 0.4399377703666687,
+      "learning_rate": 0.0001,
+      "loss": 1.8676,
+      "step": 4157
+    },
+    {
+      "epoch": 0.47762908506116825,
+      "grad_norm": 0.4146808683872223,
+      "learning_rate": 0.0001,
+      "loss": 1.4918,
+      "step": 4158
+    },
+    {
+      "epoch": 0.47774395497099537,
+      "grad_norm": 0.40414994955062866,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 4159
+    },
+    {
+      "epoch": 0.4778588248808225,
+      "grad_norm": 0.44139429926872253,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4779736947906496,
+      "grad_norm": 0.35786864161491394,
+      "learning_rate": 0.0001,
+      "loss": 1.3425,
+      "step": 4161
+    },
+    {
+      "epoch": 0.4780885647004767,
+      "grad_norm": 0.4114287793636322,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 4162
+    },
+    {
+      "epoch": 0.47820343461030385,
+      "grad_norm": 0.436565101146698,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 4163
+    },
+    {
+      "epoch": 0.47831830452013097,
+      "grad_norm": 0.42503979802131653,
+      "learning_rate": 0.0001,
+      "loss": 1.6941,
+      "step": 4164
+    },
+    {
+      "epoch": 0.4784331744299581,
+      "grad_norm": 0.4688960313796997,
+      "learning_rate": 0.0001,
+      "loss": 1.6121,
+      "step": 4165
+    },
+    {
+      "epoch": 0.4785480443397852,
+      "grad_norm": 0.42915526032447815,
+      "learning_rate": 0.0001,
+      "loss": 1.7311,
+      "step": 4166
+    },
+    {
+      "epoch": 0.4786629142496123,
+      "grad_norm": 0.42920050024986267,
+      "learning_rate": 0.0001,
+      "loss": 1.8337,
+      "step": 4167
+    },
+    {
+      "epoch": 0.47877778415943945,
+      "grad_norm": 0.4068220257759094,
+      "learning_rate": 0.0001,
+      "loss": 1.2956,
+      "step": 4168
+    },
+    {
+      "epoch": 0.47889265406926657,
+      "grad_norm": 0.38509076833724976,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 4169
+    },
+    {
+      "epoch": 0.4790075239790937,
+      "grad_norm": 0.43463391065597534,
+      "learning_rate": 0.0001,
+      "loss": 1.8903,
+      "step": 4170
+    },
+    {
+      "epoch": 0.4791223938889208,
+      "grad_norm": 0.4463506042957306,
+      "learning_rate": 0.0001,
+      "loss": 1.8181,
+      "step": 4171
+    },
+    {
+      "epoch": 0.4792372637987479,
+      "grad_norm": 0.42022788524627686,
+      "learning_rate": 0.0001,
+      "loss": 1.8081,
+      "step": 4172
+    },
+    {
+      "epoch": 0.47935213370857505,
+      "grad_norm": 0.4187469780445099,
+      "learning_rate": 0.0001,
+      "loss": 1.6547,
+      "step": 4173
+    },
+    {
+      "epoch": 0.47946700361840217,
+      "grad_norm": 0.3888328969478607,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 4174
+    },
+    {
+      "epoch": 0.4795818735282293,
+      "grad_norm": 0.4130741059780121,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 4175
+    },
+    {
+      "epoch": 0.4796967434380564,
+      "grad_norm": 0.3948152959346771,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 4176
+    },
+    {
+      "epoch": 0.4798116133478835,
+      "grad_norm": 0.3814520537853241,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 4177
+    },
+    {
+      "epoch": 0.47992648325771065,
+      "grad_norm": 0.4245845079421997,
+      "learning_rate": 0.0001,
+      "loss": 1.6257,
+      "step": 4178
+    },
+    {
+      "epoch": 0.48004135316753777,
+      "grad_norm": 0.39747804403305054,
+      "learning_rate": 0.0001,
+      "loss": 1.4679,
+      "step": 4179
+    },
+    {
+      "epoch": 0.4801562230773649,
+      "grad_norm": 0.41911858320236206,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 4180
+    },
+    {
+      "epoch": 0.480271092987192,
+      "grad_norm": 0.4305039048194885,
+      "learning_rate": 0.0001,
+      "loss": 1.7368,
+      "step": 4181
+    },
+    {
+      "epoch": 0.4803859628970191,
+      "grad_norm": 0.3938789665699005,
+      "learning_rate": 0.0001,
+      "loss": 1.4452,
+      "step": 4182
+    },
+    {
+      "epoch": 0.48050083280684625,
+      "grad_norm": 0.45596420764923096,
+      "learning_rate": 0.0001,
+      "loss": 1.697,
+      "step": 4183
+    },
+    {
+      "epoch": 0.48061570271667337,
+      "grad_norm": 0.3951970040798187,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 4184
+    },
+    {
+      "epoch": 0.4807305726265005,
+      "grad_norm": 0.40332433581352234,
+      "learning_rate": 0.0001,
+      "loss": 1.7118,
+      "step": 4185
+    },
+    {
+      "epoch": 0.4808454425363276,
+      "grad_norm": 0.4350488781929016,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 4186
+    },
+    {
+      "epoch": 0.4809603124461547,
+      "grad_norm": 0.4046489894390106,
+      "learning_rate": 0.0001,
+      "loss": 1.2882,
+      "step": 4187
+    },
+    {
+      "epoch": 0.48107518235598185,
+      "grad_norm": 0.4263879358768463,
+      "learning_rate": 0.0001,
+      "loss": 1.6298,
+      "step": 4188
+    },
+    {
+      "epoch": 0.48119005226580897,
+      "grad_norm": 0.42358294129371643,
+      "learning_rate": 0.0001,
+      "loss": 1.791,
+      "step": 4189
+    },
+    {
+      "epoch": 0.4813049221756361,
+      "grad_norm": 0.3843347430229187,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 4190
+    },
+    {
+      "epoch": 0.4814197920854632,
+      "grad_norm": 0.43530306220054626,
+      "learning_rate": 0.0001,
+      "loss": 1.7777,
+      "step": 4191
+    },
+    {
+      "epoch": 0.4815346619952903,
+      "grad_norm": 0.4170604646205902,
+      "learning_rate": 0.0001,
+      "loss": 1.6756,
+      "step": 4192
+    },
+    {
+      "epoch": 0.48164953190511745,
+      "grad_norm": 0.39548924565315247,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 4193
+    },
+    {
+      "epoch": 0.48176440181494457,
+      "grad_norm": 0.38482916355133057,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 4194
+    },
+    {
+      "epoch": 0.4818792717247717,
+      "grad_norm": 0.38123294711112976,
+      "learning_rate": 0.0001,
+      "loss": 1.4144,
+      "step": 4195
+    },
+    {
+      "epoch": 0.4819941416345988,
+      "grad_norm": 0.4207984209060669,
+      "learning_rate": 0.0001,
+      "loss": 1.6567,
+      "step": 4196
+    },
+    {
+      "epoch": 0.4821090115444259,
+      "grad_norm": 0.40305450558662415,
+      "learning_rate": 0.0001,
+      "loss": 1.6715,
+      "step": 4197
+    },
+    {
+      "epoch": 0.48222388145425304,
+      "grad_norm": 0.3921899199485779,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 4198
+    },
+    {
+      "epoch": 0.48233875136408016,
+      "grad_norm": 0.3937564790248871,
+      "learning_rate": 0.0001,
+      "loss": 1.4346,
+      "step": 4199
+    },
+    {
+      "epoch": 0.4824536212739073,
+      "grad_norm": 0.44775843620300293,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 4200
+    },
+    {
+      "epoch": 0.4825684911837344,
+      "grad_norm": 0.424165278673172,
+      "learning_rate": 0.0001,
+      "loss": 1.632,
+      "step": 4201
+    },
+    {
+      "epoch": 0.4826833610935615,
+      "grad_norm": 0.4116969406604767,
+      "learning_rate": 0.0001,
+      "loss": 1.7062,
+      "step": 4202
+    },
+    {
+      "epoch": 0.48279823100338864,
+      "grad_norm": 0.4231579303741455,
+      "learning_rate": 0.0001,
+      "loss": 1.8064,
+      "step": 4203
+    },
+    {
+      "epoch": 0.48291310091321576,
+      "grad_norm": 0.4327984154224396,
+      "learning_rate": 0.0001,
+      "loss": 1.7143,
+      "step": 4204
+    },
+    {
+      "epoch": 0.4830279708230429,
+      "grad_norm": 0.45221564173698425,
+      "learning_rate": 0.0001,
+      "loss": 1.7857,
+      "step": 4205
+    },
+    {
+      "epoch": 0.48314284073287,
+      "grad_norm": 0.4953210651874542,
+      "learning_rate": 0.0001,
+      "loss": 1.861,
+      "step": 4206
+    },
+    {
+      "epoch": 0.4832577106426971,
+      "grad_norm": 0.38952094316482544,
+      "learning_rate": 0.0001,
+      "loss": 1.4247,
+      "step": 4207
+    },
+    {
+      "epoch": 0.48337258055252424,
+      "grad_norm": 0.4227844774723053,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 4208
+    },
+    {
+      "epoch": 0.48348745046235136,
+      "grad_norm": 0.4414394199848175,
+      "learning_rate": 0.0001,
+      "loss": 1.9377,
+      "step": 4209
+    },
+    {
+      "epoch": 0.4836023203721785,
+      "grad_norm": 0.4291560649871826,
+      "learning_rate": 0.0001,
+      "loss": 1.4077,
+      "step": 4210
+    },
+    {
+      "epoch": 0.4837171902820056,
+      "grad_norm": 0.40819114446640015,
+      "learning_rate": 0.0001,
+      "loss": 1.8186,
+      "step": 4211
+    },
+    {
+      "epoch": 0.4838320601918327,
+      "grad_norm": 0.40871623158454895,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 4212
+    },
+    {
+      "epoch": 0.48394693010165984,
+      "grad_norm": 0.41948193311691284,
+      "learning_rate": 0.0001,
+      "loss": 1.7082,
+      "step": 4213
+    },
+    {
+      "epoch": 0.48406180001148696,
+      "grad_norm": 0.40412595868110657,
+      "learning_rate": 0.0001,
+      "loss": 1.6625,
+      "step": 4214
+    },
+    {
+      "epoch": 0.4841766699213141,
+      "grad_norm": 0.4160853624343872,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 4215
+    },
+    {
+      "epoch": 0.48429153983114126,
+      "grad_norm": 0.3837149143218994,
+      "learning_rate": 0.0001,
+      "loss": 1.4121,
+      "step": 4216
+    },
+    {
+      "epoch": 0.4844064097409684,
+      "grad_norm": 0.443243145942688,
+      "learning_rate": 0.0001,
+      "loss": 1.8064,
+      "step": 4217
+    },
+    {
+      "epoch": 0.4845212796507955,
+      "grad_norm": 0.38426709175109863,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 4218
+    },
+    {
+      "epoch": 0.4846361495606226,
+      "grad_norm": 0.40559226274490356,
+      "learning_rate": 0.0001,
+      "loss": 1.6694,
+      "step": 4219
+    },
+    {
+      "epoch": 0.48475101947044974,
+      "grad_norm": 0.4205404222011566,
+      "learning_rate": 0.0001,
+      "loss": 1.8064,
+      "step": 4220
+    },
+    {
+      "epoch": 0.48486588938027686,
+      "grad_norm": 0.4019322693347931,
+      "learning_rate": 0.0001,
+      "loss": 1.6447,
+      "step": 4221
+    },
+    {
+      "epoch": 0.484980759290104,
+      "grad_norm": 0.43610960245132446,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 4222
+    },
+    {
+      "epoch": 0.4850956291999311,
+      "grad_norm": 0.45350828766822815,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 4223
+    },
+    {
+      "epoch": 0.4852104991097582,
+      "grad_norm": 0.420280396938324,
+      "learning_rate": 0.0001,
+      "loss": 1.8019,
+      "step": 4224
+    },
+    {
+      "epoch": 0.48532536901958534,
+      "grad_norm": 0.4381343424320221,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 4225
+    },
+    {
+      "epoch": 0.48544023892941246,
+      "grad_norm": 0.38392412662506104,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 4226
+    },
+    {
+      "epoch": 0.4855551088392396,
+      "grad_norm": 0.4646781086921692,
+      "learning_rate": 0.0001,
+      "loss": 1.8663,
+      "step": 4227
+    },
+    {
+      "epoch": 0.4856699787490667,
+      "grad_norm": 0.39009374380111694,
+      "learning_rate": 0.0001,
+      "loss": 1.5679,
+      "step": 4228
+    },
+    {
+      "epoch": 0.4857848486588938,
+      "grad_norm": 0.44190630316734314,
+      "learning_rate": 0.0001,
+      "loss": 1.891,
+      "step": 4229
+    },
+    {
+      "epoch": 0.48589971856872094,
+      "grad_norm": 0.38428959250450134,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 4230
+    },
+    {
+      "epoch": 0.48601458847854806,
+      "grad_norm": 0.3903064429759979,
+      "learning_rate": 0.0001,
+      "loss": 1.4688,
+      "step": 4231
+    },
+    {
+      "epoch": 0.4861294583883752,
+      "grad_norm": 0.39627987146377563,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 4232
+    },
+    {
+      "epoch": 0.4862443282982023,
+      "grad_norm": 0.3723553419113159,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 4233
+    },
+    {
+      "epoch": 0.4863591982080294,
+      "grad_norm": 0.4561592936515808,
+      "learning_rate": 0.0001,
+      "loss": 1.8449,
+      "step": 4234
+    },
+    {
+      "epoch": 0.48647406811785654,
+      "grad_norm": 0.42943283915519714,
+      "learning_rate": 0.0001,
+      "loss": 1.6589,
+      "step": 4235
+    },
+    {
+      "epoch": 0.48658893802768366,
+      "grad_norm": 0.39778202772140503,
+      "learning_rate": 0.0001,
+      "loss": 1.4218,
+      "step": 4236
+    },
+    {
+      "epoch": 0.4867038079375108,
+      "grad_norm": 0.4467572569847107,
+      "learning_rate": 0.0001,
+      "loss": 1.7572,
+      "step": 4237
+    },
+    {
+      "epoch": 0.4868186778473379,
+      "grad_norm": 0.393715500831604,
+      "learning_rate": 0.0001,
+      "loss": 1.56,
+      "step": 4238
+    },
+    {
+      "epoch": 0.486933547757165,
+      "grad_norm": 0.43201369047164917,
+      "learning_rate": 0.0001,
+      "loss": 1.8393,
+      "step": 4239
+    },
+    {
+      "epoch": 0.48704841766699214,
+      "grad_norm": 0.3823792040348053,
+      "learning_rate": 0.0001,
+      "loss": 1.3971,
+      "step": 4240
+    },
+    {
+      "epoch": 0.48716328757681926,
+      "grad_norm": 0.38682422041893005,
+      "learning_rate": 0.0001,
+      "loss": 1.3362,
+      "step": 4241
+    },
+    {
+      "epoch": 0.4872781574866464,
+      "grad_norm": 0.405185341835022,
+      "learning_rate": 0.0001,
+      "loss": 1.5936,
+      "step": 4242
+    },
+    {
+      "epoch": 0.4873930273964735,
+      "grad_norm": 0.4169776141643524,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 4243
+    },
+    {
+      "epoch": 0.4875078973063006,
+      "grad_norm": 0.399635910987854,
+      "learning_rate": 0.0001,
+      "loss": 1.4352,
+      "step": 4244
+    },
+    {
+      "epoch": 0.48762276721612774,
+      "grad_norm": 0.39507734775543213,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 4245
+    },
+    {
+      "epoch": 0.48773763712595486,
+      "grad_norm": 0.42992520332336426,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 4246
+    },
+    {
+      "epoch": 0.487852507035782,
+      "grad_norm": 0.4018322825431824,
+      "learning_rate": 0.0001,
+      "loss": 1.8032,
+      "step": 4247
+    },
+    {
+      "epoch": 0.4879673769456091,
+      "grad_norm": 0.40367868542671204,
+      "learning_rate": 0.0001,
+      "loss": 1.6862,
+      "step": 4248
+    },
+    {
+      "epoch": 0.4880822468554362,
+      "grad_norm": 0.4355872571468353,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 4249
+    },
+    {
+      "epoch": 0.48819711676526334,
+      "grad_norm": 0.4191339313983917,
+      "learning_rate": 0.0001,
+      "loss": 1.8388,
+      "step": 4250
+    },
+    {
+      "epoch": 0.48831198667509046,
+      "grad_norm": 0.39485347270965576,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 4251
+    },
+    {
+      "epoch": 0.4884268565849176,
+      "grad_norm": 0.40303096175193787,
+      "learning_rate": 0.0001,
+      "loss": 1.718,
+      "step": 4252
+    },
+    {
+      "epoch": 0.4885417264947447,
+      "grad_norm": 0.4248945116996765,
+      "learning_rate": 0.0001,
+      "loss": 1.6762,
+      "step": 4253
+    },
+    {
+      "epoch": 0.4886565964045718,
+      "grad_norm": 0.45680445432662964,
+      "learning_rate": 0.0001,
+      "loss": 1.8161,
+      "step": 4254
+    },
+    {
+      "epoch": 0.48877146631439894,
+      "grad_norm": 0.40725842118263245,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 4255
+    },
+    {
+      "epoch": 0.48888633622422606,
+      "grad_norm": 0.4762290418148041,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 4256
+    },
+    {
+      "epoch": 0.4890012061340532,
+      "grad_norm": 0.42993873357772827,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 4257
+    },
+    {
+      "epoch": 0.4891160760438803,
+      "grad_norm": 0.38512858748435974,
+      "learning_rate": 0.0001,
+      "loss": 1.6684,
+      "step": 4258
+    },
+    {
+      "epoch": 0.4892309459537074,
+      "grad_norm": 0.4009222984313965,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 4259
+    },
+    {
+      "epoch": 0.48934581586353454,
+      "grad_norm": 0.3873811662197113,
+      "learning_rate": 0.0001,
+      "loss": 1.6713,
+      "step": 4260
+    },
+    {
+      "epoch": 0.48946068577336166,
+      "grad_norm": 0.39984700083732605,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 4261
+    },
+    {
+      "epoch": 0.4895755556831888,
+      "grad_norm": 0.4596845209598541,
+      "learning_rate": 0.0001,
+      "loss": 1.7131,
+      "step": 4262
+    },
+    {
+      "epoch": 0.4896904255930159,
+      "grad_norm": 0.4038037657737732,
+      "learning_rate": 0.0001,
+      "loss": 1.7656,
+      "step": 4263
+    },
+    {
+      "epoch": 0.489805295502843,
+      "grad_norm": 0.39577624201774597,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 4264
+    },
+    {
+      "epoch": 0.48992016541267014,
+      "grad_norm": 0.38728025555610657,
+      "learning_rate": 0.0001,
+      "loss": 1.5229,
+      "step": 4265
+    },
+    {
+      "epoch": 0.49003503532249726,
+      "grad_norm": 0.4493556618690491,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 4266
+    },
+    {
+      "epoch": 0.4901499052323244,
+      "grad_norm": 0.4298556447029114,
+      "learning_rate": 0.0001,
+      "loss": 1.8001,
+      "step": 4267
+    },
+    {
+      "epoch": 0.4902647751421515,
+      "grad_norm": 0.4090805947780609,
+      "learning_rate": 0.0001,
+      "loss": 1.6775,
+      "step": 4268
+    },
+    {
+      "epoch": 0.4903796450519786,
+      "grad_norm": 0.4103395640850067,
+      "learning_rate": 0.0001,
+      "loss": 1.6592,
+      "step": 4269
+    },
+    {
+      "epoch": 0.49049451496180574,
+      "grad_norm": 0.41509976983070374,
+      "learning_rate": 0.0001,
+      "loss": 1.6015,
+      "step": 4270
+    },
+    {
+      "epoch": 0.49060938487163286,
+      "grad_norm": 0.4312727153301239,
+      "learning_rate": 0.0001,
+      "loss": 1.7256,
+      "step": 4271
+    },
+    {
+      "epoch": 0.49072425478146,
+      "grad_norm": 0.42118600010871887,
+      "learning_rate": 0.0001,
+      "loss": 1.7583,
+      "step": 4272
+    },
+    {
+      "epoch": 0.4908391246912871,
+      "grad_norm": 0.4323917329311371,
+      "learning_rate": 0.0001,
+      "loss": 1.6377,
+      "step": 4273
+    },
+    {
+      "epoch": 0.4909539946011142,
+      "grad_norm": 0.4115605354309082,
+      "learning_rate": 0.0001,
+      "loss": 1.7165,
+      "step": 4274
+    },
+    {
+      "epoch": 0.49106886451094134,
+      "grad_norm": 0.41110193729400635,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 4275
+    },
+    {
+      "epoch": 0.49118373442076846,
+      "grad_norm": 0.43548524379730225,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 4276
+    },
+    {
+      "epoch": 0.4912986043305956,
+      "grad_norm": 0.4107424020767212,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 4277
+    },
+    {
+      "epoch": 0.4914134742404227,
+      "grad_norm": 0.4132782220840454,
+      "learning_rate": 0.0001,
+      "loss": 1.7099,
+      "step": 4278
+    },
+    {
+      "epoch": 0.4915283441502498,
+      "grad_norm": 0.4486086070537567,
+      "learning_rate": 0.0001,
+      "loss": 1.9551,
+      "step": 4279
+    },
+    {
+      "epoch": 0.49164321406007694,
+      "grad_norm": 0.41517210006713867,
+      "learning_rate": 0.0001,
+      "loss": 1.7067,
+      "step": 4280
+    },
+    {
+      "epoch": 0.49175808396990406,
+      "grad_norm": 0.37078627943992615,
+      "learning_rate": 0.0001,
+      "loss": 1.4954,
+      "step": 4281
+    },
+    {
+      "epoch": 0.4918729538797312,
+      "grad_norm": 0.41808491945266724,
+      "learning_rate": 0.0001,
+      "loss": 1.5452,
+      "step": 4282
+    },
+    {
+      "epoch": 0.4919878237895583,
+      "grad_norm": 0.3952369689941406,
+      "learning_rate": 0.0001,
+      "loss": 1.7146,
+      "step": 4283
+    },
+    {
+      "epoch": 0.4921026936993854,
+      "grad_norm": 0.43479466438293457,
+      "learning_rate": 0.0001,
+      "loss": 1.7984,
+      "step": 4284
+    },
+    {
+      "epoch": 0.4922175636092126,
+      "grad_norm": 0.44229763746261597,
+      "learning_rate": 0.0001,
+      "loss": 1.877,
+      "step": 4285
+    },
+    {
+      "epoch": 0.4923324335190397,
+      "grad_norm": 0.4501185715198517,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 4286
+    },
+    {
+      "epoch": 0.49244730342886683,
+      "grad_norm": 0.43694865703582764,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 4287
+    },
+    {
+      "epoch": 0.49256217333869395,
+      "grad_norm": 0.4092878997325897,
+      "learning_rate": 0.0001,
+      "loss": 1.7253,
+      "step": 4288
+    },
+    {
+      "epoch": 0.4926770432485211,
+      "grad_norm": 0.4009423851966858,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 4289
+    },
+    {
+      "epoch": 0.4927919131583482,
+      "grad_norm": 0.4281792640686035,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 4290
+    },
+    {
+      "epoch": 0.4929067830681753,
+      "grad_norm": 0.49694761633872986,
+      "learning_rate": 0.0001,
+      "loss": 1.7392,
+      "step": 4291
+    },
+    {
+      "epoch": 0.49302165297800243,
+      "grad_norm": 0.44447091221809387,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 4292
+    },
+    {
+      "epoch": 0.49313652288782955,
+      "grad_norm": 0.3885389566421509,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 4293
+    },
+    {
+      "epoch": 0.4932513927976567,
+      "grad_norm": 0.3895886242389679,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 4294
+    },
+    {
+      "epoch": 0.4933662627074838,
+      "grad_norm": 0.4553350806236267,
+      "learning_rate": 0.0001,
+      "loss": 1.793,
+      "step": 4295
+    },
+    {
+      "epoch": 0.4934811326173109,
+      "grad_norm": 0.384520560503006,
+      "learning_rate": 0.0001,
+      "loss": 1.4896,
+      "step": 4296
+    },
+    {
+      "epoch": 0.49359600252713803,
+      "grad_norm": 0.4104674160480499,
+      "learning_rate": 0.0001,
+      "loss": 1.6698,
+      "step": 4297
+    },
+    {
+      "epoch": 0.49371087243696515,
+      "grad_norm": 0.4543203115463257,
+      "learning_rate": 0.0001,
+      "loss": 1.7958,
+      "step": 4298
+    },
+    {
+      "epoch": 0.4938257423467923,
+      "grad_norm": 0.40907683968544006,
+      "learning_rate": 0.0001,
+      "loss": 1.6956,
+      "step": 4299
+    },
+    {
+      "epoch": 0.4939406122566194,
+      "grad_norm": 0.43240875005722046,
+      "learning_rate": 0.0001,
+      "loss": 1.4569,
+      "step": 4300
+    },
+    {
+      "epoch": 0.4940554821664465,
+      "grad_norm": 0.41229525208473206,
+      "learning_rate": 0.0001,
+      "loss": 1.7708,
+      "step": 4301
+    },
+    {
+      "epoch": 0.49417035207627363,
+      "grad_norm": 0.4220798909664154,
+      "learning_rate": 0.0001,
+      "loss": 1.6977,
+      "step": 4302
+    },
+    {
+      "epoch": 0.49428522198610075,
+      "grad_norm": 0.3954547345638275,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 4303
+    },
+    {
+      "epoch": 0.4944000918959279,
+      "grad_norm": 0.41736140847206116,
+      "learning_rate": 0.0001,
+      "loss": 1.7265,
+      "step": 4304
+    },
+    {
+      "epoch": 0.494514961805755,
+      "grad_norm": 0.39642804861068726,
+      "learning_rate": 0.0001,
+      "loss": 1.5365,
+      "step": 4305
+    },
+    {
+      "epoch": 0.4946298317155821,
+      "grad_norm": 0.4558791220188141,
+      "learning_rate": 0.0001,
+      "loss": 1.6012,
+      "step": 4306
+    },
+    {
+      "epoch": 0.49474470162540923,
+      "grad_norm": 0.39758846163749695,
+      "learning_rate": 0.0001,
+      "loss": 1.6934,
+      "step": 4307
+    },
+    {
+      "epoch": 0.49485957153523635,
+      "grad_norm": 0.43392619490623474,
+      "learning_rate": 0.0001,
+      "loss": 1.7523,
+      "step": 4308
+    },
+    {
+      "epoch": 0.4949744414450635,
+      "grad_norm": 0.3947283625602722,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 4309
+    },
+    {
+      "epoch": 0.4950893113548906,
+      "grad_norm": 0.3992256224155426,
+      "learning_rate": 0.0001,
+      "loss": 1.7432,
+      "step": 4310
+    },
+    {
+      "epoch": 0.4952041812647177,
+      "grad_norm": 0.4191253185272217,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 4311
+    },
+    {
+      "epoch": 0.49531905117454483,
+      "grad_norm": 0.41294851899147034,
+      "learning_rate": 0.0001,
+      "loss": 1.8674,
+      "step": 4312
+    },
+    {
+      "epoch": 0.49543392108437195,
+      "grad_norm": 0.4245891869068146,
+      "learning_rate": 0.0001,
+      "loss": 1.5189,
+      "step": 4313
+    },
+    {
+      "epoch": 0.4955487909941991,
+      "grad_norm": 0.44946831464767456,
+      "learning_rate": 0.0001,
+      "loss": 1.8442,
+      "step": 4314
+    },
+    {
+      "epoch": 0.4956636609040262,
+      "grad_norm": 0.4152979254722595,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 4315
+    },
+    {
+      "epoch": 0.4957785308138533,
+      "grad_norm": 0.3948223888874054,
+      "learning_rate": 0.0001,
+      "loss": 1.6683,
+      "step": 4316
+    },
+    {
+      "epoch": 0.49589340072368043,
+      "grad_norm": 0.423066109418869,
+      "learning_rate": 0.0001,
+      "loss": 1.6065,
+      "step": 4317
+    },
+    {
+      "epoch": 0.49600827063350755,
+      "grad_norm": 0.3875940144062042,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 4318
+    },
+    {
+      "epoch": 0.4961231405433347,
+      "grad_norm": 0.4109596014022827,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 4319
+    },
+    {
+      "epoch": 0.4962380104531618,
+      "grad_norm": 0.42905092239379883,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 4320
+    },
+    {
+      "epoch": 0.4963528803629889,
+      "grad_norm": 0.38710954785346985,
+      "learning_rate": 0.0001,
+      "loss": 1.3936,
+      "step": 4321
+    },
+    {
+      "epoch": 0.49646775027281603,
+      "grad_norm": 0.4148370623588562,
+      "learning_rate": 0.0001,
+      "loss": 1.3909,
+      "step": 4322
+    },
+    {
+      "epoch": 0.49658262018264315,
+      "grad_norm": 0.4470541775226593,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 4323
+    },
+    {
+      "epoch": 0.4966974900924703,
+      "grad_norm": 0.3968261480331421,
+      "learning_rate": 0.0001,
+      "loss": 1.5145,
+      "step": 4324
+    },
+    {
+      "epoch": 0.4968123600022974,
+      "grad_norm": 0.43433794379234314,
+      "learning_rate": 0.0001,
+      "loss": 1.7366,
+      "step": 4325
+    },
+    {
+      "epoch": 0.4969272299121245,
+      "grad_norm": 0.39728692173957825,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 4326
+    },
+    {
+      "epoch": 0.49704209982195163,
+      "grad_norm": 0.3991283178329468,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 4327
+    },
+    {
+      "epoch": 0.49715696973177875,
+      "grad_norm": 0.4111589193344116,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 4328
+    },
+    {
+      "epoch": 0.4972718396416059,
+      "grad_norm": 0.41188472509384155,
+      "learning_rate": 0.0001,
+      "loss": 1.7465,
+      "step": 4329
+    },
+    {
+      "epoch": 0.497386709551433,
+      "grad_norm": 0.435154527425766,
+      "learning_rate": 0.0001,
+      "loss": 1.6414,
+      "step": 4330
+    },
+    {
+      "epoch": 0.4975015794612601,
+      "grad_norm": 0.40389662981033325,
+      "learning_rate": 0.0001,
+      "loss": 1.7158,
+      "step": 4331
+    },
+    {
+      "epoch": 0.49761644937108723,
+      "grad_norm": 0.4168131947517395,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 4332
+    },
+    {
+      "epoch": 0.49773131928091435,
+      "grad_norm": 0.43243077397346497,
+      "learning_rate": 0.0001,
+      "loss": 1.8349,
+      "step": 4333
+    },
+    {
+      "epoch": 0.4978461891907415,
+      "grad_norm": 0.42854100465774536,
+      "learning_rate": 0.0001,
+      "loss": 1.732,
+      "step": 4334
+    },
+    {
+      "epoch": 0.4979610591005686,
+      "grad_norm": 0.41653481125831604,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 4335
+    },
+    {
+      "epoch": 0.4980759290103957,
+      "grad_norm": 0.45638951659202576,
+      "learning_rate": 0.0001,
+      "loss": 1.844,
+      "step": 4336
+    },
+    {
+      "epoch": 0.49819079892022283,
+      "grad_norm": 0.4429507553577423,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 4337
+    },
+    {
+      "epoch": 0.49830566883004995,
+      "grad_norm": 0.45221859216690063,
+      "learning_rate": 0.0001,
+      "loss": 1.8052,
+      "step": 4338
+    },
+    {
+      "epoch": 0.4984205387398771,
+      "grad_norm": 0.42000943422317505,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 4339
+    },
+    {
+      "epoch": 0.4985354086497042,
+      "grad_norm": 0.44999954104423523,
+      "learning_rate": 0.0001,
+      "loss": 1.7288,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4986502785595313,
+      "grad_norm": 0.44586285948753357,
+      "learning_rate": 0.0001,
+      "loss": 1.8559,
+      "step": 4341
+    },
+    {
+      "epoch": 0.49876514846935843,
+      "grad_norm": 0.41515326499938965,
+      "learning_rate": 0.0001,
+      "loss": 1.6989,
+      "step": 4342
+    },
+    {
+      "epoch": 0.49888001837918555,
+      "grad_norm": 0.40862441062927246,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 4343
+    },
+    {
+      "epoch": 0.4989948882890127,
+      "grad_norm": 0.4498591423034668,
+      "learning_rate": 0.0001,
+      "loss": 1.7579,
+      "step": 4344
+    },
+    {
+      "epoch": 0.4991097581988398,
+      "grad_norm": 0.4223852753639221,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 4345
+    },
+    {
+      "epoch": 0.4992246281086669,
+      "grad_norm": 0.39054638147354126,
+      "learning_rate": 0.0001,
+      "loss": 1.3893,
+      "step": 4346
+    },
+    {
+      "epoch": 0.49933949801849403,
+      "grad_norm": 0.4462035298347473,
+      "learning_rate": 0.0001,
+      "loss": 1.7637,
+      "step": 4347
+    },
+    {
+      "epoch": 0.49945436792832115,
+      "grad_norm": 0.4029097855091095,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 4348
+    },
+    {
+      "epoch": 0.4995692378381483,
+      "grad_norm": 0.4124845862388611,
+      "learning_rate": 0.0001,
+      "loss": 1.7638,
+      "step": 4349
+    },
+    {
+      "epoch": 0.4996841077479754,
+      "grad_norm": 0.37937289476394653,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 4350
+    },
+    {
+      "epoch": 0.4997989776578025,
+      "grad_norm": 0.40648961067199707,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 4351
+    },
+    {
+      "epoch": 0.49991384756762963,
+      "grad_norm": 0.4147469997406006,
+      "learning_rate": 0.0001,
+      "loss": 1.5053,
+      "step": 4352
+    },
+    {
+      "epoch": 0.5000287174774568,
+      "grad_norm": 0.42763659358024597,
+      "learning_rate": 0.0001,
+      "loss": 1.6607,
+      "step": 4353
+    },
+    {
+      "epoch": 0.5001435873872839,
+      "grad_norm": 0.4175349771976471,
+      "learning_rate": 0.0001,
+      "loss": 1.675,
+      "step": 4354
+    },
+    {
+      "epoch": 0.500258457297111,
+      "grad_norm": 0.4236915707588196,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 4355
+    },
+    {
+      "epoch": 0.5003733272069382,
+      "grad_norm": 0.41402342915534973,
+      "learning_rate": 0.0001,
+      "loss": 1.8008,
+      "step": 4356
+    },
+    {
+      "epoch": 0.5004881971167653,
+      "grad_norm": 0.4136095345020294,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 4357
+    },
+    {
+      "epoch": 0.5006030670265924,
+      "grad_norm": 0.43916845321655273,
+      "learning_rate": 0.0001,
+      "loss": 1.8298,
+      "step": 4358
+    },
+    {
+      "epoch": 0.5007179369364195,
+      "grad_norm": 0.41625019907951355,
+      "learning_rate": 0.0001,
+      "loss": 1.6969,
+      "step": 4359
+    },
+    {
+      "epoch": 0.5008328068462466,
+      "grad_norm": 0.422838419675827,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 4360
+    },
+    {
+      "epoch": 0.5009476767560738,
+      "grad_norm": 0.356978178024292,
+      "learning_rate": 0.0001,
+      "loss": 1.3617,
+      "step": 4361
+    },
+    {
+      "epoch": 0.5010625466659009,
+      "grad_norm": 0.4372119903564453,
+      "learning_rate": 0.0001,
+      "loss": 1.7264,
+      "step": 4362
+    },
+    {
+      "epoch": 0.501177416575728,
+      "grad_norm": 0.43347251415252686,
+      "learning_rate": 0.0001,
+      "loss": 1.6472,
+      "step": 4363
+    },
+    {
+      "epoch": 0.5012922864855551,
+      "grad_norm": 0.38658055663108826,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 4364
+    },
+    {
+      "epoch": 0.5014071563953822,
+      "grad_norm": 0.3823000192642212,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 4365
+    },
+    {
+      "epoch": 0.5015220263052094,
+      "grad_norm": 0.44494375586509705,
+      "learning_rate": 0.0001,
+      "loss": 1.884,
+      "step": 4366
+    },
+    {
+      "epoch": 0.5016368962150365,
+      "grad_norm": 0.37995579838752747,
+      "learning_rate": 0.0001,
+      "loss": 1.5818,
+      "step": 4367
+    },
+    {
+      "epoch": 0.5017517661248636,
+      "grad_norm": 0.416050523519516,
+      "learning_rate": 0.0001,
+      "loss": 1.8128,
+      "step": 4368
+    },
+    {
+      "epoch": 0.5018666360346907,
+      "grad_norm": 0.41402295231819153,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 4369
+    },
+    {
+      "epoch": 0.5019815059445178,
+      "grad_norm": 0.4046735167503357,
+      "learning_rate": 0.0001,
+      "loss": 1.6491,
+      "step": 4370
+    },
+    {
+      "epoch": 0.502096375854345,
+      "grad_norm": 0.4079616367816925,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 4371
+    },
+    {
+      "epoch": 0.5022112457641721,
+      "grad_norm": 0.42762479186058044,
+      "learning_rate": 0.0001,
+      "loss": 1.7243,
+      "step": 4372
+    },
+    {
+      "epoch": 0.5023261156739992,
+      "grad_norm": 0.42116057872772217,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 4373
+    },
+    {
+      "epoch": 0.5024409855838263,
+      "grad_norm": 0.43111321330070496,
+      "learning_rate": 0.0001,
+      "loss": 1.6998,
+      "step": 4374
+    },
+    {
+      "epoch": 0.5025558554936534,
+      "grad_norm": 0.43248212337493896,
+      "learning_rate": 0.0001,
+      "loss": 1.839,
+      "step": 4375
+    },
+    {
+      "epoch": 0.5026707254034806,
+      "grad_norm": 0.40893498063087463,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 4376
+    },
+    {
+      "epoch": 0.5027855953133077,
+      "grad_norm": 0.43067118525505066,
+      "learning_rate": 0.0001,
+      "loss": 1.5609,
+      "step": 4377
+    },
+    {
+      "epoch": 0.5029004652231348,
+      "grad_norm": 0.3844601809978485,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 4378
+    },
+    {
+      "epoch": 0.5030153351329619,
+      "grad_norm": 0.41813233494758606,
+      "learning_rate": 0.0001,
+      "loss": 1.7656,
+      "step": 4379
+    },
+    {
+      "epoch": 0.503130205042789,
+      "grad_norm": 0.4095008373260498,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 4380
+    },
+    {
+      "epoch": 0.5032450749526162,
+      "grad_norm": 0.4308955669403076,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 4381
+    },
+    {
+      "epoch": 0.5033599448624433,
+      "grad_norm": 0.41021478176116943,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 4382
+    },
+    {
+      "epoch": 0.5034748147722704,
+      "grad_norm": 0.37911084294319153,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 4383
+    },
+    {
+      "epoch": 0.5035896846820975,
+      "grad_norm": 0.3981432020664215,
+      "learning_rate": 0.0001,
+      "loss": 1.7432,
+      "step": 4384
+    },
+    {
+      "epoch": 0.5037045545919246,
+      "grad_norm": 0.3934857249259949,
+      "learning_rate": 0.0001,
+      "loss": 1.7143,
+      "step": 4385
+    },
+    {
+      "epoch": 0.5038194245017518,
+      "grad_norm": 0.40038371086120605,
+      "learning_rate": 0.0001,
+      "loss": 1.7377,
+      "step": 4386
+    },
+    {
+      "epoch": 0.5039342944115789,
+      "grad_norm": 0.422829806804657,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 4387
+    },
+    {
+      "epoch": 0.504049164321406,
+      "grad_norm": 0.3928106725215912,
+      "learning_rate": 0.0001,
+      "loss": 1.6269,
+      "step": 4388
+    },
+    {
+      "epoch": 0.5041640342312331,
+      "grad_norm": 0.4150887131690979,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 4389
+    },
+    {
+      "epoch": 0.5042789041410602,
+      "grad_norm": 0.40696507692337036,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 4390
+    },
+    {
+      "epoch": 0.5043937740508874,
+      "grad_norm": 0.43781277537345886,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 4391
+    },
+    {
+      "epoch": 0.5045086439607145,
+      "grad_norm": 0.4133634865283966,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 4392
+    },
+    {
+      "epoch": 0.5046235138705416,
+      "grad_norm": 0.3970431387424469,
+      "learning_rate": 0.0001,
+      "loss": 1.2777,
+      "step": 4393
+    },
+    {
+      "epoch": 0.5047383837803687,
+      "grad_norm": 0.44825881719589233,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 4394
+    },
+    {
+      "epoch": 0.5048532536901958,
+      "grad_norm": 0.38739416003227234,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 4395
+    },
+    {
+      "epoch": 0.504968123600023,
+      "grad_norm": 0.41295453906059265,
+      "learning_rate": 0.0001,
+      "loss": 1.7472,
+      "step": 4396
+    },
+    {
+      "epoch": 0.5050829935098501,
+      "grad_norm": 0.3822581470012665,
+      "learning_rate": 0.0001,
+      "loss": 1.5052,
+      "step": 4397
+    },
+    {
+      "epoch": 0.5051978634196772,
+      "grad_norm": 0.40491965413093567,
+      "learning_rate": 0.0001,
+      "loss": 1.49,
+      "step": 4398
+    },
+    {
+      "epoch": 0.5053127333295043,
+      "grad_norm": 0.48093438148498535,
+      "learning_rate": 0.0001,
+      "loss": 1.7826,
+      "step": 4399
+    },
+    {
+      "epoch": 0.5054276032393314,
+      "grad_norm": 0.4143761694431305,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 4400
+    },
+    {
+      "epoch": 0.5055424731491586,
+      "grad_norm": 0.4251968264579773,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 4401
+    },
+    {
+      "epoch": 0.5056573430589857,
+      "grad_norm": 0.4139546751976013,
+      "learning_rate": 0.0001,
+      "loss": 1.4851,
+      "step": 4402
+    },
+    {
+      "epoch": 0.5057722129688128,
+      "grad_norm": 0.4102967381477356,
+      "learning_rate": 0.0001,
+      "loss": 1.6109,
+      "step": 4403
+    },
+    {
+      "epoch": 0.5058870828786399,
+      "grad_norm": 0.411199688911438,
+      "learning_rate": 0.0001,
+      "loss": 1.3911,
+      "step": 4404
+    },
+    {
+      "epoch": 0.506001952788467,
+      "grad_norm": 0.43025514483451843,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 4405
+    },
+    {
+      "epoch": 0.5061168226982942,
+      "grad_norm": 0.4648951292037964,
+      "learning_rate": 0.0001,
+      "loss": 1.8235,
+      "step": 4406
+    },
+    {
+      "epoch": 0.5062316926081213,
+      "grad_norm": 0.4003261923789978,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 4407
+    },
+    {
+      "epoch": 0.5063465625179484,
+      "grad_norm": 0.4348773956298828,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 4408
+    },
+    {
+      "epoch": 0.5064614324277755,
+      "grad_norm": 0.3802754580974579,
+      "learning_rate": 0.0001,
+      "loss": 1.3187,
+      "step": 4409
+    },
+    {
+      "epoch": 0.5065763023376026,
+      "grad_norm": 0.4715876579284668,
+      "learning_rate": 0.0001,
+      "loss": 1.9138,
+      "step": 4410
+    },
+    {
+      "epoch": 0.5066911722474298,
+      "grad_norm": 0.42883962392807007,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 4411
+    },
+    {
+      "epoch": 0.5068060421572569,
+      "grad_norm": 0.3982695937156677,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 4412
+    },
+    {
+      "epoch": 0.506920912067084,
+      "grad_norm": 0.4575648605823517,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 4413
+    },
+    {
+      "epoch": 0.5070357819769111,
+      "grad_norm": 0.422573983669281,
+      "learning_rate": 0.0001,
+      "loss": 1.7639,
+      "step": 4414
+    },
+    {
+      "epoch": 0.5071506518867382,
+      "grad_norm": 0.41428160667419434,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 4415
+    },
+    {
+      "epoch": 0.5072655217965654,
+      "grad_norm": 0.39599475264549255,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 4416
+    },
+    {
+      "epoch": 0.5073803917063925,
+      "grad_norm": 0.41915363073349,
+      "learning_rate": 0.0001,
+      "loss": 1.6944,
+      "step": 4417
+    },
+    {
+      "epoch": 0.5074952616162196,
+      "grad_norm": 0.4034424126148224,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 4418
+    },
+    {
+      "epoch": 0.5076101315260467,
+      "grad_norm": 0.4440549612045288,
+      "learning_rate": 0.0001,
+      "loss": 1.9008,
+      "step": 4419
+    },
+    {
+      "epoch": 0.5077250014358738,
+      "grad_norm": 0.43546631932258606,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 4420
+    },
+    {
+      "epoch": 0.507839871345701,
+      "grad_norm": 0.4137895107269287,
+      "learning_rate": 0.0001,
+      "loss": 1.7555,
+      "step": 4421
+    },
+    {
+      "epoch": 0.5079547412555281,
+      "grad_norm": 0.4285581111907959,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 4422
+    },
+    {
+      "epoch": 0.5080696111653552,
+      "grad_norm": 0.39121928811073303,
+      "learning_rate": 0.0001,
+      "loss": 1.5603,
+      "step": 4423
+    },
+    {
+      "epoch": 0.5081844810751823,
+      "grad_norm": 0.45524442195892334,
+      "learning_rate": 0.0001,
+      "loss": 1.6992,
+      "step": 4424
+    },
+    {
+      "epoch": 0.5082993509850094,
+      "grad_norm": 0.42169350385665894,
+      "learning_rate": 0.0001,
+      "loss": 1.4717,
+      "step": 4425
+    },
+    {
+      "epoch": 0.5084142208948366,
+      "grad_norm": 0.44150853157043457,
+      "learning_rate": 0.0001,
+      "loss": 1.735,
+      "step": 4426
+    },
+    {
+      "epoch": 0.5085290908046637,
+      "grad_norm": 0.4405604600906372,
+      "learning_rate": 0.0001,
+      "loss": 1.7817,
+      "step": 4427
+    },
+    {
+      "epoch": 0.5086439607144908,
+      "grad_norm": 0.4324178099632263,
+      "learning_rate": 0.0001,
+      "loss": 1.8098,
+      "step": 4428
+    },
+    {
+      "epoch": 0.5087588306243179,
+      "grad_norm": 0.3942691683769226,
+      "learning_rate": 0.0001,
+      "loss": 1.5467,
+      "step": 4429
+    },
+    {
+      "epoch": 0.508873700534145,
+      "grad_norm": 0.3946244716644287,
+      "learning_rate": 0.0001,
+      "loss": 1.5341,
+      "step": 4430
+    },
+    {
+      "epoch": 0.5089885704439722,
+      "grad_norm": 0.4541240930557251,
+      "learning_rate": 0.0001,
+      "loss": 1.8342,
+      "step": 4431
+    },
+    {
+      "epoch": 0.5091034403537993,
+      "grad_norm": 0.4114912450313568,
+      "learning_rate": 0.0001,
+      "loss": 1.6808,
+      "step": 4432
+    },
+    {
+      "epoch": 0.5092183102636264,
+      "grad_norm": 0.40291449427604675,
+      "learning_rate": 0.0001,
+      "loss": 1.4467,
+      "step": 4433
+    },
+    {
+      "epoch": 0.5093331801734535,
+      "grad_norm": 0.43536198139190674,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 4434
+    },
+    {
+      "epoch": 0.5094480500832806,
+      "grad_norm": 0.4481135606765747,
+      "learning_rate": 0.0001,
+      "loss": 1.8252,
+      "step": 4435
+    },
+    {
+      "epoch": 0.5095629199931078,
+      "grad_norm": 0.41042807698249817,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 4436
+    },
+    {
+      "epoch": 0.5096777899029349,
+      "grad_norm": 0.44647055864334106,
+      "learning_rate": 0.0001,
+      "loss": 1.7893,
+      "step": 4437
+    },
+    {
+      "epoch": 0.509792659812762,
+      "grad_norm": 0.4402143061161041,
+      "learning_rate": 0.0001,
+      "loss": 1.7,
+      "step": 4438
+    },
+    {
+      "epoch": 0.5099075297225891,
+      "grad_norm": 0.44720834493637085,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 4439
+    },
+    {
+      "epoch": 0.5100223996324162,
+      "grad_norm": 0.43396279215812683,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 4440
+    },
+    {
+      "epoch": 0.5101372695422434,
+      "grad_norm": 0.42125704884529114,
+      "learning_rate": 0.0001,
+      "loss": 1.6251,
+      "step": 4441
+    },
+    {
+      "epoch": 0.5102521394520705,
+      "grad_norm": 0.4178660213947296,
+      "learning_rate": 0.0001,
+      "loss": 1.7153,
+      "step": 4442
+    },
+    {
+      "epoch": 0.5103670093618976,
+      "grad_norm": 0.490590363740921,
+      "learning_rate": 0.0001,
+      "loss": 1.6246,
+      "step": 4443
+    },
+    {
+      "epoch": 0.5104818792717247,
+      "grad_norm": 0.40945470333099365,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 4444
+    },
+    {
+      "epoch": 0.5105967491815518,
+      "grad_norm": 0.44269421696662903,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 4445
+    },
+    {
+      "epoch": 0.510711619091379,
+      "grad_norm": 0.41479024291038513,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 4446
+    },
+    {
+      "epoch": 0.5108264890012061,
+      "grad_norm": 0.3876959979534149,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 4447
+    },
+    {
+      "epoch": 0.5109413589110332,
+      "grad_norm": 0.4491494297981262,
+      "learning_rate": 0.0001,
+      "loss": 1.681,
+      "step": 4448
+    },
+    {
+      "epoch": 0.5110562288208603,
+      "grad_norm": 0.41451436281204224,
+      "learning_rate": 0.0001,
+      "loss": 1.6657,
+      "step": 4449
+    },
+    {
+      "epoch": 0.5111710987306874,
+      "grad_norm": 0.39285629987716675,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 4450
+    },
+    {
+      "epoch": 0.5112859686405146,
+      "grad_norm": 0.39508745074272156,
+      "learning_rate": 0.0001,
+      "loss": 1.4939,
+      "step": 4451
+    },
+    {
+      "epoch": 0.5114008385503417,
+      "grad_norm": 0.44860512018203735,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 4452
+    },
+    {
+      "epoch": 0.5115157084601688,
+      "grad_norm": 0.4382397532463074,
+      "learning_rate": 0.0001,
+      "loss": 1.6823,
+      "step": 4453
+    },
+    {
+      "epoch": 0.5116305783699959,
+      "grad_norm": 0.40884271264076233,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 4454
+    },
+    {
+      "epoch": 0.511745448279823,
+      "grad_norm": 0.43925192952156067,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 4455
+    },
+    {
+      "epoch": 0.5118603181896503,
+      "grad_norm": 0.42059096693992615,
+      "learning_rate": 0.0001,
+      "loss": 1.6763,
+      "step": 4456
+    },
+    {
+      "epoch": 0.5119751880994774,
+      "grad_norm": 0.4306631088256836,
+      "learning_rate": 0.0001,
+      "loss": 1.5764,
+      "step": 4457
+    },
+    {
+      "epoch": 0.5120900580093045,
+      "grad_norm": 0.41633883118629456,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 4458
+    },
+    {
+      "epoch": 0.5122049279191316,
+      "grad_norm": 0.4185430705547333,
+      "learning_rate": 0.0001,
+      "loss": 1.6695,
+      "step": 4459
+    },
+    {
+      "epoch": 0.5123197978289588,
+      "grad_norm": 0.38397789001464844,
+      "learning_rate": 0.0001,
+      "loss": 1.3248,
+      "step": 4460
+    },
+    {
+      "epoch": 0.5124346677387859,
+      "grad_norm": 0.42769157886505127,
+      "learning_rate": 0.0001,
+      "loss": 1.6386,
+      "step": 4461
+    },
+    {
+      "epoch": 0.512549537648613,
+      "grad_norm": 0.40281736850738525,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 4462
+    },
+    {
+      "epoch": 0.5126644075584401,
+      "grad_norm": 0.48889243602752686,
+      "learning_rate": 0.0001,
+      "loss": 1.8824,
+      "step": 4463
+    },
+    {
+      "epoch": 0.5127792774682672,
+      "grad_norm": 0.4288654625415802,
+      "learning_rate": 0.0001,
+      "loss": 1.6987,
+      "step": 4464
+    },
+    {
+      "epoch": 0.5128941473780944,
+      "grad_norm": 0.4444088637828827,
+      "learning_rate": 0.0001,
+      "loss": 1.7862,
+      "step": 4465
+    },
+    {
+      "epoch": 0.5130090172879215,
+      "grad_norm": 0.3586702346801758,
+      "learning_rate": 0.0001,
+      "loss": 1.4236,
+      "step": 4466
+    },
+    {
+      "epoch": 0.5131238871977486,
+      "grad_norm": 0.41832202672958374,
+      "learning_rate": 0.0001,
+      "loss": 1.6599,
+      "step": 4467
+    },
+    {
+      "epoch": 0.5132387571075757,
+      "grad_norm": 0.41678473353385925,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 4468
+    },
+    {
+      "epoch": 0.5133536270174028,
+      "grad_norm": 0.39657461643218994,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 4469
+    },
+    {
+      "epoch": 0.51346849692723,
+      "grad_norm": 0.4119437336921692,
+      "learning_rate": 0.0001,
+      "loss": 1.631,
+      "step": 4470
+    },
+    {
+      "epoch": 0.5135833668370571,
+      "grad_norm": 0.4205493927001953,
+      "learning_rate": 0.0001,
+      "loss": 1.5289,
+      "step": 4471
+    },
+    {
+      "epoch": 0.5136982367468842,
+      "grad_norm": 0.4292590618133545,
+      "learning_rate": 0.0001,
+      "loss": 1.6888,
+      "step": 4472
+    },
+    {
+      "epoch": 0.5138131066567113,
+      "grad_norm": 0.4054730534553528,
+      "learning_rate": 0.0001,
+      "loss": 1.6858,
+      "step": 4473
+    },
+    {
+      "epoch": 0.5139279765665384,
+      "grad_norm": 0.5104745626449585,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 4474
+    },
+    {
+      "epoch": 0.5140428464763656,
+      "grad_norm": 0.4039926528930664,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 4475
+    },
+    {
+      "epoch": 0.5141577163861927,
+      "grad_norm": 0.38494694232940674,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 4476
+    },
+    {
+      "epoch": 0.5142725862960198,
+      "grad_norm": 0.4241875112056732,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 4477
+    },
+    {
+      "epoch": 0.5143874562058469,
+      "grad_norm": 0.41834181547164917,
+      "learning_rate": 0.0001,
+      "loss": 1.4409,
+      "step": 4478
+    },
+    {
+      "epoch": 0.514502326115674,
+      "grad_norm": 0.38371768593788147,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 4479
+    },
+    {
+      "epoch": 0.5146171960255012,
+      "grad_norm": 0.42289024591445923,
+      "learning_rate": 0.0001,
+      "loss": 1.8331,
+      "step": 4480
+    },
+    {
+      "epoch": 0.5147320659353283,
+      "grad_norm": 0.41524538397789,
+      "learning_rate": 0.0001,
+      "loss": 1.6312,
+      "step": 4481
+    },
+    {
+      "epoch": 0.5148469358451554,
+      "grad_norm": 0.4520059823989868,
+      "learning_rate": 0.0001,
+      "loss": 1.7944,
+      "step": 4482
+    },
+    {
+      "epoch": 0.5149618057549825,
+      "grad_norm": 0.46914947032928467,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 4483
+    },
+    {
+      "epoch": 0.5150766756648096,
+      "grad_norm": 0.4201009273529053,
+      "learning_rate": 0.0001,
+      "loss": 1.7413,
+      "step": 4484
+    },
+    {
+      "epoch": 0.5151915455746368,
+      "grad_norm": 0.43544942140579224,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 4485
+    },
+    {
+      "epoch": 0.5153064154844639,
+      "grad_norm": 0.3825484812259674,
+      "learning_rate": 0.0001,
+      "loss": 1.543,
+      "step": 4486
+    },
+    {
+      "epoch": 0.515421285394291,
+      "grad_norm": 0.46979475021362305,
+      "learning_rate": 0.0001,
+      "loss": 1.765,
+      "step": 4487
+    },
+    {
+      "epoch": 0.5155361553041181,
+      "grad_norm": 0.4438627064228058,
+      "learning_rate": 0.0001,
+      "loss": 1.796,
+      "step": 4488
+    },
+    {
+      "epoch": 0.5156510252139452,
+      "grad_norm": 0.4040473699569702,
+      "learning_rate": 0.0001,
+      "loss": 1.6934,
+      "step": 4489
+    },
+    {
+      "epoch": 0.5157658951237724,
+      "grad_norm": 0.3772587776184082,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 4490
+    },
+    {
+      "epoch": 0.5158807650335995,
+      "grad_norm": 0.4254496693611145,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 4491
+    },
+    {
+      "epoch": 0.5159956349434266,
+      "grad_norm": 0.43310895562171936,
+      "learning_rate": 0.0001,
+      "loss": 1.7606,
+      "step": 4492
+    },
+    {
+      "epoch": 0.5161105048532537,
+      "grad_norm": 0.43428242206573486,
+      "learning_rate": 0.0001,
+      "loss": 1.8536,
+      "step": 4493
+    },
+    {
+      "epoch": 0.5162253747630808,
+      "grad_norm": 0.4234102964401245,
+      "learning_rate": 0.0001,
+      "loss": 1.6754,
+      "step": 4494
+    },
+    {
+      "epoch": 0.516340244672908,
+      "grad_norm": 0.4157521724700928,
+      "learning_rate": 0.0001,
+      "loss": 1.717,
+      "step": 4495
+    },
+    {
+      "epoch": 0.5164551145827351,
+      "grad_norm": 0.4052300751209259,
+      "learning_rate": 0.0001,
+      "loss": 1.4448,
+      "step": 4496
+    },
+    {
+      "epoch": 0.5165699844925622,
+      "grad_norm": 0.4426092207431793,
+      "learning_rate": 0.0001,
+      "loss": 1.683,
+      "step": 4497
+    },
+    {
+      "epoch": 0.5166848544023893,
+      "grad_norm": 0.41626426577568054,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 4498
+    },
+    {
+      "epoch": 0.5167997243122164,
+      "grad_norm": 0.38398024439811707,
+      "learning_rate": 0.0001,
+      "loss": 1.6062,
+      "step": 4499
+    },
+    {
+      "epoch": 0.5169145942220436,
+      "grad_norm": 0.4056454598903656,
+      "learning_rate": 0.0001,
+      "loss": 1.5062,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5170294641318707,
+      "grad_norm": 0.4996061325073242,
+      "learning_rate": 0.0001,
+      "loss": 1.7317,
+      "step": 4501
+    },
+    {
+      "epoch": 0.5171443340416978,
+      "grad_norm": 0.40144461393356323,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 4502
+    },
+    {
+      "epoch": 0.5172592039515249,
+      "grad_norm": 0.4855468273162842,
+      "learning_rate": 0.0001,
+      "loss": 1.6999,
+      "step": 4503
+    },
+    {
+      "epoch": 0.517374073861352,
+      "grad_norm": 0.4194789230823517,
+      "learning_rate": 0.0001,
+      "loss": 1.6659,
+      "step": 4504
+    },
+    {
+      "epoch": 0.5174889437711792,
+      "grad_norm": 0.3837035596370697,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 4505
+    },
+    {
+      "epoch": 0.5176038136810063,
+      "grad_norm": 0.39867258071899414,
+      "learning_rate": 0.0001,
+      "loss": 1.3699,
+      "step": 4506
+    },
+    {
+      "epoch": 0.5177186835908334,
+      "grad_norm": 0.4189174473285675,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 4507
+    },
+    {
+      "epoch": 0.5178335535006605,
+      "grad_norm": 0.43066421151161194,
+      "learning_rate": 0.0001,
+      "loss": 1.7007,
+      "step": 4508
+    },
+    {
+      "epoch": 0.5179484234104876,
+      "grad_norm": 0.40682274103164673,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 4509
+    },
+    {
+      "epoch": 0.5180632933203148,
+      "grad_norm": 0.4145391583442688,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 4510
+    },
+    {
+      "epoch": 0.5181781632301419,
+      "grad_norm": 0.42484042048454285,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 4511
+    },
+    {
+      "epoch": 0.518293033139969,
+      "grad_norm": 0.48949721455574036,
+      "learning_rate": 0.0001,
+      "loss": 1.7115,
+      "step": 4512
+    },
+    {
+      "epoch": 0.5184079030497961,
+      "grad_norm": 0.46473968029022217,
+      "learning_rate": 0.0001,
+      "loss": 1.778,
+      "step": 4513
+    },
+    {
+      "epoch": 0.5185227729596232,
+      "grad_norm": 0.4237271845340729,
+      "learning_rate": 0.0001,
+      "loss": 1.7289,
+      "step": 4514
+    },
+    {
+      "epoch": 0.5186376428694504,
+      "grad_norm": 0.4288172721862793,
+      "learning_rate": 0.0001,
+      "loss": 1.8229,
+      "step": 4515
+    },
+    {
+      "epoch": 0.5187525127792775,
+      "grad_norm": 0.39855021238327026,
+      "learning_rate": 0.0001,
+      "loss": 1.7383,
+      "step": 4516
+    },
+    {
+      "epoch": 0.5188673826891046,
+      "grad_norm": 0.4043690860271454,
+      "learning_rate": 0.0001,
+      "loss": 1.7437,
+      "step": 4517
+    },
+    {
+      "epoch": 0.5189822525989317,
+      "grad_norm": 0.4174858331680298,
+      "learning_rate": 0.0001,
+      "loss": 1.4415,
+      "step": 4518
+    },
+    {
+      "epoch": 0.5190971225087588,
+      "grad_norm": 0.4092039167881012,
+      "learning_rate": 0.0001,
+      "loss": 1.7563,
+      "step": 4519
+    },
+    {
+      "epoch": 0.519211992418586,
+      "grad_norm": 0.44929736852645874,
+      "learning_rate": 0.0001,
+      "loss": 1.8381,
+      "step": 4520
+    },
+    {
+      "epoch": 0.5193268623284131,
+      "grad_norm": 0.38499635457992554,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 4521
+    },
+    {
+      "epoch": 0.5194417322382402,
+      "grad_norm": 0.4187855124473572,
+      "learning_rate": 0.0001,
+      "loss": 1.6891,
+      "step": 4522
+    },
+    {
+      "epoch": 0.5195566021480673,
+      "grad_norm": 0.43714576959609985,
+      "learning_rate": 0.0001,
+      "loss": 1.7564,
+      "step": 4523
+    },
+    {
+      "epoch": 0.5196714720578944,
+      "grad_norm": 0.42951980233192444,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 4524
+    },
+    {
+      "epoch": 0.5197863419677216,
+      "grad_norm": 0.4339199662208557,
+      "learning_rate": 0.0001,
+      "loss": 1.5095,
+      "step": 4525
+    },
+    {
+      "epoch": 0.5199012118775487,
+      "grad_norm": 0.434725821018219,
+      "learning_rate": 0.0001,
+      "loss": 1.6993,
+      "step": 4526
+    },
+    {
+      "epoch": 0.5200160817873758,
+      "grad_norm": 0.42468908429145813,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 4527
+    },
+    {
+      "epoch": 0.5201309516972029,
+      "grad_norm": 0.42453402280807495,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 4528
+    },
+    {
+      "epoch": 0.52024582160703,
+      "grad_norm": 0.4131234586238861,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 4529
+    },
+    {
+      "epoch": 0.5203606915168572,
+      "grad_norm": 0.40424486994743347,
+      "learning_rate": 0.0001,
+      "loss": 1.5853,
+      "step": 4530
+    },
+    {
+      "epoch": 0.5204755614266843,
+      "grad_norm": 0.45649197697639465,
+      "learning_rate": 0.0001,
+      "loss": 1.8034,
+      "step": 4531
+    },
+    {
+      "epoch": 0.5205904313365114,
+      "grad_norm": 0.412251740694046,
+      "learning_rate": 0.0001,
+      "loss": 1.6628,
+      "step": 4532
+    },
+    {
+      "epoch": 0.5207053012463385,
+      "grad_norm": 0.4387061297893524,
+      "learning_rate": 0.0001,
+      "loss": 1.8548,
+      "step": 4533
+    },
+    {
+      "epoch": 0.5208201711561656,
+      "grad_norm": 0.4186844229698181,
+      "learning_rate": 0.0001,
+      "loss": 1.7774,
+      "step": 4534
+    },
+    {
+      "epoch": 0.5209350410659928,
+      "grad_norm": 0.41150832176208496,
+      "learning_rate": 0.0001,
+      "loss": 1.6646,
+      "step": 4535
+    },
+    {
+      "epoch": 0.5210499109758199,
+      "grad_norm": 0.4150691628456116,
+      "learning_rate": 0.0001,
+      "loss": 1.579,
+      "step": 4536
+    },
+    {
+      "epoch": 0.521164780885647,
+      "grad_norm": 0.4016995131969452,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 4537
+    },
+    {
+      "epoch": 0.5212796507954741,
+      "grad_norm": 0.4465637505054474,
+      "learning_rate": 0.0001,
+      "loss": 1.8251,
+      "step": 4538
+    },
+    {
+      "epoch": 0.5213945207053012,
+      "grad_norm": 0.38491567969322205,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 4539
+    },
+    {
+      "epoch": 0.5215093906151284,
+      "grad_norm": 0.3878459632396698,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 4540
+    },
+    {
+      "epoch": 0.5216242605249555,
+      "grad_norm": 0.38580122590065,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 4541
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 0.3909989595413208,
+      "learning_rate": 0.0001,
+      "loss": 1.6571,
+      "step": 4542
+    },
+    {
+      "epoch": 0.5218540003446097,
+      "grad_norm": 0.49450787901878357,
+      "learning_rate": 0.0001,
+      "loss": 1.744,
+      "step": 4543
+    },
+    {
+      "epoch": 0.5219688702544368,
+      "grad_norm": 0.43028536438941956,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 4544
+    },
+    {
+      "epoch": 0.522083740164264,
+      "grad_norm": 0.4146410822868347,
+      "learning_rate": 0.0001,
+      "loss": 1.5019,
+      "step": 4545
+    },
+    {
+      "epoch": 0.5221986100740911,
+      "grad_norm": 0.40828803181648254,
+      "learning_rate": 0.0001,
+      "loss": 1.696,
+      "step": 4546
+    },
+    {
+      "epoch": 0.5223134799839182,
+      "grad_norm": 0.38090986013412476,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 4547
+    },
+    {
+      "epoch": 0.5224283498937453,
+      "grad_norm": 0.3952951431274414,
+      "learning_rate": 0.0001,
+      "loss": 1.524,
+      "step": 4548
+    },
+    {
+      "epoch": 0.5225432198035724,
+      "grad_norm": 0.4446341395378113,
+      "learning_rate": 0.0001,
+      "loss": 1.7717,
+      "step": 4549
+    },
+    {
+      "epoch": 0.5226580897133996,
+      "grad_norm": 0.4435203969478607,
+      "learning_rate": 0.0001,
+      "loss": 1.6232,
+      "step": 4550
+    },
+    {
+      "epoch": 0.5227729596232267,
+      "grad_norm": 0.4518386423587799,
+      "learning_rate": 0.0001,
+      "loss": 1.6649,
+      "step": 4551
+    },
+    {
+      "epoch": 0.5228878295330538,
+      "grad_norm": 0.4210054576396942,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 4552
+    },
+    {
+      "epoch": 0.5230026994428809,
+      "grad_norm": 0.464871346950531,
+      "learning_rate": 0.0001,
+      "loss": 1.7242,
+      "step": 4553
+    },
+    {
+      "epoch": 0.523117569352708,
+      "grad_norm": 0.4055049419403076,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 4554
+    },
+    {
+      "epoch": 0.5232324392625352,
+      "grad_norm": 0.4289097189903259,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 4555
+    },
+    {
+      "epoch": 0.5233473091723623,
+      "grad_norm": 0.40787026286125183,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 4556
+    },
+    {
+      "epoch": 0.5234621790821894,
+      "grad_norm": 0.41342809796333313,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 4557
+    },
+    {
+      "epoch": 0.5235770489920165,
+      "grad_norm": 0.3959392011165619,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 4558
+    },
+    {
+      "epoch": 0.5236919189018436,
+      "grad_norm": 0.41905930638313293,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 4559
+    },
+    {
+      "epoch": 0.5238067888116708,
+      "grad_norm": 0.4155671000480652,
+      "learning_rate": 0.0001,
+      "loss": 1.6784,
+      "step": 4560
+    },
+    {
+      "epoch": 0.5239216587214979,
+      "grad_norm": 0.3972344398498535,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 4561
+    },
+    {
+      "epoch": 0.524036528631325,
+      "grad_norm": 0.40942510962486267,
+      "learning_rate": 0.0001,
+      "loss": 1.648,
+      "step": 4562
+    },
+    {
+      "epoch": 0.5241513985411521,
+      "grad_norm": 0.43989264965057373,
+      "learning_rate": 0.0001,
+      "loss": 1.8291,
+      "step": 4563
+    },
+    {
+      "epoch": 0.5242662684509792,
+      "grad_norm": 0.38364219665527344,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 4564
+    },
+    {
+      "epoch": 0.5243811383608064,
+      "grad_norm": 0.3964410424232483,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 4565
+    },
+    {
+      "epoch": 0.5244960082706335,
+      "grad_norm": 0.3990534842014313,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 4566
+    },
+    {
+      "epoch": 0.5246108781804606,
+      "grad_norm": 0.4154457747936249,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 4567
+    },
+    {
+      "epoch": 0.5247257480902877,
+      "grad_norm": 0.4000798761844635,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 4568
+    },
+    {
+      "epoch": 0.5248406180001148,
+      "grad_norm": 0.4398530423641205,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 4569
+    },
+    {
+      "epoch": 0.524955487909942,
+      "grad_norm": 0.41386568546295166,
+      "learning_rate": 0.0001,
+      "loss": 1.7705,
+      "step": 4570
+    },
+    {
+      "epoch": 0.5250703578197691,
+      "grad_norm": 0.48047158122062683,
+      "learning_rate": 0.0001,
+      "loss": 1.8748,
+      "step": 4571
+    },
+    {
+      "epoch": 0.5251852277295962,
+      "grad_norm": 0.38870546221733093,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 4572
+    },
+    {
+      "epoch": 0.5253000976394233,
+      "grad_norm": 0.48835334181785583,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 4573
+    },
+    {
+      "epoch": 0.5254149675492504,
+      "grad_norm": 0.39779576659202576,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 4574
+    },
+    {
+      "epoch": 0.5255298374590776,
+      "grad_norm": 0.3713628947734833,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 4575
+    },
+    {
+      "epoch": 0.5256447073689047,
+      "grad_norm": 0.39374494552612305,
+      "learning_rate": 0.0001,
+      "loss": 1.5818,
+      "step": 4576
+    },
+    {
+      "epoch": 0.5257595772787318,
+      "grad_norm": 0.40991827845573425,
+      "learning_rate": 0.0001,
+      "loss": 1.6499,
+      "step": 4577
+    },
+    {
+      "epoch": 0.5258744471885589,
+      "grad_norm": 0.4329836964607239,
+      "learning_rate": 0.0001,
+      "loss": 1.7166,
+      "step": 4578
+    },
+    {
+      "epoch": 0.525989317098386,
+      "grad_norm": 0.44446802139282227,
+      "learning_rate": 0.0001,
+      "loss": 1.7636,
+      "step": 4579
+    },
+    {
+      "epoch": 0.5261041870082132,
+      "grad_norm": 0.42868572473526,
+      "learning_rate": 0.0001,
+      "loss": 1.8013,
+      "step": 4580
+    },
+    {
+      "epoch": 0.5262190569180403,
+      "grad_norm": 0.38663357496261597,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 4581
+    },
+    {
+      "epoch": 0.5263339268278674,
+      "grad_norm": 0.4398927390575409,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 4582
+    },
+    {
+      "epoch": 0.5264487967376945,
+      "grad_norm": 0.4069356620311737,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 4583
+    },
+    {
+      "epoch": 0.5265636666475216,
+      "grad_norm": 0.40244022011756897,
+      "learning_rate": 0.0001,
+      "loss": 1.5808,
+      "step": 4584
+    },
+    {
+      "epoch": 0.5266785365573488,
+      "grad_norm": 0.42762094736099243,
+      "learning_rate": 0.0001,
+      "loss": 1.7083,
+      "step": 4585
+    },
+    {
+      "epoch": 0.5267934064671759,
+      "grad_norm": 0.4556163251399994,
+      "learning_rate": 0.0001,
+      "loss": 1.824,
+      "step": 4586
+    },
+    {
+      "epoch": 0.526908276377003,
+      "grad_norm": 0.4342927634716034,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 4587
+    },
+    {
+      "epoch": 0.5270231462868301,
+      "grad_norm": 0.4018527865409851,
+      "learning_rate": 0.0001,
+      "loss": 1.6768,
+      "step": 4588
+    },
+    {
+      "epoch": 0.5271380161966572,
+      "grad_norm": 0.3828233480453491,
+      "learning_rate": 0.0001,
+      "loss": 1.31,
+      "step": 4589
+    },
+    {
+      "epoch": 0.5272528861064844,
+      "grad_norm": 0.4347113072872162,
+      "learning_rate": 0.0001,
+      "loss": 1.7758,
+      "step": 4590
+    },
+    {
+      "epoch": 0.5273677560163115,
+      "grad_norm": 0.38657423853874207,
+      "learning_rate": 0.0001,
+      "loss": 1.6179,
+      "step": 4591
+    },
+    {
+      "epoch": 0.5274826259261386,
+      "grad_norm": 0.42126983404159546,
+      "learning_rate": 0.0001,
+      "loss": 1.4377,
+      "step": 4592
+    },
+    {
+      "epoch": 0.5275974958359658,
+      "grad_norm": 0.45303142070770264,
+      "learning_rate": 0.0001,
+      "loss": 1.7843,
+      "step": 4593
+    },
+    {
+      "epoch": 0.5277123657457929,
+      "grad_norm": 0.395085871219635,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 4594
+    },
+    {
+      "epoch": 0.5278272356556201,
+      "grad_norm": 0.3973239064216614,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 4595
+    },
+    {
+      "epoch": 0.5279421055654472,
+      "grad_norm": 0.3768349587917328,
+      "learning_rate": 0.0001,
+      "loss": 1.325,
+      "step": 4596
+    },
+    {
+      "epoch": 0.5280569754752743,
+      "grad_norm": 0.43345212936401367,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 4597
+    },
+    {
+      "epoch": 0.5281718453851014,
+      "grad_norm": 0.4553665816783905,
+      "learning_rate": 0.0001,
+      "loss": 1.7439,
+      "step": 4598
+    },
+    {
+      "epoch": 0.5282867152949285,
+      "grad_norm": 0.4854236841201782,
+      "learning_rate": 0.0001,
+      "loss": 1.9161,
+      "step": 4599
+    },
+    {
+      "epoch": 0.5284015852047557,
+      "grad_norm": 0.421282559633255,
+      "learning_rate": 0.0001,
+      "loss": 1.5079,
+      "step": 4600
+    },
+    {
+      "epoch": 0.5285164551145828,
+      "grad_norm": 0.4243543744087219,
+      "learning_rate": 0.0001,
+      "loss": 1.6191,
+      "step": 4601
+    },
+    {
+      "epoch": 0.5286313250244099,
+      "grad_norm": 0.4060072600841522,
+      "learning_rate": 0.0001,
+      "loss": 1.5374,
+      "step": 4602
+    },
+    {
+      "epoch": 0.528746194934237,
+      "grad_norm": 0.4349260628223419,
+      "learning_rate": 0.0001,
+      "loss": 1.8664,
+      "step": 4603
+    },
+    {
+      "epoch": 0.5288610648440641,
+      "grad_norm": 0.40768831968307495,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 4604
+    },
+    {
+      "epoch": 0.5289759347538913,
+      "grad_norm": 0.39794713258743286,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 4605
+    },
+    {
+      "epoch": 0.5290908046637184,
+      "grad_norm": 0.4161005914211273,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 4606
+    },
+    {
+      "epoch": 0.5292056745735455,
+      "grad_norm": 0.44163885712623596,
+      "learning_rate": 0.0001,
+      "loss": 1.8116,
+      "step": 4607
+    },
+    {
+      "epoch": 0.5293205444833726,
+      "grad_norm": 0.5268819332122803,
+      "learning_rate": 0.0001,
+      "loss": 1.8569,
+      "step": 4608
+    },
+    {
+      "epoch": 0.5294354143931997,
+      "grad_norm": 0.46857014298439026,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 4609
+    },
+    {
+      "epoch": 0.5295502843030269,
+      "grad_norm": 0.401594340801239,
+      "learning_rate": 0.0001,
+      "loss": 1.5846,
+      "step": 4610
+    },
+    {
+      "epoch": 0.529665154212854,
+      "grad_norm": 0.42373210191726685,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 4611
+    },
+    {
+      "epoch": 0.5297800241226811,
+      "grad_norm": 0.40101566910743713,
+      "learning_rate": 0.0001,
+      "loss": 1.6537,
+      "step": 4612
+    },
+    {
+      "epoch": 0.5298948940325082,
+      "grad_norm": 0.4653320014476776,
+      "learning_rate": 0.0001,
+      "loss": 1.6741,
+      "step": 4613
+    },
+    {
+      "epoch": 0.5300097639423353,
+      "grad_norm": 0.4147876501083374,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 4614
+    },
+    {
+      "epoch": 0.5301246338521625,
+      "grad_norm": 0.4062972068786621,
+      "learning_rate": 0.0001,
+      "loss": 1.5747,
+      "step": 4615
+    },
+    {
+      "epoch": 0.5302395037619896,
+      "grad_norm": 0.4018900394439697,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 4616
+    },
+    {
+      "epoch": 0.5303543736718167,
+      "grad_norm": 0.41791051626205444,
+      "learning_rate": 0.0001,
+      "loss": 1.7225,
+      "step": 4617
+    },
+    {
+      "epoch": 0.5304692435816438,
+      "grad_norm": 0.41558316349983215,
+      "learning_rate": 0.0001,
+      "loss": 1.8203,
+      "step": 4618
+    },
+    {
+      "epoch": 0.5305841134914709,
+      "grad_norm": 0.4309985935688019,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 4619
+    },
+    {
+      "epoch": 0.5306989834012981,
+      "grad_norm": 0.4364014267921448,
+      "learning_rate": 0.0001,
+      "loss": 1.8082,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5308138533111252,
+      "grad_norm": 0.41828298568725586,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 4621
+    },
+    {
+      "epoch": 0.5309287232209523,
+      "grad_norm": 0.3910794258117676,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 4622
+    },
+    {
+      "epoch": 0.5310435931307794,
+      "grad_norm": 0.4651833772659302,
+      "learning_rate": 0.0001,
+      "loss": 1.7743,
+      "step": 4623
+    },
+    {
+      "epoch": 0.5311584630406065,
+      "grad_norm": 0.44561001658439636,
+      "learning_rate": 0.0001,
+      "loss": 1.7388,
+      "step": 4624
+    },
+    {
+      "epoch": 0.5312733329504337,
+      "grad_norm": 0.3969671130180359,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 4625
+    },
+    {
+      "epoch": 0.5313882028602608,
+      "grad_norm": 0.4239133894443512,
+      "learning_rate": 0.0001,
+      "loss": 1.6436,
+      "step": 4626
+    },
+    {
+      "epoch": 0.5315030727700879,
+      "grad_norm": 0.3926708698272705,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 4627
+    },
+    {
+      "epoch": 0.531617942679915,
+      "grad_norm": 0.4034792184829712,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 4628
+    },
+    {
+      "epoch": 0.5317328125897421,
+      "grad_norm": 0.4075685143470764,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 4629
+    },
+    {
+      "epoch": 0.5318476824995693,
+      "grad_norm": 0.41798871755599976,
+      "learning_rate": 0.0001,
+      "loss": 1.7057,
+      "step": 4630
+    },
+    {
+      "epoch": 0.5319625524093964,
+      "grad_norm": 0.44857409596443176,
+      "learning_rate": 0.0001,
+      "loss": 1.7721,
+      "step": 4631
+    },
+    {
+      "epoch": 0.5320774223192235,
+      "grad_norm": 0.39396408200263977,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 4632
+    },
+    {
+      "epoch": 0.5321922922290506,
+      "grad_norm": 0.4028686285018921,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 4633
+    },
+    {
+      "epoch": 0.5323071621388777,
+      "grad_norm": 0.4449384808540344,
+      "learning_rate": 0.0001,
+      "loss": 1.6526,
+      "step": 4634
+    },
+    {
+      "epoch": 0.5324220320487049,
+      "grad_norm": 0.4593904912471771,
+      "learning_rate": 0.0001,
+      "loss": 1.8295,
+      "step": 4635
+    },
+    {
+      "epoch": 0.532536901958532,
+      "grad_norm": 0.40952321887016296,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 4636
+    },
+    {
+      "epoch": 0.5326517718683591,
+      "grad_norm": 0.421491801738739,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 4637
+    },
+    {
+      "epoch": 0.5327666417781862,
+      "grad_norm": 0.5018342733383179,
+      "learning_rate": 0.0001,
+      "loss": 1.7603,
+      "step": 4638
+    },
+    {
+      "epoch": 0.5328815116880133,
+      "grad_norm": 0.394999235868454,
+      "learning_rate": 0.0001,
+      "loss": 1.4546,
+      "step": 4639
+    },
+    {
+      "epoch": 0.5329963815978405,
+      "grad_norm": 0.3952369689941406,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5331112515076676,
+      "grad_norm": 0.44612810015678406,
+      "learning_rate": 0.0001,
+      "loss": 1.8486,
+      "step": 4641
+    },
+    {
+      "epoch": 0.5332261214174947,
+      "grad_norm": 0.4116518497467041,
+      "learning_rate": 0.0001,
+      "loss": 1.7686,
+      "step": 4642
+    },
+    {
+      "epoch": 0.5333409913273218,
+      "grad_norm": 0.4047883152961731,
+      "learning_rate": 0.0001,
+      "loss": 1.6333,
+      "step": 4643
+    },
+    {
+      "epoch": 0.5334558612371489,
+      "grad_norm": 0.4536430239677429,
+      "learning_rate": 0.0001,
+      "loss": 1.8459,
+      "step": 4644
+    },
+    {
+      "epoch": 0.5335707311469761,
+      "grad_norm": 0.4192769527435303,
+      "learning_rate": 0.0001,
+      "loss": 1.7317,
+      "step": 4645
+    },
+    {
+      "epoch": 0.5336856010568032,
+      "grad_norm": 0.426587849855423,
+      "learning_rate": 0.0001,
+      "loss": 1.8026,
+      "step": 4646
+    },
+    {
+      "epoch": 0.5338004709666303,
+      "grad_norm": 0.40313446521759033,
+      "learning_rate": 0.0001,
+      "loss": 1.6022,
+      "step": 4647
+    },
+    {
+      "epoch": 0.5339153408764574,
+      "grad_norm": 0.37919333577156067,
+      "learning_rate": 0.0001,
+      "loss": 1.3846,
+      "step": 4648
+    },
+    {
+      "epoch": 0.5340302107862845,
+      "grad_norm": 0.4192160367965698,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 4649
+    },
+    {
+      "epoch": 0.5341450806961117,
+      "grad_norm": 0.42048901319503784,
+      "learning_rate": 0.0001,
+      "loss": 1.6995,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5342599506059388,
+      "grad_norm": 0.4094361960887909,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 4651
+    },
+    {
+      "epoch": 0.5343748205157659,
+      "grad_norm": 0.4340735971927643,
+      "learning_rate": 0.0001,
+      "loss": 1.6967,
+      "step": 4652
+    },
+    {
+      "epoch": 0.534489690425593,
+      "grad_norm": 0.41635647416114807,
+      "learning_rate": 0.0001,
+      "loss": 1.7417,
+      "step": 4653
+    },
+    {
+      "epoch": 0.5346045603354201,
+      "grad_norm": 0.42730745673179626,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 4654
+    },
+    {
+      "epoch": 0.5347194302452473,
+      "grad_norm": 0.43178364634513855,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 4655
+    },
+    {
+      "epoch": 0.5348343001550744,
+      "grad_norm": 0.4151526391506195,
+      "learning_rate": 0.0001,
+      "loss": 1.7846,
+      "step": 4656
+    },
+    {
+      "epoch": 0.5349491700649015,
+      "grad_norm": 0.4182411730289459,
+      "learning_rate": 0.0001,
+      "loss": 1.6844,
+      "step": 4657
+    },
+    {
+      "epoch": 0.5350640399747286,
+      "grad_norm": 0.423403263092041,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 4658
+    },
+    {
+      "epoch": 0.5351789098845557,
+      "grad_norm": 0.41044798493385315,
+      "learning_rate": 0.0001,
+      "loss": 1.6801,
+      "step": 4659
+    },
+    {
+      "epoch": 0.5352937797943829,
+      "grad_norm": 0.4252939224243164,
+      "learning_rate": 0.0001,
+      "loss": 1.7553,
+      "step": 4660
+    },
+    {
+      "epoch": 0.53540864970421,
+      "grad_norm": 0.4315437078475952,
+      "learning_rate": 0.0001,
+      "loss": 1.7822,
+      "step": 4661
+    },
+    {
+      "epoch": 0.5355235196140371,
+      "grad_norm": 0.4264442026615143,
+      "learning_rate": 0.0001,
+      "loss": 1.7567,
+      "step": 4662
+    },
+    {
+      "epoch": 0.5356383895238642,
+      "grad_norm": 0.3990129232406616,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 4663
+    },
+    {
+      "epoch": 0.5357532594336913,
+      "grad_norm": 0.4486422836780548,
+      "learning_rate": 0.0001,
+      "loss": 1.9106,
+      "step": 4664
+    },
+    {
+      "epoch": 0.5358681293435185,
+      "grad_norm": 0.4086044430732727,
+      "learning_rate": 0.0001,
+      "loss": 1.7818,
+      "step": 4665
+    },
+    {
+      "epoch": 0.5359829992533456,
+      "grad_norm": 0.39344853162765503,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 4666
+    },
+    {
+      "epoch": 0.5360978691631727,
+      "grad_norm": 0.4144117534160614,
+      "learning_rate": 0.0001,
+      "loss": 1.4561,
+      "step": 4667
+    },
+    {
+      "epoch": 0.5362127390729998,
+      "grad_norm": 0.40345409512519836,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 4668
+    },
+    {
+      "epoch": 0.5363276089828269,
+      "grad_norm": 0.4093291461467743,
+      "learning_rate": 0.0001,
+      "loss": 1.5001,
+      "step": 4669
+    },
+    {
+      "epoch": 0.5364424788926541,
+      "grad_norm": 0.39114460349082947,
+      "learning_rate": 0.0001,
+      "loss": 1.4858,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5365573488024812,
+      "grad_norm": 0.4037148058414459,
+      "learning_rate": 0.0001,
+      "loss": 1.5966,
+      "step": 4671
+    },
+    {
+      "epoch": 0.5366722187123083,
+      "grad_norm": 0.3919554054737091,
+      "learning_rate": 0.0001,
+      "loss": 1.5426,
+      "step": 4672
+    },
+    {
+      "epoch": 0.5367870886221354,
+      "grad_norm": 0.45879948139190674,
+      "learning_rate": 0.0001,
+      "loss": 1.8282,
+      "step": 4673
+    },
+    {
+      "epoch": 0.5369019585319625,
+      "grad_norm": 0.4263037443161011,
+      "learning_rate": 0.0001,
+      "loss": 1.7901,
+      "step": 4674
+    },
+    {
+      "epoch": 0.5370168284417897,
+      "grad_norm": 0.4132154583930969,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 4675
+    },
+    {
+      "epoch": 0.5371316983516168,
+      "grad_norm": 0.4135374426841736,
+      "learning_rate": 0.0001,
+      "loss": 1.6075,
+      "step": 4676
+    },
+    {
+      "epoch": 0.5372465682614439,
+      "grad_norm": 0.43650367856025696,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 4677
+    },
+    {
+      "epoch": 0.537361438171271,
+      "grad_norm": 0.5079533457756042,
+      "learning_rate": 0.0001,
+      "loss": 1.7532,
+      "step": 4678
+    },
+    {
+      "epoch": 0.5374763080810981,
+      "grad_norm": 0.46438872814178467,
+      "learning_rate": 0.0001,
+      "loss": 1.7692,
+      "step": 4679
+    },
+    {
+      "epoch": 0.5375911779909253,
+      "grad_norm": 0.43697604537010193,
+      "learning_rate": 0.0001,
+      "loss": 1.7386,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5377060479007524,
+      "grad_norm": 0.41365137696266174,
+      "learning_rate": 0.0001,
+      "loss": 1.5255,
+      "step": 4681
+    },
+    {
+      "epoch": 0.5378209178105795,
+      "grad_norm": 0.42320525646209717,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 4682
+    },
+    {
+      "epoch": 0.5379357877204066,
+      "grad_norm": 0.40122270584106445,
+      "learning_rate": 0.0001,
+      "loss": 1.4893,
+      "step": 4683
+    },
+    {
+      "epoch": 0.5380506576302337,
+      "grad_norm": 0.434253990650177,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 4684
+    },
+    {
+      "epoch": 0.5381655275400609,
+      "grad_norm": 0.4121977686882019,
+      "learning_rate": 0.0001,
+      "loss": 1.639,
+      "step": 4685
+    },
+    {
+      "epoch": 0.538280397449888,
+      "grad_norm": 0.4040560722351074,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 4686
+    },
+    {
+      "epoch": 0.5383952673597151,
+      "grad_norm": 0.4145204722881317,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 4687
+    },
+    {
+      "epoch": 0.5385101372695422,
+      "grad_norm": 0.39439859986305237,
+      "learning_rate": 0.0001,
+      "loss": 1.4481,
+      "step": 4688
+    },
+    {
+      "epoch": 0.5386250071793693,
+      "grad_norm": 0.4014344811439514,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 4689
+    },
+    {
+      "epoch": 0.5387398770891965,
+      "grad_norm": 0.42101871967315674,
+      "learning_rate": 0.0001,
+      "loss": 1.6869,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5388547469990236,
+      "grad_norm": 0.45102766156196594,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 4691
+    },
+    {
+      "epoch": 0.5389696169088507,
+      "grad_norm": 0.43720656633377075,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 4692
+    },
+    {
+      "epoch": 0.5390844868186778,
+      "grad_norm": 0.4264827072620392,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 4693
+    },
+    {
+      "epoch": 0.5391993567285049,
+      "grad_norm": 0.4725572466850281,
+      "learning_rate": 0.0001,
+      "loss": 1.7914,
+      "step": 4694
+    },
+    {
+      "epoch": 0.5393142266383321,
+      "grad_norm": 0.4113958477973938,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 4695
+    },
+    {
+      "epoch": 0.5394290965481592,
+      "grad_norm": 0.4475966691970825,
+      "learning_rate": 0.0001,
+      "loss": 1.7054,
+      "step": 4696
+    },
+    {
+      "epoch": 0.5395439664579863,
+      "grad_norm": 0.43861621618270874,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 4697
+    },
+    {
+      "epoch": 0.5396588363678134,
+      "grad_norm": 0.4491176903247833,
+      "learning_rate": 0.0001,
+      "loss": 1.5547,
+      "step": 4698
+    },
+    {
+      "epoch": 0.5397737062776405,
+      "grad_norm": 0.4262305498123169,
+      "learning_rate": 0.0001,
+      "loss": 1.7657,
+      "step": 4699
+    },
+    {
+      "epoch": 0.5398885761874677,
+      "grad_norm": 0.4179171025753021,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 4700
+    },
+    {
+      "epoch": 0.5400034460972948,
+      "grad_norm": 0.39834854006767273,
+      "learning_rate": 0.0001,
+      "loss": 1.5324,
+      "step": 4701
+    },
+    {
+      "epoch": 0.5401183160071219,
+      "grad_norm": 0.42186471819877625,
+      "learning_rate": 0.0001,
+      "loss": 1.8005,
+      "step": 4702
+    },
+    {
+      "epoch": 0.540233185916949,
+      "grad_norm": 0.42857295274734497,
+      "learning_rate": 0.0001,
+      "loss": 1.7399,
+      "step": 4703
+    },
+    {
+      "epoch": 0.5403480558267761,
+      "grad_norm": 0.4083864390850067,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 4704
+    },
+    {
+      "epoch": 0.5404629257366033,
+      "grad_norm": 0.4472416341304779,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 4705
+    },
+    {
+      "epoch": 0.5405777956464304,
+      "grad_norm": 0.41945722699165344,
+      "learning_rate": 0.0001,
+      "loss": 1.5004,
+      "step": 4706
+    },
+    {
+      "epoch": 0.5406926655562575,
+      "grad_norm": 0.41861411929130554,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 4707
+    },
+    {
+      "epoch": 0.5408075354660846,
+      "grad_norm": 0.45254752039909363,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 4708
+    },
+    {
+      "epoch": 0.5409224053759117,
+      "grad_norm": 0.42132070660591125,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 4709
+    },
+    {
+      "epoch": 0.5410372752857389,
+      "grad_norm": 0.4404768943786621,
+      "learning_rate": 0.0001,
+      "loss": 1.6902,
+      "step": 4710
+    },
+    {
+      "epoch": 0.541152145195566,
+      "grad_norm": 0.42923909425735474,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 4711
+    },
+    {
+      "epoch": 0.5412670151053931,
+      "grad_norm": 0.4722338020801544,
+      "learning_rate": 0.0001,
+      "loss": 1.8949,
+      "step": 4712
+    },
+    {
+      "epoch": 0.5413818850152202,
+      "grad_norm": 0.4394582211971283,
+      "learning_rate": 0.0001,
+      "loss": 1.7403,
+      "step": 4713
+    },
+    {
+      "epoch": 0.5414967549250473,
+      "grad_norm": 0.3982559144496918,
+      "learning_rate": 0.0001,
+      "loss": 1.594,
+      "step": 4714
+    },
+    {
+      "epoch": 0.5416116248348745,
+      "grad_norm": 0.42785125970840454,
+      "learning_rate": 0.0001,
+      "loss": 1.7779,
+      "step": 4715
+    },
+    {
+      "epoch": 0.5417264947447016,
+      "grad_norm": 0.38715437054634094,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 4716
+    },
+    {
+      "epoch": 0.5418413646545287,
+      "grad_norm": 0.4336237609386444,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 4717
+    },
+    {
+      "epoch": 0.5419562345643558,
+      "grad_norm": 0.4250262975692749,
+      "learning_rate": 0.0001,
+      "loss": 1.6197,
+      "step": 4718
+    },
+    {
+      "epoch": 0.5420711044741829,
+      "grad_norm": 4.473871231079102,
+      "learning_rate": 0.0001,
+      "loss": 1.6598,
+      "step": 4719
+    },
+    {
+      "epoch": 0.5421859743840101,
+      "grad_norm": 0.4131799042224884,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 4720
+    },
+    {
+      "epoch": 0.5423008442938372,
+      "grad_norm": 0.43611302971839905,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 4721
+    },
+    {
+      "epoch": 0.5424157142036643,
+      "grad_norm": 0.4363075792789459,
+      "learning_rate": 0.0001,
+      "loss": 1.5381,
+      "step": 4722
+    },
+    {
+      "epoch": 0.5425305841134914,
+      "grad_norm": 0.37694451212882996,
+      "learning_rate": 0.0001,
+      "loss": 1.4295,
+      "step": 4723
+    },
+    {
+      "epoch": 0.5426454540233185,
+      "grad_norm": 0.4214974641799927,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 4724
+    },
+    {
+      "epoch": 0.5427603239331457,
+      "grad_norm": 0.46396970748901367,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 4725
+    },
+    {
+      "epoch": 0.5428751938429728,
+      "grad_norm": 0.45121777057647705,
+      "learning_rate": 0.0001,
+      "loss": 1.7544,
+      "step": 4726
+    },
+    {
+      "epoch": 0.5429900637527999,
+      "grad_norm": 0.40204212069511414,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 4727
+    },
+    {
+      "epoch": 0.543104933662627,
+      "grad_norm": 0.4438580572605133,
+      "learning_rate": 0.0001,
+      "loss": 1.7001,
+      "step": 4728
+    },
+    {
+      "epoch": 0.5432198035724541,
+      "grad_norm": 0.41876137256622314,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 4729
+    },
+    {
+      "epoch": 0.5433346734822814,
+      "grad_norm": 0.4323340952396393,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 4730
+    },
+    {
+      "epoch": 0.5434495433921085,
+      "grad_norm": 0.4222297668457031,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 4731
+    },
+    {
+      "epoch": 0.5435644133019356,
+      "grad_norm": 0.4332796633243561,
+      "learning_rate": 0.0001,
+      "loss": 1.6464,
+      "step": 4732
+    },
+    {
+      "epoch": 0.5436792832117627,
+      "grad_norm": 0.4457077383995056,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 4733
+    },
+    {
+      "epoch": 0.5437941531215899,
+      "grad_norm": 0.40646892786026,
+      "learning_rate": 0.0001,
+      "loss": 1.3511,
+      "step": 4734
+    },
+    {
+      "epoch": 0.543909023031417,
+      "grad_norm": 0.42264634370803833,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 4735
+    },
+    {
+      "epoch": 0.5440238929412441,
+      "grad_norm": 7.803822994232178,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 4736
+    },
+    {
+      "epoch": 0.5441387628510712,
+      "grad_norm": 0.4260199964046478,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 4737
+    },
+    {
+      "epoch": 0.5442536327608983,
+      "grad_norm": 0.4420236051082611,
+      "learning_rate": 0.0001,
+      "loss": 1.8486,
+      "step": 4738
+    },
+    {
+      "epoch": 0.5443685026707255,
+      "grad_norm": 0.39475566148757935,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 4739
+    },
+    {
+      "epoch": 0.5444833725805526,
+      "grad_norm": 0.48185980319976807,
+      "learning_rate": 0.0001,
+      "loss": 1.9924,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5445982424903797,
+      "grad_norm": 0.4522726833820343,
+      "learning_rate": 0.0001,
+      "loss": 1.7875,
+      "step": 4741
+    },
+    {
+      "epoch": 0.5447131124002068,
+      "grad_norm": 0.45154809951782227,
+      "learning_rate": 0.0001,
+      "loss": 1.5931,
+      "step": 4742
+    },
+    {
+      "epoch": 0.5448279823100339,
+      "grad_norm": 0.39332887530326843,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 4743
+    },
+    {
+      "epoch": 0.544942852219861,
+      "grad_norm": 0.4441022574901581,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 4744
+    },
+    {
+      "epoch": 0.5450577221296882,
+      "grad_norm": 0.42498818039894104,
+      "learning_rate": 0.0001,
+      "loss": 1.7215,
+      "step": 4745
+    },
+    {
+      "epoch": 0.5451725920395153,
+      "grad_norm": 0.39692699909210205,
+      "learning_rate": 0.0001,
+      "loss": 1.5205,
+      "step": 4746
+    },
+    {
+      "epoch": 0.5452874619493424,
+      "grad_norm": 0.4348728060722351,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 4747
+    },
+    {
+      "epoch": 0.5454023318591695,
+      "grad_norm": 0.4652867615222931,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 4748
+    },
+    {
+      "epoch": 0.5455172017689967,
+      "grad_norm": 3.736886739730835,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 4749
+    },
+    {
+      "epoch": 0.5456320716788238,
+      "grad_norm": 0.4334099292755127,
+      "learning_rate": 0.0001,
+      "loss": 1.7789,
+      "step": 4750
+    },
+    {
+      "epoch": 0.5457469415886509,
+      "grad_norm": 0.4378454387187958,
+      "learning_rate": 0.0001,
+      "loss": 1.6962,
+      "step": 4751
+    },
+    {
+      "epoch": 0.545861811498478,
+      "grad_norm": 0.46604326367378235,
+      "learning_rate": 0.0001,
+      "loss": 1.6623,
+      "step": 4752
+    },
+    {
+      "epoch": 0.5459766814083051,
+      "grad_norm": 13.21650505065918,
+      "learning_rate": 0.0001,
+      "loss": 2.8761,
+      "step": 4753
+    },
+    {
+      "epoch": 0.5460915513181323,
+      "grad_norm": 0.6902241706848145,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 4754
+    },
+    {
+      "epoch": 0.5462064212279594,
+      "grad_norm": 14.453927040100098,
+      "learning_rate": 0.0001,
+      "loss": 1.5613,
+      "step": 4755
+    },
+    {
+      "epoch": 0.5463212911377865,
+      "grad_norm": 1.930003046989441,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 4756
+    },
+    {
+      "epoch": 0.5464361610476136,
+      "grad_norm": 0.43484631180763245,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 4757
+    },
+    {
+      "epoch": 0.5465510309574407,
+      "grad_norm": 0.4576379656791687,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 4758
+    },
+    {
+      "epoch": 0.5466659008672679,
+      "grad_norm": 0.4986869990825653,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 4759
+    },
+    {
+      "epoch": 0.546780770777095,
+      "grad_norm": 0.4431321620941162,
+      "learning_rate": 0.0001,
+      "loss": 1.5434,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5468956406869221,
+      "grad_norm": 0.9614063501358032,
+      "learning_rate": 0.0001,
+      "loss": 1.7638,
+      "step": 4761
+    },
+    {
+      "epoch": 0.5470105105967492,
+      "grad_norm": 0.505508303642273,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 4762
+    },
+    {
+      "epoch": 0.5471253805065763,
+      "grad_norm": 0.44963207840919495,
+      "learning_rate": 0.0001,
+      "loss": 1.465,
+      "step": 4763
+    },
+    {
+      "epoch": 0.5472402504164035,
+      "grad_norm": 0.4592747092247009,
+      "learning_rate": 0.0001,
+      "loss": 1.7584,
+      "step": 4764
+    },
+    {
+      "epoch": 0.5473551203262306,
+      "grad_norm": 0.5370006561279297,
+      "learning_rate": 0.0001,
+      "loss": 1.744,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5474699902360577,
+      "grad_norm": 0.4782159924507141,
+      "learning_rate": 0.0001,
+      "loss": 1.5303,
+      "step": 4766
+    },
+    {
+      "epoch": 0.5475848601458848,
+      "grad_norm": 0.48043233156204224,
+      "learning_rate": 0.0001,
+      "loss": 1.886,
+      "step": 4767
+    },
+    {
+      "epoch": 0.5476997300557119,
+      "grad_norm": 0.47774550318717957,
+      "learning_rate": 0.0001,
+      "loss": 1.6583,
+      "step": 4768
+    },
+    {
+      "epoch": 0.547814599965539,
+      "grad_norm": 0.4518395662307739,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 4769
+    },
+    {
+      "epoch": 0.5479294698753662,
+      "grad_norm": 0.4505947530269623,
+      "learning_rate": 0.0001,
+      "loss": 1.6369,
+      "step": 4770
+    },
+    {
+      "epoch": 0.5480443397851933,
+      "grad_norm": 0.480541467666626,
+      "learning_rate": 0.0001,
+      "loss": 1.8852,
+      "step": 4771
+    },
+    {
+      "epoch": 0.5481592096950204,
+      "grad_norm": 0.48814818263053894,
+      "learning_rate": 0.0001,
+      "loss": 1.6918,
+      "step": 4772
+    },
+    {
+      "epoch": 0.5482740796048475,
+      "grad_norm": 0.5180444121360779,
+      "learning_rate": 0.0001,
+      "loss": 1.684,
+      "step": 4773
+    },
+    {
+      "epoch": 0.5483889495146747,
+      "grad_norm": 0.4888029396533966,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 4774
+    },
+    {
+      "epoch": 0.5485038194245018,
+      "grad_norm": 0.5578180551528931,
+      "learning_rate": 0.0001,
+      "loss": 1.5176,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5486186893343289,
+      "grad_norm": 0.47964030504226685,
+      "learning_rate": 0.0001,
+      "loss": 1.7186,
+      "step": 4776
+    },
+    {
+      "epoch": 0.548733559244156,
+      "grad_norm": 0.45499885082244873,
+      "learning_rate": 0.0001,
+      "loss": 1.6304,
+      "step": 4777
+    },
+    {
+      "epoch": 0.5488484291539831,
+      "grad_norm": 0.4086899757385254,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 4778
+    },
+    {
+      "epoch": 0.5489632990638103,
+      "grad_norm": 0.4332214593887329,
+      "learning_rate": 0.0001,
+      "loss": 1.6879,
+      "step": 4779
+    },
+    {
+      "epoch": 0.5490781689736374,
+      "grad_norm": 0.4639904499053955,
+      "learning_rate": 0.0001,
+      "loss": 1.741,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5491930388834645,
+      "grad_norm": 0.4308798313140869,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 4781
+    },
+    {
+      "epoch": 0.5493079087932916,
+      "grad_norm": 0.4329991042613983,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 4782
+    },
+    {
+      "epoch": 0.5494227787031187,
+      "grad_norm": 0.48464497923851013,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 4783
+    },
+    {
+      "epoch": 0.5495376486129459,
+      "grad_norm": 0.44216474890708923,
+      "learning_rate": 0.0001,
+      "loss": 1.744,
+      "step": 4784
+    },
+    {
+      "epoch": 0.549652518522773,
+      "grad_norm": 0.44045913219451904,
+      "learning_rate": 0.0001,
+      "loss": 1.6436,
+      "step": 4785
+    },
+    {
+      "epoch": 0.5497673884326001,
+      "grad_norm": 0.4664250910282135,
+      "learning_rate": 0.0001,
+      "loss": 1.634,
+      "step": 4786
+    },
+    {
+      "epoch": 0.5498822583424272,
+      "grad_norm": 0.41855260729789734,
+      "learning_rate": 0.0001,
+      "loss": 1.7692,
+      "step": 4787
+    },
+    {
+      "epoch": 0.5499971282522543,
+      "grad_norm": 0.41777145862579346,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 4788
+    },
+    {
+      "epoch": 0.5501119981620815,
+      "grad_norm": 0.4096580147743225,
+      "learning_rate": 0.0001,
+      "loss": 1.7369,
+      "step": 4789
+    },
+    {
+      "epoch": 0.5502268680719086,
+      "grad_norm": 0.4547279477119446,
+      "learning_rate": 0.0001,
+      "loss": 1.7846,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5503417379817357,
+      "grad_norm": 0.4305475354194641,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 4791
+    },
+    {
+      "epoch": 0.5504566078915628,
+      "grad_norm": 0.461092084646225,
+      "learning_rate": 0.0001,
+      "loss": 1.5893,
+      "step": 4792
+    },
+    {
+      "epoch": 0.5505714778013899,
+      "grad_norm": 0.43664494156837463,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 4793
+    },
+    {
+      "epoch": 0.550686347711217,
+      "grad_norm": 0.4284107983112335,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 4794
+    },
+    {
+      "epoch": 0.5508012176210442,
+      "grad_norm": 0.47756850719451904,
+      "learning_rate": 0.0001,
+      "loss": 1.7553,
+      "step": 4795
+    },
+    {
+      "epoch": 0.5509160875308713,
+      "grad_norm": 0.39941316843032837,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 4796
+    },
+    {
+      "epoch": 0.5510309574406984,
+      "grad_norm": 0.4312814474105835,
+      "learning_rate": 0.0001,
+      "loss": 1.4235,
+      "step": 4797
+    },
+    {
+      "epoch": 0.5511458273505255,
+      "grad_norm": 0.4587937891483307,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 4798
+    },
+    {
+      "epoch": 0.5512606972603527,
+      "grad_norm": 0.43980634212493896,
+      "learning_rate": 0.0001,
+      "loss": 1.7708,
+      "step": 4799
+    },
+    {
+      "epoch": 0.5513755671701798,
+      "grad_norm": 0.4155730903148651,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5514904370800069,
+      "grad_norm": 0.4062232971191406,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 4801
+    },
+    {
+      "epoch": 0.551605306989834,
+      "grad_norm": 0.43724343180656433,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 4802
+    },
+    {
+      "epoch": 0.5517201768996611,
+      "grad_norm": 0.4811611473560333,
+      "learning_rate": 0.0001,
+      "loss": 1.4776,
+      "step": 4803
+    },
+    {
+      "epoch": 0.5518350468094882,
+      "grad_norm": 0.3989713191986084,
+      "learning_rate": 0.0001,
+      "loss": 1.2717,
+      "step": 4804
+    },
+    {
+      "epoch": 0.5519499167193154,
+      "grad_norm": 0.43779945373535156,
+      "learning_rate": 0.0001,
+      "loss": 1.7521,
+      "step": 4805
+    },
+    {
+      "epoch": 0.5520647866291425,
+      "grad_norm": 0.46011802554130554,
+      "learning_rate": 0.0001,
+      "loss": 1.7228,
+      "step": 4806
+    },
+    {
+      "epoch": 0.5521796565389696,
+      "grad_norm": 0.3978514075279236,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 4807
+    },
+    {
+      "epoch": 0.5522945264487967,
+      "grad_norm": 0.4245090186595917,
+      "learning_rate": 0.0001,
+      "loss": 1.6901,
+      "step": 4808
+    },
+    {
+      "epoch": 0.5524093963586238,
+      "grad_norm": 0.46046197414398193,
+      "learning_rate": 0.0001,
+      "loss": 1.6541,
+      "step": 4809
+    },
+    {
+      "epoch": 0.552524266268451,
+      "grad_norm": 0.40693244338035583,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5526391361782781,
+      "grad_norm": 0.4510886073112488,
+      "learning_rate": 0.0001,
+      "loss": 1.6634,
+      "step": 4811
+    },
+    {
+      "epoch": 0.5527540060881052,
+      "grad_norm": 0.43159717321395874,
+      "learning_rate": 0.0001,
+      "loss": 1.7195,
+      "step": 4812
+    },
+    {
+      "epoch": 0.5528688759979323,
+      "grad_norm": 0.6375383734703064,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 4813
+    },
+    {
+      "epoch": 0.5529837459077594,
+      "grad_norm": 0.4993058741092682,
+      "learning_rate": 0.0001,
+      "loss": 1.8643,
+      "step": 4814
+    },
+    {
+      "epoch": 0.5530986158175866,
+      "grad_norm": 0.44548937678337097,
+      "learning_rate": 0.0001,
+      "loss": 1.6889,
+      "step": 4815
+    },
+    {
+      "epoch": 0.5532134857274137,
+      "grad_norm": 0.416087806224823,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 4816
+    },
+    {
+      "epoch": 0.5533283556372408,
+      "grad_norm": 0.4004303216934204,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 4817
+    },
+    {
+      "epoch": 0.5534432255470679,
+      "grad_norm": 0.46140408515930176,
+      "learning_rate": 0.0001,
+      "loss": 1.8474,
+      "step": 4818
+    },
+    {
+      "epoch": 0.553558095456895,
+      "grad_norm": 0.41780680418014526,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 4819
+    },
+    {
+      "epoch": 0.5536729653667222,
+      "grad_norm": 0.4246324598789215,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 4820
+    },
+    {
+      "epoch": 0.5537878352765493,
+      "grad_norm": 0.4020591080188751,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 4821
+    },
+    {
+      "epoch": 0.5539027051863764,
+      "grad_norm": 0.4175722301006317,
+      "learning_rate": 0.0001,
+      "loss": 1.6808,
+      "step": 4822
+    },
+    {
+      "epoch": 0.5540175750962035,
+      "grad_norm": 0.41285720467567444,
+      "learning_rate": 0.0001,
+      "loss": 1.7206,
+      "step": 4823
+    },
+    {
+      "epoch": 0.5541324450060306,
+      "grad_norm": 0.3958023190498352,
+      "learning_rate": 0.0001,
+      "loss": 1.4307,
+      "step": 4824
+    },
+    {
+      "epoch": 0.5542473149158578,
+      "grad_norm": 0.43072789907455444,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 4825
+    },
+    {
+      "epoch": 0.5543621848256849,
+      "grad_norm": 0.4155644178390503,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 4826
+    },
+    {
+      "epoch": 0.554477054735512,
+      "grad_norm": 0.4563823640346527,
+      "learning_rate": 0.0001,
+      "loss": 1.503,
+      "step": 4827
+    },
+    {
+      "epoch": 0.5545919246453391,
+      "grad_norm": 0.40075716376304626,
+      "learning_rate": 0.0001,
+      "loss": 1.4557,
+      "step": 4828
+    },
+    {
+      "epoch": 0.5547067945551662,
+      "grad_norm": 0.3869268000125885,
+      "learning_rate": 0.0001,
+      "loss": 1.375,
+      "step": 4829
+    },
+    {
+      "epoch": 0.5548216644649934,
+      "grad_norm": 0.4496128559112549,
+      "learning_rate": 0.0001,
+      "loss": 1.6762,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5549365343748205,
+      "grad_norm": 0.42789193987846375,
+      "learning_rate": 0.0001,
+      "loss": 1.4755,
+      "step": 4831
+    },
+    {
+      "epoch": 0.5550514042846476,
+      "grad_norm": 0.43896588683128357,
+      "learning_rate": 0.0001,
+      "loss": 1.861,
+      "step": 4832
+    },
+    {
+      "epoch": 0.5551662741944747,
+      "grad_norm": 0.42781296372413635,
+      "learning_rate": 0.0001,
+      "loss": 1.4889,
+      "step": 4833
+    },
+    {
+      "epoch": 0.5552811441043018,
+      "grad_norm": 0.40033093094825745,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 4834
+    },
+    {
+      "epoch": 0.555396014014129,
+      "grad_norm": 0.43149489164352417,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 4835
+    },
+    {
+      "epoch": 0.5555108839239561,
+      "grad_norm": 0.4161946177482605,
+      "learning_rate": 0.0001,
+      "loss": 1.4264,
+      "step": 4836
+    },
+    {
+      "epoch": 0.5556257538337832,
+      "grad_norm": 0.45024940371513367,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 4837
+    },
+    {
+      "epoch": 0.5557406237436103,
+      "grad_norm": 0.48352208733558655,
+      "learning_rate": 0.0001,
+      "loss": 1.7129,
+      "step": 4838
+    },
+    {
+      "epoch": 0.5558554936534374,
+      "grad_norm": 0.40999725461006165,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 4839
+    },
+    {
+      "epoch": 0.5559703635632646,
+      "grad_norm": 0.43245741724967957,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5560852334730917,
+      "grad_norm": 0.4161476194858551,
+      "learning_rate": 0.0001,
+      "loss": 1.5808,
+      "step": 4841
+    },
+    {
+      "epoch": 0.5562001033829188,
+      "grad_norm": 0.39406320452690125,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 4842
+    },
+    {
+      "epoch": 0.5563149732927459,
+      "grad_norm": 0.43421339988708496,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 4843
+    },
+    {
+      "epoch": 0.556429843202573,
+      "grad_norm": 0.4208906888961792,
+      "learning_rate": 0.0001,
+      "loss": 1.4881,
+      "step": 4844
+    },
+    {
+      "epoch": 0.5565447131124002,
+      "grad_norm": 0.43722692131996155,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 4845
+    },
+    {
+      "epoch": 0.5566595830222273,
+      "grad_norm": 0.4307270348072052,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 4846
+    },
+    {
+      "epoch": 0.5567744529320544,
+      "grad_norm": 0.44914358854293823,
+      "learning_rate": 0.0001,
+      "loss": 1.5319,
+      "step": 4847
+    },
+    {
+      "epoch": 0.5568893228418815,
+      "grad_norm": 0.4444207549095154,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 4848
+    },
+    {
+      "epoch": 0.5570041927517086,
+      "grad_norm": 0.46169236302375793,
+      "learning_rate": 0.0001,
+      "loss": 1.7459,
+      "step": 4849
+    },
+    {
+      "epoch": 0.5571190626615358,
+      "grad_norm": 0.45408234000205994,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5572339325713629,
+      "grad_norm": 0.4497997760772705,
+      "learning_rate": 0.0001,
+      "loss": 1.7392,
+      "step": 4851
+    },
+    {
+      "epoch": 0.55734880248119,
+      "grad_norm": 0.4248064160346985,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 4852
+    },
+    {
+      "epoch": 0.5574636723910171,
+      "grad_norm": 0.436135470867157,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 4853
+    },
+    {
+      "epoch": 0.5575785423008442,
+      "grad_norm": 0.4491892158985138,
+      "learning_rate": 0.0001,
+      "loss": 1.7014,
+      "step": 4854
+    },
+    {
+      "epoch": 0.5576934122106714,
+      "grad_norm": 0.395537793636322,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 4855
+    },
+    {
+      "epoch": 0.5578082821204985,
+      "grad_norm": 0.46674367785453796,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 4856
+    },
+    {
+      "epoch": 0.5579231520303256,
+      "grad_norm": 0.5027632713317871,
+      "learning_rate": 0.0001,
+      "loss": 1.4204,
+      "step": 4857
+    },
+    {
+      "epoch": 0.5580380219401527,
+      "grad_norm": 0.39720940589904785,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 4858
+    },
+    {
+      "epoch": 0.5581528918499798,
+      "grad_norm": 0.4141447842121124,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 4859
+    },
+    {
+      "epoch": 0.558267761759807,
+      "grad_norm": 0.4240556061267853,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5583826316696341,
+      "grad_norm": 0.47687891125679016,
+      "learning_rate": 0.0001,
+      "loss": 1.9204,
+      "step": 4861
+    },
+    {
+      "epoch": 0.5584975015794612,
+      "grad_norm": 0.4583745002746582,
+      "learning_rate": 0.0001,
+      "loss": 1.5725,
+      "step": 4862
+    },
+    {
+      "epoch": 0.5586123714892883,
+      "grad_norm": 0.43266603350639343,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 4863
+    },
+    {
+      "epoch": 0.5587272413991154,
+      "grad_norm": 0.415556401014328,
+      "learning_rate": 0.0001,
+      "loss": 1.5649,
+      "step": 4864
+    },
+    {
+      "epoch": 0.5588421113089426,
+      "grad_norm": 0.4445532262325287,
+      "learning_rate": 0.0001,
+      "loss": 1.7728,
+      "step": 4865
+    },
+    {
+      "epoch": 0.5589569812187697,
+      "grad_norm": 0.43825092911720276,
+      "learning_rate": 0.0001,
+      "loss": 1.7384,
+      "step": 4866
+    },
+    {
+      "epoch": 0.5590718511285969,
+      "grad_norm": 0.4609036147594452,
+      "learning_rate": 0.0001,
+      "loss": 1.4781,
+      "step": 4867
+    },
+    {
+      "epoch": 0.559186721038424,
+      "grad_norm": 0.44946613907814026,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 4868
+    },
+    {
+      "epoch": 0.5593015909482512,
+      "grad_norm": 0.44441309571266174,
+      "learning_rate": 0.0001,
+      "loss": 1.7457,
+      "step": 4869
+    },
+    {
+      "epoch": 0.5594164608580783,
+      "grad_norm": 0.42556867003440857,
+      "learning_rate": 0.0001,
+      "loss": 1.7197,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5595313307679054,
+      "grad_norm": 0.47983217239379883,
+      "learning_rate": 0.0001,
+      "loss": 1.8489,
+      "step": 4871
+    },
+    {
+      "epoch": 0.5596462006777325,
+      "grad_norm": 0.46023857593536377,
+      "learning_rate": 0.0001,
+      "loss": 1.7739,
+      "step": 4872
+    },
+    {
+      "epoch": 0.5597610705875596,
+      "grad_norm": 0.4268305003643036,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 4873
+    },
+    {
+      "epoch": 0.5598759404973868,
+      "grad_norm": 0.43025729060173035,
+      "learning_rate": 0.0001,
+      "loss": 1.3838,
+      "step": 4874
+    },
+    {
+      "epoch": 0.5599908104072139,
+      "grad_norm": 0.4198823869228363,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 4875
+    },
+    {
+      "epoch": 0.560105680317041,
+      "grad_norm": 0.39771756529808044,
+      "learning_rate": 0.0001,
+      "loss": 1.3502,
+      "step": 4876
+    },
+    {
+      "epoch": 0.5602205502268681,
+      "grad_norm": 0.42094194889068604,
+      "learning_rate": 0.0001,
+      "loss": 1.7735,
+      "step": 4877
+    },
+    {
+      "epoch": 0.5603354201366952,
+      "grad_norm": 0.4145548343658447,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 4878
+    },
+    {
+      "epoch": 0.5604502900465224,
+      "grad_norm": 0.4257977306842804,
+      "learning_rate": 0.0001,
+      "loss": 1.7172,
+      "step": 4879
+    },
+    {
+      "epoch": 0.5605651599563495,
+      "grad_norm": 0.43406781554222107,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5606800298661766,
+      "grad_norm": 0.4150993525981903,
+      "learning_rate": 0.0001,
+      "loss": 1.5706,
+      "step": 4881
+    },
+    {
+      "epoch": 0.5607948997760037,
+      "grad_norm": 0.4342502951622009,
+      "learning_rate": 0.0001,
+      "loss": 1.7698,
+      "step": 4882
+    },
+    {
+      "epoch": 0.5609097696858308,
+      "grad_norm": 0.46102142333984375,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 4883
+    },
+    {
+      "epoch": 0.561024639595658,
+      "grad_norm": 0.42644503712654114,
+      "learning_rate": 0.0001,
+      "loss": 1.7872,
+      "step": 4884
+    },
+    {
+      "epoch": 0.5611395095054851,
+      "grad_norm": 0.4223010241985321,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 4885
+    },
+    {
+      "epoch": 0.5612543794153122,
+      "grad_norm": 0.4439496695995331,
+      "learning_rate": 0.0001,
+      "loss": 1.6689,
+      "step": 4886
+    },
+    {
+      "epoch": 0.5613692493251393,
+      "grad_norm": 0.4849221408367157,
+      "learning_rate": 0.0001,
+      "loss": 1.7259,
+      "step": 4887
+    },
+    {
+      "epoch": 0.5614841192349664,
+      "grad_norm": 0.4383675158023834,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 4888
+    },
+    {
+      "epoch": 0.5615989891447936,
+      "grad_norm": 0.4355122745037079,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 4889
+    },
+    {
+      "epoch": 0.5617138590546207,
+      "grad_norm": 0.4068525731563568,
+      "learning_rate": 0.0001,
+      "loss": 1.3924,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5618287289644478,
+      "grad_norm": 0.4126929044723511,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 4891
+    },
+    {
+      "epoch": 0.5619435988742749,
+      "grad_norm": 0.43039217591285706,
+      "learning_rate": 0.0001,
+      "loss": 1.5134,
+      "step": 4892
+    },
+    {
+      "epoch": 0.562058468784102,
+      "grad_norm": 0.4148006737232208,
+      "learning_rate": 0.0001,
+      "loss": 1.4452,
+      "step": 4893
+    },
+    {
+      "epoch": 0.5621733386939292,
+      "grad_norm": 0.467264860868454,
+      "learning_rate": 0.0001,
+      "loss": 1.7233,
+      "step": 4894
+    },
+    {
+      "epoch": 0.5622882086037563,
+      "grad_norm": 0.43316879868507385,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 4895
+    },
+    {
+      "epoch": 0.5624030785135834,
+      "grad_norm": 0.4467410147190094,
+      "learning_rate": 0.0001,
+      "loss": 1.7144,
+      "step": 4896
+    },
+    {
+      "epoch": 0.5625179484234105,
+      "grad_norm": 0.45276862382888794,
+      "learning_rate": 0.0001,
+      "loss": 1.7343,
+      "step": 4897
+    },
+    {
+      "epoch": 0.5626328183332376,
+      "grad_norm": 0.4133438169956207,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 4898
+    },
+    {
+      "epoch": 0.5627476882430648,
+      "grad_norm": 0.4188348352909088,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 4899
+    },
+    {
+      "epoch": 0.5628625581528919,
+      "grad_norm": 0.44320085644721985,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 4900
+    },
+    {
+      "epoch": 0.562977428062719,
+      "grad_norm": 0.43440455198287964,
+      "learning_rate": 0.0001,
+      "loss": 1.6843,
+      "step": 4901
+    },
+    {
+      "epoch": 0.5630922979725461,
+      "grad_norm": 0.45006540417671204,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 4902
+    },
+    {
+      "epoch": 0.5632071678823732,
+      "grad_norm": 0.4506426453590393,
+      "learning_rate": 0.0001,
+      "loss": 1.6472,
+      "step": 4903
+    },
+    {
+      "epoch": 0.5633220377922004,
+      "grad_norm": 0.43011045455932617,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 4904
+    },
+    {
+      "epoch": 0.5634369077020275,
+      "grad_norm": 0.4250010848045349,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5635517776118546,
+      "grad_norm": 0.4180524945259094,
+      "learning_rate": 0.0001,
+      "loss": 1.5528,
+      "step": 4906
+    },
+    {
+      "epoch": 0.5636666475216817,
+      "grad_norm": 0.4589906334877014,
+      "learning_rate": 0.0001,
+      "loss": 1.69,
+      "step": 4907
+    },
+    {
+      "epoch": 0.5637815174315088,
+      "grad_norm": 0.40024542808532715,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 4908
+    },
+    {
+      "epoch": 0.563896387341336,
+      "grad_norm": 0.4209577739238739,
+      "learning_rate": 0.0001,
+      "loss": 1.7709,
+      "step": 4909
+    },
+    {
+      "epoch": 0.5640112572511631,
+      "grad_norm": 0.42181211709976196,
+      "learning_rate": 0.0001,
+      "loss": 1.6891,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5641261271609902,
+      "grad_norm": 0.44479602575302124,
+      "learning_rate": 0.0001,
+      "loss": 1.8052,
+      "step": 4911
+    },
+    {
+      "epoch": 0.5642409970708173,
+      "grad_norm": 0.41980117559432983,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 4912
+    },
+    {
+      "epoch": 0.5643558669806444,
+      "grad_norm": 0.45046502351760864,
+      "learning_rate": 0.0001,
+      "loss": 1.6636,
+      "step": 4913
+    },
+    {
+      "epoch": 0.5644707368904716,
+      "grad_norm": 0.4410959482192993,
+      "learning_rate": 0.0001,
+      "loss": 1.751,
+      "step": 4914
+    },
+    {
+      "epoch": 0.5645856068002987,
+      "grad_norm": 0.4019821286201477,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5647004767101258,
+      "grad_norm": 0.4647962152957916,
+      "learning_rate": 0.0001,
+      "loss": 1.7969,
+      "step": 4916
+    },
+    {
+      "epoch": 0.5648153466199529,
+      "grad_norm": 0.4058244824409485,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 4917
+    },
+    {
+      "epoch": 0.56493021652978,
+      "grad_norm": 0.41798049211502075,
+      "learning_rate": 0.0001,
+      "loss": 1.5339,
+      "step": 4918
+    },
+    {
+      "epoch": 0.5650450864396072,
+      "grad_norm": 0.39506152272224426,
+      "learning_rate": 0.0001,
+      "loss": 1.2809,
+      "step": 4919
+    },
+    {
+      "epoch": 0.5651599563494343,
+      "grad_norm": 0.41670095920562744,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5652748262592614,
+      "grad_norm": 0.41250118613243103,
+      "learning_rate": 0.0001,
+      "loss": 1.5185,
+      "step": 4921
+    },
+    {
+      "epoch": 0.5653896961690885,
+      "grad_norm": 0.4120524227619171,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 4922
+    },
+    {
+      "epoch": 0.5655045660789156,
+      "grad_norm": 0.3931626081466675,
+      "learning_rate": 0.0001,
+      "loss": 1.4581,
+      "step": 4923
+    },
+    {
+      "epoch": 0.5656194359887428,
+      "grad_norm": 0.42506247758865356,
+      "learning_rate": 0.0001,
+      "loss": 1.6842,
+      "step": 4924
+    },
+    {
+      "epoch": 0.5657343058985699,
+      "grad_norm": 0.4446874260902405,
+      "learning_rate": 0.0001,
+      "loss": 1.6743,
+      "step": 4925
+    },
+    {
+      "epoch": 0.565849175808397,
+      "grad_norm": 0.3943353295326233,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 4926
+    },
+    {
+      "epoch": 0.5659640457182241,
+      "grad_norm": 0.41571202874183655,
+      "learning_rate": 0.0001,
+      "loss": 1.3642,
+      "step": 4927
+    },
+    {
+      "epoch": 0.5660789156280512,
+      "grad_norm": 0.4377326965332031,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 4928
+    },
+    {
+      "epoch": 0.5661937855378784,
+      "grad_norm": 0.4330956041812897,
+      "learning_rate": 0.0001,
+      "loss": 1.6974,
+      "step": 4929
+    },
+    {
+      "epoch": 0.5663086554477055,
+      "grad_norm": 0.4336974024772644,
+      "learning_rate": 0.0001,
+      "loss": 1.6062,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5664235253575326,
+      "grad_norm": 0.43469882011413574,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 4931
+    },
+    {
+      "epoch": 0.5665383952673597,
+      "grad_norm": 0.42299988865852356,
+      "learning_rate": 0.0001,
+      "loss": 1.6504,
+      "step": 4932
+    },
+    {
+      "epoch": 0.5666532651771868,
+      "grad_norm": 0.4548715651035309,
+      "learning_rate": 0.0001,
+      "loss": 1.771,
+      "step": 4933
+    },
+    {
+      "epoch": 0.566768135087014,
+      "grad_norm": 0.45423921942710876,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 4934
+    },
+    {
+      "epoch": 0.5668830049968411,
+      "grad_norm": 0.4271089434623718,
+      "learning_rate": 0.0001,
+      "loss": 1.6531,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5669978749066682,
+      "grad_norm": 0.4774452745914459,
+      "learning_rate": 0.0001,
+      "loss": 1.7197,
+      "step": 4936
+    },
+    {
+      "epoch": 0.5671127448164953,
+      "grad_norm": 0.4133590757846832,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 4937
+    },
+    {
+      "epoch": 0.5672276147263224,
+      "grad_norm": 0.4909397065639496,
+      "learning_rate": 0.0001,
+      "loss": 1.7266,
+      "step": 4938
+    },
+    {
+      "epoch": 0.5673424846361496,
+      "grad_norm": 0.47322049736976624,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 4939
+    },
+    {
+      "epoch": 0.5674573545459767,
+      "grad_norm": 0.42633056640625,
+      "learning_rate": 0.0001,
+      "loss": 1.5971,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5675722244558038,
+      "grad_norm": 0.43367111682891846,
+      "learning_rate": 0.0001,
+      "loss": 1.5723,
+      "step": 4941
+    },
+    {
+      "epoch": 0.5676870943656309,
+      "grad_norm": 0.4305420219898224,
+      "learning_rate": 0.0001,
+      "loss": 1.6753,
+      "step": 4942
+    },
+    {
+      "epoch": 0.567801964275458,
+      "grad_norm": 0.4496208727359772,
+      "learning_rate": 0.0001,
+      "loss": 1.8287,
+      "step": 4943
+    },
+    {
+      "epoch": 0.5679168341852852,
+      "grad_norm": 0.4456053376197815,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 4944
+    },
+    {
+      "epoch": 0.5680317040951123,
+      "grad_norm": 0.41399866342544556,
+      "learning_rate": 0.0001,
+      "loss": 1.6282,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5681465740049394,
+      "grad_norm": 0.41621148586273193,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 4946
+    },
+    {
+      "epoch": 0.5682614439147665,
+      "grad_norm": 0.4587562084197998,
+      "learning_rate": 0.0001,
+      "loss": 1.8734,
+      "step": 4947
+    },
+    {
+      "epoch": 0.5683763138245936,
+      "grad_norm": 0.4021860361099243,
+      "learning_rate": 0.0001,
+      "loss": 1.3734,
+      "step": 4948
+    },
+    {
+      "epoch": 0.5684911837344208,
+      "grad_norm": 0.4463704228401184,
+      "learning_rate": 0.0001,
+      "loss": 1.8134,
+      "step": 4949
+    },
+    {
+      "epoch": 0.5686060536442479,
+      "grad_norm": 0.424570769071579,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 4950
+    },
+    {
+      "epoch": 0.568720923554075,
+      "grad_norm": 0.42182692885398865,
+      "learning_rate": 0.0001,
+      "loss": 1.4935,
+      "step": 4951
+    },
+    {
+      "epoch": 0.5688357934639021,
+      "grad_norm": 0.4009837806224823,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 4952
+    },
+    {
+      "epoch": 0.5689506633737292,
+      "grad_norm": 0.394553542137146,
+      "learning_rate": 0.0001,
+      "loss": 1.5031,
+      "step": 4953
+    },
+    {
+      "epoch": 0.5690655332835564,
+      "grad_norm": 0.4050164520740509,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 4954
+    },
+    {
+      "epoch": 0.5691804031933835,
+      "grad_norm": 0.4225456714630127,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 4955
+    },
+    {
+      "epoch": 0.5692952731032106,
+      "grad_norm": 0.4110758304595947,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 4956
+    },
+    {
+      "epoch": 0.5694101430130377,
+      "grad_norm": 0.434222549200058,
+      "learning_rate": 0.0001,
+      "loss": 1.6659,
+      "step": 4957
+    },
+    {
+      "epoch": 0.5695250129228648,
+      "grad_norm": 0.4162403345108032,
+      "learning_rate": 0.0001,
+      "loss": 1.6895,
+      "step": 4958
+    },
+    {
+      "epoch": 0.569639882832692,
+      "grad_norm": 0.43041545152664185,
+      "learning_rate": 0.0001,
+      "loss": 1.6472,
+      "step": 4959
+    },
+    {
+      "epoch": 0.5697547527425191,
+      "grad_norm": 0.4041105806827545,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5698696226523462,
+      "grad_norm": 0.4230792820453644,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 4961
+    },
+    {
+      "epoch": 0.5699844925621733,
+      "grad_norm": 0.4470837414264679,
+      "learning_rate": 0.0001,
+      "loss": 1.9244,
+      "step": 4962
+    },
+    {
+      "epoch": 0.5700993624720004,
+      "grad_norm": 0.40653735399246216,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 4963
+    },
+    {
+      "epoch": 0.5702142323818276,
+      "grad_norm": 0.42262402176856995,
+      "learning_rate": 0.0001,
+      "loss": 1.6871,
+      "step": 4964
+    },
+    {
+      "epoch": 0.5703291022916547,
+      "grad_norm": 0.4305260181427002,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 4965
+    },
+    {
+      "epoch": 0.5704439722014818,
+      "grad_norm": 0.4613392651081085,
+      "learning_rate": 0.0001,
+      "loss": 1.7875,
+      "step": 4966
+    },
+    {
+      "epoch": 0.5705588421113089,
+      "grad_norm": 0.446806937456131,
+      "learning_rate": 0.0001,
+      "loss": 1.7816,
+      "step": 4967
+    },
+    {
+      "epoch": 0.570673712021136,
+      "grad_norm": 0.40673378109931946,
+      "learning_rate": 0.0001,
+      "loss": 1.5248,
+      "step": 4968
+    },
+    {
+      "epoch": 0.5707885819309632,
+      "grad_norm": 0.46812543272972107,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 4969
+    },
+    {
+      "epoch": 0.5709034518407903,
+      "grad_norm": 0.406392902135849,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5710183217506174,
+      "grad_norm": 0.4076795279979706,
+      "learning_rate": 0.0001,
+      "loss": 1.4139,
+      "step": 4971
+    },
+    {
+      "epoch": 0.5711331916604445,
+      "grad_norm": 0.4175947308540344,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 4972
+    },
+    {
+      "epoch": 0.5712480615702716,
+      "grad_norm": 0.4206025004386902,
+      "learning_rate": 0.0001,
+      "loss": 1.7126,
+      "step": 4973
+    },
+    {
+      "epoch": 0.5713629314800988,
+      "grad_norm": 0.41030603647232056,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 4974
+    },
+    {
+      "epoch": 0.5714778013899259,
+      "grad_norm": 0.42624354362487793,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 4975
+    },
+    {
+      "epoch": 0.571592671299753,
+      "grad_norm": 0.39379024505615234,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 4976
+    },
+    {
+      "epoch": 0.5717075412095801,
+      "grad_norm": 0.42009881138801575,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 4977
+    },
+    {
+      "epoch": 0.5718224111194072,
+      "grad_norm": 0.4457103908061981,
+      "learning_rate": 0.0001,
+      "loss": 1.4205,
+      "step": 4978
+    },
+    {
+      "epoch": 0.5719372810292344,
+      "grad_norm": 0.4116220474243164,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 4979
+    },
+    {
+      "epoch": 0.5720521509390615,
+      "grad_norm": 0.44594505429267883,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5721670208488886,
+      "grad_norm": 0.4323948919773102,
+      "learning_rate": 0.0001,
+      "loss": 1.4976,
+      "step": 4981
+    },
+    {
+      "epoch": 0.5722818907587157,
+      "grad_norm": 0.38895127177238464,
+      "learning_rate": 0.0001,
+      "loss": 1.4075,
+      "step": 4982
+    },
+    {
+      "epoch": 0.5723967606685428,
+      "grad_norm": 0.4395938515663147,
+      "learning_rate": 0.0001,
+      "loss": 1.5436,
+      "step": 4983
+    },
+    {
+      "epoch": 0.57251163057837,
+      "grad_norm": 0.43774986267089844,
+      "learning_rate": 0.0001,
+      "loss": 1.4892,
+      "step": 4984
+    },
+    {
+      "epoch": 0.5726265004881971,
+      "grad_norm": 0.4183255136013031,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 4985
+    },
+    {
+      "epoch": 0.5727413703980242,
+      "grad_norm": 0.439797580242157,
+      "learning_rate": 0.0001,
+      "loss": 1.4632,
+      "step": 4986
+    },
+    {
+      "epoch": 0.5728562403078513,
+      "grad_norm": 0.413779616355896,
+      "learning_rate": 0.0001,
+      "loss": 1.3779,
+      "step": 4987
+    },
+    {
+      "epoch": 0.5729711102176784,
+      "grad_norm": 0.4428330361843109,
+      "learning_rate": 0.0001,
+      "loss": 1.7449,
+      "step": 4988
+    },
+    {
+      "epoch": 0.5730859801275056,
+      "grad_norm": 0.41953855752944946,
+      "learning_rate": 0.0001,
+      "loss": 1.6953,
+      "step": 4989
+    },
+    {
+      "epoch": 0.5732008500373327,
+      "grad_norm": 0.4438232183456421,
+      "learning_rate": 0.0001,
+      "loss": 1.7179,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5733157199471598,
+      "grad_norm": 0.4678000807762146,
+      "learning_rate": 0.0001,
+      "loss": 1.8077,
+      "step": 4991
+    },
+    {
+      "epoch": 0.5734305898569869,
+      "grad_norm": 0.48187345266342163,
+      "learning_rate": 0.0001,
+      "loss": 1.7687,
+      "step": 4992
+    },
+    {
+      "epoch": 0.573545459766814,
+      "grad_norm": 0.4310360848903656,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 4993
+    },
+    {
+      "epoch": 0.5736603296766412,
+      "grad_norm": 0.4390687346458435,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 4994
+    },
+    {
+      "epoch": 0.5737751995864683,
+      "grad_norm": 0.44566598534584045,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 4995
+    },
+    {
+      "epoch": 0.5738900694962954,
+      "grad_norm": 0.399006724357605,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 4996
+    },
+    {
+      "epoch": 0.5740049394061225,
+      "grad_norm": 0.42733508348464966,
+      "learning_rate": 0.0001,
+      "loss": 1.7704,
+      "step": 4997
+    },
+    {
+      "epoch": 0.5741198093159496,
+      "grad_norm": 0.4273543655872345,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 4998
+    },
+    {
+      "epoch": 0.5742346792257768,
+      "grad_norm": 0.43150708079338074,
+      "learning_rate": 0.0001,
+      "loss": 1.7348,
+      "step": 4999
+    },
+    {
+      "epoch": 0.5743495491356039,
+      "grad_norm": 0.4143429398536682,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 5000
+    },
+    {
+      "epoch": 0.574464419045431,
+      "grad_norm": 0.41837653517723083,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 5001
+    },
+    {
+      "epoch": 0.5745792889552581,
+      "grad_norm": 0.4500226080417633,
+      "learning_rate": 0.0001,
+      "loss": 1.8321,
+      "step": 5002
+    },
+    {
+      "epoch": 0.5746941588650852,
+      "grad_norm": 0.43939411640167236,
+      "learning_rate": 0.0001,
+      "loss": 1.7294,
+      "step": 5003
+    },
+    {
+      "epoch": 0.5748090287749125,
+      "grad_norm": 0.4168783128261566,
+      "learning_rate": 0.0001,
+      "loss": 1.7016,
+      "step": 5004
+    },
+    {
+      "epoch": 0.5749238986847396,
+      "grad_norm": 0.4361497759819031,
+      "learning_rate": 0.0001,
+      "loss": 1.7038,
+      "step": 5005
+    },
+    {
+      "epoch": 0.5750387685945667,
+      "grad_norm": 0.4564398527145386,
+      "learning_rate": 0.0001,
+      "loss": 1.738,
+      "step": 5006
+    },
+    {
+      "epoch": 0.5751536385043938,
+      "grad_norm": 0.4513961672782898,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 5007
+    },
+    {
+      "epoch": 0.575268508414221,
+      "grad_norm": 0.4509972631931305,
+      "learning_rate": 0.0001,
+      "loss": 1.6103,
+      "step": 5008
+    },
+    {
+      "epoch": 0.5753833783240481,
+      "grad_norm": 0.42755764722824097,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 5009
+    },
+    {
+      "epoch": 0.5754982482338752,
+      "grad_norm": 0.4352383017539978,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5756131181437023,
+      "grad_norm": 0.4069184362888336,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 5011
+    },
+    {
+      "epoch": 0.5757279880535294,
+      "grad_norm": 0.4504094421863556,
+      "learning_rate": 0.0001,
+      "loss": 1.7465,
+      "step": 5012
+    },
+    {
+      "epoch": 0.5758428579633565,
+      "grad_norm": 0.4137848913669586,
+      "learning_rate": 0.0001,
+      "loss": 1.419,
+      "step": 5013
+    },
+    {
+      "epoch": 0.5759577278731837,
+      "grad_norm": 0.43577203154563904,
+      "learning_rate": 0.0001,
+      "loss": 1.8129,
+      "step": 5014
+    },
+    {
+      "epoch": 0.5760725977830108,
+      "grad_norm": 0.45914286375045776,
+      "learning_rate": 0.0001,
+      "loss": 1.8664,
+      "step": 5015
+    },
+    {
+      "epoch": 0.5761874676928379,
+      "grad_norm": 0.4018266499042511,
+      "learning_rate": 0.0001,
+      "loss": 1.4698,
+      "step": 5016
+    },
+    {
+      "epoch": 0.576302337602665,
+      "grad_norm": 0.4395134449005127,
+      "learning_rate": 0.0001,
+      "loss": 1.7316,
+      "step": 5017
+    },
+    {
+      "epoch": 0.5764172075124921,
+      "grad_norm": 0.42709699273109436,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 5018
+    },
+    {
+      "epoch": 0.5765320774223193,
+      "grad_norm": 0.40391069650650024,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 5019
+    },
+    {
+      "epoch": 0.5766469473321464,
+      "grad_norm": 0.4373941421508789,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5767618172419735,
+      "grad_norm": 0.4111486077308655,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 5021
+    },
+    {
+      "epoch": 0.5768766871518006,
+      "grad_norm": 0.44314584136009216,
+      "learning_rate": 0.0001,
+      "loss": 1.7859,
+      "step": 5022
+    },
+    {
+      "epoch": 0.5769915570616277,
+      "grad_norm": 0.38819631934165955,
+      "learning_rate": 0.0001,
+      "loss": 1.3416,
+      "step": 5023
+    },
+    {
+      "epoch": 0.5771064269714549,
+      "grad_norm": 0.4158842861652374,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 5024
+    },
+    {
+      "epoch": 0.577221296881282,
+      "grad_norm": 0.41062501072883606,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 5025
+    },
+    {
+      "epoch": 0.5773361667911091,
+      "grad_norm": 0.4440518617630005,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 5026
+    },
+    {
+      "epoch": 0.5774510367009362,
+      "grad_norm": 0.3918763995170593,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 5027
+    },
+    {
+      "epoch": 0.5775659066107633,
+      "grad_norm": 0.4936619997024536,
+      "learning_rate": 0.0001,
+      "loss": 1.9151,
+      "step": 5028
+    },
+    {
+      "epoch": 0.5776807765205905,
+      "grad_norm": 0.4350280463695526,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 5029
+    },
+    {
+      "epoch": 0.5777956464304176,
+      "grad_norm": 0.41786620020866394,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5779105163402447,
+      "grad_norm": 0.44200530648231506,
+      "learning_rate": 0.0001,
+      "loss": 1.8449,
+      "step": 5031
+    },
+    {
+      "epoch": 0.5780253862500718,
+      "grad_norm": 0.49857649207115173,
+      "learning_rate": 0.0001,
+      "loss": 1.7837,
+      "step": 5032
+    },
+    {
+      "epoch": 0.578140256159899,
+      "grad_norm": 0.40374937653541565,
+      "learning_rate": 0.0001,
+      "loss": 1.3166,
+      "step": 5033
+    },
+    {
+      "epoch": 0.5782551260697261,
+      "grad_norm": 0.425289124250412,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 5034
+    },
+    {
+      "epoch": 0.5783699959795532,
+      "grad_norm": 0.42704570293426514,
+      "learning_rate": 0.0001,
+      "loss": 1.6993,
+      "step": 5035
+    },
+    {
+      "epoch": 0.5784848658893803,
+      "grad_norm": 0.46401023864746094,
+      "learning_rate": 0.0001,
+      "loss": 1.7889,
+      "step": 5036
+    },
+    {
+      "epoch": 0.5785997357992074,
+      "grad_norm": 0.45945799350738525,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 5037
+    },
+    {
+      "epoch": 0.5787146057090345,
+      "grad_norm": 0.4467889070510864,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 5038
+    },
+    {
+      "epoch": 0.5788294756188617,
+      "grad_norm": 0.418954998254776,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 5039
+    },
+    {
+      "epoch": 0.5789443455286888,
+      "grad_norm": 0.4095263183116913,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 5040
+    },
+    {
+      "epoch": 0.5790592154385159,
+      "grad_norm": 0.39436352252960205,
+      "learning_rate": 0.0001,
+      "loss": 1.3867,
+      "step": 5041
+    },
+    {
+      "epoch": 0.579174085348343,
+      "grad_norm": 0.43269625306129456,
+      "learning_rate": 0.0001,
+      "loss": 1.686,
+      "step": 5042
+    },
+    {
+      "epoch": 0.5792889552581701,
+      "grad_norm": 0.4262961745262146,
+      "learning_rate": 0.0001,
+      "loss": 1.7065,
+      "step": 5043
+    },
+    {
+      "epoch": 0.5794038251679973,
+      "grad_norm": 0.42846837639808655,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 5044
+    },
+    {
+      "epoch": 0.5795186950778244,
+      "grad_norm": 0.4515441656112671,
+      "learning_rate": 0.0001,
+      "loss": 1.7085,
+      "step": 5045
+    },
+    {
+      "epoch": 0.5796335649876515,
+      "grad_norm": 0.4548323452472687,
+      "learning_rate": 0.0001,
+      "loss": 1.9127,
+      "step": 5046
+    },
+    {
+      "epoch": 0.5797484348974786,
+      "grad_norm": 0.4489242136478424,
+      "learning_rate": 0.0001,
+      "loss": 1.7654,
+      "step": 5047
+    },
+    {
+      "epoch": 0.5798633048073057,
+      "grad_norm": 0.4277878403663635,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 5048
+    },
+    {
+      "epoch": 0.5799781747171329,
+      "grad_norm": 0.4465863108634949,
+      "learning_rate": 0.0001,
+      "loss": 1.4391,
+      "step": 5049
+    },
+    {
+      "epoch": 0.58009304462696,
+      "grad_norm": 0.4445357322692871,
+      "learning_rate": 0.0001,
+      "loss": 1.5621,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5802079145367871,
+      "grad_norm": 0.41989558935165405,
+      "learning_rate": 0.0001,
+      "loss": 1.5528,
+      "step": 5051
+    },
+    {
+      "epoch": 0.5803227844466142,
+      "grad_norm": 0.4065076410770416,
+      "learning_rate": 0.0001,
+      "loss": 1.5414,
+      "step": 5052
+    },
+    {
+      "epoch": 0.5804376543564413,
+      "grad_norm": 0.42565080523490906,
+      "learning_rate": 0.0001,
+      "loss": 1.6824,
+      "step": 5053
+    },
+    {
+      "epoch": 0.5805525242662685,
+      "grad_norm": 0.4369344115257263,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 5054
+    },
+    {
+      "epoch": 0.5806673941760956,
+      "grad_norm": 0.4437665343284607,
+      "learning_rate": 0.0001,
+      "loss": 1.702,
+      "step": 5055
+    },
+    {
+      "epoch": 0.5807822640859227,
+      "grad_norm": 0.4107830822467804,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 5056
+    },
+    {
+      "epoch": 0.5808971339957498,
+      "grad_norm": 0.4161568582057953,
+      "learning_rate": 0.0001,
+      "loss": 1.6756,
+      "step": 5057
+    },
+    {
+      "epoch": 0.581012003905577,
+      "grad_norm": 0.4613957405090332,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 5058
+    },
+    {
+      "epoch": 0.5811268738154041,
+      "grad_norm": 0.44085410237312317,
+      "learning_rate": 0.0001,
+      "loss": 1.7637,
+      "step": 5059
+    },
+    {
+      "epoch": 0.5812417437252312,
+      "grad_norm": 0.4186076819896698,
+      "learning_rate": 0.0001,
+      "loss": 1.571,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5813566136350583,
+      "grad_norm": 0.4560279846191406,
+      "learning_rate": 0.0001,
+      "loss": 1.7631,
+      "step": 5061
+    },
+    {
+      "epoch": 0.5814714835448854,
+      "grad_norm": 0.44726184010505676,
+      "learning_rate": 0.0001,
+      "loss": 1.8003,
+      "step": 5062
+    },
+    {
+      "epoch": 0.5815863534547125,
+      "grad_norm": 0.4400385320186615,
+      "learning_rate": 0.0001,
+      "loss": 1.7035,
+      "step": 5063
+    },
+    {
+      "epoch": 0.5817012233645397,
+      "grad_norm": 0.42711982131004333,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 5064
+    },
+    {
+      "epoch": 0.5818160932743668,
+      "grad_norm": 0.40731683373451233,
+      "learning_rate": 0.0001,
+      "loss": 1.715,
+      "step": 5065
+    },
+    {
+      "epoch": 0.5819309631841939,
+      "grad_norm": 0.4120251536369324,
+      "learning_rate": 0.0001,
+      "loss": 1.4904,
+      "step": 5066
+    },
+    {
+      "epoch": 0.582045833094021,
+      "grad_norm": 0.4294714033603668,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 5067
+    },
+    {
+      "epoch": 0.5821607030038481,
+      "grad_norm": 0.4543513059616089,
+      "learning_rate": 0.0001,
+      "loss": 1.7307,
+      "step": 5068
+    },
+    {
+      "epoch": 0.5822755729136753,
+      "grad_norm": 0.4606270492076874,
+      "learning_rate": 0.0001,
+      "loss": 1.7362,
+      "step": 5069
+    },
+    {
+      "epoch": 0.5823904428235024,
+      "grad_norm": 0.47500404715538025,
+      "learning_rate": 0.0001,
+      "loss": 1.7057,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5825053127333295,
+      "grad_norm": 0.39924708008766174,
+      "learning_rate": 0.0001,
+      "loss": 1.4902,
+      "step": 5071
+    },
+    {
+      "epoch": 0.5826201826431566,
+      "grad_norm": 0.4525192379951477,
+      "learning_rate": 0.0001,
+      "loss": 1.5355,
+      "step": 5072
+    },
+    {
+      "epoch": 0.5827350525529837,
+      "grad_norm": 0.462971568107605,
+      "learning_rate": 0.0001,
+      "loss": 1.5256,
+      "step": 5073
+    },
+    {
+      "epoch": 0.5828499224628109,
+      "grad_norm": 0.3806494474411011,
+      "learning_rate": 0.0001,
+      "loss": 1.3727,
+      "step": 5074
+    },
+    {
+      "epoch": 0.582964792372638,
+      "grad_norm": 0.4327925741672516,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 5075
+    },
+    {
+      "epoch": 0.5830796622824651,
+      "grad_norm": 0.39075303077697754,
+      "learning_rate": 0.0001,
+      "loss": 1.2952,
+      "step": 5076
+    },
+    {
+      "epoch": 0.5831945321922922,
+      "grad_norm": 0.4309884011745453,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 5077
+    },
+    {
+      "epoch": 0.5833094021021193,
+      "grad_norm": 0.4743507206439972,
+      "learning_rate": 0.0001,
+      "loss": 1.7537,
+      "step": 5078
+    },
+    {
+      "epoch": 0.5834242720119465,
+      "grad_norm": 0.39666301012039185,
+      "learning_rate": 0.0001,
+      "loss": 1.4356,
+      "step": 5079
+    },
+    {
+      "epoch": 0.5835391419217736,
+      "grad_norm": 0.431244432926178,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5836540118316007,
+      "grad_norm": 0.4098447859287262,
+      "learning_rate": 0.0001,
+      "loss": 1.4811,
+      "step": 5081
+    },
+    {
+      "epoch": 0.5837688817414278,
+      "grad_norm": 0.44986143708229065,
+      "learning_rate": 0.0001,
+      "loss": 1.6683,
+      "step": 5082
+    },
+    {
+      "epoch": 0.583883751651255,
+      "grad_norm": 0.40975865721702576,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 5083
+    },
+    {
+      "epoch": 0.5839986215610821,
+      "grad_norm": 0.42169931530952454,
+      "learning_rate": 0.0001,
+      "loss": 1.6373,
+      "step": 5084
+    },
+    {
+      "epoch": 0.5841134914709092,
+      "grad_norm": 0.4497613310813904,
+      "learning_rate": 0.0001,
+      "loss": 1.7151,
+      "step": 5085
+    },
+    {
+      "epoch": 0.5842283613807363,
+      "grad_norm": 0.4520888030529022,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 5086
+    },
+    {
+      "epoch": 0.5843432312905634,
+      "grad_norm": 0.42139726877212524,
+      "learning_rate": 0.0001,
+      "loss": 1.7561,
+      "step": 5087
+    },
+    {
+      "epoch": 0.5844581012003905,
+      "grad_norm": 0.4443289041519165,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 5088
+    },
+    {
+      "epoch": 0.5845729711102177,
+      "grad_norm": 0.3971986770629883,
+      "learning_rate": 0.0001,
+      "loss": 1.5313,
+      "step": 5089
+    },
+    {
+      "epoch": 0.5846878410200448,
+      "grad_norm": 0.4299372434616089,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5848027109298719,
+      "grad_norm": 0.40114906430244446,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 5091
+    },
+    {
+      "epoch": 0.584917580839699,
+      "grad_norm": 0.4438854157924652,
+      "learning_rate": 0.0001,
+      "loss": 1.6968,
+      "step": 5092
+    },
+    {
+      "epoch": 0.5850324507495261,
+      "grad_norm": 0.44349342584609985,
+      "learning_rate": 0.0001,
+      "loss": 1.6948,
+      "step": 5093
+    },
+    {
+      "epoch": 0.5851473206593533,
+      "grad_norm": 0.42675167322158813,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 5094
+    },
+    {
+      "epoch": 0.5852621905691804,
+      "grad_norm": 0.43637698888778687,
+      "learning_rate": 0.0001,
+      "loss": 1.7436,
+      "step": 5095
+    },
+    {
+      "epoch": 0.5853770604790075,
+      "grad_norm": 0.44595828652381897,
+      "learning_rate": 0.0001,
+      "loss": 1.7227,
+      "step": 5096
+    },
+    {
+      "epoch": 0.5854919303888346,
+      "grad_norm": 0.4193721413612366,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 5097
+    },
+    {
+      "epoch": 0.5856068002986617,
+      "grad_norm": 0.42000266909599304,
+      "learning_rate": 0.0001,
+      "loss": 1.6702,
+      "step": 5098
+    },
+    {
+      "epoch": 0.5857216702084889,
+      "grad_norm": 0.4284018874168396,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 5099
+    },
+    {
+      "epoch": 0.585836540118316,
+      "grad_norm": 0.40313681960105896,
+      "learning_rate": 0.0001,
+      "loss": 1.3916,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5859514100281431,
+      "grad_norm": 0.4598637521266937,
+      "learning_rate": 0.0001,
+      "loss": 1.7123,
+      "step": 5101
+    },
+    {
+      "epoch": 0.5860662799379702,
+      "grad_norm": 0.45406660437583923,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 5102
+    },
+    {
+      "epoch": 0.5861811498477973,
+      "grad_norm": 0.40444087982177734,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 5103
+    },
+    {
+      "epoch": 0.5862960197576245,
+      "grad_norm": 0.4442274272441864,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 5104
+    },
+    {
+      "epoch": 0.5864108896674516,
+      "grad_norm": 0.43179601430892944,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 5105
+    },
+    {
+      "epoch": 0.5865257595772787,
+      "grad_norm": 0.5134614706039429,
+      "learning_rate": 0.0001,
+      "loss": 1.8153,
+      "step": 5106
+    },
+    {
+      "epoch": 0.5866406294871058,
+      "grad_norm": 0.4080960750579834,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 5107
+    },
+    {
+      "epoch": 0.586755499396933,
+      "grad_norm": 0.4044763743877411,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 5108
+    },
+    {
+      "epoch": 0.5868703693067601,
+      "grad_norm": 0.40748360753059387,
+      "learning_rate": 0.0001,
+      "loss": 1.3913,
+      "step": 5109
+    },
+    {
+      "epoch": 0.5869852392165872,
+      "grad_norm": 0.46773701906204224,
+      "learning_rate": 0.0001,
+      "loss": 1.724,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5871001091264143,
+      "grad_norm": 0.46918338537216187,
+      "learning_rate": 0.0001,
+      "loss": 1.8717,
+      "step": 5111
+    },
+    {
+      "epoch": 0.5872149790362414,
+      "grad_norm": 0.45832809805870056,
+      "learning_rate": 0.0001,
+      "loss": 1.8843,
+      "step": 5112
+    },
+    {
+      "epoch": 0.5873298489460685,
+      "grad_norm": 0.4251342713832855,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 5113
+    },
+    {
+      "epoch": 0.5874447188558957,
+      "grad_norm": 0.4477440118789673,
+      "learning_rate": 0.0001,
+      "loss": 1.729,
+      "step": 5114
+    },
+    {
+      "epoch": 0.5875595887657228,
+      "grad_norm": 0.47431480884552,
+      "learning_rate": 0.0001,
+      "loss": 1.8087,
+      "step": 5115
+    },
+    {
+      "epoch": 0.5876744586755499,
+      "grad_norm": 0.4649442434310913,
+      "learning_rate": 0.0001,
+      "loss": 1.8614,
+      "step": 5116
+    },
+    {
+      "epoch": 0.587789328585377,
+      "grad_norm": 0.40479037165641785,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 5117
+    },
+    {
+      "epoch": 0.5879041984952041,
+      "grad_norm": 0.4071318805217743,
+      "learning_rate": 0.0001,
+      "loss": 1.6169,
+      "step": 5118
+    },
+    {
+      "epoch": 0.5880190684050313,
+      "grad_norm": 0.3959406316280365,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 5119
+    },
+    {
+      "epoch": 0.5881339383148584,
+      "grad_norm": 0.4508780241012573,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5882488082246855,
+      "grad_norm": 0.41295212507247925,
+      "learning_rate": 0.0001,
+      "loss": 1.4935,
+      "step": 5121
+    },
+    {
+      "epoch": 0.5883636781345126,
+      "grad_norm": 0.39473357796669006,
+      "learning_rate": 0.0001,
+      "loss": 1.4775,
+      "step": 5122
+    },
+    {
+      "epoch": 0.5884785480443397,
+      "grad_norm": 0.4360009431838989,
+      "learning_rate": 0.0001,
+      "loss": 1.8118,
+      "step": 5123
+    },
+    {
+      "epoch": 0.5885934179541669,
+      "grad_norm": 0.4192604720592499,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 5124
+    },
+    {
+      "epoch": 0.588708287863994,
+      "grad_norm": 0.44447702169418335,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 5125
+    },
+    {
+      "epoch": 0.5888231577738211,
+      "grad_norm": 0.4330410063266754,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 5126
+    },
+    {
+      "epoch": 0.5889380276836482,
+      "grad_norm": 0.44558635354042053,
+      "learning_rate": 0.0001,
+      "loss": 1.8103,
+      "step": 5127
+    },
+    {
+      "epoch": 0.5890528975934753,
+      "grad_norm": 0.4943927526473999,
+      "learning_rate": 0.0001,
+      "loss": 1.77,
+      "step": 5128
+    },
+    {
+      "epoch": 0.5891677675033025,
+      "grad_norm": 0.4345741271972656,
+      "learning_rate": 0.0001,
+      "loss": 1.5057,
+      "step": 5129
+    },
+    {
+      "epoch": 0.5892826374131296,
+      "grad_norm": 0.4273737967014313,
+      "learning_rate": 0.0001,
+      "loss": 1.6475,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5893975073229567,
+      "grad_norm": 0.4054570198059082,
+      "learning_rate": 0.0001,
+      "loss": 1.7624,
+      "step": 5131
+    },
+    {
+      "epoch": 0.5895123772327838,
+      "grad_norm": 0.43333131074905396,
+      "learning_rate": 0.0001,
+      "loss": 1.434,
+      "step": 5132
+    },
+    {
+      "epoch": 0.5896272471426109,
+      "grad_norm": 0.4401973485946655,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 5133
+    },
+    {
+      "epoch": 0.5897421170524381,
+      "grad_norm": 0.43175458908081055,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 5134
+    },
+    {
+      "epoch": 0.5898569869622652,
+      "grad_norm": 0.4577193558216095,
+      "learning_rate": 0.0001,
+      "loss": 1.825,
+      "step": 5135
+    },
+    {
+      "epoch": 0.5899718568720923,
+      "grad_norm": 0.45421233773231506,
+      "learning_rate": 0.0001,
+      "loss": 1.7356,
+      "step": 5136
+    },
+    {
+      "epoch": 0.5900867267819194,
+      "grad_norm": 0.4477955102920532,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 5137
+    },
+    {
+      "epoch": 0.5902015966917465,
+      "grad_norm": 0.41615384817123413,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 5138
+    },
+    {
+      "epoch": 0.5903164666015737,
+      "grad_norm": 0.4334186613559723,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 5139
+    },
+    {
+      "epoch": 0.5904313365114008,
+      "grad_norm": 0.4008978605270386,
+      "learning_rate": 0.0001,
+      "loss": 1.6154,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5905462064212279,
+      "grad_norm": 0.4319349229335785,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 5141
+    },
+    {
+      "epoch": 0.5906610763310551,
+      "grad_norm": 0.4281979501247406,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 5142
+    },
+    {
+      "epoch": 0.5907759462408823,
+      "grad_norm": 0.4367743134498596,
+      "learning_rate": 0.0001,
+      "loss": 1.7232,
+      "step": 5143
+    },
+    {
+      "epoch": 0.5908908161507094,
+      "grad_norm": 0.4409448504447937,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 5144
+    },
+    {
+      "epoch": 0.5910056860605365,
+      "grad_norm": 0.4476580023765564,
+      "learning_rate": 0.0001,
+      "loss": 1.6214,
+      "step": 5145
+    },
+    {
+      "epoch": 0.5911205559703636,
+      "grad_norm": 0.43535661697387695,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 5146
+    },
+    {
+      "epoch": 0.5912354258801907,
+      "grad_norm": 0.4497022330760956,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 5147
+    },
+    {
+      "epoch": 0.5913502957900179,
+      "grad_norm": 0.42515555024147034,
+      "learning_rate": 0.0001,
+      "loss": 1.4534,
+      "step": 5148
+    },
+    {
+      "epoch": 0.591465165699845,
+      "grad_norm": 0.4538893699645996,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 5149
+    },
+    {
+      "epoch": 0.5915800356096721,
+      "grad_norm": 0.4361916184425354,
+      "learning_rate": 0.0001,
+      "loss": 1.6162,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5916949055194992,
+      "grad_norm": 0.43373700976371765,
+      "learning_rate": 0.0001,
+      "loss": 1.7429,
+      "step": 5151
+    },
+    {
+      "epoch": 0.5918097754293263,
+      "grad_norm": 0.46759071946144104,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 5152
+    },
+    {
+      "epoch": 0.5919246453391535,
+      "grad_norm": 0.45296040177345276,
+      "learning_rate": 0.0001,
+      "loss": 1.6969,
+      "step": 5153
+    },
+    {
+      "epoch": 0.5920395152489806,
+      "grad_norm": 0.3908827602863312,
+      "learning_rate": 0.0001,
+      "loss": 1.5251,
+      "step": 5154
+    },
+    {
+      "epoch": 0.5921543851588077,
+      "grad_norm": 0.41083312034606934,
+      "learning_rate": 0.0001,
+      "loss": 1.53,
+      "step": 5155
+    },
+    {
+      "epoch": 0.5922692550686348,
+      "grad_norm": 0.4350758492946625,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 5156
+    },
+    {
+      "epoch": 0.5923841249784619,
+      "grad_norm": 0.44962963461875916,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 5157
+    },
+    {
+      "epoch": 0.592498994888289,
+      "grad_norm": 0.47061699628829956,
+      "learning_rate": 0.0001,
+      "loss": 1.9426,
+      "step": 5158
+    },
+    {
+      "epoch": 0.5926138647981162,
+      "grad_norm": 0.4288279712200165,
+      "learning_rate": 0.0001,
+      "loss": 1.4317,
+      "step": 5159
+    },
+    {
+      "epoch": 0.5927287347079433,
+      "grad_norm": 0.4379502534866333,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5928436046177704,
+      "grad_norm": 0.41820356249809265,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 5161
+    },
+    {
+      "epoch": 0.5929584745275975,
+      "grad_norm": 0.4359080493450165,
+      "learning_rate": 0.0001,
+      "loss": 1.7123,
+      "step": 5162
+    },
+    {
+      "epoch": 0.5930733444374247,
+      "grad_norm": 0.40916740894317627,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 5163
+    },
+    {
+      "epoch": 0.5931882143472518,
+      "grad_norm": 0.4382950961589813,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 5164
+    },
+    {
+      "epoch": 0.5933030842570789,
+      "grad_norm": 0.4391472041606903,
+      "learning_rate": 0.0001,
+      "loss": 1.8258,
+      "step": 5165
+    },
+    {
+      "epoch": 0.593417954166906,
+      "grad_norm": 0.4115656316280365,
+      "learning_rate": 0.0001,
+      "loss": 1.5322,
+      "step": 5166
+    },
+    {
+      "epoch": 0.5935328240767331,
+      "grad_norm": 0.4537063241004944,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 5167
+    },
+    {
+      "epoch": 0.5936476939865603,
+      "grad_norm": 0.4141421616077423,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 5168
+    },
+    {
+      "epoch": 0.5937625638963874,
+      "grad_norm": 0.42686209082603455,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 5169
+    },
+    {
+      "epoch": 0.5938774338062145,
+      "grad_norm": 0.4201597273349762,
+      "learning_rate": 0.0001,
+      "loss": 1.3904,
+      "step": 5170
+    },
+    {
+      "epoch": 0.5939923037160416,
+      "grad_norm": 0.43558233976364136,
+      "learning_rate": 0.0001,
+      "loss": 1.529,
+      "step": 5171
+    },
+    {
+      "epoch": 0.5941071736258687,
+      "grad_norm": 0.43399038910865784,
+      "learning_rate": 0.0001,
+      "loss": 1.6307,
+      "step": 5172
+    },
+    {
+      "epoch": 0.5942220435356959,
+      "grad_norm": 0.40567871928215027,
+      "learning_rate": 0.0001,
+      "loss": 1.4648,
+      "step": 5173
+    },
+    {
+      "epoch": 0.594336913445523,
+      "grad_norm": 0.44226759672164917,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 5174
+    },
+    {
+      "epoch": 0.5944517833553501,
+      "grad_norm": 0.4169192910194397,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 5175
+    },
+    {
+      "epoch": 0.5945666532651772,
+      "grad_norm": 0.4189157485961914,
+      "learning_rate": 0.0001,
+      "loss": 1.6832,
+      "step": 5176
+    },
+    {
+      "epoch": 0.5946815231750043,
+      "grad_norm": 0.4361801743507385,
+      "learning_rate": 0.0001,
+      "loss": 1.53,
+      "step": 5177
+    },
+    {
+      "epoch": 0.5947963930848315,
+      "grad_norm": 0.41981688141822815,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 5178
+    },
+    {
+      "epoch": 0.5949112629946586,
+      "grad_norm": 0.4301121234893799,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 5179
+    },
+    {
+      "epoch": 0.5950261329044857,
+      "grad_norm": 0.42863231897354126,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5951410028143128,
+      "grad_norm": 0.4355190694332123,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 5181
+    },
+    {
+      "epoch": 0.5952558727241399,
+      "grad_norm": 0.4153485894203186,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 5182
+    },
+    {
+      "epoch": 0.595370742633967,
+      "grad_norm": 0.44687628746032715,
+      "learning_rate": 0.0001,
+      "loss": 1.7095,
+      "step": 5183
+    },
+    {
+      "epoch": 0.5954856125437942,
+      "grad_norm": 0.4202732443809509,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 5184
+    },
+    {
+      "epoch": 0.5956004824536213,
+      "grad_norm": 0.4697241187095642,
+      "learning_rate": 0.0001,
+      "loss": 1.7244,
+      "step": 5185
+    },
+    {
+      "epoch": 0.5957153523634484,
+      "grad_norm": 0.428061306476593,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 5186
+    },
+    {
+      "epoch": 0.5958302222732755,
+      "grad_norm": 0.4194773733615875,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 5187
+    },
+    {
+      "epoch": 0.5959450921831027,
+      "grad_norm": 0.45670002698898315,
+      "learning_rate": 0.0001,
+      "loss": 1.6503,
+      "step": 5188
+    },
+    {
+      "epoch": 0.5960599620929298,
+      "grad_norm": 0.43582040071487427,
+      "learning_rate": 0.0001,
+      "loss": 1.6735,
+      "step": 5189
+    },
+    {
+      "epoch": 0.5961748320027569,
+      "grad_norm": 0.4162776470184326,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 5190
+    },
+    {
+      "epoch": 0.596289701912584,
+      "grad_norm": 0.4055253565311432,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 5191
+    },
+    {
+      "epoch": 0.5964045718224111,
+      "grad_norm": 0.4150947332382202,
+      "learning_rate": 0.0001,
+      "loss": 1.6579,
+      "step": 5192
+    },
+    {
+      "epoch": 0.5965194417322383,
+      "grad_norm": 0.4119030237197876,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 5193
+    },
+    {
+      "epoch": 0.5966343116420654,
+      "grad_norm": 0.43656405806541443,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 5194
+    },
+    {
+      "epoch": 0.5967491815518925,
+      "grad_norm": 0.4502866864204407,
+      "learning_rate": 0.0001,
+      "loss": 1.7509,
+      "step": 5195
+    },
+    {
+      "epoch": 0.5968640514617196,
+      "grad_norm": 0.4506380558013916,
+      "learning_rate": 0.0001,
+      "loss": 1.6351,
+      "step": 5196
+    },
+    {
+      "epoch": 0.5969789213715467,
+      "grad_norm": 0.48698195815086365,
+      "learning_rate": 0.0001,
+      "loss": 1.7874,
+      "step": 5197
+    },
+    {
+      "epoch": 0.5970937912813739,
+      "grad_norm": 0.42721307277679443,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 5198
+    },
+    {
+      "epoch": 0.597208661191201,
+      "grad_norm": 0.428030788898468,
+      "learning_rate": 0.0001,
+      "loss": 1.752,
+      "step": 5199
+    },
+    {
+      "epoch": 0.5973235311010281,
+      "grad_norm": 0.43251699209213257,
+      "learning_rate": 0.0001,
+      "loss": 1.8308,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5974384010108552,
+      "grad_norm": 0.4091329872608185,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 5201
+    },
+    {
+      "epoch": 0.5975532709206823,
+      "grad_norm": 0.4518309235572815,
+      "learning_rate": 0.0001,
+      "loss": 1.6743,
+      "step": 5202
+    },
+    {
+      "epoch": 0.5976681408305095,
+      "grad_norm": 0.4639250338077545,
+      "learning_rate": 0.0001,
+      "loss": 1.6573,
+      "step": 5203
+    },
+    {
+      "epoch": 0.5977830107403366,
+      "grad_norm": 0.41630759835243225,
+      "learning_rate": 0.0001,
+      "loss": 1.6118,
+      "step": 5204
+    },
+    {
+      "epoch": 0.5978978806501637,
+      "grad_norm": 0.47292619943618774,
+      "learning_rate": 0.0001,
+      "loss": 1.8113,
+      "step": 5205
+    },
+    {
+      "epoch": 0.5980127505599908,
+      "grad_norm": 0.4484859108924866,
+      "learning_rate": 0.0001,
+      "loss": 1.6676,
+      "step": 5206
+    },
+    {
+      "epoch": 0.5981276204698179,
+      "grad_norm": 0.42833200097084045,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 5207
+    },
+    {
+      "epoch": 0.598242490379645,
+      "grad_norm": 0.40881600975990295,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 5208
+    },
+    {
+      "epoch": 0.5983573602894722,
+      "grad_norm": 0.412045955657959,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 5209
+    },
+    {
+      "epoch": 0.5984722301992993,
+      "grad_norm": 0.4817086160182953,
+      "learning_rate": 0.0001,
+      "loss": 1.7922,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5985871001091264,
+      "grad_norm": 0.46558666229248047,
+      "learning_rate": 0.0001,
+      "loss": 1.7812,
+      "step": 5211
+    },
+    {
+      "epoch": 0.5987019700189535,
+      "grad_norm": 0.43379732966423035,
+      "learning_rate": 0.0001,
+      "loss": 1.7301,
+      "step": 5212
+    },
+    {
+      "epoch": 0.5988168399287807,
+      "grad_norm": 0.4646340310573578,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 5213
+    },
+    {
+      "epoch": 0.5989317098386078,
+      "grad_norm": 0.4490647315979004,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 5214
+    },
+    {
+      "epoch": 0.5990465797484349,
+      "grad_norm": 0.404638409614563,
+      "learning_rate": 0.0001,
+      "loss": 1.6963,
+      "step": 5215
+    },
+    {
+      "epoch": 0.599161449658262,
+      "grad_norm": 0.4501255452632904,
+      "learning_rate": 0.0001,
+      "loss": 1.7297,
+      "step": 5216
+    },
+    {
+      "epoch": 0.5992763195680891,
+      "grad_norm": 0.39730724692344666,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 5217
+    },
+    {
+      "epoch": 0.5993911894779163,
+      "grad_norm": 0.5064243674278259,
+      "learning_rate": 0.0001,
+      "loss": 1.8652,
+      "step": 5218
+    },
+    {
+      "epoch": 0.5995060593877434,
+      "grad_norm": 0.41334420442581177,
+      "learning_rate": 0.0001,
+      "loss": 1.7021,
+      "step": 5219
+    },
+    {
+      "epoch": 0.5996209292975705,
+      "grad_norm": 0.44307664036750793,
+      "learning_rate": 0.0001,
+      "loss": 1.7261,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5997357992073976,
+      "grad_norm": 0.4100275933742523,
+      "learning_rate": 0.0001,
+      "loss": 1.5175,
+      "step": 5221
+    },
+    {
+      "epoch": 0.5998506691172247,
+      "grad_norm": 0.4305122494697571,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 5222
+    },
+    {
+      "epoch": 0.5999655390270519,
+      "grad_norm": 0.8222983479499817,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 5223
+    },
+    {
+      "epoch": 0.600080408936879,
+      "grad_norm": 0.46560120582580566,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 5224
+    },
+    {
+      "epoch": 0.6001952788467061,
+      "grad_norm": 0.413687139749527,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 5225
+    },
+    {
+      "epoch": 0.6003101487565332,
+      "grad_norm": 0.4451950192451477,
+      "learning_rate": 0.0001,
+      "loss": 1.6965,
+      "step": 5226
+    },
+    {
+      "epoch": 0.6004250186663603,
+      "grad_norm": 0.42893052101135254,
+      "learning_rate": 0.0001,
+      "loss": 1.4295,
+      "step": 5227
+    },
+    {
+      "epoch": 0.6005398885761875,
+      "grad_norm": 0.4320535659790039,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 5228
+    },
+    {
+      "epoch": 0.6006547584860146,
+      "grad_norm": 0.4111475944519043,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 5229
+    },
+    {
+      "epoch": 0.6007696283958417,
+      "grad_norm": 0.4269995093345642,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 5230
+    },
+    {
+      "epoch": 0.6008844983056688,
+      "grad_norm": 0.4943247139453888,
+      "learning_rate": 0.0001,
+      "loss": 1.4476,
+      "step": 5231
+    },
+    {
+      "epoch": 0.6009993682154959,
+      "grad_norm": 0.43460339307785034,
+      "learning_rate": 0.0001,
+      "loss": 1.738,
+      "step": 5232
+    },
+    {
+      "epoch": 0.601114238125323,
+      "grad_norm": 0.45859336853027344,
+      "learning_rate": 0.0001,
+      "loss": 1.7815,
+      "step": 5233
+    },
+    {
+      "epoch": 0.6012291080351502,
+      "grad_norm": 0.4369470775127411,
+      "learning_rate": 0.0001,
+      "loss": 1.745,
+      "step": 5234
+    },
+    {
+      "epoch": 0.6013439779449773,
+      "grad_norm": 0.39994218945503235,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 5235
+    },
+    {
+      "epoch": 0.6014588478548044,
+      "grad_norm": 0.4235687851905823,
+      "learning_rate": 0.0001,
+      "loss": 1.6046,
+      "step": 5236
+    },
+    {
+      "epoch": 0.6015737177646315,
+      "grad_norm": 0.45614656805992126,
+      "learning_rate": 0.0001,
+      "loss": 1.6661,
+      "step": 5237
+    },
+    {
+      "epoch": 0.6016885876744587,
+      "grad_norm": 0.4340771436691284,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 5238
+    },
+    {
+      "epoch": 0.6018034575842858,
+      "grad_norm": 0.458238422870636,
+      "learning_rate": 0.0001,
+      "loss": 1.796,
+      "step": 5239
+    },
+    {
+      "epoch": 0.6019183274941129,
+      "grad_norm": 0.42685821652412415,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 5240
+    },
+    {
+      "epoch": 0.60203319740394,
+      "grad_norm": 0.4220592975616455,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 5241
+    },
+    {
+      "epoch": 0.6021480673137671,
+      "grad_norm": 0.41730085015296936,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 5242
+    },
+    {
+      "epoch": 0.6022629372235943,
+      "grad_norm": 0.4415682852268219,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 5243
+    },
+    {
+      "epoch": 0.6023778071334214,
+      "grad_norm": 0.41799354553222656,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 5244
+    },
+    {
+      "epoch": 0.6024926770432485,
+      "grad_norm": 0.47606340050697327,
+      "learning_rate": 0.0001,
+      "loss": 1.8567,
+      "step": 5245
+    },
+    {
+      "epoch": 0.6026075469530756,
+      "grad_norm": 0.4121111333370209,
+      "learning_rate": 0.0001,
+      "loss": 1.5148,
+      "step": 5246
+    },
+    {
+      "epoch": 0.6027224168629027,
+      "grad_norm": 0.43468672037124634,
+      "learning_rate": 0.0001,
+      "loss": 1.5985,
+      "step": 5247
+    },
+    {
+      "epoch": 0.6028372867727299,
+      "grad_norm": 0.3789305090904236,
+      "learning_rate": 0.0001,
+      "loss": 1.2786,
+      "step": 5248
+    },
+    {
+      "epoch": 0.602952156682557,
+      "grad_norm": 0.3968721330165863,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 5249
+    },
+    {
+      "epoch": 0.6030670265923841,
+      "grad_norm": 0.46839994192123413,
+      "learning_rate": 0.0001,
+      "loss": 1.9271,
+      "step": 5250
+    },
+    {
+      "epoch": 0.6031818965022112,
+      "grad_norm": 0.4195137023925781,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 5251
+    },
+    {
+      "epoch": 0.6032967664120383,
+      "grad_norm": 0.4006030261516571,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 5252
+    },
+    {
+      "epoch": 0.6034116363218655,
+      "grad_norm": 0.4589271545410156,
+      "learning_rate": 0.0001,
+      "loss": 1.791,
+      "step": 5253
+    },
+    {
+      "epoch": 0.6035265062316926,
+      "grad_norm": 0.4122845232486725,
+      "learning_rate": 0.0001,
+      "loss": 1.6732,
+      "step": 5254
+    },
+    {
+      "epoch": 0.6036413761415197,
+      "grad_norm": 0.4685841500759125,
+      "learning_rate": 0.0001,
+      "loss": 1.7381,
+      "step": 5255
+    },
+    {
+      "epoch": 0.6037562460513468,
+      "grad_norm": 0.455402672290802,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 5256
+    },
+    {
+      "epoch": 0.6038711159611739,
+      "grad_norm": 0.4526597261428833,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 5257
+    },
+    {
+      "epoch": 0.603985985871001,
+      "grad_norm": 0.42606300115585327,
+      "learning_rate": 0.0001,
+      "loss": 1.4687,
+      "step": 5258
+    },
+    {
+      "epoch": 0.6041008557808282,
+      "grad_norm": 0.45859774947166443,
+      "learning_rate": 0.0001,
+      "loss": 1.7105,
+      "step": 5259
+    },
+    {
+      "epoch": 0.6042157256906553,
+      "grad_norm": 0.4001207947731018,
+      "learning_rate": 0.0001,
+      "loss": 1.446,
+      "step": 5260
+    },
+    {
+      "epoch": 0.6043305956004824,
+      "grad_norm": 0.4063998758792877,
+      "learning_rate": 0.0001,
+      "loss": 1.6282,
+      "step": 5261
+    },
+    {
+      "epoch": 0.6044454655103095,
+      "grad_norm": 0.4181976616382599,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 5262
+    },
+    {
+      "epoch": 0.6045603354201367,
+      "grad_norm": 0.4905684292316437,
+      "learning_rate": 0.0001,
+      "loss": 1.8801,
+      "step": 5263
+    },
+    {
+      "epoch": 0.6046752053299638,
+      "grad_norm": 0.4444495737552643,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 5264
+    },
+    {
+      "epoch": 0.6047900752397909,
+      "grad_norm": 0.4244244694709778,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 5265
+    },
+    {
+      "epoch": 0.604904945149618,
+      "grad_norm": 0.48309603333473206,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 5266
+    },
+    {
+      "epoch": 0.6050198150594451,
+      "grad_norm": 0.4249451756477356,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 5267
+    },
+    {
+      "epoch": 0.6051346849692723,
+      "grad_norm": 0.4714745283126831,
+      "learning_rate": 0.0001,
+      "loss": 1.6689,
+      "step": 5268
+    },
+    {
+      "epoch": 0.6052495548790994,
+      "grad_norm": 0.4268541634082794,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 5269
+    },
+    {
+      "epoch": 0.6053644247889265,
+      "grad_norm": 0.4523731768131256,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 5270
+    },
+    {
+      "epoch": 0.6054792946987536,
+      "grad_norm": 0.47980424761772156,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 5271
+    },
+    {
+      "epoch": 0.6055941646085807,
+      "grad_norm": 0.43983185291290283,
+      "learning_rate": 0.0001,
+      "loss": 1.4902,
+      "step": 5272
+    },
+    {
+      "epoch": 0.6057090345184079,
+      "grad_norm": 0.41765522956848145,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 5273
+    },
+    {
+      "epoch": 0.605823904428235,
+      "grad_norm": 0.4522044360637665,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 5274
+    },
+    {
+      "epoch": 0.6059387743380621,
+      "grad_norm": 0.4616985321044922,
+      "learning_rate": 0.0001,
+      "loss": 1.714,
+      "step": 5275
+    },
+    {
+      "epoch": 0.6060536442478892,
+      "grad_norm": 0.4109836518764496,
+      "learning_rate": 0.0001,
+      "loss": 1.6499,
+      "step": 5276
+    },
+    {
+      "epoch": 0.6061685141577163,
+      "grad_norm": 0.4296359717845917,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 5277
+    },
+    {
+      "epoch": 0.6062833840675435,
+      "grad_norm": 0.469368577003479,
+      "learning_rate": 0.0001,
+      "loss": 1.7734,
+      "step": 5278
+    },
+    {
+      "epoch": 0.6063982539773707,
+      "grad_norm": 0.41937753558158875,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 5279
+    },
+    {
+      "epoch": 0.6065131238871978,
+      "grad_norm": 0.4088084399700165,
+      "learning_rate": 0.0001,
+      "loss": 1.3934,
+      "step": 5280
+    },
+    {
+      "epoch": 0.6066279937970249,
+      "grad_norm": 0.4353596568107605,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 5281
+    },
+    {
+      "epoch": 0.606742863706852,
+      "grad_norm": 0.4809739589691162,
+      "learning_rate": 0.0001,
+      "loss": 1.8213,
+      "step": 5282
+    },
+    {
+      "epoch": 0.6068577336166792,
+      "grad_norm": 0.4168318808078766,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 5283
+    },
+    {
+      "epoch": 0.6069726035265063,
+      "grad_norm": 0.4438597857952118,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 5284
+    },
+    {
+      "epoch": 0.6070874734363334,
+      "grad_norm": 0.4545292556285858,
+      "learning_rate": 0.0001,
+      "loss": 1.4997,
+      "step": 5285
+    },
+    {
+      "epoch": 0.6072023433461605,
+      "grad_norm": 0.4365883469581604,
+      "learning_rate": 0.0001,
+      "loss": 1.4618,
+      "step": 5286
+    },
+    {
+      "epoch": 0.6073172132559876,
+      "grad_norm": 0.4614627957344055,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 5287
+    },
+    {
+      "epoch": 0.6074320831658148,
+      "grad_norm": 0.4482710659503937,
+      "learning_rate": 0.0001,
+      "loss": 1.7119,
+      "step": 5288
+    },
+    {
+      "epoch": 0.6075469530756419,
+      "grad_norm": 0.404892235994339,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 5289
+    },
+    {
+      "epoch": 0.607661822985469,
+      "grad_norm": 0.4186989367008209,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 5290
+    },
+    {
+      "epoch": 0.6077766928952961,
+      "grad_norm": 0.39953678846359253,
+      "learning_rate": 0.0001,
+      "loss": 1.5322,
+      "step": 5291
+    },
+    {
+      "epoch": 0.6078915628051232,
+      "grad_norm": 0.4357908070087433,
+      "learning_rate": 0.0001,
+      "loss": 1.7891,
+      "step": 5292
+    },
+    {
+      "epoch": 0.6080064327149504,
+      "grad_norm": 0.4001089334487915,
+      "learning_rate": 0.0001,
+      "loss": 1.489,
+      "step": 5293
+    },
+    {
+      "epoch": 0.6081213026247775,
+      "grad_norm": 0.44134122133255005,
+      "learning_rate": 0.0001,
+      "loss": 1.5262,
+      "step": 5294
+    },
+    {
+      "epoch": 0.6082361725346046,
+      "grad_norm": 0.4329252243041992,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 5295
+    },
+    {
+      "epoch": 0.6083510424444317,
+      "grad_norm": 0.41654008626937866,
+      "learning_rate": 0.0001,
+      "loss": 1.5814,
+      "step": 5296
+    },
+    {
+      "epoch": 0.6084659123542588,
+      "grad_norm": 0.43320518732070923,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 5297
+    },
+    {
+      "epoch": 0.608580782264086,
+      "grad_norm": 0.4640411138534546,
+      "learning_rate": 0.0001,
+      "loss": 1.7619,
+      "step": 5298
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 0.4832577705383301,
+      "learning_rate": 0.0001,
+      "loss": 1.886,
+      "step": 5299
+    },
+    {
+      "epoch": 0.6088105220837402,
+      "grad_norm": 0.4550410509109497,
+      "learning_rate": 0.0001,
+      "loss": 1.6412,
+      "step": 5300
+    },
+    {
+      "epoch": 0.6089253919935673,
+      "grad_norm": 0.4446395933628082,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 5301
+    },
+    {
+      "epoch": 0.6090402619033944,
+      "grad_norm": 0.44712090492248535,
+      "learning_rate": 0.0001,
+      "loss": 1.7342,
+      "step": 5302
+    },
+    {
+      "epoch": 0.6091551318132216,
+      "grad_norm": 0.46963006258010864,
+      "learning_rate": 0.0001,
+      "loss": 1.7253,
+      "step": 5303
+    },
+    {
+      "epoch": 0.6092700017230487,
+      "grad_norm": 0.4730952978134155,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 5304
+    },
+    {
+      "epoch": 0.6093848716328758,
+      "grad_norm": 0.4162120521068573,
+      "learning_rate": 0.0001,
+      "loss": 1.3652,
+      "step": 5305
+    },
+    {
+      "epoch": 0.6094997415427029,
+      "grad_norm": 0.4251774251461029,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 5306
+    },
+    {
+      "epoch": 0.60961461145253,
+      "grad_norm": 0.4354092478752136,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 5307
+    },
+    {
+      "epoch": 0.6097294813623572,
+      "grad_norm": 0.41313931345939636,
+      "learning_rate": 0.0001,
+      "loss": 1.6076,
+      "step": 5308
+    },
+    {
+      "epoch": 0.6098443512721843,
+      "grad_norm": 0.47585734724998474,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 5309
+    },
+    {
+      "epoch": 0.6099592211820114,
+      "grad_norm": 0.41451969742774963,
+      "learning_rate": 0.0001,
+      "loss": 1.5368,
+      "step": 5310
+    },
+    {
+      "epoch": 0.6100740910918385,
+      "grad_norm": 0.42730268836021423,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 5311
+    },
+    {
+      "epoch": 0.6101889610016656,
+      "grad_norm": 0.4963441789150238,
+      "learning_rate": 0.0001,
+      "loss": 1.719,
+      "step": 5312
+    },
+    {
+      "epoch": 0.6103038309114928,
+      "grad_norm": 0.46581605076789856,
+      "learning_rate": 0.0001,
+      "loss": 1.8281,
+      "step": 5313
+    },
+    {
+      "epoch": 0.6104187008213199,
+      "grad_norm": 0.4398311972618103,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 5314
+    },
+    {
+      "epoch": 0.610533570731147,
+      "grad_norm": 0.44360533356666565,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 5315
+    },
+    {
+      "epoch": 0.6106484406409741,
+      "grad_norm": 0.4808873236179352,
+      "learning_rate": 0.0001,
+      "loss": 1.7675,
+      "step": 5316
+    },
+    {
+      "epoch": 0.6107633105508012,
+      "grad_norm": 0.40180516242980957,
+      "learning_rate": 0.0001,
+      "loss": 1.4326,
+      "step": 5317
+    },
+    {
+      "epoch": 0.6108781804606284,
+      "grad_norm": 0.4601239860057831,
+      "learning_rate": 0.0001,
+      "loss": 1.5387,
+      "step": 5318
+    },
+    {
+      "epoch": 0.6109930503704555,
+      "grad_norm": 0.4126811921596527,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 5319
+    },
+    {
+      "epoch": 0.6111079202802826,
+      "grad_norm": 0.45912793278694153,
+      "learning_rate": 0.0001,
+      "loss": 1.716,
+      "step": 5320
+    },
+    {
+      "epoch": 0.6112227901901097,
+      "grad_norm": 0.43586671352386475,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 5321
+    },
+    {
+      "epoch": 0.6113376600999368,
+      "grad_norm": 0.4181722402572632,
+      "learning_rate": 0.0001,
+      "loss": 1.6628,
+      "step": 5322
+    },
+    {
+      "epoch": 0.611452530009764,
+      "grad_norm": 0.4227443039417267,
+      "learning_rate": 0.0001,
+      "loss": 1.4071,
+      "step": 5323
+    },
+    {
+      "epoch": 0.6115673999195911,
+      "grad_norm": 0.42024925351142883,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 5324
+    },
+    {
+      "epoch": 0.6116822698294182,
+      "grad_norm": 0.42331016063690186,
+      "learning_rate": 0.0001,
+      "loss": 1.6499,
+      "step": 5325
+    },
+    {
+      "epoch": 0.6117971397392453,
+      "grad_norm": 0.4389086663722992,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 5326
+    },
+    {
+      "epoch": 0.6119120096490724,
+      "grad_norm": 0.42871832847595215,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 5327
+    },
+    {
+      "epoch": 0.6120268795588996,
+      "grad_norm": 0.4434123635292053,
+      "learning_rate": 0.0001,
+      "loss": 1.358,
+      "step": 5328
+    },
+    {
+      "epoch": 0.6121417494687267,
+      "grad_norm": 0.460252046585083,
+      "learning_rate": 0.0001,
+      "loss": 1.5934,
+      "step": 5329
+    },
+    {
+      "epoch": 0.6122566193785538,
+      "grad_norm": 0.4134620428085327,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 5330
+    },
+    {
+      "epoch": 0.6123714892883809,
+      "grad_norm": 0.4522944986820221,
+      "learning_rate": 0.0001,
+      "loss": 1.754,
+      "step": 5331
+    },
+    {
+      "epoch": 0.612486359198208,
+      "grad_norm": 0.47153130173683167,
+      "learning_rate": 0.0001,
+      "loss": 1.6896,
+      "step": 5332
+    },
+    {
+      "epoch": 0.6126012291080352,
+      "grad_norm": 0.4305824041366577,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 5333
+    },
+    {
+      "epoch": 0.6127160990178623,
+      "grad_norm": 0.4275292158126831,
+      "learning_rate": 0.0001,
+      "loss": 1.6582,
+      "step": 5334
+    },
+    {
+      "epoch": 0.6128309689276894,
+      "grad_norm": 0.44580233097076416,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 5335
+    },
+    {
+      "epoch": 0.6129458388375165,
+      "grad_norm": 0.42193603515625,
+      "learning_rate": 0.0001,
+      "loss": 1.5258,
+      "step": 5336
+    },
+    {
+      "epoch": 0.6130607087473436,
+      "grad_norm": 0.4233403503894806,
+      "learning_rate": 0.0001,
+      "loss": 1.6786,
+      "step": 5337
+    },
+    {
+      "epoch": 0.6131755786571708,
+      "grad_norm": 0.45567867159843445,
+      "learning_rate": 0.0001,
+      "loss": 1.76,
+      "step": 5338
+    },
+    {
+      "epoch": 0.6132904485669979,
+      "grad_norm": 0.4384849965572357,
+      "learning_rate": 0.0001,
+      "loss": 1.4506,
+      "step": 5339
+    },
+    {
+      "epoch": 0.613405318476825,
+      "grad_norm": 0.442428320646286,
+      "learning_rate": 0.0001,
+      "loss": 1.7631,
+      "step": 5340
+    },
+    {
+      "epoch": 0.6135201883866521,
+      "grad_norm": 0.42773962020874023,
+      "learning_rate": 0.0001,
+      "loss": 1.6109,
+      "step": 5341
+    },
+    {
+      "epoch": 0.6136350582964792,
+      "grad_norm": 0.4419418275356293,
+      "learning_rate": 0.0001,
+      "loss": 1.6981,
+      "step": 5342
+    },
+    {
+      "epoch": 0.6137499282063064,
+      "grad_norm": 0.4556467533111572,
+      "learning_rate": 0.0001,
+      "loss": 1.726,
+      "step": 5343
+    },
+    {
+      "epoch": 0.6138647981161335,
+      "grad_norm": 0.41763773560523987,
+      "learning_rate": 0.0001,
+      "loss": 1.7347,
+      "step": 5344
+    },
+    {
+      "epoch": 0.6139796680259606,
+      "grad_norm": 0.39312151074409485,
+      "learning_rate": 0.0001,
+      "loss": 1.4951,
+      "step": 5345
+    },
+    {
+      "epoch": 0.6140945379357877,
+      "grad_norm": 0.4212815761566162,
+      "learning_rate": 0.0001,
+      "loss": 1.711,
+      "step": 5346
+    },
+    {
+      "epoch": 0.6142094078456148,
+      "grad_norm": 0.425071656703949,
+      "learning_rate": 0.0001,
+      "loss": 1.5043,
+      "step": 5347
+    },
+    {
+      "epoch": 0.614324277755442,
+      "grad_norm": 0.4074556827545166,
+      "learning_rate": 0.0001,
+      "loss": 1.5206,
+      "step": 5348
+    },
+    {
+      "epoch": 0.6144391476652691,
+      "grad_norm": 0.45755401253700256,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 5349
+    },
+    {
+      "epoch": 0.6145540175750962,
+      "grad_norm": 0.43361806869506836,
+      "learning_rate": 0.0001,
+      "loss": 1.6676,
+      "step": 5350
+    },
+    {
+      "epoch": 0.6146688874849233,
+      "grad_norm": 0.45359665155410767,
+      "learning_rate": 0.0001,
+      "loss": 1.7329,
+      "step": 5351
+    },
+    {
+      "epoch": 0.6147837573947504,
+      "grad_norm": 0.4411814510822296,
+      "learning_rate": 0.0001,
+      "loss": 1.5939,
+      "step": 5352
+    },
+    {
+      "epoch": 0.6148986273045776,
+      "grad_norm": 0.4266383647918701,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 5353
+    },
+    {
+      "epoch": 0.6150134972144047,
+      "grad_norm": 0.4333922564983368,
+      "learning_rate": 0.0001,
+      "loss": 1.7112,
+      "step": 5354
+    },
+    {
+      "epoch": 0.6151283671242318,
+      "grad_norm": 0.48001882433891296,
+      "learning_rate": 0.0001,
+      "loss": 1.8559,
+      "step": 5355
+    },
+    {
+      "epoch": 0.6152432370340589,
+      "grad_norm": 0.43073657155036926,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 5356
+    },
+    {
+      "epoch": 0.615358106943886,
+      "grad_norm": 0.42971518635749817,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 5357
+    },
+    {
+      "epoch": 0.6154729768537132,
+      "grad_norm": 0.41858428716659546,
+      "learning_rate": 0.0001,
+      "loss": 1.7038,
+      "step": 5358
+    },
+    {
+      "epoch": 0.6155878467635403,
+      "grad_norm": 0.4066700041294098,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 5359
+    },
+    {
+      "epoch": 0.6157027166733674,
+      "grad_norm": 0.42960700392723083,
+      "learning_rate": 0.0001,
+      "loss": 1.7926,
+      "step": 5360
+    },
+    {
+      "epoch": 0.6158175865831945,
+      "grad_norm": 0.425981342792511,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 5361
+    },
+    {
+      "epoch": 0.6159324564930216,
+      "grad_norm": 0.44844427704811096,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 5362
+    },
+    {
+      "epoch": 0.6160473264028488,
+      "grad_norm": 0.44269078969955444,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 5363
+    },
+    {
+      "epoch": 0.6161621963126759,
+      "grad_norm": 0.4440799951553345,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 5364
+    },
+    {
+      "epoch": 0.616277066222503,
+      "grad_norm": 0.46443289518356323,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 5365
+    },
+    {
+      "epoch": 0.6163919361323301,
+      "grad_norm": 0.43236956000328064,
+      "learning_rate": 0.0001,
+      "loss": 1.4569,
+      "step": 5366
+    },
+    {
+      "epoch": 0.6165068060421572,
+      "grad_norm": 0.4475458860397339,
+      "learning_rate": 0.0001,
+      "loss": 1.6878,
+      "step": 5367
+    },
+    {
+      "epoch": 0.6166216759519844,
+      "grad_norm": 0.4264468252658844,
+      "learning_rate": 0.0001,
+      "loss": 1.6352,
+      "step": 5368
+    },
+    {
+      "epoch": 0.6167365458618115,
+      "grad_norm": 0.4327240586280823,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 5369
+    },
+    {
+      "epoch": 0.6168514157716386,
+      "grad_norm": 0.43093448877334595,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 5370
+    },
+    {
+      "epoch": 0.6169662856814657,
+      "grad_norm": 0.4652967154979706,
+      "learning_rate": 0.0001,
+      "loss": 1.8944,
+      "step": 5371
+    },
+    {
+      "epoch": 0.6170811555912928,
+      "grad_norm": 0.40782034397125244,
+      "learning_rate": 0.0001,
+      "loss": 1.4373,
+      "step": 5372
+    },
+    {
+      "epoch": 0.61719602550112,
+      "grad_norm": 0.3958606421947479,
+      "learning_rate": 0.0001,
+      "loss": 1.4501,
+      "step": 5373
+    },
+    {
+      "epoch": 0.6173108954109471,
+      "grad_norm": 0.4088459312915802,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 5374
+    },
+    {
+      "epoch": 0.6174257653207742,
+      "grad_norm": 0.42610761523246765,
+      "learning_rate": 0.0001,
+      "loss": 1.7281,
+      "step": 5375
+    },
+    {
+      "epoch": 0.6175406352306013,
+      "grad_norm": 0.4308398365974426,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 5376
+    },
+    {
+      "epoch": 0.6176555051404284,
+      "grad_norm": 0.4212062358856201,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 5377
+    },
+    {
+      "epoch": 0.6177703750502556,
+      "grad_norm": 0.4277915358543396,
+      "learning_rate": 0.0001,
+      "loss": 1.4436,
+      "step": 5378
+    },
+    {
+      "epoch": 0.6178852449600827,
+      "grad_norm": 0.45164668560028076,
+      "learning_rate": 0.0001,
+      "loss": 1.3541,
+      "step": 5379
+    },
+    {
+      "epoch": 0.6180001148699098,
+      "grad_norm": 0.46653538942337036,
+      "learning_rate": 0.0001,
+      "loss": 1.7403,
+      "step": 5380
+    },
+    {
+      "epoch": 0.6181149847797369,
+      "grad_norm": 0.4493342936038971,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 5381
+    },
+    {
+      "epoch": 0.618229854689564,
+      "grad_norm": 0.46748727560043335,
+      "learning_rate": 0.0001,
+      "loss": 1.7097,
+      "step": 5382
+    },
+    {
+      "epoch": 0.6183447245993912,
+      "grad_norm": 0.46008816361427307,
+      "learning_rate": 0.0001,
+      "loss": 1.7876,
+      "step": 5383
+    },
+    {
+      "epoch": 0.6184595945092183,
+      "grad_norm": 0.4166385531425476,
+      "learning_rate": 0.0001,
+      "loss": 1.666,
+      "step": 5384
+    },
+    {
+      "epoch": 0.6185744644190454,
+      "grad_norm": 0.46595972776412964,
+      "learning_rate": 0.0001,
+      "loss": 1.7317,
+      "step": 5385
+    },
+    {
+      "epoch": 0.6186893343288725,
+      "grad_norm": 0.41361644864082336,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 5386
+    },
+    {
+      "epoch": 0.6188042042386996,
+      "grad_norm": 0.4521470069885254,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 5387
+    },
+    {
+      "epoch": 0.6189190741485268,
+      "grad_norm": 0.4314734637737274,
+      "learning_rate": 0.0001,
+      "loss": 1.7027,
+      "step": 5388
+    },
+    {
+      "epoch": 0.6190339440583539,
+      "grad_norm": 0.49784350395202637,
+      "learning_rate": 0.0001,
+      "loss": 1.473,
+      "step": 5389
+    },
+    {
+      "epoch": 0.619148813968181,
+      "grad_norm": 0.4353158473968506,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 5390
+    },
+    {
+      "epoch": 0.6192636838780081,
+      "grad_norm": 0.405387818813324,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 5391
+    },
+    {
+      "epoch": 0.6193785537878352,
+      "grad_norm": 0.42280814051628113,
+      "learning_rate": 0.0001,
+      "loss": 1.6584,
+      "step": 5392
+    },
+    {
+      "epoch": 0.6194934236976624,
+      "grad_norm": 0.42651262879371643,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 5393
+    },
+    {
+      "epoch": 0.6196082936074895,
+      "grad_norm": 0.4402959644794464,
+      "learning_rate": 0.0001,
+      "loss": 1.5753,
+      "step": 5394
+    },
+    {
+      "epoch": 0.6197231635173166,
+      "grad_norm": 0.41979432106018066,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 5395
+    },
+    {
+      "epoch": 0.6198380334271437,
+      "grad_norm": 0.40594765543937683,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 5396
+    },
+    {
+      "epoch": 0.6199529033369708,
+      "grad_norm": 0.4297271966934204,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 5397
+    },
+    {
+      "epoch": 0.620067773246798,
+      "grad_norm": 0.47684258222579956,
+      "learning_rate": 0.0001,
+      "loss": 1.7715,
+      "step": 5398
+    },
+    {
+      "epoch": 0.6201826431566251,
+      "grad_norm": 0.44073134660720825,
+      "learning_rate": 0.0001,
+      "loss": 1.6191,
+      "step": 5399
+    },
+    {
+      "epoch": 0.6202975130664522,
+      "grad_norm": 0.4722387194633484,
+      "learning_rate": 0.0001,
+      "loss": 1.6578,
+      "step": 5400
+    },
+    {
+      "epoch": 0.6204123829762793,
+      "grad_norm": 0.4637337327003479,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 5401
+    },
+    {
+      "epoch": 0.6205272528861064,
+      "grad_norm": 0.43157336115837097,
+      "learning_rate": 0.0001,
+      "loss": 1.7081,
+      "step": 5402
+    },
+    {
+      "epoch": 0.6206421227959336,
+      "grad_norm": 0.4264506995677948,
+      "learning_rate": 0.0001,
+      "loss": 1.5326,
+      "step": 5403
+    },
+    {
+      "epoch": 0.6207569927057607,
+      "grad_norm": 0.42598217725753784,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 5404
+    },
+    {
+      "epoch": 0.6208718626155878,
+      "grad_norm": 0.4123907685279846,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 5405
+    },
+    {
+      "epoch": 0.6209867325254149,
+      "grad_norm": 0.4402943551540375,
+      "learning_rate": 0.0001,
+      "loss": 1.7162,
+      "step": 5406
+    },
+    {
+      "epoch": 0.621101602435242,
+      "grad_norm": 0.4304845929145813,
+      "learning_rate": 0.0001,
+      "loss": 1.7301,
+      "step": 5407
+    },
+    {
+      "epoch": 0.6212164723450692,
+      "grad_norm": 0.43135395646095276,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 5408
+    },
+    {
+      "epoch": 0.6213313422548963,
+      "grad_norm": 0.42243319749832153,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 5409
+    },
+    {
+      "epoch": 0.6214462121647234,
+      "grad_norm": 0.4920620024204254,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 5410
+    },
+    {
+      "epoch": 0.6215610820745505,
+      "grad_norm": 0.41864511370658875,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 5411
+    },
+    {
+      "epoch": 0.6216759519843776,
+      "grad_norm": 0.4509679675102234,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 5412
+    },
+    {
+      "epoch": 0.6217908218942048,
+      "grad_norm": 0.4350128769874573,
+      "learning_rate": 0.0001,
+      "loss": 1.7858,
+      "step": 5413
+    },
+    {
+      "epoch": 0.6219056918040319,
+      "grad_norm": 0.4748633801937103,
+      "learning_rate": 0.0001,
+      "loss": 1.7502,
+      "step": 5414
+    },
+    {
+      "epoch": 0.622020561713859,
+      "grad_norm": 0.47240012884140015,
+      "learning_rate": 0.0001,
+      "loss": 1.88,
+      "step": 5415
+    },
+    {
+      "epoch": 0.6221354316236862,
+      "grad_norm": 0.42443564534187317,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 5416
+    },
+    {
+      "epoch": 0.6222503015335134,
+      "grad_norm": 0.4257674217224121,
+      "learning_rate": 0.0001,
+      "loss": 1.726,
+      "step": 5417
+    },
+    {
+      "epoch": 0.6223651714433405,
+      "grad_norm": 0.4025082588195801,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 5418
+    },
+    {
+      "epoch": 0.6224800413531676,
+      "grad_norm": 0.4154866337776184,
+      "learning_rate": 0.0001,
+      "loss": 1.6686,
+      "step": 5419
+    },
+    {
+      "epoch": 0.6225949112629947,
+      "grad_norm": 0.41281425952911377,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 5420
+    },
+    {
+      "epoch": 0.6227097811728218,
+      "grad_norm": 0.44021958112716675,
+      "learning_rate": 0.0001,
+      "loss": 1.6801,
+      "step": 5421
+    },
+    {
+      "epoch": 0.622824651082649,
+      "grad_norm": 0.4224892556667328,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 5422
+    },
+    {
+      "epoch": 0.6229395209924761,
+      "grad_norm": 0.44819357991218567,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 5423
+    },
+    {
+      "epoch": 0.6230543909023032,
+      "grad_norm": 0.4217240512371063,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 5424
+    },
+    {
+      "epoch": 0.6231692608121303,
+      "grad_norm": 0.45825833082199097,
+      "learning_rate": 0.0001,
+      "loss": 1.6292,
+      "step": 5425
+    },
+    {
+      "epoch": 0.6232841307219574,
+      "grad_norm": 0.427919864654541,
+      "learning_rate": 0.0001,
+      "loss": 1.6708,
+      "step": 5426
+    },
+    {
+      "epoch": 0.6233990006317846,
+      "grad_norm": 0.42998164892196655,
+      "learning_rate": 0.0001,
+      "loss": 1.4521,
+      "step": 5427
+    },
+    {
+      "epoch": 0.6235138705416117,
+      "grad_norm": 0.4373745024204254,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 5428
+    },
+    {
+      "epoch": 0.6236287404514388,
+      "grad_norm": 0.4681239426136017,
+      "learning_rate": 0.0001,
+      "loss": 1.7811,
+      "step": 5429
+    },
+    {
+      "epoch": 0.6237436103612659,
+      "grad_norm": 0.4634624719619751,
+      "learning_rate": 0.0001,
+      "loss": 1.8239,
+      "step": 5430
+    },
+    {
+      "epoch": 0.623858480271093,
+      "grad_norm": 0.4383297562599182,
+      "learning_rate": 0.0001,
+      "loss": 1.5753,
+      "step": 5431
+    },
+    {
+      "epoch": 0.6239733501809201,
+      "grad_norm": 0.46512508392333984,
+      "learning_rate": 0.0001,
+      "loss": 1.8332,
+      "step": 5432
+    },
+    {
+      "epoch": 0.6240882200907473,
+      "grad_norm": 0.4566185474395752,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 5433
+    },
+    {
+      "epoch": 0.6242030900005744,
+      "grad_norm": 0.3922889828681946,
+      "learning_rate": 0.0001,
+      "loss": 1.525,
+      "step": 5434
+    },
+    {
+      "epoch": 0.6243179599104015,
+      "grad_norm": 0.4435923993587494,
+      "learning_rate": 0.0001,
+      "loss": 1.7022,
+      "step": 5435
+    },
+    {
+      "epoch": 0.6244328298202286,
+      "grad_norm": 0.39931339025497437,
+      "learning_rate": 0.0001,
+      "loss": 1.4173,
+      "step": 5436
+    },
+    {
+      "epoch": 0.6245476997300557,
+      "grad_norm": 0.44321221113204956,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 5437
+    },
+    {
+      "epoch": 0.6246625696398829,
+      "grad_norm": 0.40193137526512146,
+      "learning_rate": 0.0001,
+      "loss": 1.5338,
+      "step": 5438
+    },
+    {
+      "epoch": 0.62477743954971,
+      "grad_norm": 0.45055443048477173,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 5439
+    },
+    {
+      "epoch": 0.6248923094595371,
+      "grad_norm": 0.38443759083747864,
+      "learning_rate": 0.0001,
+      "loss": 1.4288,
+      "step": 5440
+    },
+    {
+      "epoch": 0.6250071793693642,
+      "grad_norm": 0.47293904423713684,
+      "learning_rate": 0.0001,
+      "loss": 1.7437,
+      "step": 5441
+    },
+    {
+      "epoch": 0.6251220492791913,
+      "grad_norm": 0.4477282464504242,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 5442
+    },
+    {
+      "epoch": 0.6252369191890185,
+      "grad_norm": 0.4328649938106537,
+      "learning_rate": 0.0001,
+      "loss": 1.4047,
+      "step": 5443
+    },
+    {
+      "epoch": 0.6253517890988456,
+      "grad_norm": 0.4671032428741455,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 5444
+    },
+    {
+      "epoch": 0.6254666590086727,
+      "grad_norm": 0.4175901412963867,
+      "learning_rate": 0.0001,
+      "loss": 1.5299,
+      "step": 5445
+    },
+    {
+      "epoch": 0.6255815289184998,
+      "grad_norm": 0.45132726430892944,
+      "learning_rate": 0.0001,
+      "loss": 1.5784,
+      "step": 5446
+    },
+    {
+      "epoch": 0.625696398828327,
+      "grad_norm": 0.45002493262290955,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 5447
+    },
+    {
+      "epoch": 0.6258112687381541,
+      "grad_norm": 0.49596941471099854,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 5448
+    },
+    {
+      "epoch": 0.6259261386479812,
+      "grad_norm": 0.5087515711784363,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 5449
+    },
+    {
+      "epoch": 0.6260410085578083,
+      "grad_norm": 0.43160462379455566,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 5450
+    },
+    {
+      "epoch": 0.6261558784676354,
+      "grad_norm": 0.41861292719841003,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 5451
+    },
+    {
+      "epoch": 0.6262707483774625,
+      "grad_norm": 0.3896733820438385,
+      "learning_rate": 0.0001,
+      "loss": 1.4015,
+      "step": 5452
+    },
+    {
+      "epoch": 0.6263856182872897,
+      "grad_norm": 0.4674106538295746,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 5453
+    },
+    {
+      "epoch": 0.6265004881971168,
+      "grad_norm": 0.4450713098049164,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 5454
+    },
+    {
+      "epoch": 0.6266153581069439,
+      "grad_norm": 0.5066030621528625,
+      "learning_rate": 0.0001,
+      "loss": 1.7224,
+      "step": 5455
+    },
+    {
+      "epoch": 0.626730228016771,
+      "grad_norm": 0.41346675157546997,
+      "learning_rate": 0.0001,
+      "loss": 1.6268,
+      "step": 5456
+    },
+    {
+      "epoch": 0.6268450979265981,
+      "grad_norm": 0.45128896832466125,
+      "learning_rate": 0.0001,
+      "loss": 1.5715,
+      "step": 5457
+    },
+    {
+      "epoch": 0.6269599678364253,
+      "grad_norm": 0.42013174295425415,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 5458
+    },
+    {
+      "epoch": 0.6270748377462524,
+      "grad_norm": 0.4740227460861206,
+      "learning_rate": 0.0001,
+      "loss": 1.6902,
+      "step": 5459
+    },
+    {
+      "epoch": 0.6271897076560795,
+      "grad_norm": 0.4719151258468628,
+      "learning_rate": 0.0001,
+      "loss": 1.905,
+      "step": 5460
+    },
+    {
+      "epoch": 0.6273045775659066,
+      "grad_norm": 0.46546176075935364,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 5461
+    },
+    {
+      "epoch": 0.6274194474757337,
+      "grad_norm": 0.44801339507102966,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 5462
+    },
+    {
+      "epoch": 0.6275343173855609,
+      "grad_norm": 0.49131953716278076,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 5463
+    },
+    {
+      "epoch": 0.627649187295388,
+      "grad_norm": 0.4395316243171692,
+      "learning_rate": 0.0001,
+      "loss": 1.5935,
+      "step": 5464
+    },
+    {
+      "epoch": 0.6277640572052151,
+      "grad_norm": 0.43068498373031616,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 5465
+    },
+    {
+      "epoch": 0.6278789271150422,
+      "grad_norm": 0.41301921010017395,
+      "learning_rate": 0.0001,
+      "loss": 1.41,
+      "step": 5466
+    },
+    {
+      "epoch": 0.6279937970248693,
+      "grad_norm": 0.44340062141418457,
+      "learning_rate": 0.0001,
+      "loss": 1.7067,
+      "step": 5467
+    },
+    {
+      "epoch": 0.6281086669346965,
+      "grad_norm": 0.43528372049331665,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 5468
+    },
+    {
+      "epoch": 0.6282235368445236,
+      "grad_norm": 0.4367297291755676,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 5469
+    },
+    {
+      "epoch": 0.6283384067543507,
+      "grad_norm": 0.4735720455646515,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 5470
+    },
+    {
+      "epoch": 0.6284532766641778,
+      "grad_norm": 0.43283382058143616,
+      "learning_rate": 0.0001,
+      "loss": 1.6992,
+      "step": 5471
+    },
+    {
+      "epoch": 0.628568146574005,
+      "grad_norm": 0.446654349565506,
+      "learning_rate": 0.0001,
+      "loss": 1.4944,
+      "step": 5472
+    },
+    {
+      "epoch": 0.6286830164838321,
+      "grad_norm": 0.4372828006744385,
+      "learning_rate": 0.0001,
+      "loss": 1.4353,
+      "step": 5473
+    },
+    {
+      "epoch": 0.6287978863936592,
+      "grad_norm": 0.4139867424964905,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 5474
+    },
+    {
+      "epoch": 0.6289127563034863,
+      "grad_norm": 0.4681108891963959,
+      "learning_rate": 0.0001,
+      "loss": 1.7943,
+      "step": 5475
+    },
+    {
+      "epoch": 0.6290276262133134,
+      "grad_norm": 0.42920827865600586,
+      "learning_rate": 0.0001,
+      "loss": 1.5649,
+      "step": 5476
+    },
+    {
+      "epoch": 0.6291424961231405,
+      "grad_norm": 0.42748481035232544,
+      "learning_rate": 0.0001,
+      "loss": 1.5491,
+      "step": 5477
+    },
+    {
+      "epoch": 0.6292573660329677,
+      "grad_norm": 0.4460189938545227,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 5478
+    },
+    {
+      "epoch": 0.6293722359427948,
+      "grad_norm": 0.43394696712493896,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 5479
+    },
+    {
+      "epoch": 0.6294871058526219,
+      "grad_norm": 0.429083913564682,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 5480
+    },
+    {
+      "epoch": 0.629601975762449,
+      "grad_norm": 0.4215359687805176,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 5481
+    },
+    {
+      "epoch": 0.6297168456722761,
+      "grad_norm": 0.45450299978256226,
+      "learning_rate": 0.0001,
+      "loss": 1.3634,
+      "step": 5482
+    },
+    {
+      "epoch": 0.6298317155821033,
+      "grad_norm": 0.4605744481086731,
+      "learning_rate": 0.0001,
+      "loss": 1.7417,
+      "step": 5483
+    },
+    {
+      "epoch": 0.6299465854919304,
+      "grad_norm": 0.4325323700904846,
+      "learning_rate": 0.0001,
+      "loss": 1.4853,
+      "step": 5484
+    },
+    {
+      "epoch": 0.6300614554017575,
+      "grad_norm": 0.441438764333725,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 5485
+    },
+    {
+      "epoch": 0.6301763253115846,
+      "grad_norm": 0.5233652591705322,
+      "learning_rate": 0.0001,
+      "loss": 1.9394,
+      "step": 5486
+    },
+    {
+      "epoch": 0.6302911952214117,
+      "grad_norm": 0.44866594672203064,
+      "learning_rate": 0.0001,
+      "loss": 1.8056,
+      "step": 5487
+    },
+    {
+      "epoch": 0.6304060651312389,
+      "grad_norm": 0.4310179352760315,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 5488
+    },
+    {
+      "epoch": 0.630520935041066,
+      "grad_norm": 0.46676987409591675,
+      "learning_rate": 0.0001,
+      "loss": 1.7363,
+      "step": 5489
+    },
+    {
+      "epoch": 0.6306358049508931,
+      "grad_norm": 0.45726069808006287,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 5490
+    },
+    {
+      "epoch": 0.6307506748607202,
+      "grad_norm": 0.46568796038627625,
+      "learning_rate": 0.0001,
+      "loss": 1.772,
+      "step": 5491
+    },
+    {
+      "epoch": 0.6308655447705473,
+      "grad_norm": 0.43564847111701965,
+      "learning_rate": 0.0001,
+      "loss": 1.7335,
+      "step": 5492
+    },
+    {
+      "epoch": 0.6309804146803745,
+      "grad_norm": 0.43100833892822266,
+      "learning_rate": 0.0001,
+      "loss": 1.5743,
+      "step": 5493
+    },
+    {
+      "epoch": 0.6310952845902016,
+      "grad_norm": 0.48715853691101074,
+      "learning_rate": 0.0001,
+      "loss": 1.6999,
+      "step": 5494
+    },
+    {
+      "epoch": 0.6312101545000287,
+      "grad_norm": 0.4202541708946228,
+      "learning_rate": 0.0001,
+      "loss": 1.4744,
+      "step": 5495
+    },
+    {
+      "epoch": 0.6313250244098558,
+      "grad_norm": 0.49194201827049255,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 5496
+    },
+    {
+      "epoch": 0.631439894319683,
+      "grad_norm": 0.4225814938545227,
+      "learning_rate": 0.0001,
+      "loss": 1.6509,
+      "step": 5497
+    },
+    {
+      "epoch": 0.6315547642295101,
+      "grad_norm": 0.41638097167015076,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 5498
+    },
+    {
+      "epoch": 0.6316696341393372,
+      "grad_norm": 0.44772642850875854,
+      "learning_rate": 0.0001,
+      "loss": 1.8187,
+      "step": 5499
+    },
+    {
+      "epoch": 0.6317845040491643,
+      "grad_norm": 0.4238649904727936,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 5500
+    },
+    {
+      "epoch": 0.6318993739589914,
+      "grad_norm": 0.4199509918689728,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 5501
+    },
+    {
+      "epoch": 0.6320142438688185,
+      "grad_norm": 0.4321916699409485,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 5502
+    },
+    {
+      "epoch": 0.6321291137786457,
+      "grad_norm": 0.4737303853034973,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 5503
+    },
+    {
+      "epoch": 0.6322439836884728,
+      "grad_norm": 0.450839102268219,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 5504
+    },
+    {
+      "epoch": 0.6323588535982999,
+      "grad_norm": 0.4439679682254791,
+      "learning_rate": 0.0001,
+      "loss": 1.6076,
+      "step": 5505
+    },
+    {
+      "epoch": 0.632473723508127,
+      "grad_norm": 0.438607782125473,
+      "learning_rate": 0.0001,
+      "loss": 1.6273,
+      "step": 5506
+    },
+    {
+      "epoch": 0.6325885934179541,
+      "grad_norm": 0.44356659054756165,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 5507
+    },
+    {
+      "epoch": 0.6327034633277813,
+      "grad_norm": 0.47536197304725647,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 5508
+    },
+    {
+      "epoch": 0.6328183332376084,
+      "grad_norm": 0.4841341972351074,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 5509
+    },
+    {
+      "epoch": 0.6329332031474355,
+      "grad_norm": 0.4180241823196411,
+      "learning_rate": 0.0001,
+      "loss": 1.4014,
+      "step": 5510
+    },
+    {
+      "epoch": 0.6330480730572626,
+      "grad_norm": 0.46159014105796814,
+      "learning_rate": 0.0001,
+      "loss": 1.7947,
+      "step": 5511
+    },
+    {
+      "epoch": 0.6331629429670897,
+      "grad_norm": 0.43560919165611267,
+      "learning_rate": 0.0001,
+      "loss": 1.6293,
+      "step": 5512
+    },
+    {
+      "epoch": 0.6332778128769169,
+      "grad_norm": 0.41563135385513306,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 5513
+    },
+    {
+      "epoch": 0.633392682786744,
+      "grad_norm": 0.4384070634841919,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 5514
+    },
+    {
+      "epoch": 0.6335075526965711,
+      "grad_norm": 0.45419734716415405,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 5515
+    },
+    {
+      "epoch": 0.6336224226063982,
+      "grad_norm": 0.438769668340683,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 5516
+    },
+    {
+      "epoch": 0.6337372925162253,
+      "grad_norm": 0.4500299394130707,
+      "learning_rate": 0.0001,
+      "loss": 1.7686,
+      "step": 5517
+    },
+    {
+      "epoch": 0.6338521624260525,
+      "grad_norm": 0.543213963508606,
+      "learning_rate": 0.0001,
+      "loss": 1.7704,
+      "step": 5518
+    },
+    {
+      "epoch": 0.6339670323358796,
+      "grad_norm": 0.4491806924343109,
+      "learning_rate": 0.0001,
+      "loss": 1.7672,
+      "step": 5519
+    },
+    {
+      "epoch": 0.6340819022457067,
+      "grad_norm": 0.4237130284309387,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 5520
+    },
+    {
+      "epoch": 0.6341967721555338,
+      "grad_norm": 0.4394038915634155,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 5521
+    },
+    {
+      "epoch": 0.634311642065361,
+      "grad_norm": 0.4348239302635193,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 5522
+    },
+    {
+      "epoch": 0.6344265119751881,
+      "grad_norm": 0.45309606194496155,
+      "learning_rate": 0.0001,
+      "loss": 1.6889,
+      "step": 5523
+    },
+    {
+      "epoch": 0.6345413818850152,
+      "grad_norm": 0.40822720527648926,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 5524
+    },
+    {
+      "epoch": 0.6346562517948423,
+      "grad_norm": 0.48253944516181946,
+      "learning_rate": 0.0001,
+      "loss": 1.727,
+      "step": 5525
+    },
+    {
+      "epoch": 0.6347711217046694,
+      "grad_norm": 0.4541143476963043,
+      "learning_rate": 0.0001,
+      "loss": 1.8615,
+      "step": 5526
+    },
+    {
+      "epoch": 0.6348859916144965,
+      "grad_norm": 0.4158259630203247,
+      "learning_rate": 0.0001,
+      "loss": 1.4537,
+      "step": 5527
+    },
+    {
+      "epoch": 0.6350008615243237,
+      "grad_norm": 0.41893231868743896,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 5528
+    },
+    {
+      "epoch": 0.6351157314341508,
+      "grad_norm": 0.4827512502670288,
+      "learning_rate": 0.0001,
+      "loss": 1.7608,
+      "step": 5529
+    },
+    {
+      "epoch": 0.6352306013439779,
+      "grad_norm": 0.44435709714889526,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 5530
+    },
+    {
+      "epoch": 0.635345471253805,
+      "grad_norm": 0.4805011451244354,
+      "learning_rate": 0.0001,
+      "loss": 1.7446,
+      "step": 5531
+    },
+    {
+      "epoch": 0.6354603411636321,
+      "grad_norm": 0.43575429916381836,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 5532
+    },
+    {
+      "epoch": 0.6355752110734593,
+      "grad_norm": 0.4281097650527954,
+      "learning_rate": 0.0001,
+      "loss": 1.7776,
+      "step": 5533
+    },
+    {
+      "epoch": 0.6356900809832864,
+      "grad_norm": 0.41298815608024597,
+      "learning_rate": 0.0001,
+      "loss": 1.5776,
+      "step": 5534
+    },
+    {
+      "epoch": 0.6358049508931135,
+      "grad_norm": 0.43981999158859253,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 5535
+    },
+    {
+      "epoch": 0.6359198208029406,
+      "grad_norm": 0.46365123987197876,
+      "learning_rate": 0.0001,
+      "loss": 1.7387,
+      "step": 5536
+    },
+    {
+      "epoch": 0.6360346907127677,
+      "grad_norm": 0.40434473752975464,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 5537
+    },
+    {
+      "epoch": 0.6361495606225949,
+      "grad_norm": 0.42286887764930725,
+      "learning_rate": 0.0001,
+      "loss": 1.4093,
+      "step": 5538
+    },
+    {
+      "epoch": 0.636264430532422,
+      "grad_norm": 0.4497700035572052,
+      "learning_rate": 0.0001,
+      "loss": 1.8652,
+      "step": 5539
+    },
+    {
+      "epoch": 0.6363793004422491,
+      "grad_norm": 0.4267425537109375,
+      "learning_rate": 0.0001,
+      "loss": 1.704,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6364941703520762,
+      "grad_norm": 0.448066383600235,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 5541
+    },
+    {
+      "epoch": 0.6366090402619033,
+      "grad_norm": 0.46805888414382935,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 5542
+    },
+    {
+      "epoch": 0.6367239101717305,
+      "grad_norm": 0.42206695675849915,
+      "learning_rate": 0.0001,
+      "loss": 1.3684,
+      "step": 5543
+    },
+    {
+      "epoch": 0.6368387800815576,
+      "grad_norm": 0.39510831236839294,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 5544
+    },
+    {
+      "epoch": 0.6369536499913847,
+      "grad_norm": 0.4754282236099243,
+      "learning_rate": 0.0001,
+      "loss": 1.8412,
+      "step": 5545
+    },
+    {
+      "epoch": 0.6370685199012118,
+      "grad_norm": 0.42278730869293213,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 5546
+    },
+    {
+      "epoch": 0.637183389811039,
+      "grad_norm": 0.4810383915901184,
+      "learning_rate": 0.0001,
+      "loss": 1.6435,
+      "step": 5547
+    },
+    {
+      "epoch": 0.6372982597208661,
+      "grad_norm": 0.4392484128475189,
+      "learning_rate": 0.0001,
+      "loss": 1.4485,
+      "step": 5548
+    },
+    {
+      "epoch": 0.6374131296306932,
+      "grad_norm": 0.46849682927131653,
+      "learning_rate": 0.0001,
+      "loss": 1.7611,
+      "step": 5549
+    },
+    {
+      "epoch": 0.6375279995405203,
+      "grad_norm": 0.428786039352417,
+      "learning_rate": 0.0001,
+      "loss": 1.7185,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6376428694503474,
+      "grad_norm": 0.41656240820884705,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 5551
+    },
+    {
+      "epoch": 0.6377577393601745,
+      "grad_norm": 0.447986364364624,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 5552
+    },
+    {
+      "epoch": 0.6378726092700018,
+      "grad_norm": 0.4375683665275574,
+      "learning_rate": 0.0001,
+      "loss": 1.7039,
+      "step": 5553
+    },
+    {
+      "epoch": 0.6379874791798289,
+      "grad_norm": 0.4190763831138611,
+      "learning_rate": 0.0001,
+      "loss": 1.5491,
+      "step": 5554
+    },
+    {
+      "epoch": 0.638102349089656,
+      "grad_norm": 0.42789924144744873,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 5555
+    },
+    {
+      "epoch": 0.6382172189994831,
+      "grad_norm": 0.42377930879592896,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 5556
+    },
+    {
+      "epoch": 0.6383320889093103,
+      "grad_norm": 0.43022698163986206,
+      "learning_rate": 0.0001,
+      "loss": 1.5902,
+      "step": 5557
+    },
+    {
+      "epoch": 0.6384469588191374,
+      "grad_norm": 0.40309348702430725,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 5558
+    },
+    {
+      "epoch": 0.6385618287289645,
+      "grad_norm": 0.4346220791339874,
+      "learning_rate": 0.0001,
+      "loss": 1.598,
+      "step": 5559
+    },
+    {
+      "epoch": 0.6386766986387916,
+      "grad_norm": 0.45233309268951416,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 5560
+    },
+    {
+      "epoch": 0.6387915685486187,
+      "grad_norm": 0.44141674041748047,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 5561
+    },
+    {
+      "epoch": 0.6389064384584459,
+      "grad_norm": 0.45340585708618164,
+      "learning_rate": 0.0001,
+      "loss": 1.7477,
+      "step": 5562
+    },
+    {
+      "epoch": 0.639021308368273,
+      "grad_norm": 0.46206724643707275,
+      "learning_rate": 0.0001,
+      "loss": 1.8006,
+      "step": 5563
+    },
+    {
+      "epoch": 0.6391361782781001,
+      "grad_norm": 0.42310798168182373,
+      "learning_rate": 0.0001,
+      "loss": 1.7606,
+      "step": 5564
+    },
+    {
+      "epoch": 0.6392510481879272,
+      "grad_norm": 0.40606430172920227,
+      "learning_rate": 0.0001,
+      "loss": 1.4893,
+      "step": 5565
+    },
+    {
+      "epoch": 0.6393659180977543,
+      "grad_norm": 0.4227276146411896,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 5566
+    },
+    {
+      "epoch": 0.6394807880075815,
+      "grad_norm": 0.42999976873397827,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 5567
+    },
+    {
+      "epoch": 0.6395956579174086,
+      "grad_norm": 0.4548937678337097,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 5568
+    },
+    {
+      "epoch": 0.6397105278272357,
+      "grad_norm": 0.4163641333580017,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 5569
+    },
+    {
+      "epoch": 0.6398253977370628,
+      "grad_norm": 0.41462451219558716,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6399402676468899,
+      "grad_norm": 0.4123389422893524,
+      "learning_rate": 0.0001,
+      "loss": 1.5769,
+      "step": 5571
+    },
+    {
+      "epoch": 0.6400551375567171,
+      "grad_norm": 0.447255939245224,
+      "learning_rate": 0.0001,
+      "loss": 1.6402,
+      "step": 5572
+    },
+    {
+      "epoch": 0.6401700074665442,
+      "grad_norm": 0.46185576915740967,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 5573
+    },
+    {
+      "epoch": 0.6402848773763713,
+      "grad_norm": 0.46707770228385925,
+      "learning_rate": 0.0001,
+      "loss": 1.7298,
+      "step": 5574
+    },
+    {
+      "epoch": 0.6403997472861984,
+      "grad_norm": 0.4091823399066925,
+      "learning_rate": 0.0001,
+      "loss": 1.4606,
+      "step": 5575
+    },
+    {
+      "epoch": 0.6405146171960255,
+      "grad_norm": 0.45736071467399597,
+      "learning_rate": 0.0001,
+      "loss": 1.764,
+      "step": 5576
+    },
+    {
+      "epoch": 0.6406294871058527,
+      "grad_norm": 0.4426022469997406,
+      "learning_rate": 0.0001,
+      "loss": 1.6942,
+      "step": 5577
+    },
+    {
+      "epoch": 0.6407443570156798,
+      "grad_norm": 0.42908066511154175,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 5578
+    },
+    {
+      "epoch": 0.6408592269255069,
+      "grad_norm": 0.4217868149280548,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 5579
+    },
+    {
+      "epoch": 0.640974096835334,
+      "grad_norm": 0.4545440673828125,
+      "learning_rate": 0.0001,
+      "loss": 1.7952,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6410889667451611,
+      "grad_norm": 0.42396777868270874,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 5581
+    },
+    {
+      "epoch": 0.6412038366549883,
+      "grad_norm": 0.43753355741500854,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 5582
+    },
+    {
+      "epoch": 0.6413187065648154,
+      "grad_norm": 0.46905073523521423,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 5583
+    },
+    {
+      "epoch": 0.6414335764746425,
+      "grad_norm": 0.4515920877456665,
+      "learning_rate": 0.0001,
+      "loss": 1.4264,
+      "step": 5584
+    },
+    {
+      "epoch": 0.6415484463844696,
+      "grad_norm": 0.434110552072525,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 5585
+    },
+    {
+      "epoch": 0.6416633162942967,
+      "grad_norm": 0.4196917414665222,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 5586
+    },
+    {
+      "epoch": 0.6417781862041239,
+      "grad_norm": 0.4676187336444855,
+      "learning_rate": 0.0001,
+      "loss": 1.7113,
+      "step": 5587
+    },
+    {
+      "epoch": 0.641893056113951,
+      "grad_norm": 0.4227234423160553,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 5588
+    },
+    {
+      "epoch": 0.6420079260237781,
+      "grad_norm": 0.45535120368003845,
+      "learning_rate": 0.0001,
+      "loss": 1.6142,
+      "step": 5589
+    },
+    {
+      "epoch": 0.6421227959336052,
+      "grad_norm": 0.4416712820529938,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6422376658434323,
+      "grad_norm": 0.454698383808136,
+      "learning_rate": 0.0001,
+      "loss": 1.835,
+      "step": 5591
+    },
+    {
+      "epoch": 0.6423525357532595,
+      "grad_norm": 0.429575115442276,
+      "learning_rate": 0.0001,
+      "loss": 1.6046,
+      "step": 5592
+    },
+    {
+      "epoch": 0.6424674056630866,
+      "grad_norm": 0.47407978773117065,
+      "learning_rate": 0.0001,
+      "loss": 1.7451,
+      "step": 5593
+    },
+    {
+      "epoch": 0.6425822755729137,
+      "grad_norm": 0.4476969838142395,
+      "learning_rate": 0.0001,
+      "loss": 1.6611,
+      "step": 5594
+    },
+    {
+      "epoch": 0.6426971454827408,
+      "grad_norm": 0.48670604825019836,
+      "learning_rate": 0.0001,
+      "loss": 1.8238,
+      "step": 5595
+    },
+    {
+      "epoch": 0.6428120153925679,
+      "grad_norm": 0.4715462923049927,
+      "learning_rate": 0.0001,
+      "loss": 1.5888,
+      "step": 5596
+    },
+    {
+      "epoch": 0.6429268853023951,
+      "grad_norm": 0.4310706555843353,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 5597
+    },
+    {
+      "epoch": 0.6430417552122222,
+      "grad_norm": 0.4375722408294678,
+      "learning_rate": 0.0001,
+      "loss": 1.7016,
+      "step": 5598
+    },
+    {
+      "epoch": 0.6431566251220493,
+      "grad_norm": 0.44785434007644653,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 5599
+    },
+    {
+      "epoch": 0.6432714950318764,
+      "grad_norm": 0.43473127484321594,
+      "learning_rate": 0.0001,
+      "loss": 1.5382,
+      "step": 5600
+    },
+    {
+      "epoch": 0.6433863649417035,
+      "grad_norm": 0.5235137939453125,
+      "learning_rate": 0.0001,
+      "loss": 1.7376,
+      "step": 5601
+    },
+    {
+      "epoch": 0.6435012348515307,
+      "grad_norm": 0.47962141036987305,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 5602
+    },
+    {
+      "epoch": 0.6436161047613578,
+      "grad_norm": 0.42864152789115906,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 5603
+    },
+    {
+      "epoch": 0.6437309746711849,
+      "grad_norm": 0.4481607675552368,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 5604
+    },
+    {
+      "epoch": 0.643845844581012,
+      "grad_norm": 0.503777027130127,
+      "learning_rate": 0.0001,
+      "loss": 1.9532,
+      "step": 5605
+    },
+    {
+      "epoch": 0.6439607144908391,
+      "grad_norm": 0.44352078437805176,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 5606
+    },
+    {
+      "epoch": 0.6440755844006663,
+      "grad_norm": 0.43461933732032776,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 5607
+    },
+    {
+      "epoch": 0.6441904543104934,
+      "grad_norm": 0.41317033767700195,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 5608
+    },
+    {
+      "epoch": 0.6443053242203205,
+      "grad_norm": 0.47428956627845764,
+      "learning_rate": 0.0001,
+      "loss": 1.768,
+      "step": 5609
+    },
+    {
+      "epoch": 0.6444201941301476,
+      "grad_norm": 0.4190300703048706,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 5610
+    },
+    {
+      "epoch": 0.6445350640399747,
+      "grad_norm": 0.4992254376411438,
+      "learning_rate": 0.0001,
+      "loss": 1.9133,
+      "step": 5611
+    },
+    {
+      "epoch": 0.6446499339498019,
+      "grad_norm": 0.4316772222518921,
+      "learning_rate": 0.0001,
+      "loss": 1.3204,
+      "step": 5612
+    },
+    {
+      "epoch": 0.644764803859629,
+      "grad_norm": 0.4693394899368286,
+      "learning_rate": 0.0001,
+      "loss": 1.8395,
+      "step": 5613
+    },
+    {
+      "epoch": 0.6448796737694561,
+      "grad_norm": 0.4208371043205261,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 5614
+    },
+    {
+      "epoch": 0.6449945436792832,
+      "grad_norm": 0.44202813506126404,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 5615
+    },
+    {
+      "epoch": 0.6451094135891103,
+      "grad_norm": 0.4268895387649536,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 5616
+    },
+    {
+      "epoch": 0.6452242834989375,
+      "grad_norm": 0.4346774220466614,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 5617
+    },
+    {
+      "epoch": 0.6453391534087646,
+      "grad_norm": 0.46032214164733887,
+      "learning_rate": 0.0001,
+      "loss": 1.6249,
+      "step": 5618
+    },
+    {
+      "epoch": 0.6454540233185917,
+      "grad_norm": 0.4409688413143158,
+      "learning_rate": 0.0001,
+      "loss": 1.5832,
+      "step": 5619
+    },
+    {
+      "epoch": 0.6455688932284188,
+      "grad_norm": 0.4379858374595642,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6456837631382459,
+      "grad_norm": 0.43202537298202515,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 5621
+    },
+    {
+      "epoch": 0.6457986330480731,
+      "grad_norm": 0.44553086161613464,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 5622
+    },
+    {
+      "epoch": 0.6459135029579002,
+      "grad_norm": 0.42607608437538147,
+      "learning_rate": 0.0001,
+      "loss": 1.6773,
+      "step": 5623
+    },
+    {
+      "epoch": 0.6460283728677273,
+      "grad_norm": 0.45914703607559204,
+      "learning_rate": 0.0001,
+      "loss": 1.8061,
+      "step": 5624
+    },
+    {
+      "epoch": 0.6461432427775544,
+      "grad_norm": 0.46458396315574646,
+      "learning_rate": 0.0001,
+      "loss": 1.701,
+      "step": 5625
+    },
+    {
+      "epoch": 0.6462581126873815,
+      "grad_norm": 0.4472190737724304,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 5626
+    },
+    {
+      "epoch": 0.6463729825972087,
+      "grad_norm": 0.4329933524131775,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 5627
+    },
+    {
+      "epoch": 0.6464878525070358,
+      "grad_norm": 0.4526257812976837,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 5628
+    },
+    {
+      "epoch": 0.6466027224168629,
+      "grad_norm": 0.4655599892139435,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 5629
+    },
+    {
+      "epoch": 0.64671759232669,
+      "grad_norm": 0.40564316511154175,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 5630
+    },
+    {
+      "epoch": 0.6468324622365171,
+      "grad_norm": 0.41868773102760315,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 5631
+    },
+    {
+      "epoch": 0.6469473321463443,
+      "grad_norm": 0.4080910384654999,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 5632
+    },
+    {
+      "epoch": 0.6470622020561714,
+      "grad_norm": 0.45903560519218445,
+      "learning_rate": 0.0001,
+      "loss": 1.8245,
+      "step": 5633
+    },
+    {
+      "epoch": 0.6471770719659985,
+      "grad_norm": 0.4414862394332886,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 5634
+    },
+    {
+      "epoch": 0.6472919418758256,
+      "grad_norm": 0.43176713585853577,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 5635
+    },
+    {
+      "epoch": 0.6474068117856527,
+      "grad_norm": 0.41198432445526123,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 5636
+    },
+    {
+      "epoch": 0.6475216816954799,
+      "grad_norm": 0.47420576214790344,
+      "learning_rate": 0.0001,
+      "loss": 1.7435,
+      "step": 5637
+    },
+    {
+      "epoch": 0.647636551605307,
+      "grad_norm": 0.44818076491355896,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 5638
+    },
+    {
+      "epoch": 0.6477514215151341,
+      "grad_norm": 0.43701666593551636,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 5639
+    },
+    {
+      "epoch": 0.6478662914249612,
+      "grad_norm": 0.4181070923805237,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 5640
+    },
+    {
+      "epoch": 0.6479811613347883,
+      "grad_norm": 0.4229847192764282,
+      "learning_rate": 0.0001,
+      "loss": 1.594,
+      "step": 5641
+    },
+    {
+      "epoch": 0.6480960312446155,
+      "grad_norm": 0.4145675003528595,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 5642
+    },
+    {
+      "epoch": 0.6482109011544426,
+      "grad_norm": 0.43566563725471497,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 5643
+    },
+    {
+      "epoch": 0.6483257710642697,
+      "grad_norm": 0.4276692867279053,
+      "learning_rate": 0.0001,
+      "loss": 1.4687,
+      "step": 5644
+    },
+    {
+      "epoch": 0.6484406409740968,
+      "grad_norm": 0.3920123875141144,
+      "learning_rate": 0.0001,
+      "loss": 1.4192,
+      "step": 5645
+    },
+    {
+      "epoch": 0.6485555108839239,
+      "grad_norm": 0.4451933801174164,
+      "learning_rate": 0.0001,
+      "loss": 1.6342,
+      "step": 5646
+    },
+    {
+      "epoch": 0.6486703807937511,
+      "grad_norm": 0.404407799243927,
+      "learning_rate": 0.0001,
+      "loss": 1.5801,
+      "step": 5647
+    },
+    {
+      "epoch": 0.6487852507035782,
+      "grad_norm": 0.4093591570854187,
+      "learning_rate": 0.0001,
+      "loss": 1.4817,
+      "step": 5648
+    },
+    {
+      "epoch": 0.6489001206134053,
+      "grad_norm": 0.4147598147392273,
+      "learning_rate": 0.0001,
+      "loss": 1.7049,
+      "step": 5649
+    },
+    {
+      "epoch": 0.6490149905232324,
+      "grad_norm": 0.45576632022857666,
+      "learning_rate": 0.0001,
+      "loss": 1.6524,
+      "step": 5650
+    },
+    {
+      "epoch": 0.6491298604330595,
+      "grad_norm": 0.4459483027458191,
+      "learning_rate": 0.0001,
+      "loss": 1.7097,
+      "step": 5651
+    },
+    {
+      "epoch": 0.6492447303428867,
+      "grad_norm": 0.47733941674232483,
+      "learning_rate": 0.0001,
+      "loss": 1.8741,
+      "step": 5652
+    },
+    {
+      "epoch": 0.6493596002527138,
+      "grad_norm": 0.45850062370300293,
+      "learning_rate": 0.0001,
+      "loss": 1.7533,
+      "step": 5653
+    },
+    {
+      "epoch": 0.6494744701625409,
+      "grad_norm": 0.42292144894599915,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 5654
+    },
+    {
+      "epoch": 0.649589340072368,
+      "grad_norm": 0.4376376271247864,
+      "learning_rate": 0.0001,
+      "loss": 1.7886,
+      "step": 5655
+    },
+    {
+      "epoch": 0.6497042099821951,
+      "grad_norm": 0.4351494610309601,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 5656
+    },
+    {
+      "epoch": 0.6498190798920223,
+      "grad_norm": 0.4215450882911682,
+      "learning_rate": 0.0001,
+      "loss": 1.6642,
+      "step": 5657
+    },
+    {
+      "epoch": 0.6499339498018494,
+      "grad_norm": 0.4552718997001648,
+      "learning_rate": 0.0001,
+      "loss": 1.7277,
+      "step": 5658
+    },
+    {
+      "epoch": 0.6500488197116765,
+      "grad_norm": 0.4284704625606537,
+      "learning_rate": 0.0001,
+      "loss": 1.6748,
+      "step": 5659
+    },
+    {
+      "epoch": 0.6501636896215036,
+      "grad_norm": 0.42294004559516907,
+      "learning_rate": 0.0001,
+      "loss": 1.575,
+      "step": 5660
+    },
+    {
+      "epoch": 0.6502785595313307,
+      "grad_norm": 0.43804609775543213,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 5661
+    },
+    {
+      "epoch": 0.6503934294411579,
+      "grad_norm": 0.4440459907054901,
+      "learning_rate": 0.0001,
+      "loss": 1.7683,
+      "step": 5662
+    },
+    {
+      "epoch": 0.650508299350985,
+      "grad_norm": 0.4592325687408447,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 5663
+    },
+    {
+      "epoch": 0.6506231692608121,
+      "grad_norm": 0.44203251600265503,
+      "learning_rate": 0.0001,
+      "loss": 1.6765,
+      "step": 5664
+    },
+    {
+      "epoch": 0.6507380391706392,
+      "grad_norm": 0.4115028977394104,
+      "learning_rate": 0.0001,
+      "loss": 1.3906,
+      "step": 5665
+    },
+    {
+      "epoch": 0.6508529090804663,
+      "grad_norm": 0.456676721572876,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 5666
+    },
+    {
+      "epoch": 0.6509677789902935,
+      "grad_norm": 0.4849773645401001,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 5667
+    },
+    {
+      "epoch": 0.6510826489001206,
+      "grad_norm": 0.4652778208255768,
+      "learning_rate": 0.0001,
+      "loss": 1.8482,
+      "step": 5668
+    },
+    {
+      "epoch": 0.6511975188099477,
+      "grad_norm": 0.42466142773628235,
+      "learning_rate": 0.0001,
+      "loss": 1.4337,
+      "step": 5669
+    },
+    {
+      "epoch": 0.6513123887197748,
+      "grad_norm": 0.4781624674797058,
+      "learning_rate": 0.0001,
+      "loss": 1.884,
+      "step": 5670
+    },
+    {
+      "epoch": 0.6514272586296019,
+      "grad_norm": 0.4678051769733429,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 5671
+    },
+    {
+      "epoch": 0.651542128539429,
+      "grad_norm": 0.4526500701904297,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 5672
+    },
+    {
+      "epoch": 0.6516569984492562,
+      "grad_norm": 0.44952574372291565,
+      "learning_rate": 0.0001,
+      "loss": 1.5678,
+      "step": 5673
+    },
+    {
+      "epoch": 0.6517718683590833,
+      "grad_norm": 0.4886074960231781,
+      "learning_rate": 0.0001,
+      "loss": 1.8023,
+      "step": 5674
+    },
+    {
+      "epoch": 0.6518867382689104,
+      "grad_norm": 0.43213948607444763,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 5675
+    },
+    {
+      "epoch": 0.6520016081787375,
+      "grad_norm": 0.41444772481918335,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 5676
+    },
+    {
+      "epoch": 0.6521164780885647,
+      "grad_norm": 0.46063005924224854,
+      "learning_rate": 0.0001,
+      "loss": 1.686,
+      "step": 5677
+    },
+    {
+      "epoch": 0.6522313479983918,
+      "grad_norm": 0.4700499176979065,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 5678
+    },
+    {
+      "epoch": 0.6523462179082189,
+      "grad_norm": 0.48296359181404114,
+      "learning_rate": 0.0001,
+      "loss": 1.5319,
+      "step": 5679
+    },
+    {
+      "epoch": 0.652461087818046,
+      "grad_norm": 0.4892067015171051,
+      "learning_rate": 0.0001,
+      "loss": 1.8181,
+      "step": 5680
+    },
+    {
+      "epoch": 0.6525759577278731,
+      "grad_norm": 0.426547646522522,
+      "learning_rate": 0.0001,
+      "loss": 1.7042,
+      "step": 5681
+    },
+    {
+      "epoch": 0.6526908276377003,
+      "grad_norm": 0.48687633872032166,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 5682
+    },
+    {
+      "epoch": 0.6528056975475274,
+      "grad_norm": 0.449789434671402,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 5683
+    },
+    {
+      "epoch": 0.6529205674573545,
+      "grad_norm": 0.43669000267982483,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 5684
+    },
+    {
+      "epoch": 0.6530354373671816,
+      "grad_norm": 0.4181348383426666,
+      "learning_rate": 0.0001,
+      "loss": 1.6324,
+      "step": 5685
+    },
+    {
+      "epoch": 0.6531503072770087,
+      "grad_norm": 0.4275439381599426,
+      "learning_rate": 0.0001,
+      "loss": 1.7064,
+      "step": 5686
+    },
+    {
+      "epoch": 0.6532651771868359,
+      "grad_norm": 0.44249075651168823,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 5687
+    },
+    {
+      "epoch": 0.653380047096663,
+      "grad_norm": 0.4645352363586426,
+      "learning_rate": 0.0001,
+      "loss": 1.733,
+      "step": 5688
+    },
+    {
+      "epoch": 0.6534949170064901,
+      "grad_norm": 0.47578033804893494,
+      "learning_rate": 0.0001,
+      "loss": 1.8043,
+      "step": 5689
+    },
+    {
+      "epoch": 0.6536097869163173,
+      "grad_norm": 0.4634343981742859,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 5690
+    },
+    {
+      "epoch": 0.6537246568261444,
+      "grad_norm": 0.4607868194580078,
+      "learning_rate": 0.0001,
+      "loss": 1.6108,
+      "step": 5691
+    },
+    {
+      "epoch": 0.6538395267359716,
+      "grad_norm": 0.43927955627441406,
+      "learning_rate": 0.0001,
+      "loss": 1.7641,
+      "step": 5692
+    },
+    {
+      "epoch": 0.6539543966457987,
+      "grad_norm": 0.42264971137046814,
+      "learning_rate": 0.0001,
+      "loss": 1.6051,
+      "step": 5693
+    },
+    {
+      "epoch": 0.6540692665556258,
+      "grad_norm": 0.4350726902484894,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 5694
+    },
+    {
+      "epoch": 0.6541841364654529,
+      "grad_norm": 0.45460623502731323,
+      "learning_rate": 0.0001,
+      "loss": 1.5079,
+      "step": 5695
+    },
+    {
+      "epoch": 0.65429900637528,
+      "grad_norm": 0.45324012637138367,
+      "learning_rate": 0.0001,
+      "loss": 1.458,
+      "step": 5696
+    },
+    {
+      "epoch": 0.6544138762851072,
+      "grad_norm": 0.4302757680416107,
+      "learning_rate": 0.0001,
+      "loss": 1.4266,
+      "step": 5697
+    },
+    {
+      "epoch": 0.6545287461949343,
+      "grad_norm": 0.4926580488681793,
+      "learning_rate": 0.0001,
+      "loss": 1.9029,
+      "step": 5698
+    },
+    {
+      "epoch": 0.6546436161047614,
+      "grad_norm": 0.4108467698097229,
+      "learning_rate": 0.0001,
+      "loss": 1.4296,
+      "step": 5699
+    },
+    {
+      "epoch": 0.6547584860145885,
+      "grad_norm": 0.4439946413040161,
+      "learning_rate": 0.0001,
+      "loss": 1.4945,
+      "step": 5700
+    },
+    {
+      "epoch": 0.6548733559244156,
+      "grad_norm": 0.42766717076301575,
+      "learning_rate": 0.0001,
+      "loss": 1.6439,
+      "step": 5701
+    },
+    {
+      "epoch": 0.6549882258342428,
+      "grad_norm": 0.4058894217014313,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 5702
+    },
+    {
+      "epoch": 0.6551030957440699,
+      "grad_norm": 0.42546913027763367,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 5703
+    },
+    {
+      "epoch": 0.655217965653897,
+      "grad_norm": 0.452540785074234,
+      "learning_rate": 0.0001,
+      "loss": 1.5635,
+      "step": 5704
+    },
+    {
+      "epoch": 0.6553328355637241,
+      "grad_norm": 0.4617590308189392,
+      "learning_rate": 0.0001,
+      "loss": 1.6154,
+      "step": 5705
+    },
+    {
+      "epoch": 0.6554477054735512,
+      "grad_norm": 0.41302189230918884,
+      "learning_rate": 0.0001,
+      "loss": 1.5644,
+      "step": 5706
+    },
+    {
+      "epoch": 0.6555625753833784,
+      "grad_norm": 0.43837377429008484,
+      "learning_rate": 0.0001,
+      "loss": 1.6912,
+      "step": 5707
+    },
+    {
+      "epoch": 0.6556774452932055,
+      "grad_norm": 0.4599246382713318,
+      "learning_rate": 0.0001,
+      "loss": 1.7636,
+      "step": 5708
+    },
+    {
+      "epoch": 0.6557923152030326,
+      "grad_norm": 0.44273531436920166,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 5709
+    },
+    {
+      "epoch": 0.6559071851128597,
+      "grad_norm": 0.40543317794799805,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 5710
+    },
+    {
+      "epoch": 0.6560220550226868,
+      "grad_norm": 0.44224417209625244,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 5711
+    },
+    {
+      "epoch": 0.656136924932514,
+      "grad_norm": 0.4091229736804962,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 5712
+    },
+    {
+      "epoch": 0.6562517948423411,
+      "grad_norm": 0.44809257984161377,
+      "learning_rate": 0.0001,
+      "loss": 1.7865,
+      "step": 5713
+    },
+    {
+      "epoch": 0.6563666647521682,
+      "grad_norm": 0.4188038110733032,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 5714
+    },
+    {
+      "epoch": 0.6564815346619953,
+      "grad_norm": 0.4264969229698181,
+      "learning_rate": 0.0001,
+      "loss": 1.3747,
+      "step": 5715
+    },
+    {
+      "epoch": 0.6565964045718224,
+      "grad_norm": 0.46266868710517883,
+      "learning_rate": 0.0001,
+      "loss": 1.8456,
+      "step": 5716
+    },
+    {
+      "epoch": 0.6567112744816496,
+      "grad_norm": 0.45116865634918213,
+      "learning_rate": 0.0001,
+      "loss": 1.4373,
+      "step": 5717
+    },
+    {
+      "epoch": 0.6568261443914767,
+      "grad_norm": 0.4384673833847046,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 5718
+    },
+    {
+      "epoch": 0.6569410143013038,
+      "grad_norm": 0.4380660355091095,
+      "learning_rate": 0.0001,
+      "loss": 1.5475,
+      "step": 5719
+    },
+    {
+      "epoch": 0.6570558842111309,
+      "grad_norm": 0.4457499086856842,
+      "learning_rate": 0.0001,
+      "loss": 1.5831,
+      "step": 5720
+    },
+    {
+      "epoch": 0.657170754120958,
+      "grad_norm": 0.45870745182037354,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 5721
+    },
+    {
+      "epoch": 0.6572856240307852,
+      "grad_norm": 0.44491496682167053,
+      "learning_rate": 0.0001,
+      "loss": 1.6472,
+      "step": 5722
+    },
+    {
+      "epoch": 0.6574004939406123,
+      "grad_norm": 0.4536452293395996,
+      "learning_rate": 0.0001,
+      "loss": 1.5365,
+      "step": 5723
+    },
+    {
+      "epoch": 0.6575153638504394,
+      "grad_norm": 0.4279188811779022,
+      "learning_rate": 0.0001,
+      "loss": 1.6861,
+      "step": 5724
+    },
+    {
+      "epoch": 0.6576302337602665,
+      "grad_norm": 0.49207669496536255,
+      "learning_rate": 0.0001,
+      "loss": 1.6214,
+      "step": 5725
+    },
+    {
+      "epoch": 0.6577451036700936,
+      "grad_norm": 0.47983112931251526,
+      "learning_rate": 0.0001,
+      "loss": 1.77,
+      "step": 5726
+    },
+    {
+      "epoch": 0.6578599735799208,
+      "grad_norm": 0.42303088307380676,
+      "learning_rate": 0.0001,
+      "loss": 1.439,
+      "step": 5727
+    },
+    {
+      "epoch": 0.6579748434897479,
+      "grad_norm": 0.40364500880241394,
+      "learning_rate": 0.0001,
+      "loss": 1.4152,
+      "step": 5728
+    },
+    {
+      "epoch": 0.658089713399575,
+      "grad_norm": 0.44114431738853455,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 5729
+    },
+    {
+      "epoch": 0.6582045833094021,
+      "grad_norm": 0.4176272451877594,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6583194532192292,
+      "grad_norm": 0.4215009808540344,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 5731
+    },
+    {
+      "epoch": 0.6584343231290564,
+      "grad_norm": 0.4414535164833069,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 5732
+    },
+    {
+      "epoch": 0.6585491930388835,
+      "grad_norm": 0.46415939927101135,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 5733
+    },
+    {
+      "epoch": 0.6586640629487106,
+      "grad_norm": 0.42953890562057495,
+      "learning_rate": 0.0001,
+      "loss": 1.5175,
+      "step": 5734
+    },
+    {
+      "epoch": 0.6587789328585377,
+      "grad_norm": 0.4586336016654968,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 5735
+    },
+    {
+      "epoch": 0.6588938027683648,
+      "grad_norm": 0.41848480701446533,
+      "learning_rate": 0.0001,
+      "loss": 1.724,
+      "step": 5736
+    },
+    {
+      "epoch": 0.659008672678192,
+      "grad_norm": 0.43761909008026123,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 5737
+    },
+    {
+      "epoch": 0.6591235425880191,
+      "grad_norm": 0.4685233235359192,
+      "learning_rate": 0.0001,
+      "loss": 1.7588,
+      "step": 5738
+    },
+    {
+      "epoch": 0.6592384124978462,
+      "grad_norm": 0.4934740364551544,
+      "learning_rate": 0.0001,
+      "loss": 1.7505,
+      "step": 5739
+    },
+    {
+      "epoch": 0.6593532824076733,
+      "grad_norm": 0.44371286034584045,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6594681523175004,
+      "grad_norm": 0.429153710603714,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 5741
+    },
+    {
+      "epoch": 0.6595830222273276,
+      "grad_norm": 0.46635177731513977,
+      "learning_rate": 0.0001,
+      "loss": 1.6942,
+      "step": 5742
+    },
+    {
+      "epoch": 0.6596978921371547,
+      "grad_norm": 0.39945048093795776,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 5743
+    },
+    {
+      "epoch": 0.6598127620469818,
+      "grad_norm": 0.46601754426956177,
+      "learning_rate": 0.0001,
+      "loss": 1.6118,
+      "step": 5744
+    },
+    {
+      "epoch": 0.6599276319568089,
+      "grad_norm": 0.48034870624542236,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 5745
+    },
+    {
+      "epoch": 0.660042501866636,
+      "grad_norm": 0.42302387952804565,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 5746
+    },
+    {
+      "epoch": 0.6601573717764632,
+      "grad_norm": 0.4531380236148834,
+      "learning_rate": 0.0001,
+      "loss": 1.72,
+      "step": 5747
+    },
+    {
+      "epoch": 0.6602722416862903,
+      "grad_norm": 0.43431270122528076,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 5748
+    },
+    {
+      "epoch": 0.6603871115961174,
+      "grad_norm": 0.44529253244400024,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 5749
+    },
+    {
+      "epoch": 0.6605019815059445,
+      "grad_norm": 0.43237540125846863,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 5750
+    },
+    {
+      "epoch": 0.6606168514157716,
+      "grad_norm": 0.4513912796974182,
+      "learning_rate": 0.0001,
+      "loss": 1.5776,
+      "step": 5751
+    },
+    {
+      "epoch": 0.6607317213255988,
+      "grad_norm": 0.42277923226356506,
+      "learning_rate": 0.0001,
+      "loss": 1.7517,
+      "step": 5752
+    },
+    {
+      "epoch": 0.6608465912354259,
+      "grad_norm": 0.40695422887802124,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 5753
+    },
+    {
+      "epoch": 0.660961461145253,
+      "grad_norm": 0.4478527009487152,
+      "learning_rate": 0.0001,
+      "loss": 1.7765,
+      "step": 5754
+    },
+    {
+      "epoch": 0.6610763310550801,
+      "grad_norm": 0.4173600971698761,
+      "learning_rate": 0.0001,
+      "loss": 1.5796,
+      "step": 5755
+    },
+    {
+      "epoch": 0.6611912009649072,
+      "grad_norm": 0.4463992416858673,
+      "learning_rate": 0.0001,
+      "loss": 1.6569,
+      "step": 5756
+    },
+    {
+      "epoch": 0.6613060708747344,
+      "grad_norm": 0.4375052750110626,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 5757
+    },
+    {
+      "epoch": 0.6614209407845615,
+      "grad_norm": 0.4219117760658264,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 5758
+    },
+    {
+      "epoch": 0.6615358106943886,
+      "grad_norm": 0.45714181661605835,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 5759
+    },
+    {
+      "epoch": 0.6616506806042157,
+      "grad_norm": 0.4279073178768158,
+      "learning_rate": 0.0001,
+      "loss": 1.5276,
+      "step": 5760
+    },
+    {
+      "epoch": 0.6617655505140428,
+      "grad_norm": 0.4680931568145752,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 5761
+    },
+    {
+      "epoch": 0.66188042042387,
+      "grad_norm": 0.4030158817768097,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 5762
+    },
+    {
+      "epoch": 0.6619952903336971,
+      "grad_norm": 0.4316936433315277,
+      "learning_rate": 0.0001,
+      "loss": 1.6257,
+      "step": 5763
+    },
+    {
+      "epoch": 0.6621101602435242,
+      "grad_norm": 0.45963194966316223,
+      "learning_rate": 0.0001,
+      "loss": 1.7182,
+      "step": 5764
+    },
+    {
+      "epoch": 0.6622250301533513,
+      "grad_norm": 0.4106147289276123,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 5765
+    },
+    {
+      "epoch": 0.6623399000631784,
+      "grad_norm": 0.4534708261489868,
+      "learning_rate": 0.0001,
+      "loss": 1.7293,
+      "step": 5766
+    },
+    {
+      "epoch": 0.6624547699730056,
+      "grad_norm": 0.4368360936641693,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 5767
+    },
+    {
+      "epoch": 0.6625696398828327,
+      "grad_norm": 0.40505045652389526,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 5768
+    },
+    {
+      "epoch": 0.6626845097926598,
+      "grad_norm": 0.42817744612693787,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 5769
+    },
+    {
+      "epoch": 0.6627993797024869,
+      "grad_norm": 0.46093493700027466,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 5770
+    },
+    {
+      "epoch": 0.662914249612314,
+      "grad_norm": 0.44305554032325745,
+      "learning_rate": 0.0001,
+      "loss": 1.6918,
+      "step": 5771
+    },
+    {
+      "epoch": 0.6630291195221412,
+      "grad_norm": 0.39377158880233765,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 5772
+    },
+    {
+      "epoch": 0.6631439894319683,
+      "grad_norm": 0.4344120919704437,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 5773
+    },
+    {
+      "epoch": 0.6632588593417954,
+      "grad_norm": 0.44036102294921875,
+      "learning_rate": 0.0001,
+      "loss": 1.6784,
+      "step": 5774
+    },
+    {
+      "epoch": 0.6633737292516225,
+      "grad_norm": 0.4466957449913025,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 5775
+    },
+    {
+      "epoch": 0.6634885991614496,
+      "grad_norm": 0.43694761395454407,
+      "learning_rate": 0.0001,
+      "loss": 1.6732,
+      "step": 5776
+    },
+    {
+      "epoch": 0.6636034690712768,
+      "grad_norm": 0.4368375539779663,
+      "learning_rate": 0.0001,
+      "loss": 1.5941,
+      "step": 5777
+    },
+    {
+      "epoch": 0.6637183389811039,
+      "grad_norm": 0.4765366017818451,
+      "learning_rate": 0.0001,
+      "loss": 1.7931,
+      "step": 5778
+    },
+    {
+      "epoch": 0.663833208890931,
+      "grad_norm": 0.43931543827056885,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 5779
+    },
+    {
+      "epoch": 0.6639480788007581,
+      "grad_norm": 0.43625199794769287,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6640629487105852,
+      "grad_norm": 0.4167712926864624,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 5781
+    },
+    {
+      "epoch": 0.6641778186204124,
+      "grad_norm": 0.40136420726776123,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 5782
+    },
+    {
+      "epoch": 0.6642926885302395,
+      "grad_norm": 0.4109748303890228,
+      "learning_rate": 0.0001,
+      "loss": 1.4092,
+      "step": 5783
+    },
+    {
+      "epoch": 0.6644075584400666,
+      "grad_norm": 0.4257631301879883,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 5784
+    },
+    {
+      "epoch": 0.6645224283498937,
+      "grad_norm": 0.4686290919780731,
+      "learning_rate": 0.0001,
+      "loss": 1.7784,
+      "step": 5785
+    },
+    {
+      "epoch": 0.6646372982597208,
+      "grad_norm": 0.4436352252960205,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 5786
+    },
+    {
+      "epoch": 0.664752168169548,
+      "grad_norm": 0.4808412790298462,
+      "learning_rate": 0.0001,
+      "loss": 1.7645,
+      "step": 5787
+    },
+    {
+      "epoch": 0.6648670380793751,
+      "grad_norm": 0.5230939984321594,
+      "learning_rate": 0.0001,
+      "loss": 1.8253,
+      "step": 5788
+    },
+    {
+      "epoch": 0.6649819079892022,
+      "grad_norm": 0.4324409067630768,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 5789
+    },
+    {
+      "epoch": 0.6650967778990293,
+      "grad_norm": 0.48814013600349426,
+      "learning_rate": 0.0001,
+      "loss": 1.8974,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6652116478088564,
+      "grad_norm": 0.4277442991733551,
+      "learning_rate": 0.0001,
+      "loss": 1.4797,
+      "step": 5791
+    },
+    {
+      "epoch": 0.6653265177186836,
+      "grad_norm": 0.43875277042388916,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 5792
+    },
+    {
+      "epoch": 0.6654413876285107,
+      "grad_norm": 0.47888147830963135,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 5793
+    },
+    {
+      "epoch": 0.6655562575383378,
+      "grad_norm": 0.43682193756103516,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 5794
+    },
+    {
+      "epoch": 0.6656711274481649,
+      "grad_norm": 0.4496222734451294,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 5795
+    },
+    {
+      "epoch": 0.665785997357992,
+      "grad_norm": 0.44679972529411316,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 5796
+    },
+    {
+      "epoch": 0.6659008672678192,
+      "grad_norm": 0.44004032015800476,
+      "learning_rate": 0.0001,
+      "loss": 1.6932,
+      "step": 5797
+    },
+    {
+      "epoch": 0.6660157371776463,
+      "grad_norm": 0.46193936467170715,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 5798
+    },
+    {
+      "epoch": 0.6661306070874734,
+      "grad_norm": 0.4364147186279297,
+      "learning_rate": 0.0001,
+      "loss": 1.6024,
+      "step": 5799
+    },
+    {
+      "epoch": 0.6662454769973005,
+      "grad_norm": 0.46602505445480347,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6663603469071276,
+      "grad_norm": 0.4367719888687134,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 5801
+    },
+    {
+      "epoch": 0.6664752168169548,
+      "grad_norm": 0.4234819710254669,
+      "learning_rate": 0.0001,
+      "loss": 1.654,
+      "step": 5802
+    },
+    {
+      "epoch": 0.6665900867267819,
+      "grad_norm": 0.46970421075820923,
+      "learning_rate": 0.0001,
+      "loss": 1.8436,
+      "step": 5803
+    },
+    {
+      "epoch": 0.666704956636609,
+      "grad_norm": 0.45983707904815674,
+      "learning_rate": 0.0001,
+      "loss": 1.5,
+      "step": 5804
+    },
+    {
+      "epoch": 0.6668198265464361,
+      "grad_norm": 0.4581074118614197,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6669346964562632,
+      "grad_norm": 0.4409380853176117,
+      "learning_rate": 0.0001,
+      "loss": 1.6336,
+      "step": 5806
+    },
+    {
+      "epoch": 0.6670495663660904,
+      "grad_norm": 0.46238353848457336,
+      "learning_rate": 0.0001,
+      "loss": 1.8162,
+      "step": 5807
+    },
+    {
+      "epoch": 0.6671644362759175,
+      "grad_norm": 0.4390278160572052,
+      "learning_rate": 0.0001,
+      "loss": 1.7262,
+      "step": 5808
+    },
+    {
+      "epoch": 0.6672793061857446,
+      "grad_norm": 0.44605204463005066,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 5809
+    },
+    {
+      "epoch": 0.6673941760955717,
+      "grad_norm": 0.41407981514930725,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6675090460053988,
+      "grad_norm": 0.44644859433174133,
+      "learning_rate": 0.0001,
+      "loss": 1.5268,
+      "step": 5811
+    },
+    {
+      "epoch": 0.667623915915226,
+      "grad_norm": 0.4450972080230713,
+      "learning_rate": 0.0001,
+      "loss": 1.7008,
+      "step": 5812
+    },
+    {
+      "epoch": 0.6677387858250531,
+      "grad_norm": 0.45438942313194275,
+      "learning_rate": 0.0001,
+      "loss": 1.572,
+      "step": 5813
+    },
+    {
+      "epoch": 0.6678536557348802,
+      "grad_norm": 0.4170825481414795,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 5814
+    },
+    {
+      "epoch": 0.6679685256447073,
+      "grad_norm": 0.4803465008735657,
+      "learning_rate": 0.0001,
+      "loss": 1.8373,
+      "step": 5815
+    },
+    {
+      "epoch": 0.6680833955545344,
+      "grad_norm": 0.4364113211631775,
+      "learning_rate": 0.0001,
+      "loss": 1.5873,
+      "step": 5816
+    },
+    {
+      "epoch": 0.6681982654643616,
+      "grad_norm": 0.4389718174934387,
+      "learning_rate": 0.0001,
+      "loss": 1.3752,
+      "step": 5817
+    },
+    {
+      "epoch": 0.6683131353741887,
+      "grad_norm": 0.4332119822502136,
+      "learning_rate": 0.0001,
+      "loss": 1.5222,
+      "step": 5818
+    },
+    {
+      "epoch": 0.6684280052840158,
+      "grad_norm": 0.42295655608177185,
+      "learning_rate": 0.0001,
+      "loss": 1.4948,
+      "step": 5819
+    },
+    {
+      "epoch": 0.6685428751938429,
+      "grad_norm": 0.39874276518821716,
+      "learning_rate": 0.0001,
+      "loss": 1.5184,
+      "step": 5820
+    },
+    {
+      "epoch": 0.66865774510367,
+      "grad_norm": 0.4008484482765198,
+      "learning_rate": 0.0001,
+      "loss": 1.6389,
+      "step": 5821
+    },
+    {
+      "epoch": 0.6687726150134972,
+      "grad_norm": 0.43164539337158203,
+      "learning_rate": 0.0001,
+      "loss": 1.6292,
+      "step": 5822
+    },
+    {
+      "epoch": 0.6688874849233243,
+      "grad_norm": 0.39950746297836304,
+      "learning_rate": 0.0001,
+      "loss": 1.5272,
+      "step": 5823
+    },
+    {
+      "epoch": 0.6690023548331514,
+      "grad_norm": 0.442396879196167,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 5824
+    },
+    {
+      "epoch": 0.6691172247429785,
+      "grad_norm": 0.4410782754421234,
+      "learning_rate": 0.0001,
+      "loss": 1.4986,
+      "step": 5825
+    },
+    {
+      "epoch": 0.6692320946528056,
+      "grad_norm": 0.4995570480823517,
+      "learning_rate": 0.0001,
+      "loss": 1.7851,
+      "step": 5826
+    },
+    {
+      "epoch": 0.6693469645626329,
+      "grad_norm": 0.43698373436927795,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 5827
+    },
+    {
+      "epoch": 0.66946183447246,
+      "grad_norm": 0.5518622994422913,
+      "learning_rate": 0.0001,
+      "loss": 1.6748,
+      "step": 5828
+    },
+    {
+      "epoch": 0.6695767043822871,
+      "grad_norm": 0.4583264887332916,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 5829
+    },
+    {
+      "epoch": 0.6696915742921142,
+      "grad_norm": 0.40689265727996826,
+      "learning_rate": 0.0001,
+      "loss": 1.5925,
+      "step": 5830
+    },
+    {
+      "epoch": 0.6698064442019414,
+      "grad_norm": 0.45366060733795166,
+      "learning_rate": 0.0001,
+      "loss": 1.6063,
+      "step": 5831
+    },
+    {
+      "epoch": 0.6699213141117685,
+      "grad_norm": 0.4742351472377777,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 5832
+    },
+    {
+      "epoch": 0.6700361840215956,
+      "grad_norm": 0.43246540427207947,
+      "learning_rate": 0.0001,
+      "loss": 1.524,
+      "step": 5833
+    },
+    {
+      "epoch": 0.6701510539314227,
+      "grad_norm": 0.4532145857810974,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 5834
+    },
+    {
+      "epoch": 0.6702659238412498,
+      "grad_norm": 0.4628235399723053,
+      "learning_rate": 0.0001,
+      "loss": 1.7553,
+      "step": 5835
+    },
+    {
+      "epoch": 0.670380793751077,
+      "grad_norm": 0.5169723629951477,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 5836
+    },
+    {
+      "epoch": 0.6704956636609041,
+      "grad_norm": 0.4652624726295471,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 5837
+    },
+    {
+      "epoch": 0.6706105335707312,
+      "grad_norm": 0.45965853333473206,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 5838
+    },
+    {
+      "epoch": 0.6707254034805583,
+      "grad_norm": 0.4296656548976898,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 5839
+    },
+    {
+      "epoch": 0.6708402733903854,
+      "grad_norm": 0.4188656806945801,
+      "learning_rate": 0.0001,
+      "loss": 1.594,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6709551433002126,
+      "grad_norm": 0.442221075296402,
+      "learning_rate": 0.0001,
+      "loss": 1.6074,
+      "step": 5841
+    },
+    {
+      "epoch": 0.6710700132100397,
+      "grad_norm": 0.4227312207221985,
+      "learning_rate": 0.0001,
+      "loss": 1.5262,
+      "step": 5842
+    },
+    {
+      "epoch": 0.6711848831198668,
+      "grad_norm": 0.4592028856277466,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 5843
+    },
+    {
+      "epoch": 0.6712997530296939,
+      "grad_norm": 0.4832910895347595,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 5844
+    },
+    {
+      "epoch": 0.671414622939521,
+      "grad_norm": 0.44571515917778015,
+      "learning_rate": 0.0001,
+      "loss": 1.7643,
+      "step": 5845
+    },
+    {
+      "epoch": 0.6715294928493482,
+      "grad_norm": 0.45350438356399536,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 5846
+    },
+    {
+      "epoch": 0.6716443627591753,
+      "grad_norm": 0.4490528106689453,
+      "learning_rate": 0.0001,
+      "loss": 1.6748,
+      "step": 5847
+    },
+    {
+      "epoch": 0.6717592326690024,
+      "grad_norm": 0.4567161500453949,
+      "learning_rate": 0.0001,
+      "loss": 1.6875,
+      "step": 5848
+    },
+    {
+      "epoch": 0.6718741025788295,
+      "grad_norm": 0.44406652450561523,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 5849
+    },
+    {
+      "epoch": 0.6719889724886566,
+      "grad_norm": 0.447835773229599,
+      "learning_rate": 0.0001,
+      "loss": 1.6824,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6721038423984838,
+      "grad_norm": 0.597815752029419,
+      "learning_rate": 0.0001,
+      "loss": 1.5491,
+      "step": 5851
+    },
+    {
+      "epoch": 0.6722187123083109,
+      "grad_norm": 0.4304426610469818,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 5852
+    },
+    {
+      "epoch": 0.672333582218138,
+      "grad_norm": 0.4473625719547272,
+      "learning_rate": 0.0001,
+      "loss": 1.511,
+      "step": 5853
+    },
+    {
+      "epoch": 0.6724484521279651,
+      "grad_norm": 0.4547788202762604,
+      "learning_rate": 0.0001,
+      "loss": 1.8366,
+      "step": 5854
+    },
+    {
+      "epoch": 0.6725633220377922,
+      "grad_norm": 0.43215829133987427,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 5855
+    },
+    {
+      "epoch": 0.6726781919476194,
+      "grad_norm": 0.515165388584137,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 5856
+    },
+    {
+      "epoch": 0.6727930618574465,
+      "grad_norm": 0.4932282865047455,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 5857
+    },
+    {
+      "epoch": 0.6729079317672736,
+      "grad_norm": 0.46201443672180176,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 5858
+    },
+    {
+      "epoch": 0.6730228016771007,
+      "grad_norm": 0.470007061958313,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 5859
+    },
+    {
+      "epoch": 0.6731376715869278,
+      "grad_norm": 0.4670080542564392,
+      "learning_rate": 0.0001,
+      "loss": 1.7879,
+      "step": 5860
+    },
+    {
+      "epoch": 0.673252541496755,
+      "grad_norm": 0.4259360730648041,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 5861
+    },
+    {
+      "epoch": 0.6733674114065821,
+      "grad_norm": 0.47610583901405334,
+      "learning_rate": 0.0001,
+      "loss": 1.8552,
+      "step": 5862
+    },
+    {
+      "epoch": 0.6734822813164092,
+      "grad_norm": 0.4580018222332001,
+      "learning_rate": 0.0001,
+      "loss": 1.6853,
+      "step": 5863
+    },
+    {
+      "epoch": 0.6735971512262363,
+      "grad_norm": 0.4032216966152191,
+      "learning_rate": 0.0001,
+      "loss": 1.5547,
+      "step": 5864
+    },
+    {
+      "epoch": 0.6737120211360634,
+      "grad_norm": 0.45826172828674316,
+      "learning_rate": 0.0001,
+      "loss": 1.7622,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6738268910458906,
+      "grad_norm": 0.43257811665534973,
+      "learning_rate": 0.0001,
+      "loss": 1.6669,
+      "step": 5866
+    },
+    {
+      "epoch": 0.6739417609557177,
+      "grad_norm": 0.434496134519577,
+      "learning_rate": 0.0001,
+      "loss": 1.6081,
+      "step": 5867
+    },
+    {
+      "epoch": 0.6740566308655448,
+      "grad_norm": 0.4298497438430786,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 5868
+    },
+    {
+      "epoch": 0.6741715007753719,
+      "grad_norm": 0.46281471848487854,
+      "learning_rate": 0.0001,
+      "loss": 1.7304,
+      "step": 5869
+    },
+    {
+      "epoch": 0.674286370685199,
+      "grad_norm": 0.42578932642936707,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6744012405950262,
+      "grad_norm": 0.431417316198349,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 5871
+    },
+    {
+      "epoch": 0.6745161105048533,
+      "grad_norm": 0.4810255169868469,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 5872
+    },
+    {
+      "epoch": 0.6746309804146804,
+      "grad_norm": 0.4635904133319855,
+      "learning_rate": 0.0001,
+      "loss": 1.8039,
+      "step": 5873
+    },
+    {
+      "epoch": 0.6747458503245075,
+      "grad_norm": 0.46906232833862305,
+      "learning_rate": 0.0001,
+      "loss": 1.86,
+      "step": 5874
+    },
+    {
+      "epoch": 0.6748607202343346,
+      "grad_norm": 0.44870471954345703,
+      "learning_rate": 0.0001,
+      "loss": 1.5199,
+      "step": 5875
+    },
+    {
+      "epoch": 0.6749755901441618,
+      "grad_norm": 0.4645586609840393,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 5876
+    },
+    {
+      "epoch": 0.6750904600539889,
+      "grad_norm": 0.45580583810806274,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 5877
+    },
+    {
+      "epoch": 0.675205329963816,
+      "grad_norm": 0.4602547585964203,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 5878
+    },
+    {
+      "epoch": 0.6753201998736431,
+      "grad_norm": 0.45218491554260254,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 5879
+    },
+    {
+      "epoch": 0.6754350697834702,
+      "grad_norm": 0.4633892774581909,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6755499396932974,
+      "grad_norm": 0.47411301732063293,
+      "learning_rate": 0.0001,
+      "loss": 1.6984,
+      "step": 5881
+    },
+    {
+      "epoch": 0.6756648096031245,
+      "grad_norm": 0.4305265545845032,
+      "learning_rate": 0.0001,
+      "loss": 1.6009,
+      "step": 5882
+    },
+    {
+      "epoch": 0.6757796795129516,
+      "grad_norm": 0.444837749004364,
+      "learning_rate": 0.0001,
+      "loss": 1.5941,
+      "step": 5883
+    },
+    {
+      "epoch": 0.6758945494227787,
+      "grad_norm": 0.4728662967681885,
+      "learning_rate": 0.0001,
+      "loss": 1.7626,
+      "step": 5884
+    },
+    {
+      "epoch": 0.6760094193326058,
+      "grad_norm": 0.4483180046081543,
+      "learning_rate": 0.0001,
+      "loss": 1.7023,
+      "step": 5885
+    },
+    {
+      "epoch": 0.676124289242433,
+      "grad_norm": 0.4406643211841583,
+      "learning_rate": 0.0001,
+      "loss": 1.7195,
+      "step": 5886
+    },
+    {
+      "epoch": 0.6762391591522601,
+      "grad_norm": 0.44651803374290466,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 5887
+    },
+    {
+      "epoch": 0.6763540290620872,
+      "grad_norm": 0.4229283630847931,
+      "learning_rate": 0.0001,
+      "loss": 1.7038,
+      "step": 5888
+    },
+    {
+      "epoch": 0.6764688989719143,
+      "grad_norm": 0.4413118064403534,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 5889
+    },
+    {
+      "epoch": 0.6765837688817414,
+      "grad_norm": 0.44531190395355225,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6766986387915686,
+      "grad_norm": 0.4322546720504761,
+      "learning_rate": 0.0001,
+      "loss": 1.7821,
+      "step": 5891
+    },
+    {
+      "epoch": 0.6768135087013957,
+      "grad_norm": 0.43194666504859924,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 5892
+    },
+    {
+      "epoch": 0.6769283786112228,
+      "grad_norm": 0.45931199193000793,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 5893
+    },
+    {
+      "epoch": 0.6770432485210499,
+      "grad_norm": 0.4326918423175812,
+      "learning_rate": 0.0001,
+      "loss": 1.394,
+      "step": 5894
+    },
+    {
+      "epoch": 0.677158118430877,
+      "grad_norm": 0.43630099296569824,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 5895
+    },
+    {
+      "epoch": 0.6772729883407042,
+      "grad_norm": 0.4202820360660553,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 5896
+    },
+    {
+      "epoch": 0.6773878582505313,
+      "grad_norm": 0.508823812007904,
+      "learning_rate": 0.0001,
+      "loss": 1.6324,
+      "step": 5897
+    },
+    {
+      "epoch": 0.6775027281603584,
+      "grad_norm": 0.43873631954193115,
+      "learning_rate": 0.0001,
+      "loss": 1.522,
+      "step": 5898
+    },
+    {
+      "epoch": 0.6776175980701855,
+      "grad_norm": 0.4063846170902252,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 5899
+    },
+    {
+      "epoch": 0.6777324679800126,
+      "grad_norm": 0.4242575466632843,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6778473378898398,
+      "grad_norm": 0.44178491830825806,
+      "learning_rate": 0.0001,
+      "loss": 1.4893,
+      "step": 5901
+    },
+    {
+      "epoch": 0.6779622077996669,
+      "grad_norm": 0.4485710561275482,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 5902
+    },
+    {
+      "epoch": 0.678077077709494,
+      "grad_norm": 0.4207887351512909,
+      "learning_rate": 0.0001,
+      "loss": 1.5055,
+      "step": 5903
+    },
+    {
+      "epoch": 0.6781919476193211,
+      "grad_norm": 1.451164722442627,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 5904
+    },
+    {
+      "epoch": 0.6783068175291482,
+      "grad_norm": 0.46162787079811096,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 5905
+    },
+    {
+      "epoch": 0.6784216874389754,
+      "grad_norm": 0.4590327739715576,
+      "learning_rate": 0.0001,
+      "loss": 1.703,
+      "step": 5906
+    },
+    {
+      "epoch": 0.6785365573488025,
+      "grad_norm": 0.44896572828292847,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 5907
+    },
+    {
+      "epoch": 0.6786514272586296,
+      "grad_norm": 0.43614816665649414,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 5908
+    },
+    {
+      "epoch": 0.6787662971684567,
+      "grad_norm": 0.4527941942214966,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 5909
+    },
+    {
+      "epoch": 0.6788811670782838,
+      "grad_norm": 0.4711673855781555,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 5910
+    },
+    {
+      "epoch": 0.678996036988111,
+      "grad_norm": 0.5014576315879822,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 5911
+    },
+    {
+      "epoch": 0.6791109068979381,
+      "grad_norm": 0.4469035565853119,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 5912
+    },
+    {
+      "epoch": 0.6792257768077652,
+      "grad_norm": 0.47193190455436707,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 5913
+    },
+    {
+      "epoch": 0.6793406467175923,
+      "grad_norm": 0.4429560899734497,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 5914
+    },
+    {
+      "epoch": 0.6794555166274194,
+      "grad_norm": 0.46765467524528503,
+      "learning_rate": 0.0001,
+      "loss": 1.6985,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6795703865372466,
+      "grad_norm": 0.4488067328929901,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 5916
+    },
+    {
+      "epoch": 0.6796852564470737,
+      "grad_norm": 0.4276580214500427,
+      "learning_rate": 0.0001,
+      "loss": 1.579,
+      "step": 5917
+    },
+    {
+      "epoch": 0.6798001263569008,
+      "grad_norm": 0.47405701875686646,
+      "learning_rate": 0.0001,
+      "loss": 1.7718,
+      "step": 5918
+    },
+    {
+      "epoch": 0.6799149962667279,
+      "grad_norm": 0.4314579367637634,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 5919
+    },
+    {
+      "epoch": 0.680029866176555,
+      "grad_norm": 0.4281775951385498,
+      "learning_rate": 0.0001,
+      "loss": 1.4945,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6801447360863822,
+      "grad_norm": 0.4201895594596863,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 5921
+    },
+    {
+      "epoch": 0.6802596059962093,
+      "grad_norm": 0.42386960983276367,
+      "learning_rate": 0.0001,
+      "loss": 1.5078,
+      "step": 5922
+    },
+    {
+      "epoch": 0.6803744759060364,
+      "grad_norm": 0.49620696902275085,
+      "learning_rate": 0.0001,
+      "loss": 1.8191,
+      "step": 5923
+    },
+    {
+      "epoch": 0.6804893458158635,
+      "grad_norm": 0.4446530044078827,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 5924
+    },
+    {
+      "epoch": 0.6806042157256906,
+      "grad_norm": 0.436937540769577,
+      "learning_rate": 0.0001,
+      "loss": 1.5185,
+      "step": 5925
+    },
+    {
+      "epoch": 0.6807190856355178,
+      "grad_norm": 0.41910529136657715,
+      "learning_rate": 0.0001,
+      "loss": 1.3554,
+      "step": 5926
+    },
+    {
+      "epoch": 0.6808339555453449,
+      "grad_norm": 0.4524851143360138,
+      "learning_rate": 0.0001,
+      "loss": 1.5111,
+      "step": 5927
+    },
+    {
+      "epoch": 0.680948825455172,
+      "grad_norm": 0.5210041999816895,
+      "learning_rate": 0.0001,
+      "loss": 1.8005,
+      "step": 5928
+    },
+    {
+      "epoch": 0.6810636953649991,
+      "grad_norm": 0.4216352105140686,
+      "learning_rate": 0.0001,
+      "loss": 1.5726,
+      "step": 5929
+    },
+    {
+      "epoch": 0.6811785652748262,
+      "grad_norm": 0.44295185804367065,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6812934351846534,
+      "grad_norm": 0.4396561086177826,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 5931
+    },
+    {
+      "epoch": 0.6814083050944805,
+      "grad_norm": 0.440660297870636,
+      "learning_rate": 0.0001,
+      "loss": 1.7889,
+      "step": 5932
+    },
+    {
+      "epoch": 0.6815231750043076,
+      "grad_norm": 0.4529242217540741,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 5933
+    },
+    {
+      "epoch": 0.6816380449141347,
+      "grad_norm": 0.43237507343292236,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 5934
+    },
+    {
+      "epoch": 0.6817529148239618,
+      "grad_norm": 0.41882574558258057,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 5935
+    },
+    {
+      "epoch": 0.681867784733789,
+      "grad_norm": 0.501582682132721,
+      "learning_rate": 0.0001,
+      "loss": 1.7749,
+      "step": 5936
+    },
+    {
+      "epoch": 0.6819826546436161,
+      "grad_norm": 0.4597390294075012,
+      "learning_rate": 0.0001,
+      "loss": 1.7377,
+      "step": 5937
+    },
+    {
+      "epoch": 0.6820975245534432,
+      "grad_norm": 0.42805132269859314,
+      "learning_rate": 0.0001,
+      "loss": 1.5204,
+      "step": 5938
+    },
+    {
+      "epoch": 0.6822123944632703,
+      "grad_norm": 0.4181848168373108,
+      "learning_rate": 0.0001,
+      "loss": 1.3851,
+      "step": 5939
+    },
+    {
+      "epoch": 0.6823272643730974,
+      "grad_norm": 0.5363374352455139,
+      "learning_rate": 0.0001,
+      "loss": 1.9319,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6824421342829246,
+      "grad_norm": 0.47564831376075745,
+      "learning_rate": 0.0001,
+      "loss": 1.7815,
+      "step": 5941
+    },
+    {
+      "epoch": 0.6825570041927517,
+      "grad_norm": 0.4543587267398834,
+      "learning_rate": 0.0001,
+      "loss": 1.626,
+      "step": 5942
+    },
+    {
+      "epoch": 0.6826718741025788,
+      "grad_norm": 0.43262454867362976,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 5943
+    },
+    {
+      "epoch": 0.6827867440124059,
+      "grad_norm": 0.44168633222579956,
+      "learning_rate": 0.0001,
+      "loss": 1.3962,
+      "step": 5944
+    },
+    {
+      "epoch": 0.682901613922233,
+      "grad_norm": 0.4552295207977295,
+      "learning_rate": 0.0001,
+      "loss": 1.6962,
+      "step": 5945
+    },
+    {
+      "epoch": 0.6830164838320602,
+      "grad_norm": 0.4199536442756653,
+      "learning_rate": 0.0001,
+      "loss": 1.6031,
+      "step": 5946
+    },
+    {
+      "epoch": 0.6831313537418873,
+      "grad_norm": 0.44757145643234253,
+      "learning_rate": 0.0001,
+      "loss": 1.7174,
+      "step": 5947
+    },
+    {
+      "epoch": 0.6832462236517144,
+      "grad_norm": 0.4116254448890686,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 5948
+    },
+    {
+      "epoch": 0.6833610935615415,
+      "grad_norm": 0.4747268259525299,
+      "learning_rate": 0.0001,
+      "loss": 1.6715,
+      "step": 5949
+    },
+    {
+      "epoch": 0.6834759634713686,
+      "grad_norm": 0.41993069648742676,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6835908333811958,
+      "grad_norm": 0.4138827621936798,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 5951
+    },
+    {
+      "epoch": 0.6837057032910229,
+      "grad_norm": 0.45296478271484375,
+      "learning_rate": 0.0001,
+      "loss": 1.7347,
+      "step": 5952
+    },
+    {
+      "epoch": 0.68382057320085,
+      "grad_norm": 0.510452389717102,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 5953
+    },
+    {
+      "epoch": 0.6839354431106771,
+      "grad_norm": 0.474874883890152,
+      "learning_rate": 0.0001,
+      "loss": 1.6779,
+      "step": 5954
+    },
+    {
+      "epoch": 0.6840503130205042,
+      "grad_norm": 0.45102521777153015,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 5955
+    },
+    {
+      "epoch": 0.6841651829303314,
+      "grad_norm": 0.4434516429901123,
+      "learning_rate": 0.0001,
+      "loss": 1.4887,
+      "step": 5956
+    },
+    {
+      "epoch": 0.6842800528401585,
+      "grad_norm": 0.44997087121009827,
+      "learning_rate": 0.0001,
+      "loss": 1.6192,
+      "step": 5957
+    },
+    {
+      "epoch": 0.6843949227499856,
+      "grad_norm": 0.42875492572784424,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 5958
+    },
+    {
+      "epoch": 0.6845097926598127,
+      "grad_norm": 0.4152733087539673,
+      "learning_rate": 0.0001,
+      "loss": 1.5594,
+      "step": 5959
+    },
+    {
+      "epoch": 0.6846246625696398,
+      "grad_norm": 0.43790438771247864,
+      "learning_rate": 0.0001,
+      "loss": 1.5427,
+      "step": 5960
+    },
+    {
+      "epoch": 0.684739532479467,
+      "grad_norm": 0.45538073778152466,
+      "learning_rate": 0.0001,
+      "loss": 1.8264,
+      "step": 5961
+    },
+    {
+      "epoch": 0.6848544023892941,
+      "grad_norm": 0.671297013759613,
+      "learning_rate": 0.0001,
+      "loss": 1.6119,
+      "step": 5962
+    },
+    {
+      "epoch": 0.6849692722991212,
+      "grad_norm": 0.454745888710022,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 5963
+    },
+    {
+      "epoch": 0.6850841422089484,
+      "grad_norm": 0.43612053990364075,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 5964
+    },
+    {
+      "epoch": 0.6851990121187755,
+      "grad_norm": 0.42518892884254456,
+      "learning_rate": 0.0001,
+      "loss": 1.5105,
+      "step": 5965
+    },
+    {
+      "epoch": 0.6853138820286027,
+      "grad_norm": 0.4237819015979767,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 5966
+    },
+    {
+      "epoch": 0.6854287519384298,
+      "grad_norm": 0.4369647204875946,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 5967
+    },
+    {
+      "epoch": 0.6855436218482569,
+      "grad_norm": 0.4214072525501251,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 5968
+    },
+    {
+      "epoch": 0.685658491758084,
+      "grad_norm": 0.45082470774650574,
+      "learning_rate": 0.0001,
+      "loss": 1.7264,
+      "step": 5969
+    },
+    {
+      "epoch": 0.6857733616679111,
+      "grad_norm": 0.44140127301216125,
+      "learning_rate": 0.0001,
+      "loss": 1.447,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6858882315777383,
+      "grad_norm": 0.4286753833293915,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 5971
+    },
+    {
+      "epoch": 0.6860031014875654,
+      "grad_norm": 0.43218451738357544,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 5972
+    },
+    {
+      "epoch": 0.6861179713973925,
+      "grad_norm": 0.40552157163619995,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 5973
+    },
+    {
+      "epoch": 0.6862328413072196,
+      "grad_norm": 0.449190229177475,
+      "learning_rate": 0.0001,
+      "loss": 1.5253,
+      "step": 5974
+    },
+    {
+      "epoch": 0.6863477112170467,
+      "grad_norm": 0.42103078961372375,
+      "learning_rate": 0.0001,
+      "loss": 1.485,
+      "step": 5975
+    },
+    {
+      "epoch": 0.6864625811268739,
+      "grad_norm": 0.4737277328968048,
+      "learning_rate": 0.0001,
+      "loss": 1.5635,
+      "step": 5976
+    },
+    {
+      "epoch": 0.686577451036701,
+      "grad_norm": 0.44992202520370483,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 5977
+    },
+    {
+      "epoch": 0.6866923209465281,
+      "grad_norm": 0.4830211400985718,
+      "learning_rate": 0.0001,
+      "loss": 1.3773,
+      "step": 5978
+    },
+    {
+      "epoch": 0.6868071908563552,
+      "grad_norm": 0.48300692439079285,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 5979
+    },
+    {
+      "epoch": 0.6869220607661823,
+      "grad_norm": 0.4886820614337921,
+      "learning_rate": 0.0001,
+      "loss": 1.7533,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6870369306760095,
+      "grad_norm": 0.4783158600330353,
+      "learning_rate": 0.0001,
+      "loss": 1.5382,
+      "step": 5981
+    },
+    {
+      "epoch": 0.6871518005858366,
+      "grad_norm": 0.44039803743362427,
+      "learning_rate": 0.0001,
+      "loss": 1.6981,
+      "step": 5982
+    },
+    {
+      "epoch": 0.6872666704956637,
+      "grad_norm": 0.44316425919532776,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 5983
+    },
+    {
+      "epoch": 0.6873815404054908,
+      "grad_norm": 0.442717581987381,
+      "learning_rate": 0.0001,
+      "loss": 1.6243,
+      "step": 5984
+    },
+    {
+      "epoch": 0.6874964103153179,
+      "grad_norm": 0.40507930517196655,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 5985
+    },
+    {
+      "epoch": 0.6876112802251451,
+      "grad_norm": 0.4449567496776581,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 5986
+    },
+    {
+      "epoch": 0.6877261501349722,
+      "grad_norm": 0.4152354300022125,
+      "learning_rate": 0.0001,
+      "loss": 1.5178,
+      "step": 5987
+    },
+    {
+      "epoch": 0.6878410200447993,
+      "grad_norm": 0.4327658414840698,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 5988
+    },
+    {
+      "epoch": 0.6879558899546264,
+      "grad_norm": 0.39523184299468994,
+      "learning_rate": 0.0001,
+      "loss": 1.4862,
+      "step": 5989
+    },
+    {
+      "epoch": 0.6880707598644535,
+      "grad_norm": 0.4782007336616516,
+      "learning_rate": 0.0001,
+      "loss": 1.7516,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6881856297742807,
+      "grad_norm": 0.4494955539703369,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 5991
+    },
+    {
+      "epoch": 0.6883004996841078,
+      "grad_norm": 0.44850456714630127,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 5992
+    },
+    {
+      "epoch": 0.6884153695939349,
+      "grad_norm": 0.43033161759376526,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 5993
+    },
+    {
+      "epoch": 0.688530239503762,
+      "grad_norm": 0.4402123987674713,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 5994
+    },
+    {
+      "epoch": 0.6886451094135891,
+      "grad_norm": 0.41884496808052063,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 5995
+    },
+    {
+      "epoch": 0.6887599793234163,
+      "grad_norm": 0.4290192127227783,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 5996
+    },
+    {
+      "epoch": 0.6888748492332434,
+      "grad_norm": 0.5004449486732483,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 5997
+    },
+    {
+      "epoch": 0.6889897191430705,
+      "grad_norm": 0.4282512068748474,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 5998
+    },
+    {
+      "epoch": 0.6891045890528976,
+      "grad_norm": 0.4622134566307068,
+      "learning_rate": 0.0001,
+      "loss": 1.5423,
+      "step": 5999
+    },
+    {
+      "epoch": 0.6892194589627247,
+      "grad_norm": 0.43330010771751404,
+      "learning_rate": 0.0001,
+      "loss": 1.2861,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6893343288725519,
+      "grad_norm": 0.47380784153938293,
+      "learning_rate": 0.0001,
+      "loss": 1.5154,
+      "step": 6001
+    },
+    {
+      "epoch": 0.689449198782379,
+      "grad_norm": 0.4909721314907074,
+      "learning_rate": 0.0001,
+      "loss": 1.6611,
+      "step": 6002
+    },
+    {
+      "epoch": 0.6895640686922061,
+      "grad_norm": 0.444553405046463,
+      "learning_rate": 0.0001,
+      "loss": 1.5056,
+      "step": 6003
+    },
+    {
+      "epoch": 0.6896789386020332,
+      "grad_norm": 0.4200032651424408,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 6004
+    },
+    {
+      "epoch": 0.6897938085118603,
+      "grad_norm": 0.45052191615104675,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 6005
+    },
+    {
+      "epoch": 0.6899086784216875,
+      "grad_norm": 0.44020742177963257,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 6006
+    },
+    {
+      "epoch": 0.6900235483315146,
+      "grad_norm": 0.4479757249355316,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 6007
+    },
+    {
+      "epoch": 0.6901384182413417,
+      "grad_norm": 0.40676623582839966,
+      "learning_rate": 0.0001,
+      "loss": 1.4707,
+      "step": 6008
+    },
+    {
+      "epoch": 0.6902532881511688,
+      "grad_norm": 0.44237807393074036,
+      "learning_rate": 0.0001,
+      "loss": 1.6429,
+      "step": 6009
+    },
+    {
+      "epoch": 0.6903681580609959,
+      "grad_norm": 0.5069726705551147,
+      "learning_rate": 0.0001,
+      "loss": 1.819,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6904830279708231,
+      "grad_norm": 0.4417176842689514,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 6011
+    },
+    {
+      "epoch": 0.6905978978806502,
+      "grad_norm": 0.4071502685546875,
+      "learning_rate": 0.0001,
+      "loss": 1.4589,
+      "step": 6012
+    },
+    {
+      "epoch": 0.6907127677904773,
+      "grad_norm": 0.4580749273300171,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 6013
+    },
+    {
+      "epoch": 0.6908276377003044,
+      "grad_norm": 0.4313982427120209,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 6014
+    },
+    {
+      "epoch": 0.6909425076101315,
+      "grad_norm": 0.4240618944168091,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 6015
+    },
+    {
+      "epoch": 0.6910573775199587,
+      "grad_norm": 0.6877708435058594,
+      "learning_rate": 0.0001,
+      "loss": 1.4963,
+      "step": 6016
+    },
+    {
+      "epoch": 0.6911722474297858,
+      "grad_norm": 0.4607912003993988,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 6017
+    },
+    {
+      "epoch": 0.6912871173396129,
+      "grad_norm": 0.44218817353248596,
+      "learning_rate": 0.0001,
+      "loss": 1.7323,
+      "step": 6018
+    },
+    {
+      "epoch": 0.69140198724944,
+      "grad_norm": 0.4411420226097107,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 6019
+    },
+    {
+      "epoch": 0.6915168571592671,
+      "grad_norm": 0.4698505103588104,
+      "learning_rate": 0.0001,
+      "loss": 1.8326,
+      "step": 6020
+    },
+    {
+      "epoch": 0.6916317270690943,
+      "grad_norm": 0.4248715937137604,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 6021
+    },
+    {
+      "epoch": 0.6917465969789214,
+      "grad_norm": 0.4034425616264343,
+      "learning_rate": 0.0001,
+      "loss": 1.4084,
+      "step": 6022
+    },
+    {
+      "epoch": 0.6918614668887485,
+      "grad_norm": 0.4377814829349518,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 6023
+    },
+    {
+      "epoch": 0.6919763367985756,
+      "grad_norm": 0.4794885218143463,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 6024
+    },
+    {
+      "epoch": 0.6920912067084027,
+      "grad_norm": 0.4388996958732605,
+      "learning_rate": 0.0001,
+      "loss": 1.6848,
+      "step": 6025
+    },
+    {
+      "epoch": 0.6922060766182299,
+      "grad_norm": 0.3899931311607361,
+      "learning_rate": 0.0001,
+      "loss": 1.3496,
+      "step": 6026
+    },
+    {
+      "epoch": 0.692320946528057,
+      "grad_norm": 0.49666503071784973,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 6027
+    },
+    {
+      "epoch": 0.6924358164378841,
+      "grad_norm": 0.4482400119304657,
+      "learning_rate": 0.0001,
+      "loss": 1.6767,
+      "step": 6028
+    },
+    {
+      "epoch": 0.6925506863477112,
+      "grad_norm": 0.41266587376594543,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 6029
+    },
+    {
+      "epoch": 0.6926655562575383,
+      "grad_norm": 0.3807138502597809,
+      "learning_rate": 0.0001,
+      "loss": 1.34,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6927804261673655,
+      "grad_norm": 0.46382012963294983,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 6031
+    },
+    {
+      "epoch": 0.6928952960771926,
+      "grad_norm": 0.45717358589172363,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 6032
+    },
+    {
+      "epoch": 0.6930101659870197,
+      "grad_norm": 0.44662654399871826,
+      "learning_rate": 0.0001,
+      "loss": 1.7359,
+      "step": 6033
+    },
+    {
+      "epoch": 0.6931250358968468,
+      "grad_norm": 0.4659220278263092,
+      "learning_rate": 0.0001,
+      "loss": 1.7403,
+      "step": 6034
+    },
+    {
+      "epoch": 0.6932399058066739,
+      "grad_norm": 0.4265783727169037,
+      "learning_rate": 0.0001,
+      "loss": 1.6654,
+      "step": 6035
+    },
+    {
+      "epoch": 0.6933547757165011,
+      "grad_norm": 0.45917513966560364,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 6036
+    },
+    {
+      "epoch": 0.6934696456263282,
+      "grad_norm": 0.4457080066204071,
+      "learning_rate": 0.0001,
+      "loss": 1.6824,
+      "step": 6037
+    },
+    {
+      "epoch": 0.6935845155361553,
+      "grad_norm": 0.428594172000885,
+      "learning_rate": 0.0001,
+      "loss": 1.4754,
+      "step": 6038
+    },
+    {
+      "epoch": 0.6936993854459824,
+      "grad_norm": 0.4832080602645874,
+      "learning_rate": 0.0001,
+      "loss": 1.7493,
+      "step": 6039
+    },
+    {
+      "epoch": 0.6938142553558095,
+      "grad_norm": 0.4549182653427124,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6939291252656367,
+      "grad_norm": 0.46311667561531067,
+      "learning_rate": 0.0001,
+      "loss": 1.6486,
+      "step": 6041
+    },
+    {
+      "epoch": 0.6940439951754638,
+      "grad_norm": 0.43765565752983093,
+      "learning_rate": 0.0001,
+      "loss": 1.5472,
+      "step": 6042
+    },
+    {
+      "epoch": 0.6941588650852909,
+      "grad_norm": 0.4676961600780487,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 6043
+    },
+    {
+      "epoch": 0.694273734995118,
+      "grad_norm": 0.48037439584732056,
+      "learning_rate": 0.0001,
+      "loss": 1.7783,
+      "step": 6044
+    },
+    {
+      "epoch": 0.6943886049049451,
+      "grad_norm": 0.4451299011707306,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 6045
+    },
+    {
+      "epoch": 0.6945034748147723,
+      "grad_norm": 0.4473123848438263,
+      "learning_rate": 0.0001,
+      "loss": 1.7079,
+      "step": 6046
+    },
+    {
+      "epoch": 0.6946183447245994,
+      "grad_norm": 0.4191644489765167,
+      "learning_rate": 0.0001,
+      "loss": 1.4635,
+      "step": 6047
+    },
+    {
+      "epoch": 0.6947332146344265,
+      "grad_norm": 0.5033878087997437,
+      "learning_rate": 0.0001,
+      "loss": 1.6531,
+      "step": 6048
+    },
+    {
+      "epoch": 0.6948480845442536,
+      "grad_norm": 0.4745533764362335,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 6049
+    },
+    {
+      "epoch": 0.6949629544540807,
+      "grad_norm": 0.44269341230392456,
+      "learning_rate": 0.0001,
+      "loss": 1.5863,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6950778243639079,
+      "grad_norm": 0.47186774015426636,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 6051
+    },
+    {
+      "epoch": 0.695192694273735,
+      "grad_norm": 0.5083333849906921,
+      "learning_rate": 0.0001,
+      "loss": 1.7727,
+      "step": 6052
+    },
+    {
+      "epoch": 0.6953075641835621,
+      "grad_norm": 0.4614897072315216,
+      "learning_rate": 0.0001,
+      "loss": 1.669,
+      "step": 6053
+    },
+    {
+      "epoch": 0.6954224340933892,
+      "grad_norm": 0.47114160656929016,
+      "learning_rate": 0.0001,
+      "loss": 1.6241,
+      "step": 6054
+    },
+    {
+      "epoch": 0.6955373040032163,
+      "grad_norm": 0.4435749650001526,
+      "learning_rate": 0.0001,
+      "loss": 1.5528,
+      "step": 6055
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.4462308883666992,
+      "learning_rate": 0.0001,
+      "loss": 1.5747,
+      "step": 6056
+    },
+    {
+      "epoch": 0.6957670438228706,
+      "grad_norm": 0.4508498013019562,
+      "learning_rate": 0.0001,
+      "loss": 1.772,
+      "step": 6057
+    },
+    {
+      "epoch": 0.6958819137326977,
+      "grad_norm": 0.45102837681770325,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 6058
+    },
+    {
+      "epoch": 0.6959967836425248,
+      "grad_norm": 0.427142471075058,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 6059
+    },
+    {
+      "epoch": 0.6961116535523519,
+      "grad_norm": 0.44649988412857056,
+      "learning_rate": 0.0001,
+      "loss": 1.7092,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6962265234621791,
+      "grad_norm": 0.44798871874809265,
+      "learning_rate": 0.0001,
+      "loss": 1.6849,
+      "step": 6061
+    },
+    {
+      "epoch": 0.6963413933720062,
+      "grad_norm": 0.5189127326011658,
+      "learning_rate": 0.0001,
+      "loss": 1.1203,
+      "step": 6062
+    },
+    {
+      "epoch": 0.6964562632818333,
+      "grad_norm": 0.4362175762653351,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 6063
+    },
+    {
+      "epoch": 0.6965711331916604,
+      "grad_norm": 0.4506964683532715,
+      "learning_rate": 0.0001,
+      "loss": 1.5871,
+      "step": 6064
+    },
+    {
+      "epoch": 0.6966860031014875,
+      "grad_norm": 0.4647710621356964,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 6065
+    },
+    {
+      "epoch": 0.6968008730113147,
+      "grad_norm": 0.43700385093688965,
+      "learning_rate": 0.0001,
+      "loss": 1.5093,
+      "step": 6066
+    },
+    {
+      "epoch": 0.6969157429211418,
+      "grad_norm": 0.4622689187526703,
+      "learning_rate": 0.0001,
+      "loss": 1.6878,
+      "step": 6067
+    },
+    {
+      "epoch": 0.6970306128309689,
+      "grad_norm": 0.4806528687477112,
+      "learning_rate": 0.0001,
+      "loss": 1.7705,
+      "step": 6068
+    },
+    {
+      "epoch": 0.697145482740796,
+      "grad_norm": 0.4175126254558563,
+      "learning_rate": 0.0001,
+      "loss": 1.4511,
+      "step": 6069
+    },
+    {
+      "epoch": 0.6972603526506231,
+      "grad_norm": 0.4509715735912323,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6973752225604503,
+      "grad_norm": 0.4509661793708801,
+      "learning_rate": 0.0001,
+      "loss": 1.5706,
+      "step": 6071
+    },
+    {
+      "epoch": 0.6974900924702774,
+      "grad_norm": 0.4647543728351593,
+      "learning_rate": 0.0001,
+      "loss": 1.7786,
+      "step": 6072
+    },
+    {
+      "epoch": 0.6976049623801045,
+      "grad_norm": 0.4497321546077728,
+      "learning_rate": 0.0001,
+      "loss": 1.4881,
+      "step": 6073
+    },
+    {
+      "epoch": 0.6977198322899316,
+      "grad_norm": 0.431309312582016,
+      "learning_rate": 0.0001,
+      "loss": 1.5486,
+      "step": 6074
+    },
+    {
+      "epoch": 0.6978347021997587,
+      "grad_norm": 0.423828661441803,
+      "learning_rate": 0.0001,
+      "loss": 1.717,
+      "step": 6075
+    },
+    {
+      "epoch": 0.6979495721095859,
+      "grad_norm": 0.4408908784389496,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 6076
+    },
+    {
+      "epoch": 0.698064442019413,
+      "grad_norm": 0.4263383448123932,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 6077
+    },
+    {
+      "epoch": 0.6981793119292401,
+      "grad_norm": 0.4382713735103607,
+      "learning_rate": 0.0001,
+      "loss": 1.6121,
+      "step": 6078
+    },
+    {
+      "epoch": 0.6982941818390672,
+      "grad_norm": 0.4006011486053467,
+      "learning_rate": 0.0001,
+      "loss": 1.4744,
+      "step": 6079
+    },
+    {
+      "epoch": 0.6984090517488943,
+      "grad_norm": 0.4670262038707733,
+      "learning_rate": 0.0001,
+      "loss": 1.7082,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6985239216587215,
+      "grad_norm": 0.46280163526535034,
+      "learning_rate": 0.0001,
+      "loss": 1.7092,
+      "step": 6081
+    },
+    {
+      "epoch": 0.6986387915685486,
+      "grad_norm": 0.4739822745323181,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 6082
+    },
+    {
+      "epoch": 0.6987536614783757,
+      "grad_norm": 0.44407156109809875,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 6083
+    },
+    {
+      "epoch": 0.6988685313882028,
+      "grad_norm": 0.44577091932296753,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 6084
+    },
+    {
+      "epoch": 0.6989834012980299,
+      "grad_norm": 0.465145468711853,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 6085
+    },
+    {
+      "epoch": 0.6990982712078571,
+      "grad_norm": 0.47078871726989746,
+      "learning_rate": 0.0001,
+      "loss": 1.7108,
+      "step": 6086
+    },
+    {
+      "epoch": 0.6992131411176842,
+      "grad_norm": 0.4897402226924896,
+      "learning_rate": 0.0001,
+      "loss": 1.5122,
+      "step": 6087
+    },
+    {
+      "epoch": 0.6993280110275113,
+      "grad_norm": 0.46980687975883484,
+      "learning_rate": 0.0001,
+      "loss": 1.7254,
+      "step": 6088
+    },
+    {
+      "epoch": 0.6994428809373384,
+      "grad_norm": 0.41509878635406494,
+      "learning_rate": 0.0001,
+      "loss": 1.4403,
+      "step": 6089
+    },
+    {
+      "epoch": 0.6995577508471655,
+      "grad_norm": 0.3764444887638092,
+      "learning_rate": 0.0001,
+      "loss": 1.2123,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6996726207569927,
+      "grad_norm": 0.4393060803413391,
+      "learning_rate": 0.0001,
+      "loss": 1.4743,
+      "step": 6091
+    },
+    {
+      "epoch": 0.6997874906668198,
+      "grad_norm": 0.45182573795318604,
+      "learning_rate": 0.0001,
+      "loss": 1.7701,
+      "step": 6092
+    },
+    {
+      "epoch": 0.6999023605766469,
+      "grad_norm": 0.4436628818511963,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 6093
+    },
+    {
+      "epoch": 0.700017230486474,
+      "grad_norm": 0.44743016362190247,
+      "learning_rate": 0.0001,
+      "loss": 1.7463,
+      "step": 6094
+    },
+    {
+      "epoch": 0.7001321003963011,
+      "grad_norm": 0.4477095901966095,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 6095
+    },
+    {
+      "epoch": 0.7002469703061283,
+      "grad_norm": 0.43921157717704773,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 6096
+    },
+    {
+      "epoch": 0.7003618402159554,
+      "grad_norm": 0.4518073797225952,
+      "learning_rate": 0.0001,
+      "loss": 1.715,
+      "step": 6097
+    },
+    {
+      "epoch": 0.7004767101257825,
+      "grad_norm": 0.4737764298915863,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 6098
+    },
+    {
+      "epoch": 0.7005915800356096,
+      "grad_norm": 0.42427653074264526,
+      "learning_rate": 0.0001,
+      "loss": 1.4087,
+      "step": 6099
+    },
+    {
+      "epoch": 0.7007064499454367,
+      "grad_norm": 0.42667144536972046,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 6100
+    },
+    {
+      "epoch": 0.700821319855264,
+      "grad_norm": 0.49589815735816956,
+      "learning_rate": 0.0001,
+      "loss": 1.712,
+      "step": 6101
+    },
+    {
+      "epoch": 0.7009361897650911,
+      "grad_norm": 0.43956196308135986,
+      "learning_rate": 0.0001,
+      "loss": 1.5506,
+      "step": 6102
+    },
+    {
+      "epoch": 0.7010510596749182,
+      "grad_norm": 0.45342060923576355,
+      "learning_rate": 0.0001,
+      "loss": 1.6568,
+      "step": 6103
+    },
+    {
+      "epoch": 0.7011659295847453,
+      "grad_norm": 0.4133893847465515,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 6104
+    },
+    {
+      "epoch": 0.7012807994945724,
+      "grad_norm": 0.41374027729034424,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 6105
+    },
+    {
+      "epoch": 0.7013956694043996,
+      "grad_norm": 0.44979679584503174,
+      "learning_rate": 0.0001,
+      "loss": 1.7201,
+      "step": 6106
+    },
+    {
+      "epoch": 0.7015105393142267,
+      "grad_norm": 0.42173275351524353,
+      "learning_rate": 0.0001,
+      "loss": 1.6273,
+      "step": 6107
+    },
+    {
+      "epoch": 0.7016254092240538,
+      "grad_norm": 0.4458094835281372,
+      "learning_rate": 0.0001,
+      "loss": 1.6467,
+      "step": 6108
+    },
+    {
+      "epoch": 0.7017402791338809,
+      "grad_norm": 0.44119831919670105,
+      "learning_rate": 0.0001,
+      "loss": 1.7528,
+      "step": 6109
+    },
+    {
+      "epoch": 0.701855149043708,
+      "grad_norm": 0.4866656959056854,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 6110
+    },
+    {
+      "epoch": 0.7019700189535352,
+      "grad_norm": 0.43976983428001404,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 6111
+    },
+    {
+      "epoch": 0.7020848888633623,
+      "grad_norm": 0.4584594666957855,
+      "learning_rate": 0.0001,
+      "loss": 1.5199,
+      "step": 6112
+    },
+    {
+      "epoch": 0.7021997587731894,
+      "grad_norm": 0.4671233594417572,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 6113
+    },
+    {
+      "epoch": 0.7023146286830165,
+      "grad_norm": 0.4047556519508362,
+      "learning_rate": 0.0001,
+      "loss": 1.3305,
+      "step": 6114
+    },
+    {
+      "epoch": 0.7024294985928436,
+      "grad_norm": 0.44763243198394775,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 6115
+    },
+    {
+      "epoch": 0.7025443685026708,
+      "grad_norm": 0.42392697930336,
+      "learning_rate": 0.0001,
+      "loss": 1.5303,
+      "step": 6116
+    },
+    {
+      "epoch": 0.7026592384124979,
+      "grad_norm": 0.4174063205718994,
+      "learning_rate": 0.0001,
+      "loss": 1.4395,
+      "step": 6117
+    },
+    {
+      "epoch": 0.702774108322325,
+      "grad_norm": 0.44011473655700684,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 6118
+    },
+    {
+      "epoch": 0.7028889782321521,
+      "grad_norm": 0.467358261346817,
+      "learning_rate": 0.0001,
+      "loss": 1.7407,
+      "step": 6119
+    },
+    {
+      "epoch": 0.7030038481419792,
+      "grad_norm": 0.43859270215034485,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 6120
+    },
+    {
+      "epoch": 0.7031187180518064,
+      "grad_norm": 0.45992812514305115,
+      "learning_rate": 0.0001,
+      "loss": 1.6784,
+      "step": 6121
+    },
+    {
+      "epoch": 0.7032335879616335,
+      "grad_norm": 0.5027499198913574,
+      "learning_rate": 0.0001,
+      "loss": 1.8963,
+      "step": 6122
+    },
+    {
+      "epoch": 0.7033484578714606,
+      "grad_norm": 0.4613763689994812,
+      "learning_rate": 0.0001,
+      "loss": 1.7305,
+      "step": 6123
+    },
+    {
+      "epoch": 0.7034633277812877,
+      "grad_norm": 0.44609150290489197,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 6124
+    },
+    {
+      "epoch": 0.7035781976911148,
+      "grad_norm": 0.47288355231285095,
+      "learning_rate": 0.0001,
+      "loss": 1.5033,
+      "step": 6125
+    },
+    {
+      "epoch": 0.703693067600942,
+      "grad_norm": 0.4981932044029236,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 6126
+    },
+    {
+      "epoch": 0.7038079375107691,
+      "grad_norm": 0.43476441502571106,
+      "learning_rate": 0.0001,
+      "loss": 1.6783,
+      "step": 6127
+    },
+    {
+      "epoch": 0.7039228074205962,
+      "grad_norm": 0.4434894323348999,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 6128
+    },
+    {
+      "epoch": 0.7040376773304233,
+      "grad_norm": 0.4626059830188751,
+      "learning_rate": 0.0001,
+      "loss": 1.6981,
+      "step": 6129
+    },
+    {
+      "epoch": 0.7041525472402504,
+      "grad_norm": 0.43429237604141235,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 6130
+    },
+    {
+      "epoch": 0.7042674171500776,
+      "grad_norm": 0.4962638318538666,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 6131
+    },
+    {
+      "epoch": 0.7043822870599047,
+      "grad_norm": 0.47368475794792175,
+      "learning_rate": 0.0001,
+      "loss": 1.7509,
+      "step": 6132
+    },
+    {
+      "epoch": 0.7044971569697318,
+      "grad_norm": 0.44226646423339844,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 6133
+    },
+    {
+      "epoch": 0.7046120268795589,
+      "grad_norm": 0.4369940757751465,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 6134
+    },
+    {
+      "epoch": 0.704726896789386,
+      "grad_norm": 0.4563344120979309,
+      "learning_rate": 0.0001,
+      "loss": 1.7639,
+      "step": 6135
+    },
+    {
+      "epoch": 0.7048417666992132,
+      "grad_norm": 0.4372788965702057,
+      "learning_rate": 0.0001,
+      "loss": 1.6908,
+      "step": 6136
+    },
+    {
+      "epoch": 0.7049566366090403,
+      "grad_norm": 0.43250662088394165,
+      "learning_rate": 0.0001,
+      "loss": 1.6608,
+      "step": 6137
+    },
+    {
+      "epoch": 0.7050715065188674,
+      "grad_norm": 0.43698424100875854,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 6138
+    },
+    {
+      "epoch": 0.7051863764286945,
+      "grad_norm": 0.4592958390712738,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 6139
+    },
+    {
+      "epoch": 0.7053012463385216,
+      "grad_norm": 0.4927108883857727,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 6140
+    },
+    {
+      "epoch": 0.7054161162483488,
+      "grad_norm": 0.41103577613830566,
+      "learning_rate": 0.0001,
+      "loss": 1.4693,
+      "step": 6141
+    },
+    {
+      "epoch": 0.7055309861581759,
+      "grad_norm": 0.429688423871994,
+      "learning_rate": 0.0001,
+      "loss": 1.6462,
+      "step": 6142
+    },
+    {
+      "epoch": 0.705645856068003,
+      "grad_norm": 0.4417214095592499,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 6143
+    },
+    {
+      "epoch": 0.7057607259778301,
+      "grad_norm": 0.4288574457168579,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 6144
+    },
+    {
+      "epoch": 0.7058755958876572,
+      "grad_norm": 0.4483475089073181,
+      "learning_rate": 0.0001,
+      "loss": 1.6976,
+      "step": 6145
+    },
+    {
+      "epoch": 0.7059904657974844,
+      "grad_norm": 0.44906046986579895,
+      "learning_rate": 0.0001,
+      "loss": 1.771,
+      "step": 6146
+    },
+    {
+      "epoch": 0.7061053357073115,
+      "grad_norm": 0.43787476420402527,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 6147
+    },
+    {
+      "epoch": 0.7062202056171386,
+      "grad_norm": 0.47539424896240234,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 6148
+    },
+    {
+      "epoch": 0.7063350755269657,
+      "grad_norm": 0.4609004855155945,
+      "learning_rate": 0.0001,
+      "loss": 1.5423,
+      "step": 6149
+    },
+    {
+      "epoch": 0.7064499454367928,
+      "grad_norm": 0.443278431892395,
+      "learning_rate": 0.0001,
+      "loss": 1.5154,
+      "step": 6150
+    },
+    {
+      "epoch": 0.70656481534662,
+      "grad_norm": 0.43776434659957886,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 6151
+    },
+    {
+      "epoch": 0.7066796852564471,
+      "grad_norm": 0.4367440342903137,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 6152
+    },
+    {
+      "epoch": 0.7067945551662742,
+      "grad_norm": 0.4293498694896698,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 6153
+    },
+    {
+      "epoch": 0.7069094250761013,
+      "grad_norm": 0.46721354126930237,
+      "learning_rate": 0.0001,
+      "loss": 1.8351,
+      "step": 6154
+    },
+    {
+      "epoch": 0.7070242949859284,
+      "grad_norm": 0.4610436260700226,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 6155
+    },
+    {
+      "epoch": 0.7071391648957556,
+      "grad_norm": 0.44414547085762024,
+      "learning_rate": 0.0001,
+      "loss": 1.6412,
+      "step": 6156
+    },
+    {
+      "epoch": 0.7072540348055827,
+      "grad_norm": 0.4219599962234497,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 6157
+    },
+    {
+      "epoch": 0.7073689047154098,
+      "grad_norm": 0.45348840951919556,
+      "learning_rate": 0.0001,
+      "loss": 1.6561,
+      "step": 6158
+    },
+    {
+      "epoch": 0.7074837746252369,
+      "grad_norm": 0.44765985012054443,
+      "learning_rate": 0.0001,
+      "loss": 1.7357,
+      "step": 6159
+    },
+    {
+      "epoch": 0.707598644535064,
+      "grad_norm": 0.43484190106391907,
+      "learning_rate": 0.0001,
+      "loss": 1.6815,
+      "step": 6160
+    },
+    {
+      "epoch": 0.7077135144448912,
+      "grad_norm": 0.44024690985679626,
+      "learning_rate": 0.0001,
+      "loss": 1.5954,
+      "step": 6161
+    },
+    {
+      "epoch": 0.7078283843547183,
+      "grad_norm": 0.4765664339065552,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 6162
+    },
+    {
+      "epoch": 0.7079432542645454,
+      "grad_norm": 0.4749075174331665,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 6163
+    },
+    {
+      "epoch": 0.7080581241743725,
+      "grad_norm": 0.47169381380081177,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 6164
+    },
+    {
+      "epoch": 0.7081729940841996,
+      "grad_norm": 0.45661336183547974,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 6165
+    },
+    {
+      "epoch": 0.7082878639940268,
+      "grad_norm": 0.45333951711654663,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 6166
+    },
+    {
+      "epoch": 0.7084027339038539,
+      "grad_norm": 0.4652560353279114,
+      "learning_rate": 0.0001,
+      "loss": 1.7598,
+      "step": 6167
+    },
+    {
+      "epoch": 0.708517603813681,
+      "grad_norm": 0.46927064657211304,
+      "learning_rate": 0.0001,
+      "loss": 1.782,
+      "step": 6168
+    },
+    {
+      "epoch": 0.7086324737235081,
+      "grad_norm": 0.44694873690605164,
+      "learning_rate": 0.0001,
+      "loss": 1.7146,
+      "step": 6169
+    },
+    {
+      "epoch": 0.7087473436333352,
+      "grad_norm": 0.4289446473121643,
+      "learning_rate": 0.0001,
+      "loss": 1.6457,
+      "step": 6170
+    },
+    {
+      "epoch": 0.7088622135431624,
+      "grad_norm": 0.41569510102272034,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 6171
+    },
+    {
+      "epoch": 0.7089770834529895,
+      "grad_norm": 0.42229345440864563,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 6172
+    },
+    {
+      "epoch": 0.7090919533628166,
+      "grad_norm": 0.4784461557865143,
+      "learning_rate": 0.0001,
+      "loss": 1.727,
+      "step": 6173
+    },
+    {
+      "epoch": 0.7092068232726437,
+      "grad_norm": 0.40896356105804443,
+      "learning_rate": 0.0001,
+      "loss": 1.3902,
+      "step": 6174
+    },
+    {
+      "epoch": 0.7093216931824708,
+      "grad_norm": 0.4578010141849518,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 6175
+    },
+    {
+      "epoch": 0.709436563092298,
+      "grad_norm": 0.4804964065551758,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 6176
+    },
+    {
+      "epoch": 0.7095514330021251,
+      "grad_norm": 0.4631587266921997,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 6177
+    },
+    {
+      "epoch": 0.7096663029119522,
+      "grad_norm": 0.4732271134853363,
+      "learning_rate": 0.0001,
+      "loss": 1.7075,
+      "step": 6178
+    },
+    {
+      "epoch": 0.7097811728217793,
+      "grad_norm": 0.4341810643672943,
+      "learning_rate": 0.0001,
+      "loss": 1.6526,
+      "step": 6179
+    },
+    {
+      "epoch": 0.7098960427316064,
+      "grad_norm": 0.4346090853214264,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 6180
+    },
+    {
+      "epoch": 0.7100109126414336,
+      "grad_norm": 0.4299982190132141,
+      "learning_rate": 0.0001,
+      "loss": 1.4397,
+      "step": 6181
+    },
+    {
+      "epoch": 0.7101257825512607,
+      "grad_norm": 0.4570772647857666,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 6182
+    },
+    {
+      "epoch": 0.7102406524610878,
+      "grad_norm": 0.44084087014198303,
+      "learning_rate": 0.0001,
+      "loss": 1.664,
+      "step": 6183
+    },
+    {
+      "epoch": 0.7103555223709149,
+      "grad_norm": 0.4548969566822052,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 6184
+    },
+    {
+      "epoch": 0.710470392280742,
+      "grad_norm": 0.40112459659576416,
+      "learning_rate": 0.0001,
+      "loss": 1.4677,
+      "step": 6185
+    },
+    {
+      "epoch": 0.7105852621905692,
+      "grad_norm": 0.4894128143787384,
+      "learning_rate": 0.0001,
+      "loss": 1.7654,
+      "step": 6186
+    },
+    {
+      "epoch": 0.7107001321003963,
+      "grad_norm": 0.46307554841041565,
+      "learning_rate": 0.0001,
+      "loss": 1.731,
+      "step": 6187
+    },
+    {
+      "epoch": 0.7108150020102234,
+      "grad_norm": 0.44500499963760376,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 6188
+    },
+    {
+      "epoch": 0.7109298719200505,
+      "grad_norm": 0.46841320395469666,
+      "learning_rate": 0.0001,
+      "loss": 1.7267,
+      "step": 6189
+    },
+    {
+      "epoch": 0.7110447418298776,
+      "grad_norm": 0.452532023191452,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 6190
+    },
+    {
+      "epoch": 0.7111596117397048,
+      "grad_norm": 0.4364396929740906,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 6191
+    },
+    {
+      "epoch": 0.7112744816495319,
+      "grad_norm": 0.46478208899497986,
+      "learning_rate": 0.0001,
+      "loss": 1.7538,
+      "step": 6192
+    },
+    {
+      "epoch": 0.711389351559359,
+      "grad_norm": 0.5013893246650696,
+      "learning_rate": 0.0001,
+      "loss": 1.9132,
+      "step": 6193
+    },
+    {
+      "epoch": 0.7115042214691861,
+      "grad_norm": 0.4351854622364044,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 6194
+    },
+    {
+      "epoch": 0.7116190913790132,
+      "grad_norm": 0.4540832042694092,
+      "learning_rate": 0.0001,
+      "loss": 1.782,
+      "step": 6195
+    },
+    {
+      "epoch": 0.7117339612888404,
+      "grad_norm": 0.444970041513443,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 6196
+    },
+    {
+      "epoch": 0.7118488311986675,
+      "grad_norm": 0.4413135349750519,
+      "learning_rate": 0.0001,
+      "loss": 1.4519,
+      "step": 6197
+    },
+    {
+      "epoch": 0.7119637011084946,
+      "grad_norm": 0.4258001148700714,
+      "learning_rate": 0.0001,
+      "loss": 1.6815,
+      "step": 6198
+    },
+    {
+      "epoch": 0.7120785710183217,
+      "grad_norm": 0.4716213047504425,
+      "learning_rate": 0.0001,
+      "loss": 1.7128,
+      "step": 6199
+    },
+    {
+      "epoch": 0.7121934409281488,
+      "grad_norm": 0.4352003335952759,
+      "learning_rate": 0.0001,
+      "loss": 1.6439,
+      "step": 6200
+    },
+    {
+      "epoch": 0.712308310837976,
+      "grad_norm": 0.43441304564476013,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 6201
+    },
+    {
+      "epoch": 0.7124231807478031,
+      "grad_norm": 0.4781912565231323,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 6202
+    },
+    {
+      "epoch": 0.7125380506576302,
+      "grad_norm": 0.4544585347175598,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 6203
+    },
+    {
+      "epoch": 0.7126529205674573,
+      "grad_norm": 0.398957222700119,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 6204
+    },
+    {
+      "epoch": 0.7127677904772844,
+      "grad_norm": 0.44276830554008484,
+      "learning_rate": 0.0001,
+      "loss": 1.4976,
+      "step": 6205
+    },
+    {
+      "epoch": 0.7128826603871116,
+      "grad_norm": 0.44275638461112976,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 6206
+    },
+    {
+      "epoch": 0.7129975302969387,
+      "grad_norm": 0.5160487294197083,
+      "learning_rate": 0.0001,
+      "loss": 1.6689,
+      "step": 6207
+    },
+    {
+      "epoch": 0.7131124002067658,
+      "grad_norm": 0.44017454981803894,
+      "learning_rate": 0.0001,
+      "loss": 1.6119,
+      "step": 6208
+    },
+    {
+      "epoch": 0.7132272701165929,
+      "grad_norm": 0.4619143307209015,
+      "learning_rate": 0.0001,
+      "loss": 1.8167,
+      "step": 6209
+    },
+    {
+      "epoch": 0.71334214002642,
+      "grad_norm": 0.4661194086074829,
+      "learning_rate": 0.0001,
+      "loss": 1.5478,
+      "step": 6210
+    },
+    {
+      "epoch": 0.7134570099362472,
+      "grad_norm": 0.4605958163738251,
+      "learning_rate": 0.0001,
+      "loss": 1.7684,
+      "step": 6211
+    },
+    {
+      "epoch": 0.7135718798460743,
+      "grad_norm": 0.463900089263916,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 6212
+    },
+    {
+      "epoch": 0.7136867497559014,
+      "grad_norm": 0.48637497425079346,
+      "learning_rate": 0.0001,
+      "loss": 1.8437,
+      "step": 6213
+    },
+    {
+      "epoch": 0.7138016196657285,
+      "grad_norm": 0.40677204728126526,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 6214
+    },
+    {
+      "epoch": 0.7139164895755556,
+      "grad_norm": 0.5679096579551697,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 6215
+    },
+    {
+      "epoch": 0.7140313594853828,
+      "grad_norm": 0.4575541019439697,
+      "learning_rate": 0.0001,
+      "loss": 1.7186,
+      "step": 6216
+    },
+    {
+      "epoch": 0.7141462293952099,
+      "grad_norm": 0.4405764043331146,
+      "learning_rate": 0.0001,
+      "loss": 1.7805,
+      "step": 6217
+    },
+    {
+      "epoch": 0.714261099305037,
+      "grad_norm": 0.4583107531070709,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 6218
+    },
+    {
+      "epoch": 0.7143759692148641,
+      "grad_norm": 0.44853004813194275,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 6219
+    },
+    {
+      "epoch": 0.7144908391246912,
+      "grad_norm": 0.4117748439311981,
+      "learning_rate": 0.0001,
+      "loss": 1.4321,
+      "step": 6220
+    },
+    {
+      "epoch": 0.7146057090345184,
+      "grad_norm": 0.43110406398773193,
+      "learning_rate": 0.0001,
+      "loss": 1.483,
+      "step": 6221
+    },
+    {
+      "epoch": 0.7147205789443455,
+      "grad_norm": 0.46626466512680054,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 6222
+    },
+    {
+      "epoch": 0.7148354488541726,
+      "grad_norm": 0.4862065017223358,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 6223
+    },
+    {
+      "epoch": 0.7149503187639997,
+      "grad_norm": 0.4461599588394165,
+      "learning_rate": 0.0001,
+      "loss": 1.4846,
+      "step": 6224
+    },
+    {
+      "epoch": 0.7150651886738268,
+      "grad_norm": 0.4415801763534546,
+      "learning_rate": 0.0001,
+      "loss": 1.7068,
+      "step": 6225
+    },
+    {
+      "epoch": 0.715180058583654,
+      "grad_norm": 0.4292342960834503,
+      "learning_rate": 0.0001,
+      "loss": 1.5751,
+      "step": 6226
+    },
+    {
+      "epoch": 0.7152949284934811,
+      "grad_norm": 0.4549332857131958,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 6227
+    },
+    {
+      "epoch": 0.7154097984033082,
+      "grad_norm": 0.4223959147930145,
+      "learning_rate": 0.0001,
+      "loss": 1.4203,
+      "step": 6228
+    },
+    {
+      "epoch": 0.7155246683131353,
+      "grad_norm": 0.42485642433166504,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 6229
+    },
+    {
+      "epoch": 0.7156395382229624,
+      "grad_norm": 0.45224228501319885,
+      "learning_rate": 0.0001,
+      "loss": 1.4644,
+      "step": 6230
+    },
+    {
+      "epoch": 0.7157544081327896,
+      "grad_norm": 0.4319050908088684,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 6231
+    },
+    {
+      "epoch": 0.7158692780426167,
+      "grad_norm": 0.4529288113117218,
+      "learning_rate": 0.0001,
+      "loss": 1.7083,
+      "step": 6232
+    },
+    {
+      "epoch": 0.7159841479524438,
+      "grad_norm": 0.4431191682815552,
+      "learning_rate": 0.0001,
+      "loss": 1.5841,
+      "step": 6233
+    },
+    {
+      "epoch": 0.7160990178622709,
+      "grad_norm": 0.44510093331336975,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 6234
+    },
+    {
+      "epoch": 0.716213887772098,
+      "grad_norm": 0.43626534938812256,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 6235
+    },
+    {
+      "epoch": 0.7163287576819252,
+      "grad_norm": 0.4344061017036438,
+      "learning_rate": 0.0001,
+      "loss": 1.3115,
+      "step": 6236
+    },
+    {
+      "epoch": 0.7164436275917523,
+      "grad_norm": 0.5267593860626221,
+      "learning_rate": 0.0001,
+      "loss": 1.9086,
+      "step": 6237
+    },
+    {
+      "epoch": 0.7165584975015795,
+      "grad_norm": 0.4149724245071411,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 6238
+    },
+    {
+      "epoch": 0.7166733674114066,
+      "grad_norm": 0.4510195255279541,
+      "learning_rate": 0.0001,
+      "loss": 1.8163,
+      "step": 6239
+    },
+    {
+      "epoch": 0.7167882373212338,
+      "grad_norm": 0.4458303451538086,
+      "learning_rate": 0.0001,
+      "loss": 1.5324,
+      "step": 6240
+    },
+    {
+      "epoch": 0.7169031072310609,
+      "grad_norm": 0.4366897642612457,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 6241
+    },
+    {
+      "epoch": 0.717017977140888,
+      "grad_norm": 0.43107160925865173,
+      "learning_rate": 0.0001,
+      "loss": 1.5355,
+      "step": 6242
+    },
+    {
+      "epoch": 0.7171328470507151,
+      "grad_norm": 0.43756523728370667,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 6243
+    },
+    {
+      "epoch": 0.7172477169605422,
+      "grad_norm": 0.4852414131164551,
+      "learning_rate": 0.0001,
+      "loss": 1.8807,
+      "step": 6244
+    },
+    {
+      "epoch": 0.7173625868703694,
+      "grad_norm": 0.43804264068603516,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 6245
+    },
+    {
+      "epoch": 0.7174774567801965,
+      "grad_norm": 0.4281727075576782,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 6246
+    },
+    {
+      "epoch": 0.7175923266900236,
+      "grad_norm": 0.45637455582618713,
+      "learning_rate": 0.0001,
+      "loss": 1.6597,
+      "step": 6247
+    },
+    {
+      "epoch": 0.7177071965998507,
+      "grad_norm": 0.48942601680755615,
+      "learning_rate": 0.0001,
+      "loss": 1.7886,
+      "step": 6248
+    },
+    {
+      "epoch": 0.7178220665096778,
+      "grad_norm": 0.44699156284332275,
+      "learning_rate": 0.0001,
+      "loss": 1.6049,
+      "step": 6249
+    },
+    {
+      "epoch": 0.717936936419505,
+      "grad_norm": 0.45388197898864746,
+      "learning_rate": 0.0001,
+      "loss": 1.5923,
+      "step": 6250
+    },
+    {
+      "epoch": 0.7180518063293321,
+      "grad_norm": 0.4573292136192322,
+      "learning_rate": 0.0001,
+      "loss": 1.3832,
+      "step": 6251
+    },
+    {
+      "epoch": 0.7181666762391592,
+      "grad_norm": 0.4720889925956726,
+      "learning_rate": 0.0001,
+      "loss": 1.698,
+      "step": 6252
+    },
+    {
+      "epoch": 0.7182815461489863,
+      "grad_norm": 0.44124001264572144,
+      "learning_rate": 0.0001,
+      "loss": 1.468,
+      "step": 6253
+    },
+    {
+      "epoch": 0.7183964160588134,
+      "grad_norm": 0.48664864897727966,
+      "learning_rate": 0.0001,
+      "loss": 1.6452,
+      "step": 6254
+    },
+    {
+      "epoch": 0.7185112859686406,
+      "grad_norm": 0.44402679800987244,
+      "learning_rate": 0.0001,
+      "loss": 1.5786,
+      "step": 6255
+    },
+    {
+      "epoch": 0.7186261558784677,
+      "grad_norm": 0.46472397446632385,
+      "learning_rate": 0.0001,
+      "loss": 1.4484,
+      "step": 6256
+    },
+    {
+      "epoch": 0.7187410257882948,
+      "grad_norm": 0.49754804372787476,
+      "learning_rate": 0.0001,
+      "loss": 1.9748,
+      "step": 6257
+    },
+    {
+      "epoch": 0.7188558956981219,
+      "grad_norm": 0.439335435628891,
+      "learning_rate": 0.0001,
+      "loss": 1.695,
+      "step": 6258
+    },
+    {
+      "epoch": 0.718970765607949,
+      "grad_norm": 0.4374498426914215,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 6259
+    },
+    {
+      "epoch": 0.7190856355177762,
+      "grad_norm": 0.4089290201663971,
+      "learning_rate": 0.0001,
+      "loss": 1.4817,
+      "step": 6260
+    },
+    {
+      "epoch": 0.7192005054276033,
+      "grad_norm": 0.4366007447242737,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 6261
+    },
+    {
+      "epoch": 0.7193153753374304,
+      "grad_norm": 0.44272562861442566,
+      "learning_rate": 0.0001,
+      "loss": 1.4592,
+      "step": 6262
+    },
+    {
+      "epoch": 0.7194302452472575,
+      "grad_norm": 0.4497889280319214,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 6263
+    },
+    {
+      "epoch": 0.7195451151570846,
+      "grad_norm": 0.43780991435050964,
+      "learning_rate": 0.0001,
+      "loss": 1.5423,
+      "step": 6264
+    },
+    {
+      "epoch": 0.7196599850669118,
+      "grad_norm": 0.4534786343574524,
+      "learning_rate": 0.0001,
+      "loss": 1.7093,
+      "step": 6265
+    },
+    {
+      "epoch": 0.7197748549767389,
+      "grad_norm": 0.42275941371917725,
+      "learning_rate": 0.0001,
+      "loss": 1.4069,
+      "step": 6266
+    },
+    {
+      "epoch": 0.719889724886566,
+      "grad_norm": 0.47264930605888367,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 6267
+    },
+    {
+      "epoch": 0.7200045947963931,
+      "grad_norm": 0.4754311740398407,
+      "learning_rate": 0.0001,
+      "loss": 1.8423,
+      "step": 6268
+    },
+    {
+      "epoch": 0.7201194647062202,
+      "grad_norm": 0.4460528492927551,
+      "learning_rate": 0.0001,
+      "loss": 1.4606,
+      "step": 6269
+    },
+    {
+      "epoch": 0.7202343346160474,
+      "grad_norm": 0.43937763571739197,
+      "learning_rate": 0.0001,
+      "loss": 1.5779,
+      "step": 6270
+    },
+    {
+      "epoch": 0.7203492045258745,
+      "grad_norm": 0.4385170638561249,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 6271
+    },
+    {
+      "epoch": 0.7204640744357016,
+      "grad_norm": 0.4786663353443146,
+      "learning_rate": 0.0001,
+      "loss": 1.3311,
+      "step": 6272
+    },
+    {
+      "epoch": 0.7205789443455287,
+      "grad_norm": 0.45746463537216187,
+      "learning_rate": 0.0001,
+      "loss": 1.6661,
+      "step": 6273
+    },
+    {
+      "epoch": 0.7206938142553558,
+      "grad_norm": 0.4555452764034271,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 6274
+    },
+    {
+      "epoch": 0.720808684165183,
+      "grad_norm": 0.4532817006111145,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 6275
+    },
+    {
+      "epoch": 0.7209235540750101,
+      "grad_norm": 0.42538461089134216,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 6276
+    },
+    {
+      "epoch": 0.7210384239848372,
+      "grad_norm": 0.4184345006942749,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 6277
+    },
+    {
+      "epoch": 0.7211532938946643,
+      "grad_norm": 0.4728604853153229,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 6278
+    },
+    {
+      "epoch": 0.7212681638044914,
+      "grad_norm": 0.45347052812576294,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 6279
+    },
+    {
+      "epoch": 0.7213830337143186,
+      "grad_norm": 0.4866488575935364,
+      "learning_rate": 0.0001,
+      "loss": 1.7027,
+      "step": 6280
+    },
+    {
+      "epoch": 0.7214979036241457,
+      "grad_norm": 0.48550790548324585,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 6281
+    },
+    {
+      "epoch": 0.7216127735339728,
+      "grad_norm": 0.45319709181785583,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 6282
+    },
+    {
+      "epoch": 0.7217276434437999,
+      "grad_norm": 0.49187415838241577,
+      "learning_rate": 0.0001,
+      "loss": 1.7311,
+      "step": 6283
+    },
+    {
+      "epoch": 0.721842513353627,
+      "grad_norm": 0.4277302324771881,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 6284
+    },
+    {
+      "epoch": 0.7219573832634542,
+      "grad_norm": 0.4738651216030121,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 6285
+    },
+    {
+      "epoch": 0.7220722531732813,
+      "grad_norm": 0.4568105638027191,
+      "learning_rate": 0.0001,
+      "loss": 1.5111,
+      "step": 6286
+    },
+    {
+      "epoch": 0.7221871230831084,
+      "grad_norm": 0.41216906905174255,
+      "learning_rate": 0.0001,
+      "loss": 1.4644,
+      "step": 6287
+    },
+    {
+      "epoch": 0.7223019929929355,
+      "grad_norm": 0.4425976574420929,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 6288
+    },
+    {
+      "epoch": 0.7224168629027626,
+      "grad_norm": 0.42881807684898376,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 6289
+    },
+    {
+      "epoch": 0.7225317328125898,
+      "grad_norm": 0.41956013441085815,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 6290
+    },
+    {
+      "epoch": 0.7226466027224169,
+      "grad_norm": 0.46805068850517273,
+      "learning_rate": 0.0001,
+      "loss": 1.727,
+      "step": 6291
+    },
+    {
+      "epoch": 0.722761472632244,
+      "grad_norm": 0.432712197303772,
+      "learning_rate": 0.0001,
+      "loss": 1.4838,
+      "step": 6292
+    },
+    {
+      "epoch": 0.7228763425420711,
+      "grad_norm": 0.4402911067008972,
+      "learning_rate": 0.0001,
+      "loss": 1.7633,
+      "step": 6293
+    },
+    {
+      "epoch": 0.7229912124518982,
+      "grad_norm": 0.44642138481140137,
+      "learning_rate": 0.0001,
+      "loss": 1.5206,
+      "step": 6294
+    },
+    {
+      "epoch": 0.7231060823617254,
+      "grad_norm": 0.43020543456077576,
+      "learning_rate": 0.0001,
+      "loss": 1.7095,
+      "step": 6295
+    },
+    {
+      "epoch": 0.7232209522715525,
+      "grad_norm": 0.4489821195602417,
+      "learning_rate": 0.0001,
+      "loss": 1.5012,
+      "step": 6296
+    },
+    {
+      "epoch": 0.7233358221813796,
+      "grad_norm": 0.427055299282074,
+      "learning_rate": 0.0001,
+      "loss": 1.5981,
+      "step": 6297
+    },
+    {
+      "epoch": 0.7234506920912067,
+      "grad_norm": 0.41614022850990295,
+      "learning_rate": 0.0001,
+      "loss": 1.3573,
+      "step": 6298
+    },
+    {
+      "epoch": 0.7235655620010338,
+      "grad_norm": 0.4344491958618164,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 6299
+    },
+    {
+      "epoch": 0.723680431910861,
+      "grad_norm": 0.43843087553977966,
+      "learning_rate": 0.0001,
+      "loss": 1.1252,
+      "step": 6300
+    },
+    {
+      "epoch": 0.7237953018206881,
+      "grad_norm": 0.4700905680656433,
+      "learning_rate": 0.0001,
+      "loss": 1.48,
+      "step": 6301
+    },
+    {
+      "epoch": 0.7239101717305152,
+      "grad_norm": 0.48474234342575073,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 6302
+    },
+    {
+      "epoch": 0.7240250416403423,
+      "grad_norm": 0.44834601879119873,
+      "learning_rate": 0.0001,
+      "loss": 1.3831,
+      "step": 6303
+    },
+    {
+      "epoch": 0.7241399115501694,
+      "grad_norm": 0.4446815252304077,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 6304
+    },
+    {
+      "epoch": 0.7242547814599966,
+      "grad_norm": 0.4817836880683899,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 6305
+    },
+    {
+      "epoch": 0.7243696513698237,
+      "grad_norm": 0.5483472347259521,
+      "learning_rate": 0.0001,
+      "loss": 1.6487,
+      "step": 6306
+    },
+    {
+      "epoch": 0.7244845212796508,
+      "grad_norm": 0.42818188667297363,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 6307
+    },
+    {
+      "epoch": 0.7245993911894779,
+      "grad_norm": 0.49449649453163147,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 6308
+    },
+    {
+      "epoch": 0.724714261099305,
+      "grad_norm": 0.4369608163833618,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 6309
+    },
+    {
+      "epoch": 0.7248291310091322,
+      "grad_norm": 0.5256768465042114,
+      "learning_rate": 0.0001,
+      "loss": 1.583,
+      "step": 6310
+    },
+    {
+      "epoch": 0.7249440009189593,
+      "grad_norm": 0.44161832332611084,
+      "learning_rate": 0.0001,
+      "loss": 1.5239,
+      "step": 6311
+    },
+    {
+      "epoch": 0.7250588708287864,
+      "grad_norm": 0.4450240731239319,
+      "learning_rate": 0.0001,
+      "loss": 1.3923,
+      "step": 6312
+    },
+    {
+      "epoch": 0.7251737407386135,
+      "grad_norm": 0.45049697160720825,
+      "learning_rate": 0.0001,
+      "loss": 1.6166,
+      "step": 6313
+    },
+    {
+      "epoch": 0.7252886106484406,
+      "grad_norm": 0.43609675765037537,
+      "learning_rate": 0.0001,
+      "loss": 1.7061,
+      "step": 6314
+    },
+    {
+      "epoch": 0.7254034805582678,
+      "grad_norm": 0.516049325466156,
+      "learning_rate": 0.0001,
+      "loss": 1.7453,
+      "step": 6315
+    },
+    {
+      "epoch": 0.7255183504680949,
+      "grad_norm": 0.45927536487579346,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 6316
+    },
+    {
+      "epoch": 0.725633220377922,
+      "grad_norm": 0.4337046146392822,
+      "learning_rate": 0.0001,
+      "loss": 1.6932,
+      "step": 6317
+    },
+    {
+      "epoch": 0.7257480902877491,
+      "grad_norm": 0.44162049889564514,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 6318
+    },
+    {
+      "epoch": 0.7258629601975762,
+      "grad_norm": 0.4465755224227905,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 6319
+    },
+    {
+      "epoch": 0.7259778301074034,
+      "grad_norm": 0.4506109654903412,
+      "learning_rate": 0.0001,
+      "loss": 1.6735,
+      "step": 6320
+    },
+    {
+      "epoch": 0.7260927000172305,
+      "grad_norm": 0.45097821950912476,
+      "learning_rate": 0.0001,
+      "loss": 1.595,
+      "step": 6321
+    },
+    {
+      "epoch": 0.7262075699270576,
+      "grad_norm": 0.4603080749511719,
+      "learning_rate": 0.0001,
+      "loss": 1.7865,
+      "step": 6322
+    },
+    {
+      "epoch": 0.7263224398368847,
+      "grad_norm": 0.4264637231826782,
+      "learning_rate": 0.0001,
+      "loss": 1.7495,
+      "step": 6323
+    },
+    {
+      "epoch": 0.7264373097467118,
+      "grad_norm": 0.4358821213245392,
+      "learning_rate": 0.0001,
+      "loss": 1.7046,
+      "step": 6324
+    },
+    {
+      "epoch": 0.726552179656539,
+      "grad_norm": 0.4452092945575714,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 6325
+    },
+    {
+      "epoch": 0.7266670495663661,
+      "grad_norm": 0.5116655230522156,
+      "learning_rate": 0.0001,
+      "loss": 1.7254,
+      "step": 6326
+    },
+    {
+      "epoch": 0.7267819194761932,
+      "grad_norm": 0.4238274097442627,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 6327
+    },
+    {
+      "epoch": 0.7268967893860203,
+      "grad_norm": 0.4614119231700897,
+      "learning_rate": 0.0001,
+      "loss": 1.7239,
+      "step": 6328
+    },
+    {
+      "epoch": 0.7270116592958474,
+      "grad_norm": 0.47386687994003296,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 6329
+    },
+    {
+      "epoch": 0.7271265292056746,
+      "grad_norm": 0.4244793653488159,
+      "learning_rate": 0.0001,
+      "loss": 1.5649,
+      "step": 6330
+    },
+    {
+      "epoch": 0.7272413991155017,
+      "grad_norm": 0.44631174206733704,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 6331
+    },
+    {
+      "epoch": 0.7273562690253288,
+      "grad_norm": 0.4230101406574249,
+      "learning_rate": 0.0001,
+      "loss": 1.4192,
+      "step": 6332
+    },
+    {
+      "epoch": 0.7274711389351559,
+      "grad_norm": 0.42536047101020813,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 6333
+    },
+    {
+      "epoch": 0.727586008844983,
+      "grad_norm": 0.45973941683769226,
+      "learning_rate": 0.0001,
+      "loss": 1.7336,
+      "step": 6334
+    },
+    {
+      "epoch": 0.7277008787548102,
+      "grad_norm": 0.4620080888271332,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 6335
+    },
+    {
+      "epoch": 0.7278157486646373,
+      "grad_norm": 0.5099964737892151,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 6336
+    },
+    {
+      "epoch": 0.7279306185744644,
+      "grad_norm": 0.4300444722175598,
+      "learning_rate": 0.0001,
+      "loss": 1.3811,
+      "step": 6337
+    },
+    {
+      "epoch": 0.7280454884842915,
+      "grad_norm": 0.5135782957077026,
+      "learning_rate": 0.0001,
+      "loss": 1.8914,
+      "step": 6338
+    },
+    {
+      "epoch": 0.7281603583941186,
+      "grad_norm": 0.4119094908237457,
+      "learning_rate": 0.0001,
+      "loss": 1.5381,
+      "step": 6339
+    },
+    {
+      "epoch": 0.7282752283039458,
+      "grad_norm": 0.4423353374004364,
+      "learning_rate": 0.0001,
+      "loss": 1.6617,
+      "step": 6340
+    },
+    {
+      "epoch": 0.7283900982137729,
+      "grad_norm": 0.4462561011314392,
+      "learning_rate": 0.0001,
+      "loss": 1.6236,
+      "step": 6341
+    },
+    {
+      "epoch": 0.7285049681236,
+      "grad_norm": 0.47333210706710815,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 6342
+    },
+    {
+      "epoch": 0.7286198380334271,
+      "grad_norm": 0.46242082118988037,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 6343
+    },
+    {
+      "epoch": 0.7287347079432542,
+      "grad_norm": 0.4549300968647003,
+      "learning_rate": 0.0001,
+      "loss": 1.6149,
+      "step": 6344
+    },
+    {
+      "epoch": 0.7288495778530814,
+      "grad_norm": 0.46361663937568665,
+      "learning_rate": 0.0001,
+      "loss": 1.7257,
+      "step": 6345
+    },
+    {
+      "epoch": 0.7289644477629085,
+      "grad_norm": 0.4788486361503601,
+      "learning_rate": 0.0001,
+      "loss": 1.6031,
+      "step": 6346
+    },
+    {
+      "epoch": 0.7290793176727356,
+      "grad_norm": 0.4863881766796112,
+      "learning_rate": 0.0001,
+      "loss": 1.8181,
+      "step": 6347
+    },
+    {
+      "epoch": 0.7291941875825627,
+      "grad_norm": 0.4203285276889801,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 6348
+    },
+    {
+      "epoch": 0.7293090574923898,
+      "grad_norm": 0.4476950168609619,
+      "learning_rate": 0.0001,
+      "loss": 1.5288,
+      "step": 6349
+    },
+    {
+      "epoch": 0.729423927402217,
+      "grad_norm": 0.4235599935054779,
+      "learning_rate": 0.0001,
+      "loss": 1.3824,
+      "step": 6350
+    },
+    {
+      "epoch": 0.7295387973120441,
+      "grad_norm": 0.4462408423423767,
+      "learning_rate": 0.0001,
+      "loss": 1.5426,
+      "step": 6351
+    },
+    {
+      "epoch": 0.7296536672218712,
+      "grad_norm": 0.40480706095695496,
+      "learning_rate": 0.0001,
+      "loss": 1.4957,
+      "step": 6352
+    },
+    {
+      "epoch": 0.7297685371316983,
+      "grad_norm": 0.4449676275253296,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 6353
+    },
+    {
+      "epoch": 0.7298834070415254,
+      "grad_norm": 0.43793755769729614,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 6354
+    },
+    {
+      "epoch": 0.7299982769513526,
+      "grad_norm": 0.45420295000076294,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 6355
+    },
+    {
+      "epoch": 0.7301131468611797,
+      "grad_norm": 0.5009284615516663,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 6356
+    },
+    {
+      "epoch": 0.7302280167710068,
+      "grad_norm": 0.5149348974227905,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 6357
+    },
+    {
+      "epoch": 0.7303428866808339,
+      "grad_norm": 0.4348343014717102,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 6358
+    },
+    {
+      "epoch": 0.730457756590661,
+      "grad_norm": 0.44643691182136536,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 6359
+    },
+    {
+      "epoch": 0.7305726265004882,
+      "grad_norm": 0.42023977637290955,
+      "learning_rate": 0.0001,
+      "loss": 1.5319,
+      "step": 6360
+    },
+    {
+      "epoch": 0.7306874964103153,
+      "grad_norm": 0.49339669942855835,
+      "learning_rate": 0.0001,
+      "loss": 1.7135,
+      "step": 6361
+    },
+    {
+      "epoch": 0.7308023663201424,
+      "grad_norm": 0.44311895966529846,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 6362
+    },
+    {
+      "epoch": 0.7309172362299695,
+      "grad_norm": 0.4121879041194916,
+      "learning_rate": 0.0001,
+      "loss": 1.3989,
+      "step": 6363
+    },
+    {
+      "epoch": 0.7310321061397966,
+      "grad_norm": 0.43048325181007385,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 6364
+    },
+    {
+      "epoch": 0.7311469760496238,
+      "grad_norm": 0.4717782735824585,
+      "learning_rate": 0.0001,
+      "loss": 1.6826,
+      "step": 6365
+    },
+    {
+      "epoch": 0.7312618459594509,
+      "grad_norm": 0.46323728561401367,
+      "learning_rate": 0.0001,
+      "loss": 1.6805,
+      "step": 6366
+    },
+    {
+      "epoch": 0.731376715869278,
+      "grad_norm": 0.4346836805343628,
+      "learning_rate": 0.0001,
+      "loss": 1.5547,
+      "step": 6367
+    },
+    {
+      "epoch": 0.7314915857791051,
+      "grad_norm": 0.47042396664619446,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 6368
+    },
+    {
+      "epoch": 0.7316064556889322,
+      "grad_norm": 0.44026100635528564,
+      "learning_rate": 0.0001,
+      "loss": 1.618,
+      "step": 6369
+    },
+    {
+      "epoch": 0.7317213255987594,
+      "grad_norm": 0.428864449262619,
+      "learning_rate": 0.0001,
+      "loss": 1.4316,
+      "step": 6370
+    },
+    {
+      "epoch": 0.7318361955085865,
+      "grad_norm": 0.4537869989871979,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 6371
+    },
+    {
+      "epoch": 0.7319510654184136,
+      "grad_norm": 0.45302248001098633,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 6372
+    },
+    {
+      "epoch": 0.7320659353282407,
+      "grad_norm": 0.4309695363044739,
+      "learning_rate": 0.0001,
+      "loss": 1.6486,
+      "step": 6373
+    },
+    {
+      "epoch": 0.7321808052380678,
+      "grad_norm": 0.4645009934902191,
+      "learning_rate": 0.0001,
+      "loss": 1.5925,
+      "step": 6374
+    },
+    {
+      "epoch": 0.7322956751478951,
+      "grad_norm": 0.42796799540519714,
+      "learning_rate": 0.0001,
+      "loss": 1.6142,
+      "step": 6375
+    },
+    {
+      "epoch": 0.7324105450577222,
+      "grad_norm": 0.4301031231880188,
+      "learning_rate": 0.0001,
+      "loss": 1.4068,
+      "step": 6376
+    },
+    {
+      "epoch": 0.7325254149675493,
+      "grad_norm": 0.4349769055843353,
+      "learning_rate": 0.0001,
+      "loss": 1.5291,
+      "step": 6377
+    },
+    {
+      "epoch": 0.7326402848773764,
+      "grad_norm": 0.4733826518058777,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 6378
+    },
+    {
+      "epoch": 0.7327551547872035,
+      "grad_norm": 0.4366433322429657,
+      "learning_rate": 0.0001,
+      "loss": 1.5781,
+      "step": 6379
+    },
+    {
+      "epoch": 0.7328700246970307,
+      "grad_norm": 0.47278228402137756,
+      "learning_rate": 0.0001,
+      "loss": 1.6445,
+      "step": 6380
+    },
+    {
+      "epoch": 0.7329848946068578,
+      "grad_norm": 0.4604305922985077,
+      "learning_rate": 0.0001,
+      "loss": 1.7437,
+      "step": 6381
+    },
+    {
+      "epoch": 0.7330997645166849,
+      "grad_norm": 0.45044025778770447,
+      "learning_rate": 0.0001,
+      "loss": 1.649,
+      "step": 6382
+    },
+    {
+      "epoch": 0.733214634426512,
+      "grad_norm": 0.4326942563056946,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 6383
+    },
+    {
+      "epoch": 0.7333295043363391,
+      "grad_norm": 0.4268265664577484,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 6384
+    },
+    {
+      "epoch": 0.7334443742461663,
+      "grad_norm": 0.49201422929763794,
+      "learning_rate": 0.0001,
+      "loss": 1.8518,
+      "step": 6385
+    },
+    {
+      "epoch": 0.7335592441559934,
+      "grad_norm": 0.4553898274898529,
+      "learning_rate": 0.0001,
+      "loss": 1.8142,
+      "step": 6386
+    },
+    {
+      "epoch": 0.7336741140658205,
+      "grad_norm": 0.4906349778175354,
+      "learning_rate": 0.0001,
+      "loss": 1.8773,
+      "step": 6387
+    },
+    {
+      "epoch": 0.7337889839756476,
+      "grad_norm": 0.4062640070915222,
+      "learning_rate": 0.0001,
+      "loss": 1.4207,
+      "step": 6388
+    },
+    {
+      "epoch": 0.7339038538854747,
+      "grad_norm": 0.4532264173030853,
+      "learning_rate": 0.0001,
+      "loss": 1.7094,
+      "step": 6389
+    },
+    {
+      "epoch": 0.7340187237953019,
+      "grad_norm": 0.4875713586807251,
+      "learning_rate": 0.0001,
+      "loss": 1.835,
+      "step": 6390
+    },
+    {
+      "epoch": 0.734133593705129,
+      "grad_norm": 0.4259282052516937,
+      "learning_rate": 0.0001,
+      "loss": 1.4534,
+      "step": 6391
+    },
+    {
+      "epoch": 0.7342484636149561,
+      "grad_norm": 0.4660018980503082,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 6392
+    },
+    {
+      "epoch": 0.7343633335247832,
+      "grad_norm": 0.4709480106830597,
+      "learning_rate": 0.0001,
+      "loss": 1.7843,
+      "step": 6393
+    },
+    {
+      "epoch": 0.7344782034346103,
+      "grad_norm": 0.4932401180267334,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 6394
+    },
+    {
+      "epoch": 0.7345930733444375,
+      "grad_norm": 0.4418211281299591,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 6395
+    },
+    {
+      "epoch": 0.7347079432542646,
+      "grad_norm": 0.4617857038974762,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 6396
+    },
+    {
+      "epoch": 0.7348228131640917,
+      "grad_norm": 0.44936829805374146,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 6397
+    },
+    {
+      "epoch": 0.7349376830739188,
+      "grad_norm": 0.49701419472694397,
+      "learning_rate": 0.0001,
+      "loss": 1.7478,
+      "step": 6398
+    },
+    {
+      "epoch": 0.7350525529837459,
+      "grad_norm": 0.42285898327827454,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 6399
+    },
+    {
+      "epoch": 0.7351674228935731,
+      "grad_norm": 0.46721500158309937,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 6400
+    },
+    {
+      "epoch": 0.7352822928034002,
+      "grad_norm": 0.4647897779941559,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 6401
+    },
+    {
+      "epoch": 0.7353971627132273,
+      "grad_norm": 0.48368552327156067,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 6402
+    },
+    {
+      "epoch": 0.7355120326230544,
+      "grad_norm": 0.4766364097595215,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 6403
+    },
+    {
+      "epoch": 0.7356269025328815,
+      "grad_norm": 0.4311745762825012,
+      "learning_rate": 0.0001,
+      "loss": 1.5385,
+      "step": 6404
+    },
+    {
+      "epoch": 0.7357417724427087,
+      "grad_norm": 0.424921452999115,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 6405
+    },
+    {
+      "epoch": 0.7358566423525358,
+      "grad_norm": 0.44613373279571533,
+      "learning_rate": 0.0001,
+      "loss": 1.6768,
+      "step": 6406
+    },
+    {
+      "epoch": 0.7359715122623629,
+      "grad_norm": 0.52595055103302,
+      "learning_rate": 0.0001,
+      "loss": 1.8678,
+      "step": 6407
+    },
+    {
+      "epoch": 0.73608638217219,
+      "grad_norm": 0.4704437851905823,
+      "learning_rate": 0.0001,
+      "loss": 1.7375,
+      "step": 6408
+    },
+    {
+      "epoch": 0.7362012520820171,
+      "grad_norm": 0.4089406132698059,
+      "learning_rate": 0.0001,
+      "loss": 1.5,
+      "step": 6409
+    },
+    {
+      "epoch": 0.7363161219918443,
+      "grad_norm": 0.4442186951637268,
+      "learning_rate": 0.0001,
+      "loss": 1.4664,
+      "step": 6410
+    },
+    {
+      "epoch": 0.7364309919016714,
+      "grad_norm": 0.4903067350387573,
+      "learning_rate": 0.0001,
+      "loss": 1.7815,
+      "step": 6411
+    },
+    {
+      "epoch": 0.7365458618114985,
+      "grad_norm": 0.47199535369873047,
+      "learning_rate": 0.0001,
+      "loss": 1.7579,
+      "step": 6412
+    },
+    {
+      "epoch": 0.7366607317213256,
+      "grad_norm": 0.4860610365867615,
+      "learning_rate": 0.0001,
+      "loss": 1.7036,
+      "step": 6413
+    },
+    {
+      "epoch": 0.7367756016311527,
+      "grad_norm": 0.44982248544692993,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 6414
+    },
+    {
+      "epoch": 0.7368904715409799,
+      "grad_norm": 0.41628536581993103,
+      "learning_rate": 0.0001,
+      "loss": 1.3993,
+      "step": 6415
+    },
+    {
+      "epoch": 0.737005341450807,
+      "grad_norm": 0.4167647063732147,
+      "learning_rate": 0.0001,
+      "loss": 1.4185,
+      "step": 6416
+    },
+    {
+      "epoch": 0.7371202113606341,
+      "grad_norm": 0.44325876235961914,
+      "learning_rate": 0.0001,
+      "loss": 1.6582,
+      "step": 6417
+    },
+    {
+      "epoch": 0.7372350812704612,
+      "grad_norm": 0.4645753800868988,
+      "learning_rate": 0.0001,
+      "loss": 1.44,
+      "step": 6418
+    },
+    {
+      "epoch": 0.7373499511802883,
+      "grad_norm": 0.41918548941612244,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 6419
+    },
+    {
+      "epoch": 0.7374648210901155,
+      "grad_norm": 0.4537114202976227,
+      "learning_rate": 0.0001,
+      "loss": 1.664,
+      "step": 6420
+    },
+    {
+      "epoch": 0.7375796909999426,
+      "grad_norm": 0.4690401554107666,
+      "learning_rate": 0.0001,
+      "loss": 1.7344,
+      "step": 6421
+    },
+    {
+      "epoch": 0.7376945609097697,
+      "grad_norm": 0.5302848815917969,
+      "learning_rate": 0.0001,
+      "loss": 1.8387,
+      "step": 6422
+    },
+    {
+      "epoch": 0.7378094308195968,
+      "grad_norm": 0.49464482069015503,
+      "learning_rate": 0.0001,
+      "loss": 1.7305,
+      "step": 6423
+    },
+    {
+      "epoch": 0.7379243007294239,
+      "grad_norm": 0.48099836707115173,
+      "learning_rate": 0.0001,
+      "loss": 1.5702,
+      "step": 6424
+    },
+    {
+      "epoch": 0.7380391706392511,
+      "grad_norm": 0.4626838266849518,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 6425
+    },
+    {
+      "epoch": 0.7381540405490782,
+      "grad_norm": 0.5171931385993958,
+      "learning_rate": 0.0001,
+      "loss": 1.562,
+      "step": 6426
+    },
+    {
+      "epoch": 0.7382689104589053,
+      "grad_norm": 0.46404412388801575,
+      "learning_rate": 0.0001,
+      "loss": 1.767,
+      "step": 6427
+    },
+    {
+      "epoch": 0.7383837803687324,
+      "grad_norm": 0.44521862268447876,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 6428
+    },
+    {
+      "epoch": 0.7384986502785595,
+      "grad_norm": 0.4961501359939575,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 6429
+    },
+    {
+      "epoch": 0.7386135201883867,
+      "grad_norm": 0.43013089895248413,
+      "learning_rate": 0.0001,
+      "loss": 1.5063,
+      "step": 6430
+    },
+    {
+      "epoch": 0.7387283900982138,
+      "grad_norm": 0.43474483489990234,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 6431
+    },
+    {
+      "epoch": 0.7388432600080409,
+      "grad_norm": 0.4235991835594177,
+      "learning_rate": 0.0001,
+      "loss": 1.6437,
+      "step": 6432
+    },
+    {
+      "epoch": 0.738958129917868,
+      "grad_norm": 0.45976969599723816,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 6433
+    },
+    {
+      "epoch": 0.7390729998276951,
+      "grad_norm": 0.5007357597351074,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 6434
+    },
+    {
+      "epoch": 0.7391878697375223,
+      "grad_norm": 0.4222777783870697,
+      "learning_rate": 0.0001,
+      "loss": 1.4098,
+      "step": 6435
+    },
+    {
+      "epoch": 0.7393027396473494,
+      "grad_norm": 0.4068715274333954,
+      "learning_rate": 0.0001,
+      "loss": 1.2175,
+      "step": 6436
+    },
+    {
+      "epoch": 0.7394176095571765,
+      "grad_norm": 0.4571983516216278,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 6437
+    },
+    {
+      "epoch": 0.7395324794670036,
+      "grad_norm": 0.4495130181312561,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 6438
+    },
+    {
+      "epoch": 0.7396473493768307,
+      "grad_norm": 0.48614737391471863,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 6439
+    },
+    {
+      "epoch": 0.7397622192866579,
+      "grad_norm": 0.48480525612831116,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 6440
+    },
+    {
+      "epoch": 0.739877089196485,
+      "grad_norm": 0.46960780024528503,
+      "learning_rate": 0.0001,
+      "loss": 1.3502,
+      "step": 6441
+    },
+    {
+      "epoch": 0.7399919591063121,
+      "grad_norm": 0.45793652534484863,
+      "learning_rate": 0.0001,
+      "loss": 1.6815,
+      "step": 6442
+    },
+    {
+      "epoch": 0.7401068290161392,
+      "grad_norm": 0.46224090456962585,
+      "learning_rate": 0.0001,
+      "loss": 1.6644,
+      "step": 6443
+    },
+    {
+      "epoch": 0.7402216989259663,
+      "grad_norm": 0.41686657071113586,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 6444
+    },
+    {
+      "epoch": 0.7403365688357935,
+      "grad_norm": 0.42895323038101196,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 6445
+    },
+    {
+      "epoch": 0.7404514387456206,
+      "grad_norm": 0.42914414405822754,
+      "learning_rate": 0.0001,
+      "loss": 1.6159,
+      "step": 6446
+    },
+    {
+      "epoch": 0.7405663086554477,
+      "grad_norm": 0.4458409249782562,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 6447
+    },
+    {
+      "epoch": 0.7406811785652748,
+      "grad_norm": 0.4255877435207367,
+      "learning_rate": 0.0001,
+      "loss": 1.4395,
+      "step": 6448
+    },
+    {
+      "epoch": 0.7407960484751019,
+      "grad_norm": 0.4231339991092682,
+      "learning_rate": 0.0001,
+      "loss": 1.495,
+      "step": 6449
+    },
+    {
+      "epoch": 0.7409109183849291,
+      "grad_norm": 0.4323672652244568,
+      "learning_rate": 0.0001,
+      "loss": 1.4857,
+      "step": 6450
+    },
+    {
+      "epoch": 0.7410257882947562,
+      "grad_norm": 0.4640507996082306,
+      "learning_rate": 0.0001,
+      "loss": 1.4276,
+      "step": 6451
+    },
+    {
+      "epoch": 0.7411406582045833,
+      "grad_norm": 0.442341685295105,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 6452
+    },
+    {
+      "epoch": 0.7412555281144104,
+      "grad_norm": 0.44191497564315796,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 6453
+    },
+    {
+      "epoch": 0.7413703980242375,
+      "grad_norm": 0.4533472955226898,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 6454
+    },
+    {
+      "epoch": 0.7414852679340647,
+      "grad_norm": 0.45057639479637146,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 6455
+    },
+    {
+      "epoch": 0.7416001378438918,
+      "grad_norm": 0.5106820464134216,
+      "learning_rate": 0.0001,
+      "loss": 1.7923,
+      "step": 6456
+    },
+    {
+      "epoch": 0.7417150077537189,
+      "grad_norm": 0.4512089788913727,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 6457
+    },
+    {
+      "epoch": 0.741829877663546,
+      "grad_norm": 0.44131678342819214,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 6458
+    },
+    {
+      "epoch": 0.7419447475733731,
+      "grad_norm": 0.4136804938316345,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 6459
+    },
+    {
+      "epoch": 0.7420596174832003,
+      "grad_norm": 0.48369288444519043,
+      "learning_rate": 0.0001,
+      "loss": 1.7312,
+      "step": 6460
+    },
+    {
+      "epoch": 0.7421744873930274,
+      "grad_norm": 0.4654393792152405,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 6461
+    },
+    {
+      "epoch": 0.7422893573028545,
+      "grad_norm": 0.457107275724411,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 6462
+    },
+    {
+      "epoch": 0.7424042272126816,
+      "grad_norm": 0.46595972776412964,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 6463
+    },
+    {
+      "epoch": 0.7425190971225087,
+      "grad_norm": 0.43557512760162354,
+      "learning_rate": 0.0001,
+      "loss": 1.3404,
+      "step": 6464
+    },
+    {
+      "epoch": 0.7426339670323359,
+      "grad_norm": 0.4468023180961609,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 6465
+    },
+    {
+      "epoch": 0.742748836942163,
+      "grad_norm": 0.4448022246360779,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 6466
+    },
+    {
+      "epoch": 0.7428637068519901,
+      "grad_norm": 0.461955726146698,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 6467
+    },
+    {
+      "epoch": 0.7429785767618172,
+      "grad_norm": 0.47350478172302246,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 6468
+    },
+    {
+      "epoch": 0.7430934466716443,
+      "grad_norm": 0.506061851978302,
+      "learning_rate": 0.0001,
+      "loss": 1.9245,
+      "step": 6469
+    },
+    {
+      "epoch": 0.7432083165814715,
+      "grad_norm": 0.48322594165802,
+      "learning_rate": 0.0001,
+      "loss": 1.8228,
+      "step": 6470
+    },
+    {
+      "epoch": 0.7433231864912986,
+      "grad_norm": 0.4514816701412201,
+      "learning_rate": 0.0001,
+      "loss": 1.7976,
+      "step": 6471
+    },
+    {
+      "epoch": 0.7434380564011257,
+      "grad_norm": 0.45972806215286255,
+      "learning_rate": 0.0001,
+      "loss": 1.676,
+      "step": 6472
+    },
+    {
+      "epoch": 0.7435529263109528,
+      "grad_norm": 0.4653480350971222,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 6473
+    },
+    {
+      "epoch": 0.7436677962207799,
+      "grad_norm": 0.4564400315284729,
+      "learning_rate": 0.0001,
+      "loss": 1.4732,
+      "step": 6474
+    },
+    {
+      "epoch": 0.7437826661306071,
+      "grad_norm": 0.44110190868377686,
+      "learning_rate": 0.0001,
+      "loss": 1.4867,
+      "step": 6475
+    },
+    {
+      "epoch": 0.7438975360404342,
+      "grad_norm": 0.4426039159297943,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 6476
+    },
+    {
+      "epoch": 0.7440124059502613,
+      "grad_norm": 0.4515429735183716,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 6477
+    },
+    {
+      "epoch": 0.7441272758600884,
+      "grad_norm": 0.4343465268611908,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 6478
+    },
+    {
+      "epoch": 0.7442421457699155,
+      "grad_norm": 0.448831170797348,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 6479
+    },
+    {
+      "epoch": 0.7443570156797427,
+      "grad_norm": 0.45811012387275696,
+      "learning_rate": 0.0001,
+      "loss": 1.6288,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7444718855895698,
+      "grad_norm": 0.48360949754714966,
+      "learning_rate": 0.0001,
+      "loss": 1.6076,
+      "step": 6481
+    },
+    {
+      "epoch": 0.7445867554993969,
+      "grad_norm": 0.4996730089187622,
+      "learning_rate": 0.0001,
+      "loss": 1.8236,
+      "step": 6482
+    },
+    {
+      "epoch": 0.744701625409224,
+      "grad_norm": 0.4542221426963806,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 6483
+    },
+    {
+      "epoch": 0.7448164953190511,
+      "grad_norm": 0.4285677671432495,
+      "learning_rate": 0.0001,
+      "loss": 1.374,
+      "step": 6484
+    },
+    {
+      "epoch": 0.7449313652288783,
+      "grad_norm": 0.48036250472068787,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 6485
+    },
+    {
+      "epoch": 0.7450462351387054,
+      "grad_norm": 0.4445507824420929,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 6486
+    },
+    {
+      "epoch": 0.7451611050485325,
+      "grad_norm": 0.4675546884536743,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 6487
+    },
+    {
+      "epoch": 0.7452759749583596,
+      "grad_norm": 0.4238932132720947,
+      "learning_rate": 0.0001,
+      "loss": 1.6086,
+      "step": 6488
+    },
+    {
+      "epoch": 0.7453908448681867,
+      "grad_norm": 0.4289308786392212,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 6489
+    },
+    {
+      "epoch": 0.7455057147780139,
+      "grad_norm": 0.449756383895874,
+      "learning_rate": 0.0001,
+      "loss": 1.6866,
+      "step": 6490
+    },
+    {
+      "epoch": 0.745620584687841,
+      "grad_norm": 0.4513997435569763,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 6491
+    },
+    {
+      "epoch": 0.7457354545976681,
+      "grad_norm": 0.49391379952430725,
+      "learning_rate": 0.0001,
+      "loss": 1.7308,
+      "step": 6492
+    },
+    {
+      "epoch": 0.7458503245074952,
+      "grad_norm": 0.45236408710479736,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 6493
+    },
+    {
+      "epoch": 0.7459651944173223,
+      "grad_norm": 0.466229110956192,
+      "learning_rate": 0.0001,
+      "loss": 1.7665,
+      "step": 6494
+    },
+    {
+      "epoch": 0.7460800643271495,
+      "grad_norm": 0.44140058755874634,
+      "learning_rate": 0.0001,
+      "loss": 1.572,
+      "step": 6495
+    },
+    {
+      "epoch": 0.7461949342369766,
+      "grad_norm": 0.4668871760368347,
+      "learning_rate": 0.0001,
+      "loss": 1.7177,
+      "step": 6496
+    },
+    {
+      "epoch": 0.7463098041468037,
+      "grad_norm": 0.43868231773376465,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 6497
+    },
+    {
+      "epoch": 0.7464246740566308,
+      "grad_norm": 0.517144501209259,
+      "learning_rate": 0.0001,
+      "loss": 1.7245,
+      "step": 6498
+    },
+    {
+      "epoch": 0.7465395439664579,
+      "grad_norm": 0.4234006404876709,
+      "learning_rate": 0.0001,
+      "loss": 1.4977,
+      "step": 6499
+    },
+    {
+      "epoch": 0.7466544138762851,
+      "grad_norm": 0.44216203689575195,
+      "learning_rate": 0.0001,
+      "loss": 1.549,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7467692837861122,
+      "grad_norm": 0.43986350297927856,
+      "learning_rate": 0.0001,
+      "loss": 1.6273,
+      "step": 6501
+    },
+    {
+      "epoch": 0.7468841536959393,
+      "grad_norm": 0.4326637387275696,
+      "learning_rate": 0.0001,
+      "loss": 1.4611,
+      "step": 6502
+    },
+    {
+      "epoch": 0.7469990236057664,
+      "grad_norm": 0.43953758478164673,
+      "learning_rate": 0.0001,
+      "loss": 1.6316,
+      "step": 6503
+    },
+    {
+      "epoch": 0.7471138935155935,
+      "grad_norm": 0.44682297110557556,
+      "learning_rate": 0.0001,
+      "loss": 1.5408,
+      "step": 6504
+    },
+    {
+      "epoch": 0.7472287634254207,
+      "grad_norm": 0.4634048342704773,
+      "learning_rate": 0.0001,
+      "loss": 1.68,
+      "step": 6505
+    },
+    {
+      "epoch": 0.7473436333352478,
+      "grad_norm": 0.43683484196662903,
+      "learning_rate": 0.0001,
+      "loss": 1.6397,
+      "step": 6506
+    },
+    {
+      "epoch": 0.7474585032450749,
+      "grad_norm": 0.42723357677459717,
+      "learning_rate": 0.0001,
+      "loss": 1.4604,
+      "step": 6507
+    },
+    {
+      "epoch": 0.747573373154902,
+      "grad_norm": 0.49617254734039307,
+      "learning_rate": 0.0001,
+      "loss": 1.8581,
+      "step": 6508
+    },
+    {
+      "epoch": 0.7476882430647291,
+      "grad_norm": 0.46171995997428894,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 6509
+    },
+    {
+      "epoch": 0.7478031129745563,
+      "grad_norm": 0.4606563150882721,
+      "learning_rate": 0.0001,
+      "loss": 1.4719,
+      "step": 6510
+    },
+    {
+      "epoch": 0.7479179828843834,
+      "grad_norm": 0.47616055607795715,
+      "learning_rate": 0.0001,
+      "loss": 1.7907,
+      "step": 6511
+    },
+    {
+      "epoch": 0.7480328527942106,
+      "grad_norm": 0.42046403884887695,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 6512
+    },
+    {
+      "epoch": 0.7481477227040377,
+      "grad_norm": 0.41847604513168335,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 6513
+    },
+    {
+      "epoch": 0.7482625926138649,
+      "grad_norm": 0.45703554153442383,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 6514
+    },
+    {
+      "epoch": 0.748377462523692,
+      "grad_norm": 0.40654999017715454,
+      "learning_rate": 0.0001,
+      "loss": 1.2961,
+      "step": 6515
+    },
+    {
+      "epoch": 0.7484923324335191,
+      "grad_norm": 0.4782845377922058,
+      "learning_rate": 0.0001,
+      "loss": 1.6385,
+      "step": 6516
+    },
+    {
+      "epoch": 0.7486072023433462,
+      "grad_norm": 0.4496891498565674,
+      "learning_rate": 0.0001,
+      "loss": 1.4721,
+      "step": 6517
+    },
+    {
+      "epoch": 0.7487220722531733,
+      "grad_norm": 0.4883591830730438,
+      "learning_rate": 0.0001,
+      "loss": 1.6914,
+      "step": 6518
+    },
+    {
+      "epoch": 0.7488369421630005,
+      "grad_norm": 0.5101348161697388,
+      "learning_rate": 0.0001,
+      "loss": 1.8276,
+      "step": 6519
+    },
+    {
+      "epoch": 0.7489518120728276,
+      "grad_norm": 0.47420623898506165,
+      "learning_rate": 0.0001,
+      "loss": 1.7383,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7490666819826547,
+      "grad_norm": 0.544798731803894,
+      "learning_rate": 0.0001,
+      "loss": 1.2877,
+      "step": 6521
+    },
+    {
+      "epoch": 0.7491815518924818,
+      "grad_norm": 0.496698796749115,
+      "learning_rate": 0.0001,
+      "loss": 1.7508,
+      "step": 6522
+    },
+    {
+      "epoch": 0.7492964218023089,
+      "grad_norm": 0.42780181765556335,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 6523
+    },
+    {
+      "epoch": 0.749411291712136,
+      "grad_norm": 0.4542038142681122,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 6524
+    },
+    {
+      "epoch": 0.7495261616219632,
+      "grad_norm": 0.4443798363208771,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 6525
+    },
+    {
+      "epoch": 0.7496410315317903,
+      "grad_norm": 0.4632874131202698,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 6526
+    },
+    {
+      "epoch": 0.7497559014416174,
+      "grad_norm": 0.4470142424106598,
+      "learning_rate": 0.0001,
+      "loss": 1.6075,
+      "step": 6527
+    },
+    {
+      "epoch": 0.7498707713514445,
+      "grad_norm": 0.4853494465351105,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 6528
+    },
+    {
+      "epoch": 0.7499856412612717,
+      "grad_norm": 0.43715977668762207,
+      "learning_rate": 0.0001,
+      "loss": 1.3894,
+      "step": 6529
+    },
+    {
+      "epoch": 0.7501005111710988,
+      "grad_norm": 0.43983104825019836,
+      "learning_rate": 0.0001,
+      "loss": 1.6225,
+      "step": 6530
+    },
+    {
+      "epoch": 0.7502153810809259,
+      "grad_norm": 0.4784538745880127,
+      "learning_rate": 0.0001,
+      "loss": 1.7794,
+      "step": 6531
+    },
+    {
+      "epoch": 0.750330250990753,
+      "grad_norm": 0.47763192653656006,
+      "learning_rate": 0.0001,
+      "loss": 1.85,
+      "step": 6532
+    },
+    {
+      "epoch": 0.7504451209005801,
+      "grad_norm": 0.4692215025424957,
+      "learning_rate": 0.0001,
+      "loss": 1.7338,
+      "step": 6533
+    },
+    {
+      "epoch": 0.7505599908104073,
+      "grad_norm": 0.4766184091567993,
+      "learning_rate": 0.0001,
+      "loss": 1.8508,
+      "step": 6534
+    },
+    {
+      "epoch": 0.7506748607202344,
+      "grad_norm": 0.44639328122138977,
+      "learning_rate": 0.0001,
+      "loss": 1.7465,
+      "step": 6535
+    },
+    {
+      "epoch": 0.7507897306300615,
+      "grad_norm": 0.4912661910057068,
+      "learning_rate": 0.0001,
+      "loss": 1.7199,
+      "step": 6536
+    },
+    {
+      "epoch": 0.7509046005398886,
+      "grad_norm": 0.4561523199081421,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 6537
+    },
+    {
+      "epoch": 0.7510194704497157,
+      "grad_norm": 0.44610196352005005,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 6538
+    },
+    {
+      "epoch": 0.7511343403595429,
+      "grad_norm": 0.46128877997398376,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 6539
+    },
+    {
+      "epoch": 0.75124921026937,
+      "grad_norm": 0.4640040397644043,
+      "learning_rate": 0.0001,
+      "loss": 1.8346,
+      "step": 6540
+    },
+    {
+      "epoch": 0.7513640801791971,
+      "grad_norm": 0.4831351339817047,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 6541
+    },
+    {
+      "epoch": 0.7514789500890242,
+      "grad_norm": 0.4846135377883911,
+      "learning_rate": 0.0001,
+      "loss": 1.727,
+      "step": 6542
+    },
+    {
+      "epoch": 0.7515938199988513,
+      "grad_norm": 0.40949901938438416,
+      "learning_rate": 0.0001,
+      "loss": 1.5282,
+      "step": 6543
+    },
+    {
+      "epoch": 0.7517086899086785,
+      "grad_norm": 0.4737204909324646,
+      "learning_rate": 0.0001,
+      "loss": 1.766,
+      "step": 6544
+    },
+    {
+      "epoch": 0.7518235598185056,
+      "grad_norm": 0.4291594922542572,
+      "learning_rate": 0.0001,
+      "loss": 1.6355,
+      "step": 6545
+    },
+    {
+      "epoch": 0.7519384297283327,
+      "grad_norm": 0.4513239562511444,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 6546
+    },
+    {
+      "epoch": 0.7520532996381598,
+      "grad_norm": 0.42415329813957214,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 6547
+    },
+    {
+      "epoch": 0.7521681695479869,
+      "grad_norm": 0.4564039707183838,
+      "learning_rate": 0.0001,
+      "loss": 1.2685,
+      "step": 6548
+    },
+    {
+      "epoch": 0.752283039457814,
+      "grad_norm": 0.4450725018978119,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 6549
+    },
+    {
+      "epoch": 0.7523979093676412,
+      "grad_norm": 0.4227917492389679,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 6550
+    },
+    {
+      "epoch": 0.7525127792774683,
+      "grad_norm": 0.4478033483028412,
+      "learning_rate": 0.0001,
+      "loss": 1.6344,
+      "step": 6551
+    },
+    {
+      "epoch": 0.7526276491872954,
+      "grad_norm": 0.45714229345321655,
+      "learning_rate": 0.0001,
+      "loss": 1.7176,
+      "step": 6552
+    },
+    {
+      "epoch": 0.7527425190971225,
+      "grad_norm": 0.4998722970485687,
+      "learning_rate": 0.0001,
+      "loss": 1.4318,
+      "step": 6553
+    },
+    {
+      "epoch": 0.7528573890069497,
+      "grad_norm": 0.49186933040618896,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 6554
+    },
+    {
+      "epoch": 0.7529722589167768,
+      "grad_norm": 0.5109425187110901,
+      "learning_rate": 0.0001,
+      "loss": 1.7841,
+      "step": 6555
+    },
+    {
+      "epoch": 0.7530871288266039,
+      "grad_norm": 0.436103492975235,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 6556
+    },
+    {
+      "epoch": 0.753201998736431,
+      "grad_norm": 0.46560850739479065,
+      "learning_rate": 0.0001,
+      "loss": 1.4743,
+      "step": 6557
+    },
+    {
+      "epoch": 0.7533168686462581,
+      "grad_norm": 0.47097107768058777,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 6558
+    },
+    {
+      "epoch": 0.7534317385560853,
+      "grad_norm": 0.47928592562675476,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 6559
+    },
+    {
+      "epoch": 0.7535466084659124,
+      "grad_norm": 0.52452552318573,
+      "learning_rate": 0.0001,
+      "loss": 1.6683,
+      "step": 6560
+    },
+    {
+      "epoch": 0.7536614783757395,
+      "grad_norm": 0.47707682847976685,
+      "learning_rate": 0.0001,
+      "loss": 1.7117,
+      "step": 6561
+    },
+    {
+      "epoch": 0.7537763482855666,
+      "grad_norm": 0.4742587208747864,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 6562
+    },
+    {
+      "epoch": 0.7538912181953937,
+      "grad_norm": 0.4345073103904724,
+      "learning_rate": 0.0001,
+      "loss": 1.4673,
+      "step": 6563
+    },
+    {
+      "epoch": 0.7540060881052209,
+      "grad_norm": 0.4827091693878174,
+      "learning_rate": 0.0001,
+      "loss": 1.731,
+      "step": 6564
+    },
+    {
+      "epoch": 0.754120958015048,
+      "grad_norm": 0.4398217797279358,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 6565
+    },
+    {
+      "epoch": 0.7542358279248751,
+      "grad_norm": 0.4701884090900421,
+      "learning_rate": 0.0001,
+      "loss": 1.811,
+      "step": 6566
+    },
+    {
+      "epoch": 0.7543506978347022,
+      "grad_norm": 0.41458040475845337,
+      "learning_rate": 0.0001,
+      "loss": 1.1869,
+      "step": 6567
+    },
+    {
+      "epoch": 0.7544655677445293,
+      "grad_norm": 0.4266148507595062,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 6568
+    },
+    {
+      "epoch": 0.7545804376543565,
+      "grad_norm": 0.441137433052063,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 6569
+    },
+    {
+      "epoch": 0.7546953075641836,
+      "grad_norm": 0.511440098285675,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 6570
+    },
+    {
+      "epoch": 0.7548101774740107,
+      "grad_norm": 0.4327561855316162,
+      "learning_rate": 0.0001,
+      "loss": 1.3331,
+      "step": 6571
+    },
+    {
+      "epoch": 0.7549250473838378,
+      "grad_norm": 0.4539497494697571,
+      "learning_rate": 0.0001,
+      "loss": 1.3839,
+      "step": 6572
+    },
+    {
+      "epoch": 0.7550399172936649,
+      "grad_norm": 0.47245991230010986,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 6573
+    },
+    {
+      "epoch": 0.755154787203492,
+      "grad_norm": 0.46419230103492737,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 6574
+    },
+    {
+      "epoch": 0.7552696571133192,
+      "grad_norm": 0.47439756989479065,
+      "learning_rate": 0.0001,
+      "loss": 1.5447,
+      "step": 6575
+    },
+    {
+      "epoch": 0.7553845270231463,
+      "grad_norm": 0.4236104488372803,
+      "learning_rate": 0.0001,
+      "loss": 1.3438,
+      "step": 6576
+    },
+    {
+      "epoch": 0.7554993969329734,
+      "grad_norm": 0.4503575563430786,
+      "learning_rate": 0.0001,
+      "loss": 1.6844,
+      "step": 6577
+    },
+    {
+      "epoch": 0.7556142668428005,
+      "grad_norm": 0.4399665594100952,
+      "learning_rate": 0.0001,
+      "loss": 1.5789,
+      "step": 6578
+    },
+    {
+      "epoch": 0.7557291367526277,
+      "grad_norm": 0.45208460092544556,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 6579
+    },
+    {
+      "epoch": 0.7558440066624548,
+      "grad_norm": 0.48191729187965393,
+      "learning_rate": 0.0001,
+      "loss": 1.7206,
+      "step": 6580
+    },
+    {
+      "epoch": 0.7559588765722819,
+      "grad_norm": 0.496855765581131,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 6581
+    },
+    {
+      "epoch": 0.756073746482109,
+      "grad_norm": 0.4180924594402313,
+      "learning_rate": 0.0001,
+      "loss": 1.4212,
+      "step": 6582
+    },
+    {
+      "epoch": 0.7561886163919361,
+      "grad_norm": 0.4331256151199341,
+      "learning_rate": 0.0001,
+      "loss": 1.3518,
+      "step": 6583
+    },
+    {
+      "epoch": 0.7563034863017633,
+      "grad_norm": 0.47106215357780457,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 6584
+    },
+    {
+      "epoch": 0.7564183562115904,
+      "grad_norm": 0.4699764549732208,
+      "learning_rate": 0.0001,
+      "loss": 1.3998,
+      "step": 6585
+    },
+    {
+      "epoch": 0.7565332261214175,
+      "grad_norm": 0.44458329677581787,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 6586
+    },
+    {
+      "epoch": 0.7566480960312446,
+      "grad_norm": 0.46415066719055176,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 6587
+    },
+    {
+      "epoch": 0.7567629659410717,
+      "grad_norm": 0.44048815965652466,
+      "learning_rate": 0.0001,
+      "loss": 1.4584,
+      "step": 6588
+    },
+    {
+      "epoch": 0.7568778358508989,
+      "grad_norm": 0.43758389353752136,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 6589
+    },
+    {
+      "epoch": 0.756992705760726,
+      "grad_norm": 0.44562289118766785,
+      "learning_rate": 0.0001,
+      "loss": 1.5985,
+      "step": 6590
+    },
+    {
+      "epoch": 0.7571075756705531,
+      "grad_norm": 0.4680253267288208,
+      "learning_rate": 0.0001,
+      "loss": 1.6759,
+      "step": 6591
+    },
+    {
+      "epoch": 0.7572224455803802,
+      "grad_norm": 0.5020065307617188,
+      "learning_rate": 0.0001,
+      "loss": 1.7335,
+      "step": 6592
+    },
+    {
+      "epoch": 0.7573373154902073,
+      "grad_norm": 0.42748236656188965,
+      "learning_rate": 0.0001,
+      "loss": 1.4555,
+      "step": 6593
+    },
+    {
+      "epoch": 0.7574521854000345,
+      "grad_norm": 0.4277678430080414,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 6594
+    },
+    {
+      "epoch": 0.7575670553098616,
+      "grad_norm": 0.4632684886455536,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 6595
+    },
+    {
+      "epoch": 0.7576819252196887,
+      "grad_norm": 0.5025344491004944,
+      "learning_rate": 0.0001,
+      "loss": 1.7728,
+      "step": 6596
+    },
+    {
+      "epoch": 0.7577967951295158,
+      "grad_norm": 0.45161494612693787,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 6597
+    },
+    {
+      "epoch": 0.7579116650393429,
+      "grad_norm": 0.45215389132499695,
+      "learning_rate": 0.0001,
+      "loss": 1.659,
+      "step": 6598
+    },
+    {
+      "epoch": 0.75802653494917,
+      "grad_norm": 0.4214898347854614,
+      "learning_rate": 0.0001,
+      "loss": 1.4783,
+      "step": 6599
+    },
+    {
+      "epoch": 0.7581414048589972,
+      "grad_norm": 0.5195730924606323,
+      "learning_rate": 0.0001,
+      "loss": 1.8058,
+      "step": 6600
+    },
+    {
+      "epoch": 0.7582562747688243,
+      "grad_norm": 0.48730143904685974,
+      "learning_rate": 0.0001,
+      "loss": 1.8334,
+      "step": 6601
+    },
+    {
+      "epoch": 0.7583711446786514,
+      "grad_norm": 0.5312708020210266,
+      "learning_rate": 0.0001,
+      "loss": 1.8008,
+      "step": 6602
+    },
+    {
+      "epoch": 0.7584860145884785,
+      "grad_norm": 0.44341936707496643,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 6603
+    },
+    {
+      "epoch": 0.7586008844983056,
+      "grad_norm": 0.4990461468696594,
+      "learning_rate": 0.0001,
+      "loss": 1.7039,
+      "step": 6604
+    },
+    {
+      "epoch": 0.7587157544081328,
+      "grad_norm": 0.43165311217308044,
+      "learning_rate": 0.0001,
+      "loss": 1.6646,
+      "step": 6605
+    },
+    {
+      "epoch": 0.7588306243179599,
+      "grad_norm": 0.4335373640060425,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 6606
+    },
+    {
+      "epoch": 0.758945494227787,
+      "grad_norm": 0.4470045566558838,
+      "learning_rate": 0.0001,
+      "loss": 1.3655,
+      "step": 6607
+    },
+    {
+      "epoch": 0.7590603641376141,
+      "grad_norm": 0.46403536200523376,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 6608
+    },
+    {
+      "epoch": 0.7591752340474412,
+      "grad_norm": 0.49644750356674194,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 6609
+    },
+    {
+      "epoch": 0.7592901039572684,
+      "grad_norm": 0.5031653046607971,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 6610
+    },
+    {
+      "epoch": 0.7594049738670955,
+      "grad_norm": 0.46444204449653625,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 6611
+    },
+    {
+      "epoch": 0.7595198437769226,
+      "grad_norm": 0.4190670847892761,
+      "learning_rate": 0.0001,
+      "loss": 1.3953,
+      "step": 6612
+    },
+    {
+      "epoch": 0.7596347136867497,
+      "grad_norm": 0.4325488209724426,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 6613
+    },
+    {
+      "epoch": 0.7597495835965768,
+      "grad_norm": 0.4146260917186737,
+      "learning_rate": 0.0001,
+      "loss": 1.4447,
+      "step": 6614
+    },
+    {
+      "epoch": 0.759864453506404,
+      "grad_norm": 0.4315129816532135,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 6615
+    },
+    {
+      "epoch": 0.7599793234162311,
+      "grad_norm": 0.4837382733821869,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 6616
+    },
+    {
+      "epoch": 0.7600941933260582,
+      "grad_norm": 0.5264213681221008,
+      "learning_rate": 0.0001,
+      "loss": 1.7784,
+      "step": 6617
+    },
+    {
+      "epoch": 0.7602090632358853,
+      "grad_norm": 0.4192154109477997,
+      "learning_rate": 0.0001,
+      "loss": 1.49,
+      "step": 6618
+    },
+    {
+      "epoch": 0.7603239331457124,
+      "grad_norm": 0.4764638841152191,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 6619
+    },
+    {
+      "epoch": 0.7604388030555396,
+      "grad_norm": 0.4535200893878937,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 6620
+    },
+    {
+      "epoch": 0.7605536729653667,
+      "grad_norm": 0.40182626247406006,
+      "learning_rate": 0.0001,
+      "loss": 1.2703,
+      "step": 6621
+    },
+    {
+      "epoch": 0.7606685428751938,
+      "grad_norm": 0.4054408669471741,
+      "learning_rate": 0.0001,
+      "loss": 1.4678,
+      "step": 6622
+    },
+    {
+      "epoch": 0.7607834127850209,
+      "grad_norm": 0.4554063081741333,
+      "learning_rate": 0.0001,
+      "loss": 1.7263,
+      "step": 6623
+    },
+    {
+      "epoch": 0.760898282694848,
+      "grad_norm": 0.447591632604599,
+      "learning_rate": 0.0001,
+      "loss": 1.75,
+      "step": 6624
+    },
+    {
+      "epoch": 0.7610131526046752,
+      "grad_norm": 0.4537143409252167,
+      "learning_rate": 0.0001,
+      "loss": 1.7036,
+      "step": 6625
+    },
+    {
+      "epoch": 0.7611280225145023,
+      "grad_norm": 0.47372350096702576,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 6626
+    },
+    {
+      "epoch": 0.7612428924243294,
+      "grad_norm": 0.47313424944877625,
+      "learning_rate": 0.0001,
+      "loss": 1.7578,
+      "step": 6627
+    },
+    {
+      "epoch": 0.7613577623341565,
+      "grad_norm": 0.4805237054824829,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 6628
+    },
+    {
+      "epoch": 0.7614726322439836,
+      "grad_norm": 0.4734886884689331,
+      "learning_rate": 0.0001,
+      "loss": 1.7557,
+      "step": 6629
+    },
+    {
+      "epoch": 0.7615875021538108,
+      "grad_norm": 0.46872133016586304,
+      "learning_rate": 0.0001,
+      "loss": 1.7054,
+      "step": 6630
+    },
+    {
+      "epoch": 0.7617023720636379,
+      "grad_norm": 0.44141799211502075,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 6631
+    },
+    {
+      "epoch": 0.761817241973465,
+      "grad_norm": 0.4713442623615265,
+      "learning_rate": 0.0001,
+      "loss": 1.7963,
+      "step": 6632
+    },
+    {
+      "epoch": 0.7619321118832921,
+      "grad_norm": 0.46353405714035034,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 6633
+    },
+    {
+      "epoch": 0.7620469817931192,
+      "grad_norm": 0.9744009971618652,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 6634
+    },
+    {
+      "epoch": 0.7621618517029464,
+      "grad_norm": 0.41723188757896423,
+      "learning_rate": 0.0001,
+      "loss": 1.5305,
+      "step": 6635
+    },
+    {
+      "epoch": 0.7622767216127735,
+      "grad_norm": 0.47938770055770874,
+      "learning_rate": 0.0001,
+      "loss": 1.7722,
+      "step": 6636
+    },
+    {
+      "epoch": 0.7623915915226006,
+      "grad_norm": 0.46202394366264343,
+      "learning_rate": 0.0001,
+      "loss": 1.6189,
+      "step": 6637
+    },
+    {
+      "epoch": 0.7625064614324277,
+      "grad_norm": 0.4409821033477783,
+      "learning_rate": 0.0001,
+      "loss": 1.3921,
+      "step": 6638
+    },
+    {
+      "epoch": 0.7626213313422548,
+      "grad_norm": 0.4694823920726776,
+      "learning_rate": 0.0001,
+      "loss": 1.6386,
+      "step": 6639
+    },
+    {
+      "epoch": 0.762736201252082,
+      "grad_norm": 0.45864009857177734,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 6640
+    },
+    {
+      "epoch": 0.7628510711619091,
+      "grad_norm": 0.46855515241622925,
+      "learning_rate": 0.0001,
+      "loss": 1.4683,
+      "step": 6641
+    },
+    {
+      "epoch": 0.7629659410717362,
+      "grad_norm": 0.498154878616333,
+      "learning_rate": 0.0001,
+      "loss": 1.5162,
+      "step": 6642
+    },
+    {
+      "epoch": 0.7630808109815633,
+      "grad_norm": 0.4623854160308838,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 6643
+    },
+    {
+      "epoch": 0.7631956808913904,
+      "grad_norm": 0.42440950870513916,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 6644
+    },
+    {
+      "epoch": 0.7633105508012176,
+      "grad_norm": 0.4813312888145447,
+      "learning_rate": 0.0001,
+      "loss": 1.5837,
+      "step": 6645
+    },
+    {
+      "epoch": 0.7634254207110447,
+      "grad_norm": 0.430618554353714,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 6646
+    },
+    {
+      "epoch": 0.7635402906208718,
+      "grad_norm": 0.4468959867954254,
+      "learning_rate": 0.0001,
+      "loss": 1.5506,
+      "step": 6647
+    },
+    {
+      "epoch": 0.7636551605306989,
+      "grad_norm": 0.45413529872894287,
+      "learning_rate": 0.0001,
+      "loss": 1.5858,
+      "step": 6648
+    },
+    {
+      "epoch": 0.763770030440526,
+      "grad_norm": 0.4446607530117035,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 6649
+    },
+    {
+      "epoch": 0.7638849003503533,
+      "grad_norm": 0.4520895779132843,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 6650
+    },
+    {
+      "epoch": 0.7639997702601804,
+      "grad_norm": 0.46558019518852234,
+      "learning_rate": 0.0001,
+      "loss": 1.6506,
+      "step": 6651
+    },
+    {
+      "epoch": 0.7641146401700075,
+      "grad_norm": 0.501394510269165,
+      "learning_rate": 0.0001,
+      "loss": 1.8724,
+      "step": 6652
+    },
+    {
+      "epoch": 0.7642295100798346,
+      "grad_norm": 0.47188153862953186,
+      "learning_rate": 0.0001,
+      "loss": 1.7557,
+      "step": 6653
+    },
+    {
+      "epoch": 0.7643443799896618,
+      "grad_norm": 0.42527419328689575,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 6654
+    },
+    {
+      "epoch": 0.7644592498994889,
+      "grad_norm": 0.4962664246559143,
+      "learning_rate": 0.0001,
+      "loss": 1.7203,
+      "step": 6655
+    },
+    {
+      "epoch": 0.764574119809316,
+      "grad_norm": 0.4442030191421509,
+      "learning_rate": 0.0001,
+      "loss": 1.5981,
+      "step": 6656
+    },
+    {
+      "epoch": 0.7646889897191431,
+      "grad_norm": 0.47021228075027466,
+      "learning_rate": 0.0001,
+      "loss": 1.7917,
+      "step": 6657
+    },
+    {
+      "epoch": 0.7648038596289702,
+      "grad_norm": 0.4421970546245575,
+      "learning_rate": 0.0001,
+      "loss": 1.5717,
+      "step": 6658
+    },
+    {
+      "epoch": 0.7649187295387974,
+      "grad_norm": 0.45684003829956055,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 6659
+    },
+    {
+      "epoch": 0.7650335994486245,
+      "grad_norm": 0.47243228554725647,
+      "learning_rate": 0.0001,
+      "loss": 1.6116,
+      "step": 6660
+    },
+    {
+      "epoch": 0.7651484693584516,
+      "grad_norm": 0.44212606549263,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 6661
+    },
+    {
+      "epoch": 0.7652633392682787,
+      "grad_norm": 0.48998787999153137,
+      "learning_rate": 0.0001,
+      "loss": 1.7869,
+      "step": 6662
+    },
+    {
+      "epoch": 0.7653782091781058,
+      "grad_norm": 0.4589942693710327,
+      "learning_rate": 0.0001,
+      "loss": 1.5954,
+      "step": 6663
+    },
+    {
+      "epoch": 0.765493079087933,
+      "grad_norm": 0.45436352491378784,
+      "learning_rate": 0.0001,
+      "loss": 1.7321,
+      "step": 6664
+    },
+    {
+      "epoch": 0.7656079489977601,
+      "grad_norm": 0.45719271898269653,
+      "learning_rate": 0.0001,
+      "loss": 1.3269,
+      "step": 6665
+    },
+    {
+      "epoch": 0.7657228189075872,
+      "grad_norm": 0.4515266716480255,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 6666
+    },
+    {
+      "epoch": 0.7658376888174143,
+      "grad_norm": 0.4383704662322998,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 6667
+    },
+    {
+      "epoch": 0.7659525587272414,
+      "grad_norm": 0.5026112198829651,
+      "learning_rate": 0.0001,
+      "loss": 1.6132,
+      "step": 6668
+    },
+    {
+      "epoch": 0.7660674286370686,
+      "grad_norm": 0.4402121305465698,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 6669
+    },
+    {
+      "epoch": 0.7661822985468957,
+      "grad_norm": 0.4881230294704437,
+      "learning_rate": 0.0001,
+      "loss": 1.7567,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7662971684567228,
+      "grad_norm": 0.454129159450531,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 6671
+    },
+    {
+      "epoch": 0.7664120383665499,
+      "grad_norm": 0.45725277066230774,
+      "learning_rate": 0.0001,
+      "loss": 1.3238,
+      "step": 6672
+    },
+    {
+      "epoch": 0.766526908276377,
+      "grad_norm": 0.4708629250526428,
+      "learning_rate": 0.0001,
+      "loss": 1.5594,
+      "step": 6673
+    },
+    {
+      "epoch": 0.7666417781862042,
+      "grad_norm": 0.4380055367946625,
+      "learning_rate": 0.0001,
+      "loss": 1.6742,
+      "step": 6674
+    },
+    {
+      "epoch": 0.7667566480960313,
+      "grad_norm": 0.45258665084838867,
+      "learning_rate": 0.0001,
+      "loss": 1.4825,
+      "step": 6675
+    },
+    {
+      "epoch": 0.7668715180058584,
+      "grad_norm": 0.4417783319950104,
+      "learning_rate": 0.0001,
+      "loss": 1.469,
+      "step": 6676
+    },
+    {
+      "epoch": 0.7669863879156855,
+      "grad_norm": 0.5079810619354248,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 6677
+    },
+    {
+      "epoch": 0.7671012578255126,
+      "grad_norm": 0.4944427013397217,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 6678
+    },
+    {
+      "epoch": 0.7672161277353398,
+      "grad_norm": 0.4718817472457886,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 6679
+    },
+    {
+      "epoch": 0.7673309976451669,
+      "grad_norm": 0.4739345610141754,
+      "learning_rate": 0.0001,
+      "loss": 1.765,
+      "step": 6680
+    },
+    {
+      "epoch": 0.767445867554994,
+      "grad_norm": 0.46249300241470337,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 6681
+    },
+    {
+      "epoch": 0.7675607374648211,
+      "grad_norm": 0.5330554246902466,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 6682
+    },
+    {
+      "epoch": 0.7676756073746482,
+      "grad_norm": 0.4298166334629059,
+      "learning_rate": 0.0001,
+      "loss": 1.4984,
+      "step": 6683
+    },
+    {
+      "epoch": 0.7677904772844754,
+      "grad_norm": 0.47553664445877075,
+      "learning_rate": 0.0001,
+      "loss": 1.6892,
+      "step": 6684
+    },
+    {
+      "epoch": 0.7679053471943025,
+      "grad_norm": 0.4601770341396332,
+      "learning_rate": 0.0001,
+      "loss": 1.7258,
+      "step": 6685
+    },
+    {
+      "epoch": 0.7680202171041296,
+      "grad_norm": 0.4717545509338379,
+      "learning_rate": 0.0001,
+      "loss": 1.7111,
+      "step": 6686
+    },
+    {
+      "epoch": 0.7681350870139567,
+      "grad_norm": 0.47477278113365173,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 6687
+    },
+    {
+      "epoch": 0.7682499569237838,
+      "grad_norm": 0.45599865913391113,
+      "learning_rate": 0.0001,
+      "loss": 1.6069,
+      "step": 6688
+    },
+    {
+      "epoch": 0.768364826833611,
+      "grad_norm": 0.43916571140289307,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 6689
+    },
+    {
+      "epoch": 0.7684796967434381,
+      "grad_norm": 0.4506683945655823,
+      "learning_rate": 0.0001,
+      "loss": 1.6638,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7685945666532652,
+      "grad_norm": 0.42485401034355164,
+      "learning_rate": 0.0001,
+      "loss": 1.4915,
+      "step": 6691
+    },
+    {
+      "epoch": 0.7687094365630923,
+      "grad_norm": 0.47712936997413635,
+      "learning_rate": 0.0001,
+      "loss": 1.6449,
+      "step": 6692
+    },
+    {
+      "epoch": 0.7688243064729194,
+      "grad_norm": 0.4386729896068573,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 6693
+    },
+    {
+      "epoch": 0.7689391763827466,
+      "grad_norm": 0.5089852213859558,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 6694
+    },
+    {
+      "epoch": 0.7690540462925737,
+      "grad_norm": 0.4199332594871521,
+      "learning_rate": 0.0001,
+      "loss": 1.4757,
+      "step": 6695
+    },
+    {
+      "epoch": 0.7691689162024008,
+      "grad_norm": 0.5055813789367676,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 6696
+    },
+    {
+      "epoch": 0.7692837861122279,
+      "grad_norm": 0.4748409390449524,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 6697
+    },
+    {
+      "epoch": 0.769398656022055,
+      "grad_norm": 0.452891081571579,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 6698
+    },
+    {
+      "epoch": 0.7695135259318822,
+      "grad_norm": 0.4543799161911011,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 6699
+    },
+    {
+      "epoch": 0.7696283958417093,
+      "grad_norm": 0.47071510553359985,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7697432657515364,
+      "grad_norm": 0.46207287907600403,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 6701
+    },
+    {
+      "epoch": 0.7698581356613635,
+      "grad_norm": 0.4610616862773895,
+      "learning_rate": 0.0001,
+      "loss": 1.4635,
+      "step": 6702
+    },
+    {
+      "epoch": 0.7699730055711906,
+      "grad_norm": 0.46330535411834717,
+      "learning_rate": 0.0001,
+      "loss": 1.5052,
+      "step": 6703
+    },
+    {
+      "epoch": 0.7700878754810178,
+      "grad_norm": 0.4601458013057709,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 6704
+    },
+    {
+      "epoch": 0.7702027453908449,
+      "grad_norm": 0.4191734790802002,
+      "learning_rate": 0.0001,
+      "loss": 1.3989,
+      "step": 6705
+    },
+    {
+      "epoch": 0.770317615300672,
+      "grad_norm": 0.4650634229183197,
+      "learning_rate": 0.0001,
+      "loss": 1.6411,
+      "step": 6706
+    },
+    {
+      "epoch": 0.7704324852104991,
+      "grad_norm": 0.4624280035495758,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 6707
+    },
+    {
+      "epoch": 0.7705473551203262,
+      "grad_norm": 0.45528537034988403,
+      "learning_rate": 0.0001,
+      "loss": 1.7724,
+      "step": 6708
+    },
+    {
+      "epoch": 0.7706622250301534,
+      "grad_norm": 0.49074849486351013,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 6709
+    },
+    {
+      "epoch": 0.7707770949399805,
+      "grad_norm": 0.4072690010070801,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7708919648498076,
+      "grad_norm": 0.5691601634025574,
+      "learning_rate": 0.0001,
+      "loss": 1.4397,
+      "step": 6711
+    },
+    {
+      "epoch": 0.7710068347596347,
+      "grad_norm": 0.4546988308429718,
+      "learning_rate": 0.0001,
+      "loss": 1.5887,
+      "step": 6712
+    },
+    {
+      "epoch": 0.7711217046694618,
+      "grad_norm": 0.49612048268318176,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 6713
+    },
+    {
+      "epoch": 0.771236574579289,
+      "grad_norm": 0.5100909471511841,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 6714
+    },
+    {
+      "epoch": 0.7713514444891161,
+      "grad_norm": 0.4859296679496765,
+      "learning_rate": 0.0001,
+      "loss": 1.6302,
+      "step": 6715
+    },
+    {
+      "epoch": 0.7714663143989432,
+      "grad_norm": 0.45790478587150574,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 6716
+    },
+    {
+      "epoch": 0.7715811843087703,
+      "grad_norm": 0.4487425982952118,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 6717
+    },
+    {
+      "epoch": 0.7716960542185974,
+      "grad_norm": 0.49326303601264954,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 6718
+    },
+    {
+      "epoch": 0.7718109241284246,
+      "grad_norm": 0.43098387122154236,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 6719
+    },
+    {
+      "epoch": 0.7719257940382517,
+      "grad_norm": 0.4224907159805298,
+      "learning_rate": 0.0001,
+      "loss": 1.4381,
+      "step": 6720
+    },
+    {
+      "epoch": 0.7720406639480788,
+      "grad_norm": 0.4463758170604706,
+      "learning_rate": 0.0001,
+      "loss": 1.6655,
+      "step": 6721
+    },
+    {
+      "epoch": 0.7721555338579059,
+      "grad_norm": 0.45395219326019287,
+      "learning_rate": 0.0001,
+      "loss": 1.7321,
+      "step": 6722
+    },
+    {
+      "epoch": 0.772270403767733,
+      "grad_norm": 0.443327933549881,
+      "learning_rate": 0.0001,
+      "loss": 1.6938,
+      "step": 6723
+    },
+    {
+      "epoch": 0.7723852736775602,
+      "grad_norm": 0.44689181447029114,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 6724
+    },
+    {
+      "epoch": 0.7725001435873873,
+      "grad_norm": 0.4609023928642273,
+      "learning_rate": 0.0001,
+      "loss": 1.7214,
+      "step": 6725
+    },
+    {
+      "epoch": 0.7726150134972144,
+      "grad_norm": 0.42757725715637207,
+      "learning_rate": 0.0001,
+      "loss": 1.298,
+      "step": 6726
+    },
+    {
+      "epoch": 0.7727298834070415,
+      "grad_norm": 0.43794259428977966,
+      "learning_rate": 0.0001,
+      "loss": 1.6358,
+      "step": 6727
+    },
+    {
+      "epoch": 0.7728447533168686,
+      "grad_norm": 0.4543568193912506,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 6728
+    },
+    {
+      "epoch": 0.7729596232266958,
+      "grad_norm": 0.4287487864494324,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 6729
+    },
+    {
+      "epoch": 0.7730744931365229,
+      "grad_norm": 0.44154518842697144,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 6730
+    },
+    {
+      "epoch": 0.77318936304635,
+      "grad_norm": 0.44655200839042664,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 6731
+    },
+    {
+      "epoch": 0.7733042329561771,
+      "grad_norm": 0.5005887150764465,
+      "learning_rate": 0.0001,
+      "loss": 1.4248,
+      "step": 6732
+    },
+    {
+      "epoch": 0.7734191028660042,
+      "grad_norm": 0.5025423169136047,
+      "learning_rate": 0.0001,
+      "loss": 1.7576,
+      "step": 6733
+    },
+    {
+      "epoch": 0.7735339727758314,
+      "grad_norm": 0.5430256128311157,
+      "learning_rate": 0.0001,
+      "loss": 1.7347,
+      "step": 6734
+    },
+    {
+      "epoch": 0.7736488426856585,
+      "grad_norm": 0.4566929340362549,
+      "learning_rate": 0.0001,
+      "loss": 1.4515,
+      "step": 6735
+    },
+    {
+      "epoch": 0.7737637125954856,
+      "grad_norm": 0.45113489031791687,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 6736
+    },
+    {
+      "epoch": 0.7738785825053127,
+      "grad_norm": 0.44897618889808655,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 6737
+    },
+    {
+      "epoch": 0.7739934524151398,
+      "grad_norm": 0.4536812901496887,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 6738
+    },
+    {
+      "epoch": 0.774108322324967,
+      "grad_norm": 0.4436149299144745,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 6739
+    },
+    {
+      "epoch": 0.7742231922347941,
+      "grad_norm": 0.4435299336910248,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7743380621446212,
+      "grad_norm": 0.44554266333580017,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 6741
+    },
+    {
+      "epoch": 0.7744529320544483,
+      "grad_norm": 0.45656824111938477,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 6742
+    },
+    {
+      "epoch": 0.7745678019642754,
+      "grad_norm": 0.4641306698322296,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 6743
+    },
+    {
+      "epoch": 0.7746826718741026,
+      "grad_norm": 0.48745623230934143,
+      "learning_rate": 0.0001,
+      "loss": 1.603,
+      "step": 6744
+    },
+    {
+      "epoch": 0.7747975417839297,
+      "grad_norm": 0.4215722978115082,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 6745
+    },
+    {
+      "epoch": 0.7749124116937568,
+      "grad_norm": 0.49982690811157227,
+      "learning_rate": 0.0001,
+      "loss": 1.6318,
+      "step": 6746
+    },
+    {
+      "epoch": 0.7750272816035839,
+      "grad_norm": 0.5177478194236755,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 6747
+    },
+    {
+      "epoch": 0.775142151513411,
+      "grad_norm": 0.4642798900604248,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 6748
+    },
+    {
+      "epoch": 0.7752570214232382,
+      "grad_norm": 0.5008178949356079,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 6749
+    },
+    {
+      "epoch": 0.7753718913330653,
+      "grad_norm": 0.4520767033100128,
+      "learning_rate": 0.0001,
+      "loss": 1.4705,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7754867612428924,
+      "grad_norm": 0.4959637224674225,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 6751
+    },
+    {
+      "epoch": 0.7756016311527195,
+      "grad_norm": 0.4342813491821289,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 6752
+    },
+    {
+      "epoch": 0.7757165010625466,
+      "grad_norm": 0.45691919326782227,
+      "learning_rate": 0.0001,
+      "loss": 1.7955,
+      "step": 6753
+    },
+    {
+      "epoch": 0.7758313709723738,
+      "grad_norm": 0.44725337624549866,
+      "learning_rate": 0.0001,
+      "loss": 1.4242,
+      "step": 6754
+    },
+    {
+      "epoch": 0.7759462408822009,
+      "grad_norm": 0.4812532365322113,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 6755
+    },
+    {
+      "epoch": 0.776061110792028,
+      "grad_norm": 0.40750646591186523,
+      "learning_rate": 0.0001,
+      "loss": 1.5516,
+      "step": 6756
+    },
+    {
+      "epoch": 0.7761759807018551,
+      "grad_norm": 0.482294499874115,
+      "learning_rate": 0.0001,
+      "loss": 1.6497,
+      "step": 6757
+    },
+    {
+      "epoch": 0.7762908506116822,
+      "grad_norm": 0.4344806969165802,
+      "learning_rate": 0.0001,
+      "loss": 1.4736,
+      "step": 6758
+    },
+    {
+      "epoch": 0.7764057205215094,
+      "grad_norm": 0.46780574321746826,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 6759
+    },
+    {
+      "epoch": 0.7765205904313365,
+      "grad_norm": 0.4722866714000702,
+      "learning_rate": 0.0001,
+      "loss": 1.8075,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7766354603411636,
+      "grad_norm": 0.45282799005508423,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 6761
+    },
+    {
+      "epoch": 0.7767503302509907,
+      "grad_norm": 0.4654501974582672,
+      "learning_rate": 0.0001,
+      "loss": 1.7371,
+      "step": 6762
+    },
+    {
+      "epoch": 0.7768652001608178,
+      "grad_norm": 0.46721896529197693,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 6763
+    },
+    {
+      "epoch": 0.776980070070645,
+      "grad_norm": 0.47334960103034973,
+      "learning_rate": 0.0001,
+      "loss": 1.6024,
+      "step": 6764
+    },
+    {
+      "epoch": 0.7770949399804721,
+      "grad_norm": 0.46868133544921875,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 6765
+    },
+    {
+      "epoch": 0.7772098098902992,
+      "grad_norm": 0.45955634117126465,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 6766
+    },
+    {
+      "epoch": 0.7773246798001263,
+      "grad_norm": 0.43872690200805664,
+      "learning_rate": 0.0001,
+      "loss": 1.3424,
+      "step": 6767
+    },
+    {
+      "epoch": 0.7774395497099534,
+      "grad_norm": 0.5430575609207153,
+      "learning_rate": 0.0001,
+      "loss": 1.8176,
+      "step": 6768
+    },
+    {
+      "epoch": 0.7775544196197806,
+      "grad_norm": 0.4751816689968109,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 6769
+    },
+    {
+      "epoch": 0.7776692895296077,
+      "grad_norm": 0.42308947443962097,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7777841594394348,
+      "grad_norm": 0.4343721866607666,
+      "learning_rate": 0.0001,
+      "loss": 1.5106,
+      "step": 6771
+    },
+    {
+      "epoch": 0.7778990293492619,
+      "grad_norm": 0.4403552711009979,
+      "learning_rate": 0.0001,
+      "loss": 1.5326,
+      "step": 6772
+    },
+    {
+      "epoch": 0.778013899259089,
+      "grad_norm": 0.46488016843795776,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 6773
+    },
+    {
+      "epoch": 0.7781287691689162,
+      "grad_norm": 0.43037131428718567,
+      "learning_rate": 0.0001,
+      "loss": 1.5796,
+      "step": 6774
+    },
+    {
+      "epoch": 0.7782436390787433,
+      "grad_norm": 0.42041462659835815,
+      "learning_rate": 0.0001,
+      "loss": 1.4634,
+      "step": 6775
+    },
+    {
+      "epoch": 0.7783585089885704,
+      "grad_norm": 0.46670305728912354,
+      "learning_rate": 0.0001,
+      "loss": 1.7338,
+      "step": 6776
+    },
+    {
+      "epoch": 0.7784733788983975,
+      "grad_norm": 0.44566023349761963,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 6777
+    },
+    {
+      "epoch": 0.7785882488082246,
+      "grad_norm": 0.46468743681907654,
+      "learning_rate": 0.0001,
+      "loss": 1.4851,
+      "step": 6778
+    },
+    {
+      "epoch": 0.7787031187180518,
+      "grad_norm": 0.4546216130256653,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 6779
+    },
+    {
+      "epoch": 0.7788179886278789,
+      "grad_norm": 0.41460397839546204,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 6780
+    },
+    {
+      "epoch": 0.778932858537706,
+      "grad_norm": 0.43776392936706543,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 6781
+    },
+    {
+      "epoch": 0.7790477284475331,
+      "grad_norm": 0.42901432514190674,
+      "learning_rate": 0.0001,
+      "loss": 1.4523,
+      "step": 6782
+    },
+    {
+      "epoch": 0.7791625983573602,
+      "grad_norm": 0.4392737150192261,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 6783
+    },
+    {
+      "epoch": 0.7792774682671874,
+      "grad_norm": 0.5367296934127808,
+      "learning_rate": 0.0001,
+      "loss": 1.7535,
+      "step": 6784
+    },
+    {
+      "epoch": 0.7793923381770145,
+      "grad_norm": 0.48000702261924744,
+      "learning_rate": 0.0001,
+      "loss": 1.6611,
+      "step": 6785
+    },
+    {
+      "epoch": 0.7795072080868416,
+      "grad_norm": 0.42784345149993896,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 6786
+    },
+    {
+      "epoch": 0.7796220779966688,
+      "grad_norm": 0.4323700964450836,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 6787
+    },
+    {
+      "epoch": 0.779736947906496,
+      "grad_norm": 0.46200308203697205,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 6788
+    },
+    {
+      "epoch": 0.7798518178163231,
+      "grad_norm": 0.4896979033946991,
+      "learning_rate": 0.0001,
+      "loss": 1.698,
+      "step": 6789
+    },
+    {
+      "epoch": 0.7799666877261502,
+      "grad_norm": 0.4582618474960327,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7800815576359773,
+      "grad_norm": 0.45183730125427246,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 6791
+    },
+    {
+      "epoch": 0.7801964275458044,
+      "grad_norm": 0.45246532559394836,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 6792
+    },
+    {
+      "epoch": 0.7803112974556315,
+      "grad_norm": 0.453540176153183,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 6793
+    },
+    {
+      "epoch": 0.7804261673654587,
+      "grad_norm": 0.460021436214447,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 6794
+    },
+    {
+      "epoch": 0.7805410372752858,
+      "grad_norm": 0.5325375199317932,
+      "learning_rate": 0.0001,
+      "loss": 1.7254,
+      "step": 6795
+    },
+    {
+      "epoch": 0.7806559071851129,
+      "grad_norm": 0.41828301548957825,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 6796
+    },
+    {
+      "epoch": 0.78077077709494,
+      "grad_norm": 0.4506258964538574,
+      "learning_rate": 0.0001,
+      "loss": 1.587,
+      "step": 6797
+    },
+    {
+      "epoch": 0.7808856470047671,
+      "grad_norm": 0.4368896484375,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 6798
+    },
+    {
+      "epoch": 0.7810005169145943,
+      "grad_norm": 0.4336557388305664,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 6799
+    },
+    {
+      "epoch": 0.7811153868244214,
+      "grad_norm": 0.47496965527534485,
+      "learning_rate": 0.0001,
+      "loss": 1.7202,
+      "step": 6800
+    },
+    {
+      "epoch": 0.7812302567342485,
+      "grad_norm": 0.44038236141204834,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 6801
+    },
+    {
+      "epoch": 0.7813451266440756,
+      "grad_norm": 0.44530773162841797,
+      "learning_rate": 0.0001,
+      "loss": 1.668,
+      "step": 6802
+    },
+    {
+      "epoch": 0.7814599965539027,
+      "grad_norm": 0.4356937110424042,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 6803
+    },
+    {
+      "epoch": 0.7815748664637299,
+      "grad_norm": 0.4739866256713867,
+      "learning_rate": 0.0001,
+      "loss": 1.6232,
+      "step": 6804
+    },
+    {
+      "epoch": 0.781689736373557,
+      "grad_norm": 0.5034109950065613,
+      "learning_rate": 0.0001,
+      "loss": 1.7754,
+      "step": 6805
+    },
+    {
+      "epoch": 0.7818046062833841,
+      "grad_norm": 0.46401265263557434,
+      "learning_rate": 0.0001,
+      "loss": 1.6829,
+      "step": 6806
+    },
+    {
+      "epoch": 0.7819194761932112,
+      "grad_norm": 0.44969433546066284,
+      "learning_rate": 0.0001,
+      "loss": 1.6893,
+      "step": 6807
+    },
+    {
+      "epoch": 0.7820343461030383,
+      "grad_norm": 0.48960667848587036,
+      "learning_rate": 0.0001,
+      "loss": 1.8962,
+      "step": 6808
+    },
+    {
+      "epoch": 0.7821492160128655,
+      "grad_norm": 0.41496121883392334,
+      "learning_rate": 0.0001,
+      "loss": 1.3764,
+      "step": 6809
+    },
+    {
+      "epoch": 0.7822640859226926,
+      "grad_norm": 0.4606899917125702,
+      "learning_rate": 0.0001,
+      "loss": 1.6955,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7823789558325197,
+      "grad_norm": 0.5153933763504028,
+      "learning_rate": 0.0001,
+      "loss": 1.8612,
+      "step": 6811
+    },
+    {
+      "epoch": 0.7824938257423468,
+      "grad_norm": 0.4672335684299469,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 6812
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 0.4316304326057434,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 6813
+    },
+    {
+      "epoch": 0.7827235655620011,
+      "grad_norm": 0.4348050355911255,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 6814
+    },
+    {
+      "epoch": 0.7828384354718282,
+      "grad_norm": 0.44443196058273315,
+      "learning_rate": 0.0001,
+      "loss": 1.6767,
+      "step": 6815
+    },
+    {
+      "epoch": 0.7829533053816553,
+      "grad_norm": 0.45539575815200806,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 6816
+    },
+    {
+      "epoch": 0.7830681752914824,
+      "grad_norm": 0.42358124256134033,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 6817
+    },
+    {
+      "epoch": 0.7831830452013095,
+      "grad_norm": 0.444414347410202,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 6818
+    },
+    {
+      "epoch": 0.7832979151111367,
+      "grad_norm": 0.45530837774276733,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 6819
+    },
+    {
+      "epoch": 0.7834127850209638,
+      "grad_norm": 0.4563850462436676,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7835276549307909,
+      "grad_norm": 0.43416640162467957,
+      "learning_rate": 0.0001,
+      "loss": 1.3742,
+      "step": 6821
+    },
+    {
+      "epoch": 0.783642524840618,
+      "grad_norm": 0.44997450709342957,
+      "learning_rate": 0.0001,
+      "loss": 1.6559,
+      "step": 6822
+    },
+    {
+      "epoch": 0.7837573947504451,
+      "grad_norm": 0.4280833303928375,
+      "learning_rate": 0.0001,
+      "loss": 1.4376,
+      "step": 6823
+    },
+    {
+      "epoch": 0.7838722646602723,
+      "grad_norm": 0.4517759680747986,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 6824
+    },
+    {
+      "epoch": 0.7839871345700994,
+      "grad_norm": 0.45241299271583557,
+      "learning_rate": 0.0001,
+      "loss": 1.3445,
+      "step": 6825
+    },
+    {
+      "epoch": 0.7841020044799265,
+      "grad_norm": 0.4801664352416992,
+      "learning_rate": 0.0001,
+      "loss": 1.7256,
+      "step": 6826
+    },
+    {
+      "epoch": 0.7842168743897536,
+      "grad_norm": 0.44268131256103516,
+      "learning_rate": 0.0001,
+      "loss": 1.7158,
+      "step": 6827
+    },
+    {
+      "epoch": 0.7843317442995807,
+      "grad_norm": 0.46175646781921387,
+      "learning_rate": 0.0001,
+      "loss": 1.7005,
+      "step": 6828
+    },
+    {
+      "epoch": 0.7844466142094079,
+      "grad_norm": 0.4507410228252411,
+      "learning_rate": 0.0001,
+      "loss": 1.415,
+      "step": 6829
+    },
+    {
+      "epoch": 0.784561484119235,
+      "grad_norm": 0.42673787474632263,
+      "learning_rate": 0.0001,
+      "loss": 1.4747,
+      "step": 6830
+    },
+    {
+      "epoch": 0.7846763540290621,
+      "grad_norm": 0.42640364170074463,
+      "learning_rate": 0.0001,
+      "loss": 1.3069,
+      "step": 6831
+    },
+    {
+      "epoch": 0.7847912239388892,
+      "grad_norm": 0.46200641989707947,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 6832
+    },
+    {
+      "epoch": 0.7849060938487163,
+      "grad_norm": 0.43748313188552856,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 6833
+    },
+    {
+      "epoch": 0.7850209637585435,
+      "grad_norm": 0.49261534214019775,
+      "learning_rate": 0.0001,
+      "loss": 1.9485,
+      "step": 6834
+    },
+    {
+      "epoch": 0.7851358336683706,
+      "grad_norm": 0.4890846908092499,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 6835
+    },
+    {
+      "epoch": 0.7852507035781977,
+      "grad_norm": 0.4594542384147644,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 6836
+    },
+    {
+      "epoch": 0.7853655734880248,
+      "grad_norm": 0.434871107339859,
+      "learning_rate": 0.0001,
+      "loss": 1.5575,
+      "step": 6837
+    },
+    {
+      "epoch": 0.785480443397852,
+      "grad_norm": 0.45254603028297424,
+      "learning_rate": 0.0001,
+      "loss": 1.6604,
+      "step": 6838
+    },
+    {
+      "epoch": 0.7855953133076791,
+      "grad_norm": 0.46842509508132935,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 6839
+    },
+    {
+      "epoch": 0.7857101832175062,
+      "grad_norm": 0.45516836643218994,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7858250531273333,
+      "grad_norm": 0.4427262246608734,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 6841
+    },
+    {
+      "epoch": 0.7859399230371604,
+      "grad_norm": 0.4520769417285919,
+      "learning_rate": 0.0001,
+      "loss": 1.6729,
+      "step": 6842
+    },
+    {
+      "epoch": 0.7860547929469875,
+      "grad_norm": 0.5243545770645142,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 6843
+    },
+    {
+      "epoch": 0.7861696628568147,
+      "grad_norm": 0.4461134374141693,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 6844
+    },
+    {
+      "epoch": 0.7862845327666418,
+      "grad_norm": 0.48180684447288513,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 6845
+    },
+    {
+      "epoch": 0.7863994026764689,
+      "grad_norm": 0.4266572594642639,
+      "learning_rate": 0.0001,
+      "loss": 1.4144,
+      "step": 6846
+    },
+    {
+      "epoch": 0.786514272586296,
+      "grad_norm": 0.43107786774635315,
+      "learning_rate": 0.0001,
+      "loss": 1.5917,
+      "step": 6847
+    },
+    {
+      "epoch": 0.7866291424961231,
+      "grad_norm": 0.49343347549438477,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 6848
+    },
+    {
+      "epoch": 0.7867440124059503,
+      "grad_norm": 0.4486045837402344,
+      "learning_rate": 0.0001,
+      "loss": 1.2793,
+      "step": 6849
+    },
+    {
+      "epoch": 0.7868588823157774,
+      "grad_norm": 0.4293416440486908,
+      "learning_rate": 0.0001,
+      "loss": 1.5555,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7869737522256045,
+      "grad_norm": 0.48061510920524597,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 6851
+    },
+    {
+      "epoch": 0.7870886221354316,
+      "grad_norm": 0.4739622175693512,
+      "learning_rate": 0.0001,
+      "loss": 1.7332,
+      "step": 6852
+    },
+    {
+      "epoch": 0.7872034920452587,
+      "grad_norm": 0.4450550675392151,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 6853
+    },
+    {
+      "epoch": 0.7873183619550859,
+      "grad_norm": 0.4692426025867462,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 6854
+    },
+    {
+      "epoch": 0.787433231864913,
+      "grad_norm": 0.4459279775619507,
+      "learning_rate": 0.0001,
+      "loss": 1.4608,
+      "step": 6855
+    },
+    {
+      "epoch": 0.7875481017747401,
+      "grad_norm": 0.44502994418144226,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 6856
+    },
+    {
+      "epoch": 0.7876629716845672,
+      "grad_norm": 0.4621850550174713,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 6857
+    },
+    {
+      "epoch": 0.7877778415943943,
+      "grad_norm": 0.40475985407829285,
+      "learning_rate": 0.0001,
+      "loss": 1.3404,
+      "step": 6858
+    },
+    {
+      "epoch": 0.7878927115042215,
+      "grad_norm": 0.47660937905311584,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 6859
+    },
+    {
+      "epoch": 0.7880075814140486,
+      "grad_norm": 0.44972842931747437,
+      "learning_rate": 0.0001,
+      "loss": 1.6736,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7881224513238757,
+      "grad_norm": 0.454998642206192,
+      "learning_rate": 0.0001,
+      "loss": 1.5207,
+      "step": 6861
+    },
+    {
+      "epoch": 0.7882373212337028,
+      "grad_norm": 0.4434056580066681,
+      "learning_rate": 0.0001,
+      "loss": 1.5593,
+      "step": 6862
+    },
+    {
+      "epoch": 0.78835219114353,
+      "grad_norm": 0.4378054141998291,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 6863
+    },
+    {
+      "epoch": 0.7884670610533571,
+      "grad_norm": 0.43698400259017944,
+      "learning_rate": 0.0001,
+      "loss": 1.7174,
+      "step": 6864
+    },
+    {
+      "epoch": 0.7885819309631842,
+      "grad_norm": 0.4555806815624237,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 6865
+    },
+    {
+      "epoch": 0.7886968008730113,
+      "grad_norm": 0.44483011960983276,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 6866
+    },
+    {
+      "epoch": 0.7888116707828384,
+      "grad_norm": 0.42852985858917236,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 6867
+    },
+    {
+      "epoch": 0.7889265406926655,
+      "grad_norm": 0.4609838128089905,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 6868
+    },
+    {
+      "epoch": 0.7890414106024927,
+      "grad_norm": 0.45633402466773987,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 6869
+    },
+    {
+      "epoch": 0.7891562805123198,
+      "grad_norm": 0.44641777873039246,
+      "learning_rate": 0.0001,
+      "loss": 1.4307,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7892711504221469,
+      "grad_norm": 0.45838481187820435,
+      "learning_rate": 0.0001,
+      "loss": 1.7874,
+      "step": 6871
+    },
+    {
+      "epoch": 0.789386020331974,
+      "grad_norm": 0.44452112913131714,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 6872
+    },
+    {
+      "epoch": 0.7895008902418011,
+      "grad_norm": 0.44366639852523804,
+      "learning_rate": 0.0001,
+      "loss": 1.6561,
+      "step": 6873
+    },
+    {
+      "epoch": 0.7896157601516283,
+      "grad_norm": 0.44662636518478394,
+      "learning_rate": 0.0001,
+      "loss": 1.5177,
+      "step": 6874
+    },
+    {
+      "epoch": 0.7897306300614554,
+      "grad_norm": 0.46313855051994324,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 6875
+    },
+    {
+      "epoch": 0.7898454999712825,
+      "grad_norm": 0.4705411195755005,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 6876
+    },
+    {
+      "epoch": 0.7899603698811096,
+      "grad_norm": 0.4569123387336731,
+      "learning_rate": 0.0001,
+      "loss": 1.617,
+      "step": 6877
+    },
+    {
+      "epoch": 0.7900752397909367,
+      "grad_norm": 0.5129233002662659,
+      "learning_rate": 0.0001,
+      "loss": 1.9261,
+      "step": 6878
+    },
+    {
+      "epoch": 0.7901901097007639,
+      "grad_norm": 0.4295032322406769,
+      "learning_rate": 0.0001,
+      "loss": 1.4843,
+      "step": 6879
+    },
+    {
+      "epoch": 0.790304979610591,
+      "grad_norm": 0.44028350710868835,
+      "learning_rate": 0.0001,
+      "loss": 1.7061,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7904198495204181,
+      "grad_norm": 0.48637375235557556,
+      "learning_rate": 0.0001,
+      "loss": 1.6503,
+      "step": 6881
+    },
+    {
+      "epoch": 0.7905347194302452,
+      "grad_norm": 0.4660567045211792,
+      "learning_rate": 0.0001,
+      "loss": 1.6903,
+      "step": 6882
+    },
+    {
+      "epoch": 0.7906495893400723,
+      "grad_norm": 0.4239864945411682,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 6883
+    },
+    {
+      "epoch": 0.7907644592498995,
+      "grad_norm": 0.5081626772880554,
+      "learning_rate": 0.0001,
+      "loss": 1.8917,
+      "step": 6884
+    },
+    {
+      "epoch": 0.7908793291597266,
+      "grad_norm": 0.47426649928092957,
+      "learning_rate": 0.0001,
+      "loss": 1.7486,
+      "step": 6885
+    },
+    {
+      "epoch": 0.7909941990695537,
+      "grad_norm": 0.47770363092422485,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 6886
+    },
+    {
+      "epoch": 0.7911090689793808,
+      "grad_norm": 0.44490906596183777,
+      "learning_rate": 0.0001,
+      "loss": 1.6119,
+      "step": 6887
+    },
+    {
+      "epoch": 0.791223938889208,
+      "grad_norm": 0.4341287612915039,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 6888
+    },
+    {
+      "epoch": 0.7913388087990351,
+      "grad_norm": 0.46785807609558105,
+      "learning_rate": 0.0001,
+      "loss": 1.6912,
+      "step": 6889
+    },
+    {
+      "epoch": 0.7914536787088622,
+      "grad_norm": 0.4538176655769348,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 6890
+    },
+    {
+      "epoch": 0.7915685486186893,
+      "grad_norm": 0.4823910593986511,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 6891
+    },
+    {
+      "epoch": 0.7916834185285164,
+      "grad_norm": 0.49208977818489075,
+      "learning_rate": 0.0001,
+      "loss": 1.7127,
+      "step": 6892
+    },
+    {
+      "epoch": 0.7917982884383435,
+      "grad_norm": 0.6321043968200684,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 6893
+    },
+    {
+      "epoch": 0.7919131583481707,
+      "grad_norm": 0.43280404806137085,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 6894
+    },
+    {
+      "epoch": 0.7920280282579978,
+      "grad_norm": 0.5009095072746277,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 6895
+    },
+    {
+      "epoch": 0.7921428981678249,
+      "grad_norm": 0.4481852650642395,
+      "learning_rate": 0.0001,
+      "loss": 1.6166,
+      "step": 6896
+    },
+    {
+      "epoch": 0.792257768077652,
+      "grad_norm": 0.4593367278575897,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 6897
+    },
+    {
+      "epoch": 0.7923726379874791,
+      "grad_norm": 0.5545166730880737,
+      "learning_rate": 0.0001,
+      "loss": 1.3038,
+      "step": 6898
+    },
+    {
+      "epoch": 0.7924875078973063,
+      "grad_norm": 0.4565151631832123,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 6899
+    },
+    {
+      "epoch": 0.7926023778071334,
+      "grad_norm": 0.45123347640037537,
+      "learning_rate": 0.0001,
+      "loss": 1.7472,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7927172477169605,
+      "grad_norm": 0.46333909034729004,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 6901
+    },
+    {
+      "epoch": 0.7928321176267876,
+      "grad_norm": 0.4783911108970642,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 6902
+    },
+    {
+      "epoch": 0.7929469875366147,
+      "grad_norm": 0.48153024911880493,
+      "learning_rate": 0.0001,
+      "loss": 1.6846,
+      "step": 6903
+    },
+    {
+      "epoch": 0.7930618574464419,
+      "grad_norm": 0.45357227325439453,
+      "learning_rate": 0.0001,
+      "loss": 1.6658,
+      "step": 6904
+    },
+    {
+      "epoch": 0.793176727356269,
+      "grad_norm": 0.44956380128860474,
+      "learning_rate": 0.0001,
+      "loss": 1.6623,
+      "step": 6905
+    },
+    {
+      "epoch": 0.7932915972660961,
+      "grad_norm": 0.4557205140590668,
+      "learning_rate": 0.0001,
+      "loss": 1.4921,
+      "step": 6906
+    },
+    {
+      "epoch": 0.7934064671759232,
+      "grad_norm": 0.45987358689308167,
+      "learning_rate": 0.0001,
+      "loss": 1.6181,
+      "step": 6907
+    },
+    {
+      "epoch": 0.7935213370857503,
+      "grad_norm": 0.4671670198440552,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 6908
+    },
+    {
+      "epoch": 0.7936362069955775,
+      "grad_norm": 0.4789045453071594,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 6909
+    },
+    {
+      "epoch": 0.7937510769054046,
+      "grad_norm": 0.4139240086078644,
+      "learning_rate": 0.0001,
+      "loss": 1.4065,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7938659468152317,
+      "grad_norm": 0.43465399742126465,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 6911
+    },
+    {
+      "epoch": 0.7939808167250588,
+      "grad_norm": 0.45922383666038513,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 6912
+    },
+    {
+      "epoch": 0.7940956866348859,
+      "grad_norm": 0.5036603808403015,
+      "learning_rate": 0.0001,
+      "loss": 1.7516,
+      "step": 6913
+    },
+    {
+      "epoch": 0.7942105565447131,
+      "grad_norm": 0.4896206855773926,
+      "learning_rate": 0.0001,
+      "loss": 1.6954,
+      "step": 6914
+    },
+    {
+      "epoch": 0.7943254264545402,
+      "grad_norm": 0.4837943911552429,
+      "learning_rate": 0.0001,
+      "loss": 1.7452,
+      "step": 6915
+    },
+    {
+      "epoch": 0.7944402963643673,
+      "grad_norm": 0.4519433081150055,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 6916
+    },
+    {
+      "epoch": 0.7945551662741944,
+      "grad_norm": 0.45447391271591187,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 6917
+    },
+    {
+      "epoch": 0.7946700361840215,
+      "grad_norm": 0.4731905460357666,
+      "learning_rate": 0.0001,
+      "loss": 1.8278,
+      "step": 6918
+    },
+    {
+      "epoch": 0.7947849060938487,
+      "grad_norm": 0.44193151593208313,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 6919
+    },
+    {
+      "epoch": 0.7948997760036758,
+      "grad_norm": 0.472493439912796,
+      "learning_rate": 0.0001,
+      "loss": 1.6412,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7950146459135029,
+      "grad_norm": 0.46209755539894104,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 6921
+    },
+    {
+      "epoch": 0.79512951582333,
+      "grad_norm": 0.41559138894081116,
+      "learning_rate": 0.0001,
+      "loss": 1.6852,
+      "step": 6922
+    },
+    {
+      "epoch": 0.7952443857331571,
+      "grad_norm": 0.44188040494918823,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 6923
+    },
+    {
+      "epoch": 0.7953592556429844,
+      "grad_norm": 0.4449104070663452,
+      "learning_rate": 0.0001,
+      "loss": 1.5437,
+      "step": 6924
+    },
+    {
+      "epoch": 0.7954741255528115,
+      "grad_norm": 0.45927226543426514,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 6925
+    },
+    {
+      "epoch": 0.7955889954626386,
+      "grad_norm": 0.532455563545227,
+      "learning_rate": 0.0001,
+      "loss": 1.7761,
+      "step": 6926
+    },
+    {
+      "epoch": 0.7957038653724657,
+      "grad_norm": 0.5386255979537964,
+      "learning_rate": 0.0001,
+      "loss": 1.7247,
+      "step": 6927
+    },
+    {
+      "epoch": 0.7958187352822929,
+      "grad_norm": 0.449303537607193,
+      "learning_rate": 0.0001,
+      "loss": 1.4939,
+      "step": 6928
+    },
+    {
+      "epoch": 0.79593360519212,
+      "grad_norm": 0.43787631392478943,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 6929
+    },
+    {
+      "epoch": 0.7960484751019471,
+      "grad_norm": 0.4425432085990906,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 6930
+    },
+    {
+      "epoch": 0.7961633450117742,
+      "grad_norm": 0.47082197666168213,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 6931
+    },
+    {
+      "epoch": 0.7962782149216013,
+      "grad_norm": 0.42969945073127747,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 6932
+    },
+    {
+      "epoch": 0.7963930848314285,
+      "grad_norm": 0.4878128468990326,
+      "learning_rate": 0.0001,
+      "loss": 1.4099,
+      "step": 6933
+    },
+    {
+      "epoch": 0.7965079547412556,
+      "grad_norm": 0.4496504068374634,
+      "learning_rate": 0.0001,
+      "loss": 1.4885,
+      "step": 6934
+    },
+    {
+      "epoch": 0.7966228246510827,
+      "grad_norm": 0.48936036229133606,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 6935
+    },
+    {
+      "epoch": 0.7967376945609098,
+      "grad_norm": 0.45953816175460815,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 6936
+    },
+    {
+      "epoch": 0.7968525644707369,
+      "grad_norm": 0.4707304537296295,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 6937
+    },
+    {
+      "epoch": 0.796967434380564,
+      "grad_norm": 0.4913303256034851,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 6938
+    },
+    {
+      "epoch": 0.7970823042903912,
+      "grad_norm": 0.4540485739707947,
+      "learning_rate": 0.0001,
+      "loss": 1.5723,
+      "step": 6939
+    },
+    {
+      "epoch": 0.7971971742002183,
+      "grad_norm": 0.4464291036128998,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7973120441100454,
+      "grad_norm": 0.4318457245826721,
+      "learning_rate": 0.0001,
+      "loss": 1.5276,
+      "step": 6941
+    },
+    {
+      "epoch": 0.7974269140198725,
+      "grad_norm": 0.47109347581863403,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 6942
+    },
+    {
+      "epoch": 0.7975417839296997,
+      "grad_norm": 0.47364145517349243,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 6943
+    },
+    {
+      "epoch": 0.7976566538395268,
+      "grad_norm": 0.5263649821281433,
+      "learning_rate": 0.0001,
+      "loss": 1.3896,
+      "step": 6944
+    },
+    {
+      "epoch": 0.7977715237493539,
+      "grad_norm": 0.45865774154663086,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 6945
+    },
+    {
+      "epoch": 0.797886393659181,
+      "grad_norm": 0.4584158658981323,
+      "learning_rate": 0.0001,
+      "loss": 1.7354,
+      "step": 6946
+    },
+    {
+      "epoch": 0.7980012635690081,
+      "grad_norm": 0.4830566346645355,
+      "learning_rate": 0.0001,
+      "loss": 1.525,
+      "step": 6947
+    },
+    {
+      "epoch": 0.7981161334788353,
+      "grad_norm": 0.4785566031932831,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 6948
+    },
+    {
+      "epoch": 0.7982310033886624,
+      "grad_norm": 0.4663696587085724,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 6949
+    },
+    {
+      "epoch": 0.7983458732984895,
+      "grad_norm": 0.48032259941101074,
+      "learning_rate": 0.0001,
+      "loss": 1.3853,
+      "step": 6950
+    },
+    {
+      "epoch": 0.7984607432083166,
+      "grad_norm": 0.4568888545036316,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 6951
+    },
+    {
+      "epoch": 0.7985756131181437,
+      "grad_norm": 0.458404541015625,
+      "learning_rate": 0.0001,
+      "loss": 1.5422,
+      "step": 6952
+    },
+    {
+      "epoch": 0.7986904830279709,
+      "grad_norm": 0.45047512650489807,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 6953
+    },
+    {
+      "epoch": 0.798805352937798,
+      "grad_norm": 0.4487355053424835,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 6954
+    },
+    {
+      "epoch": 0.7989202228476251,
+      "grad_norm": 0.46647393703460693,
+      "learning_rate": 0.0001,
+      "loss": 1.5918,
+      "step": 6955
+    },
+    {
+      "epoch": 0.7990350927574522,
+      "grad_norm": 0.4552063047885895,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 6956
+    },
+    {
+      "epoch": 0.7991499626672793,
+      "grad_norm": 0.5072106719017029,
+      "learning_rate": 0.0001,
+      "loss": 1.8092,
+      "step": 6957
+    },
+    {
+      "epoch": 0.7992648325771065,
+      "grad_norm": 0.4884167015552521,
+      "learning_rate": 0.0001,
+      "loss": 1.6688,
+      "step": 6958
+    },
+    {
+      "epoch": 0.7993797024869336,
+      "grad_norm": 0.4894075095653534,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 6959
+    },
+    {
+      "epoch": 0.7994945723967607,
+      "grad_norm": 0.5430724024772644,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7996094423065878,
+      "grad_norm": 0.49636295437812805,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 6961
+    },
+    {
+      "epoch": 0.7997243122164149,
+      "grad_norm": 0.46621257066726685,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 6962
+    },
+    {
+      "epoch": 0.799839182126242,
+      "grad_norm": 0.4584997296333313,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 6963
+    },
+    {
+      "epoch": 0.7999540520360692,
+      "grad_norm": 0.4953577518463135,
+      "learning_rate": 0.0001,
+      "loss": 1.744,
+      "step": 6964
+    },
+    {
+      "epoch": 0.8000689219458963,
+      "grad_norm": 0.5007856488227844,
+      "learning_rate": 0.0001,
+      "loss": 1.6076,
+      "step": 6965
+    },
+    {
+      "epoch": 0.8001837918557234,
+      "grad_norm": 0.46211037039756775,
+      "learning_rate": 0.0001,
+      "loss": 1.6078,
+      "step": 6966
+    },
+    {
+      "epoch": 0.8002986617655505,
+      "grad_norm": 0.47474467754364014,
+      "learning_rate": 0.0001,
+      "loss": 1.7794,
+      "step": 6967
+    },
+    {
+      "epoch": 0.8004135316753777,
+      "grad_norm": 0.4621082544326782,
+      "learning_rate": 0.0001,
+      "loss": 1.6307,
+      "step": 6968
+    },
+    {
+      "epoch": 0.8005284015852048,
+      "grad_norm": 0.4746955335140228,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 6969
+    },
+    {
+      "epoch": 0.8006432714950319,
+      "grad_norm": 0.44209524989128113,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 6970
+    },
+    {
+      "epoch": 0.800758141404859,
+      "grad_norm": 0.48901766538619995,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 6971
+    },
+    {
+      "epoch": 0.8008730113146861,
+      "grad_norm": 0.5113202333450317,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 6972
+    },
+    {
+      "epoch": 0.8009878812245133,
+      "grad_norm": 0.4625761806964874,
+      "learning_rate": 0.0001,
+      "loss": 1.6895,
+      "step": 6973
+    },
+    {
+      "epoch": 0.8011027511343404,
+      "grad_norm": 0.4590431749820709,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 6974
+    },
+    {
+      "epoch": 0.8012176210441675,
+      "grad_norm": 0.4370403289794922,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 6975
+    },
+    {
+      "epoch": 0.8013324909539946,
+      "grad_norm": 0.4700826406478882,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 6976
+    },
+    {
+      "epoch": 0.8014473608638217,
+      "grad_norm": 0.5367459058761597,
+      "learning_rate": 0.0001,
+      "loss": 1.7318,
+      "step": 6977
+    },
+    {
+      "epoch": 0.8015622307736489,
+      "grad_norm": 0.4377821087837219,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 6978
+    },
+    {
+      "epoch": 0.801677100683476,
+      "grad_norm": 0.45518580079078674,
+      "learning_rate": 0.0001,
+      "loss": 1.5111,
+      "step": 6979
+    },
+    {
+      "epoch": 0.8017919705933031,
+      "grad_norm": 0.4376772344112396,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 6980
+    },
+    {
+      "epoch": 0.8019068405031302,
+      "grad_norm": 0.45350295305252075,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 6981
+    },
+    {
+      "epoch": 0.8020217104129573,
+      "grad_norm": 0.4758756458759308,
+      "learning_rate": 0.0001,
+      "loss": 1.7442,
+      "step": 6982
+    },
+    {
+      "epoch": 0.8021365803227845,
+      "grad_norm": 0.43972933292388916,
+      "learning_rate": 0.0001,
+      "loss": 1.5419,
+      "step": 6983
+    },
+    {
+      "epoch": 0.8022514502326116,
+      "grad_norm": 0.45926135778427124,
+      "learning_rate": 0.0001,
+      "loss": 1.7564,
+      "step": 6984
+    },
+    {
+      "epoch": 0.8023663201424387,
+      "grad_norm": 0.46066513657569885,
+      "learning_rate": 0.0001,
+      "loss": 1.5823,
+      "step": 6985
+    },
+    {
+      "epoch": 0.8024811900522658,
+      "grad_norm": 0.4452419579029083,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 6986
+    },
+    {
+      "epoch": 0.8025960599620929,
+      "grad_norm": 0.44738146662712097,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 6987
+    },
+    {
+      "epoch": 0.80271092987192,
+      "grad_norm": 0.4505269229412079,
+      "learning_rate": 0.0001,
+      "loss": 1.5,
+      "step": 6988
+    },
+    {
+      "epoch": 0.8028257997817472,
+      "grad_norm": 0.4531947672367096,
+      "learning_rate": 0.0001,
+      "loss": 1.6508,
+      "step": 6989
+    },
+    {
+      "epoch": 0.8029406696915743,
+      "grad_norm": 0.5062926411628723,
+      "learning_rate": 0.0001,
+      "loss": 1.8187,
+      "step": 6990
+    },
+    {
+      "epoch": 0.8030555396014014,
+      "grad_norm": 0.472827285528183,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 6991
+    },
+    {
+      "epoch": 0.8031704095112285,
+      "grad_norm": 0.5587745308876038,
+      "learning_rate": 0.0001,
+      "loss": 1.7311,
+      "step": 6992
+    },
+    {
+      "epoch": 0.8032852794210557,
+      "grad_norm": 0.43462514877319336,
+      "learning_rate": 0.0001,
+      "loss": 1.3869,
+      "step": 6993
+    },
+    {
+      "epoch": 0.8034001493308828,
+      "grad_norm": 0.45932313799858093,
+      "learning_rate": 0.0001,
+      "loss": 1.7735,
+      "step": 6994
+    },
+    {
+      "epoch": 0.8035150192407099,
+      "grad_norm": 0.47527042031288147,
+      "learning_rate": 0.0001,
+      "loss": 1.7342,
+      "step": 6995
+    },
+    {
+      "epoch": 0.803629889150537,
+      "grad_norm": 0.4415908455848694,
+      "learning_rate": 0.0001,
+      "loss": 1.4238,
+      "step": 6996
+    },
+    {
+      "epoch": 0.8037447590603641,
+      "grad_norm": 0.4424525499343872,
+      "learning_rate": 0.0001,
+      "loss": 1.8068,
+      "step": 6997
+    },
+    {
+      "epoch": 0.8038596289701913,
+      "grad_norm": 0.45409083366394043,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 6998
+    },
+    {
+      "epoch": 0.8039744988800184,
+      "grad_norm": 0.5033005475997925,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 6999
+    },
+    {
+      "epoch": 0.8040893687898455,
+      "grad_norm": 0.4428901970386505,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 7000
+    },
+    {
+      "epoch": 0.8042042386996726,
+      "grad_norm": 0.4608984589576721,
+      "learning_rate": 0.0001,
+      "loss": 1.5062,
+      "step": 7001
+    },
+    {
+      "epoch": 0.8043191086094997,
+      "grad_norm": 0.47772228717803955,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 7002
+    },
+    {
+      "epoch": 0.8044339785193269,
+      "grad_norm": 0.4590427577495575,
+      "learning_rate": 0.0001,
+      "loss": 1.7222,
+      "step": 7003
+    },
+    {
+      "epoch": 0.804548848429154,
+      "grad_norm": 0.5008092522621155,
+      "learning_rate": 0.0001,
+      "loss": 1.8124,
+      "step": 7004
+    },
+    {
+      "epoch": 0.8046637183389811,
+      "grad_norm": 0.428852915763855,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 7005
+    },
+    {
+      "epoch": 0.8047785882488082,
+      "grad_norm": 0.46282047033309937,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 7006
+    },
+    {
+      "epoch": 0.8048934581586353,
+      "grad_norm": 0.4332880675792694,
+      "learning_rate": 0.0001,
+      "loss": 1.796,
+      "step": 7007
+    },
+    {
+      "epoch": 0.8050083280684625,
+      "grad_norm": 0.4656613767147064,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 7008
+    },
+    {
+      "epoch": 0.8051231979782896,
+      "grad_norm": 0.4826306104660034,
+      "learning_rate": 0.0001,
+      "loss": 1.8894,
+      "step": 7009
+    },
+    {
+      "epoch": 0.8052380678881167,
+      "grad_norm": 0.4590745270252228,
+      "learning_rate": 0.0001,
+      "loss": 1.7591,
+      "step": 7010
+    },
+    {
+      "epoch": 0.8053529377979438,
+      "grad_norm": 0.48546114563941956,
+      "learning_rate": 0.0001,
+      "loss": 1.7585,
+      "step": 7011
+    },
+    {
+      "epoch": 0.8054678077077709,
+      "grad_norm": 0.4384916424751282,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 7012
+    },
+    {
+      "epoch": 0.805582677617598,
+      "grad_norm": 0.47794288396835327,
+      "learning_rate": 0.0001,
+      "loss": 1.7491,
+      "step": 7013
+    },
+    {
+      "epoch": 0.8056975475274252,
+      "grad_norm": 0.4551166892051697,
+      "learning_rate": 0.0001,
+      "loss": 1.6755,
+      "step": 7014
+    },
+    {
+      "epoch": 0.8058124174372523,
+      "grad_norm": 0.4861622452735901,
+      "learning_rate": 0.0001,
+      "loss": 1.5717,
+      "step": 7015
+    },
+    {
+      "epoch": 0.8059272873470794,
+      "grad_norm": 0.4588170051574707,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 7016
+    },
+    {
+      "epoch": 0.8060421572569065,
+      "grad_norm": 0.46102410554885864,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 7017
+    },
+    {
+      "epoch": 0.8061570271667337,
+      "grad_norm": 0.44939014315605164,
+      "learning_rate": 0.0001,
+      "loss": 1.6842,
+      "step": 7018
+    },
+    {
+      "epoch": 0.8062718970765608,
+      "grad_norm": 0.4488585293292999,
+      "learning_rate": 0.0001,
+      "loss": 1.485,
+      "step": 7019
+    },
+    {
+      "epoch": 0.8063867669863879,
+      "grad_norm": 0.4329974055290222,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 7020
+    },
+    {
+      "epoch": 0.806501636896215,
+      "grad_norm": 0.4898880124092102,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 7021
+    },
+    {
+      "epoch": 0.8066165068060421,
+      "grad_norm": 0.4246252477169037,
+      "learning_rate": 0.0001,
+      "loss": 1.4405,
+      "step": 7022
+    },
+    {
+      "epoch": 0.8067313767158693,
+      "grad_norm": 0.49754396080970764,
+      "learning_rate": 0.0001,
+      "loss": 1.4626,
+      "step": 7023
+    },
+    {
+      "epoch": 0.8068462466256964,
+      "grad_norm": 0.4949999749660492,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 7024
+    },
+    {
+      "epoch": 0.8069611165355235,
+      "grad_norm": 0.4603281617164612,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 7025
+    },
+    {
+      "epoch": 0.8070759864453506,
+      "grad_norm": 0.4442635476589203,
+      "learning_rate": 0.0001,
+      "loss": 1.6615,
+      "step": 7026
+    },
+    {
+      "epoch": 0.8071908563551777,
+      "grad_norm": 0.46228450536727905,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 7027
+    },
+    {
+      "epoch": 0.8073057262650049,
+      "grad_norm": 0.46300187706947327,
+      "learning_rate": 0.0001,
+      "loss": 1.5931,
+      "step": 7028
+    },
+    {
+      "epoch": 0.807420596174832,
+      "grad_norm": 0.4652291536331177,
+      "learning_rate": 0.0001,
+      "loss": 1.7684,
+      "step": 7029
+    },
+    {
+      "epoch": 0.8075354660846591,
+      "grad_norm": 0.4570204019546509,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 7030
+    },
+    {
+      "epoch": 0.8076503359944862,
+      "grad_norm": 0.453274667263031,
+      "learning_rate": 0.0001,
+      "loss": 1.5297,
+      "step": 7031
+    },
+    {
+      "epoch": 0.8077652059043133,
+      "grad_norm": 0.4465121030807495,
+      "learning_rate": 0.0001,
+      "loss": 1.4167,
+      "step": 7032
+    },
+    {
+      "epoch": 0.8078800758141405,
+      "grad_norm": 0.4628574550151825,
+      "learning_rate": 0.0001,
+      "loss": 1.5498,
+      "step": 7033
+    },
+    {
+      "epoch": 0.8079949457239676,
+      "grad_norm": 0.4569374918937683,
+      "learning_rate": 0.0001,
+      "loss": 1.4265,
+      "step": 7034
+    },
+    {
+      "epoch": 0.8081098156337947,
+      "grad_norm": 0.43578875064849854,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 7035
+    },
+    {
+      "epoch": 0.8082246855436218,
+      "grad_norm": 0.45690855383872986,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 7036
+    },
+    {
+      "epoch": 0.8083395554534489,
+      "grad_norm": 0.4643135666847229,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 7037
+    },
+    {
+      "epoch": 0.808454425363276,
+      "grad_norm": 0.4422778785228729,
+      "learning_rate": 0.0001,
+      "loss": 1.4945,
+      "step": 7038
+    },
+    {
+      "epoch": 0.8085692952731032,
+      "grad_norm": 0.5053475499153137,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 7039
+    },
+    {
+      "epoch": 0.8086841651829303,
+      "grad_norm": 0.4230154752731323,
+      "learning_rate": 0.0001,
+      "loss": 1.4409,
+      "step": 7040
+    },
+    {
+      "epoch": 0.8087990350927574,
+      "grad_norm": 0.45723220705986023,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 7041
+    },
+    {
+      "epoch": 0.8089139050025845,
+      "grad_norm": 0.43414244055747986,
+      "learning_rate": 0.0001,
+      "loss": 1.4774,
+      "step": 7042
+    },
+    {
+      "epoch": 0.8090287749124117,
+      "grad_norm": 0.45168161392211914,
+      "learning_rate": 0.0001,
+      "loss": 1.468,
+      "step": 7043
+    },
+    {
+      "epoch": 0.8091436448222388,
+      "grad_norm": 0.4413895308971405,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 7044
+    },
+    {
+      "epoch": 0.8092585147320659,
+      "grad_norm": 0.45228877663612366,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 7045
+    },
+    {
+      "epoch": 0.809373384641893,
+      "grad_norm": 0.4586316645145416,
+      "learning_rate": 0.0001,
+      "loss": 1.6476,
+      "step": 7046
+    },
+    {
+      "epoch": 0.8094882545517201,
+      "grad_norm": 0.4477645456790924,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 7047
+    },
+    {
+      "epoch": 0.8096031244615473,
+      "grad_norm": 0.5138921737670898,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 7048
+    },
+    {
+      "epoch": 0.8097179943713744,
+      "grad_norm": 0.4743337333202362,
+      "learning_rate": 0.0001,
+      "loss": 1.8095,
+      "step": 7049
+    },
+    {
+      "epoch": 0.8098328642812015,
+      "grad_norm": 0.47952887415885925,
+      "learning_rate": 0.0001,
+      "loss": 1.8992,
+      "step": 7050
+    },
+    {
+      "epoch": 0.8099477341910286,
+      "grad_norm": 0.4406343102455139,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 7051
+    },
+    {
+      "epoch": 0.8100626041008557,
+      "grad_norm": 0.5044456720352173,
+      "learning_rate": 0.0001,
+      "loss": 1.6264,
+      "step": 7052
+    },
+    {
+      "epoch": 0.8101774740106829,
+      "grad_norm": 0.45316067337989807,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 7053
+    },
+    {
+      "epoch": 0.81029234392051,
+      "grad_norm": 0.4341808259487152,
+      "learning_rate": 0.0001,
+      "loss": 1.3805,
+      "step": 7054
+    },
+    {
+      "epoch": 0.8104072138303371,
+      "grad_norm": 0.44615742564201355,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 7055
+    },
+    {
+      "epoch": 0.8105220837401642,
+      "grad_norm": 0.5084645748138428,
+      "learning_rate": 0.0001,
+      "loss": 1.4708,
+      "step": 7056
+    },
+    {
+      "epoch": 0.8106369536499913,
+      "grad_norm": 0.4395124912261963,
+      "learning_rate": 0.0001,
+      "loss": 1.4493,
+      "step": 7057
+    },
+    {
+      "epoch": 0.8107518235598185,
+      "grad_norm": 0.4524141550064087,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 7058
+    },
+    {
+      "epoch": 0.8108666934696456,
+      "grad_norm": 0.4532511532306671,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 7059
+    },
+    {
+      "epoch": 0.8109815633794727,
+      "grad_norm": 0.45477429032325745,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 7060
+    },
+    {
+      "epoch": 0.8110964332892999,
+      "grad_norm": 0.4670192003250122,
+      "learning_rate": 0.0001,
+      "loss": 1.7155,
+      "step": 7061
+    },
+    {
+      "epoch": 0.811211303199127,
+      "grad_norm": 0.47830730676651,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 7062
+    },
+    {
+      "epoch": 0.8113261731089542,
+      "grad_norm": 0.4505847692489624,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 7063
+    },
+    {
+      "epoch": 0.8114410430187813,
+      "grad_norm": 0.44358721375465393,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 7064
+    },
+    {
+      "epoch": 0.8115559129286084,
+      "grad_norm": 0.45430463552474976,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 7065
+    },
+    {
+      "epoch": 0.8116707828384355,
+      "grad_norm": 0.40954452753067017,
+      "learning_rate": 0.0001,
+      "loss": 1.4936,
+      "step": 7066
+    },
+    {
+      "epoch": 0.8117856527482626,
+      "grad_norm": 0.4465819001197815,
+      "learning_rate": 0.0001,
+      "loss": 1.641,
+      "step": 7067
+    },
+    {
+      "epoch": 0.8119005226580898,
+      "grad_norm": 0.45103660225868225,
+      "learning_rate": 0.0001,
+      "loss": 1.5577,
+      "step": 7068
+    },
+    {
+      "epoch": 0.8120153925679169,
+      "grad_norm": 0.44100210070610046,
+      "learning_rate": 0.0001,
+      "loss": 1.5393,
+      "step": 7069
+    },
+    {
+      "epoch": 0.812130262477744,
+      "grad_norm": 0.45234718918800354,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 7070
+    },
+    {
+      "epoch": 0.8122451323875711,
+      "grad_norm": 0.4711310565471649,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 7071
+    },
+    {
+      "epoch": 0.8123600022973982,
+      "grad_norm": 0.4617573618888855,
+      "learning_rate": 0.0001,
+      "loss": 1.6191,
+      "step": 7072
+    },
+    {
+      "epoch": 0.8124748722072254,
+      "grad_norm": 0.4681065082550049,
+      "learning_rate": 0.0001,
+      "loss": 1.3769,
+      "step": 7073
+    },
+    {
+      "epoch": 0.8125897421170525,
+      "grad_norm": 0.500937819480896,
+      "learning_rate": 0.0001,
+      "loss": 1.8512,
+      "step": 7074
+    },
+    {
+      "epoch": 0.8127046120268796,
+      "grad_norm": 0.5059450268745422,
+      "learning_rate": 0.0001,
+      "loss": 1.6661,
+      "step": 7075
+    },
+    {
+      "epoch": 0.8128194819367067,
+      "grad_norm": 0.49339231848716736,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 7076
+    },
+    {
+      "epoch": 0.8129343518465338,
+      "grad_norm": 0.4886321723461151,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 7077
+    },
+    {
+      "epoch": 0.813049221756361,
+      "grad_norm": 0.4459485113620758,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 7078
+    },
+    {
+      "epoch": 0.8131640916661881,
+      "grad_norm": 0.45102250576019287,
+      "learning_rate": 0.0001,
+      "loss": 1.6559,
+      "step": 7079
+    },
+    {
+      "epoch": 0.8132789615760152,
+      "grad_norm": 0.46558091044425964,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 7080
+    },
+    {
+      "epoch": 0.8133938314858423,
+      "grad_norm": 0.45526906847953796,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 7081
+    },
+    {
+      "epoch": 0.8135087013956694,
+      "grad_norm": 0.44742417335510254,
+      "learning_rate": 0.0001,
+      "loss": 1.6513,
+      "step": 7082
+    },
+    {
+      "epoch": 0.8136235713054966,
+      "grad_norm": 0.44720160961151123,
+      "learning_rate": 0.0001,
+      "loss": 1.5228,
+      "step": 7083
+    },
+    {
+      "epoch": 0.8137384412153237,
+      "grad_norm": 0.4373728632926941,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 7084
+    },
+    {
+      "epoch": 0.8138533111251508,
+      "grad_norm": 0.4450926184654236,
+      "learning_rate": 0.0001,
+      "loss": 1.4912,
+      "step": 7085
+    },
+    {
+      "epoch": 0.8139681810349779,
+      "grad_norm": 0.44571229815483093,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 7086
+    },
+    {
+      "epoch": 0.814083050944805,
+      "grad_norm": 0.49900081753730774,
+      "learning_rate": 0.0001,
+      "loss": 1.5591,
+      "step": 7087
+    },
+    {
+      "epoch": 0.8141979208546322,
+      "grad_norm": 0.440686970949173,
+      "learning_rate": 0.0001,
+      "loss": 1.6891,
+      "step": 7088
+    },
+    {
+      "epoch": 0.8143127907644593,
+      "grad_norm": 0.5091794729232788,
+      "learning_rate": 0.0001,
+      "loss": 1.7744,
+      "step": 7089
+    },
+    {
+      "epoch": 0.8144276606742864,
+      "grad_norm": 0.4285080134868622,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 7090
+    },
+    {
+      "epoch": 0.8145425305841135,
+      "grad_norm": 0.43077725172042847,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 7091
+    },
+    {
+      "epoch": 0.8146574004939406,
+      "grad_norm": 0.46352505683898926,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 7092
+    },
+    {
+      "epoch": 0.8147722704037678,
+      "grad_norm": 0.44118359684944153,
+      "learning_rate": 0.0001,
+      "loss": 1.6337,
+      "step": 7093
+    },
+    {
+      "epoch": 0.8148871403135949,
+      "grad_norm": 0.45531347393989563,
+      "learning_rate": 0.0001,
+      "loss": 1.6527,
+      "step": 7094
+    },
+    {
+      "epoch": 0.815002010223422,
+      "grad_norm": 0.5839067697525024,
+      "learning_rate": 0.0001,
+      "loss": 1.6389,
+      "step": 7095
+    },
+    {
+      "epoch": 0.8151168801332491,
+      "grad_norm": 0.4553830623626709,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 7096
+    },
+    {
+      "epoch": 0.8152317500430762,
+      "grad_norm": 0.4333423972129822,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 7097
+    },
+    {
+      "epoch": 0.8153466199529034,
+      "grad_norm": 0.4156707525253296,
+      "learning_rate": 0.0001,
+      "loss": 1.2689,
+      "step": 7098
+    },
+    {
+      "epoch": 0.8154614898627305,
+      "grad_norm": 0.5249226093292236,
+      "learning_rate": 0.0001,
+      "loss": 1.8285,
+      "step": 7099
+    },
+    {
+      "epoch": 0.8155763597725576,
+      "grad_norm": 0.45831164717674255,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 7100
+    },
+    {
+      "epoch": 0.8156912296823847,
+      "grad_norm": 0.5419580340385437,
+      "learning_rate": 0.0001,
+      "loss": 1.9262,
+      "step": 7101
+    },
+    {
+      "epoch": 0.8158060995922118,
+      "grad_norm": 0.46825796365737915,
+      "learning_rate": 0.0001,
+      "loss": 1.5954,
+      "step": 7102
+    },
+    {
+      "epoch": 0.815920969502039,
+      "grad_norm": 0.47406765818595886,
+      "learning_rate": 0.0001,
+      "loss": 1.5404,
+      "step": 7103
+    },
+    {
+      "epoch": 0.8160358394118661,
+      "grad_norm": 0.49613794684410095,
+      "learning_rate": 0.0001,
+      "loss": 1.7463,
+      "step": 7104
+    },
+    {
+      "epoch": 0.8161507093216932,
+      "grad_norm": 0.4632255733013153,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 7105
+    },
+    {
+      "epoch": 0.8162655792315203,
+      "grad_norm": 0.45750927925109863,
+      "learning_rate": 0.0001,
+      "loss": 1.6081,
+      "step": 7106
+    },
+    {
+      "epoch": 0.8163804491413474,
+      "grad_norm": 0.46657127141952515,
+      "learning_rate": 0.0001,
+      "loss": 1.7844,
+      "step": 7107
+    },
+    {
+      "epoch": 0.8164953190511746,
+      "grad_norm": 0.47650468349456787,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 7108
+    },
+    {
+      "epoch": 0.8166101889610017,
+      "grad_norm": 0.4252217411994934,
+      "learning_rate": 0.0001,
+      "loss": 1.5233,
+      "step": 7109
+    },
+    {
+      "epoch": 0.8167250588708288,
+      "grad_norm": 0.45260146260261536,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 7110
+    },
+    {
+      "epoch": 0.8168399287806559,
+      "grad_norm": 0.43198126554489136,
+      "learning_rate": 0.0001,
+      "loss": 1.5063,
+      "step": 7111
+    },
+    {
+      "epoch": 0.816954798690483,
+      "grad_norm": 0.48374143242836,
+      "learning_rate": 0.0001,
+      "loss": 1.72,
+      "step": 7112
+    },
+    {
+      "epoch": 0.8170696686003102,
+      "grad_norm": 0.5113547444343567,
+      "learning_rate": 0.0001,
+      "loss": 1.7583,
+      "step": 7113
+    },
+    {
+      "epoch": 0.8171845385101373,
+      "grad_norm": 0.46327292919158936,
+      "learning_rate": 0.0001,
+      "loss": 1.6309,
+      "step": 7114
+    },
+    {
+      "epoch": 0.8172994084199644,
+      "grad_norm": 0.4762055575847626,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 7115
+    },
+    {
+      "epoch": 0.8174142783297915,
+      "grad_norm": 0.5080848932266235,
+      "learning_rate": 0.0001,
+      "loss": 1.7955,
+      "step": 7116
+    },
+    {
+      "epoch": 0.8175291482396186,
+      "grad_norm": 0.47820234298706055,
+      "learning_rate": 0.0001,
+      "loss": 1.736,
+      "step": 7117
+    },
+    {
+      "epoch": 0.8176440181494458,
+      "grad_norm": 0.4423477351665497,
+      "learning_rate": 0.0001,
+      "loss": 1.5467,
+      "step": 7118
+    },
+    {
+      "epoch": 0.8177588880592729,
+      "grad_norm": 0.4471362829208374,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 7119
+    },
+    {
+      "epoch": 0.8178737579691,
+      "grad_norm": 0.4536069631576538,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 7120
+    },
+    {
+      "epoch": 0.8179886278789271,
+      "grad_norm": 0.4437789022922516,
+      "learning_rate": 0.0001,
+      "loss": 1.4918,
+      "step": 7121
+    },
+    {
+      "epoch": 0.8181034977887542,
+      "grad_norm": 0.4352627992630005,
+      "learning_rate": 0.0001,
+      "loss": 1.5588,
+      "step": 7122
+    },
+    {
+      "epoch": 0.8182183676985814,
+      "grad_norm": 0.4951383173465729,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 7123
+    },
+    {
+      "epoch": 0.8183332376084085,
+      "grad_norm": 0.47660359740257263,
+      "learning_rate": 0.0001,
+      "loss": 1.649,
+      "step": 7124
+    },
+    {
+      "epoch": 0.8184481075182356,
+      "grad_norm": 0.4454366862773895,
+      "learning_rate": 0.0001,
+      "loss": 1.647,
+      "step": 7125
+    },
+    {
+      "epoch": 0.8185629774280627,
+      "grad_norm": 0.47320306301116943,
+      "learning_rate": 0.0001,
+      "loss": 1.8041,
+      "step": 7126
+    },
+    {
+      "epoch": 0.8186778473378898,
+      "grad_norm": 0.4817422330379486,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 7127
+    },
+    {
+      "epoch": 0.818792717247717,
+      "grad_norm": 0.4612285792827606,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 7128
+    },
+    {
+      "epoch": 0.8189075871575441,
+      "grad_norm": 0.4710875153541565,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 7129
+    },
+    {
+      "epoch": 0.8190224570673712,
+      "grad_norm": 0.4853847324848175,
+      "learning_rate": 0.0001,
+      "loss": 1.7546,
+      "step": 7130
+    },
+    {
+      "epoch": 0.8191373269771983,
+      "grad_norm": 0.46727731823921204,
+      "learning_rate": 0.0001,
+      "loss": 1.7467,
+      "step": 7131
+    },
+    {
+      "epoch": 0.8192521968870254,
+      "grad_norm": 0.44645458459854126,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 7132
+    },
+    {
+      "epoch": 0.8193670667968526,
+      "grad_norm": 0.4743711054325104,
+      "learning_rate": 0.0001,
+      "loss": 1.7297,
+      "step": 7133
+    },
+    {
+      "epoch": 0.8194819367066797,
+      "grad_norm": 0.45293739438056946,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 7134
+    },
+    {
+      "epoch": 0.8195968066165068,
+      "grad_norm": 0.46416231989860535,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 7135
+    },
+    {
+      "epoch": 0.8197116765263339,
+      "grad_norm": 0.45859238505363464,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 7136
+    },
+    {
+      "epoch": 0.819826546436161,
+      "grad_norm": 0.45095375180244446,
+      "learning_rate": 0.0001,
+      "loss": 1.5282,
+      "step": 7137
+    },
+    {
+      "epoch": 0.8199414163459882,
+      "grad_norm": 0.5300037860870361,
+      "learning_rate": 0.0001,
+      "loss": 1.827,
+      "step": 7138
+    },
+    {
+      "epoch": 0.8200562862558153,
+      "grad_norm": 0.4848960340023041,
+      "learning_rate": 0.0001,
+      "loss": 1.4755,
+      "step": 7139
+    },
+    {
+      "epoch": 0.8201711561656424,
+      "grad_norm": 0.5703202486038208,
+      "learning_rate": 0.0001,
+      "loss": 1.2951,
+      "step": 7140
+    },
+    {
+      "epoch": 0.8202860260754695,
+      "grad_norm": 0.44500285387039185,
+      "learning_rate": 0.0001,
+      "loss": 1.5112,
+      "step": 7141
+    },
+    {
+      "epoch": 0.8204008959852966,
+      "grad_norm": 0.4689328968524933,
+      "learning_rate": 0.0001,
+      "loss": 1.4309,
+      "step": 7142
+    },
+    {
+      "epoch": 0.8205157658951238,
+      "grad_norm": 0.486074298620224,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 7143
+    },
+    {
+      "epoch": 0.8206306358049509,
+      "grad_norm": 0.4206257462501526,
+      "learning_rate": 0.0001,
+      "loss": 1.334,
+      "step": 7144
+    },
+    {
+      "epoch": 0.820745505714778,
+      "grad_norm": 0.4606837332248688,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 7145
+    },
+    {
+      "epoch": 0.8208603756246051,
+      "grad_norm": 0.5066717267036438,
+      "learning_rate": 0.0001,
+      "loss": 1.8185,
+      "step": 7146
+    },
+    {
+      "epoch": 0.8209752455344322,
+      "grad_norm": 0.46547627449035645,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 7147
+    },
+    {
+      "epoch": 0.8210901154442594,
+      "grad_norm": 0.4761962294578552,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 7148
+    },
+    {
+      "epoch": 0.8212049853540865,
+      "grad_norm": 0.459431916475296,
+      "learning_rate": 0.0001,
+      "loss": 1.66,
+      "step": 7149
+    },
+    {
+      "epoch": 0.8213198552639136,
+      "grad_norm": 0.4470955729484558,
+      "learning_rate": 0.0001,
+      "loss": 1.6065,
+      "step": 7150
+    },
+    {
+      "epoch": 0.8214347251737407,
+      "grad_norm": 0.48236003518104553,
+      "learning_rate": 0.0001,
+      "loss": 1.8159,
+      "step": 7151
+    },
+    {
+      "epoch": 0.8215495950835678,
+      "grad_norm": 0.46525686979293823,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 7152
+    },
+    {
+      "epoch": 0.821664464993395,
+      "grad_norm": 0.41665059328079224,
+      "learning_rate": 0.0001,
+      "loss": 1.4355,
+      "step": 7153
+    },
+    {
+      "epoch": 0.8217793349032221,
+      "grad_norm": 0.518457293510437,
+      "learning_rate": 0.0001,
+      "loss": 1.83,
+      "step": 7154
+    },
+    {
+      "epoch": 0.8218942048130492,
+      "grad_norm": 0.4431942105293274,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 7155
+    },
+    {
+      "epoch": 0.8220090747228763,
+      "grad_norm": 0.5172891616821289,
+      "learning_rate": 0.0001,
+      "loss": 1.6801,
+      "step": 7156
+    },
+    {
+      "epoch": 0.8221239446327034,
+      "grad_norm": 0.4258945882320404,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 7157
+    },
+    {
+      "epoch": 0.8222388145425306,
+      "grad_norm": 0.4604533016681671,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 7158
+    },
+    {
+      "epoch": 0.8223536844523577,
+      "grad_norm": 0.4496094584465027,
+      "learning_rate": 0.0001,
+      "loss": 1.4679,
+      "step": 7159
+    },
+    {
+      "epoch": 0.8224685543621848,
+      "grad_norm": 0.4688582420349121,
+      "learning_rate": 0.0001,
+      "loss": 1.668,
+      "step": 7160
+    },
+    {
+      "epoch": 0.8225834242720119,
+      "grad_norm": 0.47711849212646484,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 7161
+    },
+    {
+      "epoch": 0.822698294181839,
+      "grad_norm": 0.45683956146240234,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 7162
+    },
+    {
+      "epoch": 0.8228131640916662,
+      "grad_norm": 0.437517374753952,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 7163
+    },
+    {
+      "epoch": 0.8229280340014933,
+      "grad_norm": 0.4502145051956177,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 7164
+    },
+    {
+      "epoch": 0.8230429039113204,
+      "grad_norm": 0.4960477650165558,
+      "learning_rate": 0.0001,
+      "loss": 1.6825,
+      "step": 7165
+    },
+    {
+      "epoch": 0.8231577738211475,
+      "grad_norm": 0.4727974534034729,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 7166
+    },
+    {
+      "epoch": 0.8232726437309746,
+      "grad_norm": 0.4566989541053772,
+      "learning_rate": 0.0001,
+      "loss": 1.5257,
+      "step": 7167
+    },
+    {
+      "epoch": 0.8233875136408018,
+      "grad_norm": 0.49063435196876526,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 7168
+    },
+    {
+      "epoch": 0.8235023835506289,
+      "grad_norm": 0.4589118957519531,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 7169
+    },
+    {
+      "epoch": 0.823617253460456,
+      "grad_norm": 0.46506285667419434,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 7170
+    },
+    {
+      "epoch": 0.8237321233702831,
+      "grad_norm": 0.4581180810928345,
+      "learning_rate": 0.0001,
+      "loss": 1.6783,
+      "step": 7171
+    },
+    {
+      "epoch": 0.8238469932801102,
+      "grad_norm": 0.47631680965423584,
+      "learning_rate": 0.0001,
+      "loss": 1.6033,
+      "step": 7172
+    },
+    {
+      "epoch": 0.8239618631899374,
+      "grad_norm": 0.4644998013973236,
+      "learning_rate": 0.0001,
+      "loss": 1.6591,
+      "step": 7173
+    },
+    {
+      "epoch": 0.8240767330997645,
+      "grad_norm": 0.43831029534339905,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 7174
+    },
+    {
+      "epoch": 0.8241916030095916,
+      "grad_norm": 0.45947766304016113,
+      "learning_rate": 0.0001,
+      "loss": 1.6604,
+      "step": 7175
+    },
+    {
+      "epoch": 0.8243064729194187,
+      "grad_norm": 0.44559556245803833,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 7176
+    },
+    {
+      "epoch": 0.8244213428292458,
+      "grad_norm": 0.43358561396598816,
+      "learning_rate": 0.0001,
+      "loss": 1.5983,
+      "step": 7177
+    },
+    {
+      "epoch": 0.824536212739073,
+      "grad_norm": 0.4433882534503937,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 7178
+    },
+    {
+      "epoch": 0.8246510826489001,
+      "grad_norm": 0.48557913303375244,
+      "learning_rate": 0.0001,
+      "loss": 1.6611,
+      "step": 7179
+    },
+    {
+      "epoch": 0.8247659525587272,
+      "grad_norm": 0.4322436451911926,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 7180
+    },
+    {
+      "epoch": 0.8248808224685543,
+      "grad_norm": 0.4723256230354309,
+      "learning_rate": 0.0001,
+      "loss": 1.6417,
+      "step": 7181
+    },
+    {
+      "epoch": 0.8249956923783814,
+      "grad_norm": 0.4498218894004822,
+      "learning_rate": 0.0001,
+      "loss": 1.69,
+      "step": 7182
+    },
+    {
+      "epoch": 0.8251105622882086,
+      "grad_norm": 0.44787004590034485,
+      "learning_rate": 0.0001,
+      "loss": 1.8015,
+      "step": 7183
+    },
+    {
+      "epoch": 0.8252254321980357,
+      "grad_norm": 0.4494984745979309,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 7184
+    },
+    {
+      "epoch": 0.8253403021078628,
+      "grad_norm": 0.4542940557003021,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 7185
+    },
+    {
+      "epoch": 0.8254551720176899,
+      "grad_norm": 0.4573206901550293,
+      "learning_rate": 0.0001,
+      "loss": 1.731,
+      "step": 7186
+    },
+    {
+      "epoch": 0.825570041927517,
+      "grad_norm": 0.455084890127182,
+      "learning_rate": 0.0001,
+      "loss": 1.7468,
+      "step": 7187
+    },
+    {
+      "epoch": 0.8256849118373442,
+      "grad_norm": 0.4983527958393097,
+      "learning_rate": 0.0001,
+      "loss": 1.5253,
+      "step": 7188
+    },
+    {
+      "epoch": 0.8257997817471713,
+      "grad_norm": 0.457497239112854,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 7189
+    },
+    {
+      "epoch": 0.8259146516569984,
+      "grad_norm": 0.485844224691391,
+      "learning_rate": 0.0001,
+      "loss": 1.7502,
+      "step": 7190
+    },
+    {
+      "epoch": 0.8260295215668255,
+      "grad_norm": 0.47317448258399963,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 7191
+    },
+    {
+      "epoch": 0.8261443914766526,
+      "grad_norm": 0.4314495921134949,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 7192
+    },
+    {
+      "epoch": 0.8262592613864798,
+      "grad_norm": 0.45822349190711975,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 7193
+    },
+    {
+      "epoch": 0.8263741312963069,
+      "grad_norm": 0.4706357419490814,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 7194
+    },
+    {
+      "epoch": 0.826489001206134,
+      "grad_norm": 0.45781227946281433,
+      "learning_rate": 0.0001,
+      "loss": 1.4011,
+      "step": 7195
+    },
+    {
+      "epoch": 0.8266038711159611,
+      "grad_norm": 0.43155431747436523,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 7196
+    },
+    {
+      "epoch": 0.8267187410257882,
+      "grad_norm": 0.4829701781272888,
+      "learning_rate": 0.0001,
+      "loss": 1.6697,
+      "step": 7197
+    },
+    {
+      "epoch": 0.8268336109356155,
+      "grad_norm": 0.4185773432254791,
+      "learning_rate": 0.0001,
+      "loss": 1.4531,
+      "step": 7198
+    },
+    {
+      "epoch": 0.8269484808454426,
+      "grad_norm": 0.4704040288925171,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 7199
+    },
+    {
+      "epoch": 0.8270633507552697,
+      "grad_norm": 0.41344597935676575,
+      "learning_rate": 0.0001,
+      "loss": 1.4317,
+      "step": 7200
+    },
+    {
+      "epoch": 0.8271782206650968,
+      "grad_norm": 0.4467063546180725,
+      "learning_rate": 0.0001,
+      "loss": 1.3732,
+      "step": 7201
+    },
+    {
+      "epoch": 0.827293090574924,
+      "grad_norm": 0.4785305857658386,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 7202
+    },
+    {
+      "epoch": 0.8274079604847511,
+      "grad_norm": 0.4496002793312073,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 7203
+    },
+    {
+      "epoch": 0.8275228303945782,
+      "grad_norm": 0.45911693572998047,
+      "learning_rate": 0.0001,
+      "loss": 1.3529,
+      "step": 7204
+    },
+    {
+      "epoch": 0.8276377003044053,
+      "grad_norm": 0.4438045918941498,
+      "learning_rate": 0.0001,
+      "loss": 1.2912,
+      "step": 7205
+    },
+    {
+      "epoch": 0.8277525702142324,
+      "grad_norm": 0.4709499776363373,
+      "learning_rate": 0.0001,
+      "loss": 1.5936,
+      "step": 7206
+    },
+    {
+      "epoch": 0.8278674401240596,
+      "grad_norm": 0.4676428437232971,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 7207
+    },
+    {
+      "epoch": 0.8279823100338867,
+      "grad_norm": 0.5031962394714355,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 7208
+    },
+    {
+      "epoch": 0.8280971799437138,
+      "grad_norm": 0.49766260385513306,
+      "learning_rate": 0.0001,
+      "loss": 1.4382,
+      "step": 7209
+    },
+    {
+      "epoch": 0.8282120498535409,
+      "grad_norm": 0.5159705877304077,
+      "learning_rate": 0.0001,
+      "loss": 1.8591,
+      "step": 7210
+    },
+    {
+      "epoch": 0.828326919763368,
+      "grad_norm": 0.47950395941734314,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 7211
+    },
+    {
+      "epoch": 0.8284417896731952,
+      "grad_norm": 0.4581819772720337,
+      "learning_rate": 0.0001,
+      "loss": 1.6142,
+      "step": 7212
+    },
+    {
+      "epoch": 0.8285566595830223,
+      "grad_norm": 0.4366266429424286,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 7213
+    },
+    {
+      "epoch": 0.8286715294928494,
+      "grad_norm": 0.48760971426963806,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 7214
+    },
+    {
+      "epoch": 0.8287863994026765,
+      "grad_norm": 0.4329768419265747,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 7215
+    },
+    {
+      "epoch": 0.8289012693125036,
+      "grad_norm": 0.44413167238235474,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 7216
+    },
+    {
+      "epoch": 0.8290161392223308,
+      "grad_norm": 0.4911179840564728,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 7217
+    },
+    {
+      "epoch": 0.8291310091321579,
+      "grad_norm": 0.4415479302406311,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 7218
+    },
+    {
+      "epoch": 0.829245879041985,
+      "grad_norm": 0.5084326863288879,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 7219
+    },
+    {
+      "epoch": 0.8293607489518121,
+      "grad_norm": 0.48137006163597107,
+      "learning_rate": 0.0001,
+      "loss": 1.7523,
+      "step": 7220
+    },
+    {
+      "epoch": 0.8294756188616392,
+      "grad_norm": 0.4507937431335449,
+      "learning_rate": 0.0001,
+      "loss": 1.6697,
+      "step": 7221
+    },
+    {
+      "epoch": 0.8295904887714664,
+      "grad_norm": 0.4717406630516052,
+      "learning_rate": 0.0001,
+      "loss": 1.754,
+      "step": 7222
+    },
+    {
+      "epoch": 0.8297053586812935,
+      "grad_norm": 0.4703536927700043,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 7223
+    },
+    {
+      "epoch": 0.8298202285911206,
+      "grad_norm": 0.41335755586624146,
+      "learning_rate": 0.0001,
+      "loss": 1.4745,
+      "step": 7224
+    },
+    {
+      "epoch": 0.8299350985009477,
+      "grad_norm": 0.47170180082321167,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 7225
+    },
+    {
+      "epoch": 0.8300499684107748,
+      "grad_norm": 0.44753673672676086,
+      "learning_rate": 0.0001,
+      "loss": 1.3859,
+      "step": 7226
+    },
+    {
+      "epoch": 0.830164838320602,
+      "grad_norm": 0.44213739037513733,
+      "learning_rate": 0.0001,
+      "loss": 1.3913,
+      "step": 7227
+    },
+    {
+      "epoch": 0.8302797082304291,
+      "grad_norm": 0.46415868401527405,
+      "learning_rate": 0.0001,
+      "loss": 1.7794,
+      "step": 7228
+    },
+    {
+      "epoch": 0.8303945781402562,
+      "grad_norm": 0.47127893567085266,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 7229
+    },
+    {
+      "epoch": 0.8305094480500833,
+      "grad_norm": 0.48462972044944763,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 7230
+    },
+    {
+      "epoch": 0.8306243179599104,
+      "grad_norm": 0.43438246846199036,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 7231
+    },
+    {
+      "epoch": 0.8307391878697375,
+      "grad_norm": 0.44734856486320496,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 7232
+    },
+    {
+      "epoch": 0.8308540577795647,
+      "grad_norm": 0.4758615493774414,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 7233
+    },
+    {
+      "epoch": 0.8309689276893918,
+      "grad_norm": 0.5029293298721313,
+      "learning_rate": 0.0001,
+      "loss": 1.7695,
+      "step": 7234
+    },
+    {
+      "epoch": 0.8310837975992189,
+      "grad_norm": 0.4502490162849426,
+      "learning_rate": 0.0001,
+      "loss": 1.4654,
+      "step": 7235
+    },
+    {
+      "epoch": 0.831198667509046,
+      "grad_norm": 0.4694804251194,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 7236
+    },
+    {
+      "epoch": 0.8313135374188731,
+      "grad_norm": 0.4628579616546631,
+      "learning_rate": 0.0001,
+      "loss": 1.4772,
+      "step": 7237
+    },
+    {
+      "epoch": 0.8314284073287003,
+      "grad_norm": 0.40973448753356934,
+      "learning_rate": 0.0001,
+      "loss": 1.3017,
+      "step": 7238
+    },
+    {
+      "epoch": 0.8315432772385274,
+      "grad_norm": 0.48314985632896423,
+      "learning_rate": 0.0001,
+      "loss": 1.6824,
+      "step": 7239
+    },
+    {
+      "epoch": 0.8316581471483545,
+      "grad_norm": 0.43235471844673157,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 7240
+    },
+    {
+      "epoch": 0.8317730170581816,
+      "grad_norm": 0.45430463552474976,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 7241
+    },
+    {
+      "epoch": 0.8318878869680087,
+      "grad_norm": 0.4984131455421448,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 7242
+    },
+    {
+      "epoch": 0.8320027568778359,
+      "grad_norm": 0.46940264105796814,
+      "learning_rate": 0.0001,
+      "loss": 1.6658,
+      "step": 7243
+    },
+    {
+      "epoch": 0.832117626787663,
+      "grad_norm": 0.47424784302711487,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 7244
+    },
+    {
+      "epoch": 0.8322324966974901,
+      "grad_norm": 0.5137829780578613,
+      "learning_rate": 0.0001,
+      "loss": 1.6511,
+      "step": 7245
+    },
+    {
+      "epoch": 0.8323473666073172,
+      "grad_norm": 0.5260666012763977,
+      "learning_rate": 0.0001,
+      "loss": 1.7015,
+      "step": 7246
+    },
+    {
+      "epoch": 0.8324622365171443,
+      "grad_norm": 0.47583386301994324,
+      "learning_rate": 0.0001,
+      "loss": 1.504,
+      "step": 7247
+    },
+    {
+      "epoch": 0.8325771064269715,
+      "grad_norm": 0.4737547039985657,
+      "learning_rate": 0.0001,
+      "loss": 1.7114,
+      "step": 7248
+    },
+    {
+      "epoch": 0.8326919763367986,
+      "grad_norm": 0.49073606729507446,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 7249
+    },
+    {
+      "epoch": 0.8328068462466257,
+      "grad_norm": 0.4550478756427765,
+      "learning_rate": 0.0001,
+      "loss": 1.366,
+      "step": 7250
+    },
+    {
+      "epoch": 0.8329217161564528,
+      "grad_norm": 0.45904481410980225,
+      "learning_rate": 0.0001,
+      "loss": 1.4628,
+      "step": 7251
+    },
+    {
+      "epoch": 0.83303658606628,
+      "grad_norm": 0.4898592233657837,
+      "learning_rate": 0.0001,
+      "loss": 1.7453,
+      "step": 7252
+    },
+    {
+      "epoch": 0.8331514559761071,
+      "grad_norm": 0.46044033765792847,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 7253
+    },
+    {
+      "epoch": 0.8332663258859342,
+      "grad_norm": 0.4777674376964569,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 7254
+    },
+    {
+      "epoch": 0.8333811957957613,
+      "grad_norm": 0.44753503799438477,
+      "learning_rate": 0.0001,
+      "loss": 1.729,
+      "step": 7255
+    },
+    {
+      "epoch": 0.8334960657055884,
+      "grad_norm": 0.45856595039367676,
+      "learning_rate": 0.0001,
+      "loss": 1.5738,
+      "step": 7256
+    },
+    {
+      "epoch": 0.8336109356154155,
+      "grad_norm": 0.4056832790374756,
+      "learning_rate": 0.0001,
+      "loss": 1.2812,
+      "step": 7257
+    },
+    {
+      "epoch": 0.8337258055252427,
+      "grad_norm": 0.48094066977500916,
+      "learning_rate": 0.0001,
+      "loss": 1.5232,
+      "step": 7258
+    },
+    {
+      "epoch": 0.8338406754350698,
+      "grad_norm": 0.4848513603210449,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 7259
+    },
+    {
+      "epoch": 0.8339555453448969,
+      "grad_norm": 0.4800076484680176,
+      "learning_rate": 0.0001,
+      "loss": 1.4396,
+      "step": 7260
+    },
+    {
+      "epoch": 0.834070415254724,
+      "grad_norm": 0.47761738300323486,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 7261
+    },
+    {
+      "epoch": 0.8341852851645511,
+      "grad_norm": 0.4880167245864868,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 7262
+    },
+    {
+      "epoch": 0.8343001550743783,
+      "grad_norm": 0.464358925819397,
+      "learning_rate": 0.0001,
+      "loss": 1.7559,
+      "step": 7263
+    },
+    {
+      "epoch": 0.8344150249842054,
+      "grad_norm": 0.4656590521335602,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 7264
+    },
+    {
+      "epoch": 0.8345298948940325,
+      "grad_norm": 0.4818393290042877,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 7265
+    },
+    {
+      "epoch": 0.8346447648038596,
+      "grad_norm": 0.4994739294052124,
+      "learning_rate": 0.0001,
+      "loss": 1.4481,
+      "step": 7266
+    },
+    {
+      "epoch": 0.8347596347136867,
+      "grad_norm": 0.49908795952796936,
+      "learning_rate": 0.0001,
+      "loss": 1.6915,
+      "step": 7267
+    },
+    {
+      "epoch": 0.8348745046235139,
+      "grad_norm": 0.5030030608177185,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 7268
+    },
+    {
+      "epoch": 0.834989374533341,
+      "grad_norm": 0.4775954484939575,
+      "learning_rate": 0.0001,
+      "loss": 1.7861,
+      "step": 7269
+    },
+    {
+      "epoch": 0.8351042444431681,
+      "grad_norm": 0.41004353761672974,
+      "learning_rate": 0.0001,
+      "loss": 1.4602,
+      "step": 7270
+    },
+    {
+      "epoch": 0.8352191143529952,
+      "grad_norm": 0.49906715750694275,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 7271
+    },
+    {
+      "epoch": 0.8353339842628223,
+      "grad_norm": 0.4458923637866974,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 7272
+    },
+    {
+      "epoch": 0.8354488541726495,
+      "grad_norm": 0.4532411992549896,
+      "learning_rate": 0.0001,
+      "loss": 1.6273,
+      "step": 7273
+    },
+    {
+      "epoch": 0.8355637240824766,
+      "grad_norm": 0.4611133933067322,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 7274
+    },
+    {
+      "epoch": 0.8356785939923037,
+      "grad_norm": 0.4394255578517914,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 7275
+    },
+    {
+      "epoch": 0.8357934639021308,
+      "grad_norm": 0.4390329420566559,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 7276
+    },
+    {
+      "epoch": 0.835908333811958,
+      "grad_norm": 0.4387739300727844,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 7277
+    },
+    {
+      "epoch": 0.8360232037217851,
+      "grad_norm": 0.44605275988578796,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 7278
+    },
+    {
+      "epoch": 0.8361380736316122,
+      "grad_norm": 0.4986502528190613,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 7279
+    },
+    {
+      "epoch": 0.8362529435414393,
+      "grad_norm": 0.4799439013004303,
+      "learning_rate": 0.0001,
+      "loss": 1.7177,
+      "step": 7280
+    },
+    {
+      "epoch": 0.8363678134512664,
+      "grad_norm": 0.4725569784641266,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 7281
+    },
+    {
+      "epoch": 0.8364826833610935,
+      "grad_norm": 0.46974942088127136,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 7282
+    },
+    {
+      "epoch": 0.8365975532709207,
+      "grad_norm": 0.4600738286972046,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 7283
+    },
+    {
+      "epoch": 0.8367124231807478,
+      "grad_norm": 0.516070544719696,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 7284
+    },
+    {
+      "epoch": 0.8368272930905749,
+      "grad_norm": 0.44913583993911743,
+      "learning_rate": 0.0001,
+      "loss": 1.4985,
+      "step": 7285
+    },
+    {
+      "epoch": 0.836942163000402,
+      "grad_norm": 0.4450875222682953,
+      "learning_rate": 0.0001,
+      "loss": 1.4021,
+      "step": 7286
+    },
+    {
+      "epoch": 0.8370570329102291,
+      "grad_norm": 0.45671719312667847,
+      "learning_rate": 0.0001,
+      "loss": 1.4193,
+      "step": 7287
+    },
+    {
+      "epoch": 0.8371719028200563,
+      "grad_norm": 0.4811131954193115,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 7288
+    },
+    {
+      "epoch": 0.8372867727298834,
+      "grad_norm": 0.473019540309906,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 7289
+    },
+    {
+      "epoch": 0.8374016426397105,
+      "grad_norm": 0.4964619576931,
+      "learning_rate": 0.0001,
+      "loss": 1.4574,
+      "step": 7290
+    },
+    {
+      "epoch": 0.8375165125495376,
+      "grad_norm": 0.43981707096099854,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 7291
+    },
+    {
+      "epoch": 0.8376313824593647,
+      "grad_norm": 0.47483888268470764,
+      "learning_rate": 0.0001,
+      "loss": 1.7885,
+      "step": 7292
+    },
+    {
+      "epoch": 0.8377462523691919,
+      "grad_norm": 0.46889838576316833,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 7293
+    },
+    {
+      "epoch": 0.837861122279019,
+      "grad_norm": 0.46966785192489624,
+      "learning_rate": 0.0001,
+      "loss": 1.4182,
+      "step": 7294
+    },
+    {
+      "epoch": 0.8379759921888461,
+      "grad_norm": 0.4754125475883484,
+      "learning_rate": 0.0001,
+      "loss": 1.8329,
+      "step": 7295
+    },
+    {
+      "epoch": 0.8380908620986732,
+      "grad_norm": 0.44175228476524353,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 7296
+    },
+    {
+      "epoch": 0.8382057320085003,
+      "grad_norm": 0.45148342847824097,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 7297
+    },
+    {
+      "epoch": 0.8383206019183275,
+      "grad_norm": 0.46137553453445435,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 7298
+    },
+    {
+      "epoch": 0.8384354718281546,
+      "grad_norm": 0.49679750204086304,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 7299
+    },
+    {
+      "epoch": 0.8385503417379817,
+      "grad_norm": 0.44846269488334656,
+      "learning_rate": 0.0001,
+      "loss": 1.5111,
+      "step": 7300
+    },
+    {
+      "epoch": 0.8386652116478088,
+      "grad_norm": 0.43875476717948914,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 7301
+    },
+    {
+      "epoch": 0.838780081557636,
+      "grad_norm": 0.44996675848960876,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 7302
+    },
+    {
+      "epoch": 0.8388949514674631,
+      "grad_norm": 0.47677183151245117,
+      "learning_rate": 0.0001,
+      "loss": 1.7262,
+      "step": 7303
+    },
+    {
+      "epoch": 0.8390098213772902,
+      "grad_norm": 0.4702812433242798,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 7304
+    },
+    {
+      "epoch": 0.8391246912871173,
+      "grad_norm": 0.4662913978099823,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 7305
+    },
+    {
+      "epoch": 0.8392395611969444,
+      "grad_norm": 0.5055838227272034,
+      "learning_rate": 0.0001,
+      "loss": 1.747,
+      "step": 7306
+    },
+    {
+      "epoch": 0.8393544311067715,
+      "grad_norm": 0.47500503063201904,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 7307
+    },
+    {
+      "epoch": 0.8394693010165987,
+      "grad_norm": 0.48270219564437866,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 7308
+    },
+    {
+      "epoch": 0.8395841709264258,
+      "grad_norm": 0.48767292499542236,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 7309
+    },
+    {
+      "epoch": 0.8396990408362529,
+      "grad_norm": 0.46618691086769104,
+      "learning_rate": 0.0001,
+      "loss": 1.4696,
+      "step": 7310
+    },
+    {
+      "epoch": 0.83981391074608,
+      "grad_norm": 0.4929366111755371,
+      "learning_rate": 0.0001,
+      "loss": 1.683,
+      "step": 7311
+    },
+    {
+      "epoch": 0.8399287806559071,
+      "grad_norm": 0.4555191397666931,
+      "learning_rate": 0.0001,
+      "loss": 1.549,
+      "step": 7312
+    },
+    {
+      "epoch": 0.8400436505657343,
+      "grad_norm": 0.4752098321914673,
+      "learning_rate": 0.0001,
+      "loss": 1.8156,
+      "step": 7313
+    },
+    {
+      "epoch": 0.8401585204755614,
+      "grad_norm": 0.4447777271270752,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 7314
+    },
+    {
+      "epoch": 0.8402733903853885,
+      "grad_norm": 0.46375584602355957,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 7315
+    },
+    {
+      "epoch": 0.8403882602952156,
+      "grad_norm": 0.5160993933677673,
+      "learning_rate": 0.0001,
+      "loss": 1.7724,
+      "step": 7316
+    },
+    {
+      "epoch": 0.8405031302050427,
+      "grad_norm": 0.49875614047050476,
+      "learning_rate": 0.0001,
+      "loss": 1.7911,
+      "step": 7317
+    },
+    {
+      "epoch": 0.8406180001148699,
+      "grad_norm": 0.47104793787002563,
+      "learning_rate": 0.0001,
+      "loss": 1.5512,
+      "step": 7318
+    },
+    {
+      "epoch": 0.840732870024697,
+      "grad_norm": 0.47585365176200867,
+      "learning_rate": 0.0001,
+      "loss": 1.4807,
+      "step": 7319
+    },
+    {
+      "epoch": 0.8408477399345241,
+      "grad_norm": 0.48287901282310486,
+      "learning_rate": 0.0001,
+      "loss": 1.4909,
+      "step": 7320
+    },
+    {
+      "epoch": 0.8409626098443512,
+      "grad_norm": 0.4639113247394562,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 7321
+    },
+    {
+      "epoch": 0.8410774797541783,
+      "grad_norm": 0.4889727830886841,
+      "learning_rate": 0.0001,
+      "loss": 1.8168,
+      "step": 7322
+    },
+    {
+      "epoch": 0.8411923496640055,
+      "grad_norm": 0.44564002752304077,
+      "learning_rate": 0.0001,
+      "loss": 1.4868,
+      "step": 7323
+    },
+    {
+      "epoch": 0.8413072195738326,
+      "grad_norm": 0.4678811728954315,
+      "learning_rate": 0.0001,
+      "loss": 1.5586,
+      "step": 7324
+    },
+    {
+      "epoch": 0.8414220894836597,
+      "grad_norm": 0.483021080493927,
+      "learning_rate": 0.0001,
+      "loss": 1.7088,
+      "step": 7325
+    },
+    {
+      "epoch": 0.8415369593934868,
+      "grad_norm": 0.47490793466567993,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 7326
+    },
+    {
+      "epoch": 0.841651829303314,
+      "grad_norm": 0.4489779770374298,
+      "learning_rate": 0.0001,
+      "loss": 1.5528,
+      "step": 7327
+    },
+    {
+      "epoch": 0.8417666992131411,
+      "grad_norm": 0.4593144953250885,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 7328
+    },
+    {
+      "epoch": 0.8418815691229682,
+      "grad_norm": 0.4318476617336273,
+      "learning_rate": 0.0001,
+      "loss": 1.458,
+      "step": 7329
+    },
+    {
+      "epoch": 0.8419964390327953,
+      "grad_norm": 0.44436419010162354,
+      "learning_rate": 0.0001,
+      "loss": 1.4058,
+      "step": 7330
+    },
+    {
+      "epoch": 0.8421113089426224,
+      "grad_norm": 0.4526481032371521,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 7331
+    },
+    {
+      "epoch": 0.8422261788524495,
+      "grad_norm": 0.4639897644519806,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 7332
+    },
+    {
+      "epoch": 0.8423410487622767,
+      "grad_norm": 0.4710429012775421,
+      "learning_rate": 0.0001,
+      "loss": 1.7472,
+      "step": 7333
+    },
+    {
+      "epoch": 0.8424559186721038,
+      "grad_norm": 0.4736747443675995,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 7334
+    },
+    {
+      "epoch": 0.842570788581931,
+      "grad_norm": 0.46228286623954773,
+      "learning_rate": 0.0001,
+      "loss": 1.4623,
+      "step": 7335
+    },
+    {
+      "epoch": 0.8426856584917581,
+      "grad_norm": 0.5162138938903809,
+      "learning_rate": 0.0001,
+      "loss": 1.8545,
+      "step": 7336
+    },
+    {
+      "epoch": 0.8428005284015853,
+      "grad_norm": 0.4684411585330963,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 7337
+    },
+    {
+      "epoch": 0.8429153983114124,
+      "grad_norm": 0.45652058720588684,
+      "learning_rate": 0.0001,
+      "loss": 1.4637,
+      "step": 7338
+    },
+    {
+      "epoch": 0.8430302682212395,
+      "grad_norm": 0.49784135818481445,
+      "learning_rate": 0.0001,
+      "loss": 1.491,
+      "step": 7339
+    },
+    {
+      "epoch": 0.8431451381310666,
+      "grad_norm": 0.5035806894302368,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 7340
+    },
+    {
+      "epoch": 0.8432600080408937,
+      "grad_norm": 0.4898698925971985,
+      "learning_rate": 0.0001,
+      "loss": 1.7416,
+      "step": 7341
+    },
+    {
+      "epoch": 0.8433748779507209,
+      "grad_norm": 0.4696720242500305,
+      "learning_rate": 0.0001,
+      "loss": 1.5175,
+      "step": 7342
+    },
+    {
+      "epoch": 0.843489747860548,
+      "grad_norm": 0.4923301339149475,
+      "learning_rate": 0.0001,
+      "loss": 1.7249,
+      "step": 7343
+    },
+    {
+      "epoch": 0.8436046177703751,
+      "grad_norm": 0.49450981616973877,
+      "learning_rate": 0.0001,
+      "loss": 1.7339,
+      "step": 7344
+    },
+    {
+      "epoch": 0.8437194876802022,
+      "grad_norm": 0.43973812460899353,
+      "learning_rate": 0.0001,
+      "loss": 1.4611,
+      "step": 7345
+    },
+    {
+      "epoch": 0.8438343575900293,
+      "grad_norm": 0.48189517855644226,
+      "learning_rate": 0.0001,
+      "loss": 1.6952,
+      "step": 7346
+    },
+    {
+      "epoch": 0.8439492274998565,
+      "grad_norm": 0.4575080871582031,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 7347
+    },
+    {
+      "epoch": 0.8440640974096836,
+      "grad_norm": 0.4627358019351959,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 7348
+    },
+    {
+      "epoch": 0.8441789673195107,
+      "grad_norm": 0.4325619041919708,
+      "learning_rate": 0.0001,
+      "loss": 1.5157,
+      "step": 7349
+    },
+    {
+      "epoch": 0.8442938372293378,
+      "grad_norm": 0.4869388937950134,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 7350
+    },
+    {
+      "epoch": 0.8444087071391649,
+      "grad_norm": 0.4528372585773468,
+      "learning_rate": 0.0001,
+      "loss": 1.6697,
+      "step": 7351
+    },
+    {
+      "epoch": 0.8445235770489921,
+      "grad_norm": 0.45322051644325256,
+      "learning_rate": 0.0001,
+      "loss": 1.7075,
+      "step": 7352
+    },
+    {
+      "epoch": 0.8446384469588192,
+      "grad_norm": 0.43234097957611084,
+      "learning_rate": 0.0001,
+      "loss": 1.4102,
+      "step": 7353
+    },
+    {
+      "epoch": 0.8447533168686463,
+      "grad_norm": 0.4526819586753845,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 7354
+    },
+    {
+      "epoch": 0.8448681867784734,
+      "grad_norm": 0.46750757098197937,
+      "learning_rate": 0.0001,
+      "loss": 1.8057,
+      "step": 7355
+    },
+    {
+      "epoch": 0.8449830566883005,
+      "grad_norm": 0.47405919432640076,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 7356
+    },
+    {
+      "epoch": 0.8450979265981277,
+      "grad_norm": 0.4665440320968628,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 7357
+    },
+    {
+      "epoch": 0.8452127965079548,
+      "grad_norm": 0.4736506938934326,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 7358
+    },
+    {
+      "epoch": 0.8453276664177819,
+      "grad_norm": 0.4538653492927551,
+      "learning_rate": 0.0001,
+      "loss": 1.7792,
+      "step": 7359
+    },
+    {
+      "epoch": 0.845442536327609,
+      "grad_norm": 0.4535897672176361,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 7360
+    },
+    {
+      "epoch": 0.8455574062374361,
+      "grad_norm": 0.5242683291435242,
+      "learning_rate": 0.0001,
+      "loss": 1.6937,
+      "step": 7361
+    },
+    {
+      "epoch": 0.8456722761472633,
+      "grad_norm": 0.4639322757720947,
+      "learning_rate": 0.0001,
+      "loss": 1.6695,
+      "step": 7362
+    },
+    {
+      "epoch": 0.8457871460570904,
+      "grad_norm": 0.44305261969566345,
+      "learning_rate": 0.0001,
+      "loss": 1.3519,
+      "step": 7363
+    },
+    {
+      "epoch": 0.8459020159669175,
+      "grad_norm": 0.4527410864830017,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 7364
+    },
+    {
+      "epoch": 0.8460168858767446,
+      "grad_norm": 0.48342829942703247,
+      "learning_rate": 0.0001,
+      "loss": 1.5043,
+      "step": 7365
+    },
+    {
+      "epoch": 0.8461317557865717,
+      "grad_norm": 0.46347376704216003,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 7366
+    },
+    {
+      "epoch": 0.8462466256963989,
+      "grad_norm": 0.45847606658935547,
+      "learning_rate": 0.0001,
+      "loss": 1.5473,
+      "step": 7367
+    },
+    {
+      "epoch": 0.846361495606226,
+      "grad_norm": 0.4818578362464905,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 7368
+    },
+    {
+      "epoch": 0.8464763655160531,
+      "grad_norm": 0.48799648880958557,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 7369
+    },
+    {
+      "epoch": 0.8465912354258802,
+      "grad_norm": 0.4803750813007355,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 7370
+    },
+    {
+      "epoch": 0.8467061053357073,
+      "grad_norm": 0.48348790407180786,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 7371
+    },
+    {
+      "epoch": 0.8468209752455345,
+      "grad_norm": 0.4604859948158264,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 7372
+    },
+    {
+      "epoch": 0.8469358451553616,
+      "grad_norm": 0.459789901971817,
+      "learning_rate": 0.0001,
+      "loss": 1.4266,
+      "step": 7373
+    },
+    {
+      "epoch": 0.8470507150651887,
+      "grad_norm": 0.5160319805145264,
+      "learning_rate": 0.0001,
+      "loss": 1.7858,
+      "step": 7374
+    },
+    {
+      "epoch": 0.8471655849750158,
+      "grad_norm": 0.4354211688041687,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 7375
+    },
+    {
+      "epoch": 0.8472804548848429,
+      "grad_norm": 0.4287918508052826,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 7376
+    },
+    {
+      "epoch": 0.8473953247946701,
+      "grad_norm": 0.4390510320663452,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 7377
+    },
+    {
+      "epoch": 0.8475101947044972,
+      "grad_norm": 0.4523385465145111,
+      "learning_rate": 0.0001,
+      "loss": 1.631,
+      "step": 7378
+    },
+    {
+      "epoch": 0.8476250646143243,
+      "grad_norm": 0.4572320878505707,
+      "learning_rate": 0.0001,
+      "loss": 1.1648,
+      "step": 7379
+    },
+    {
+      "epoch": 0.8477399345241514,
+      "grad_norm": 0.5559184551239014,
+      "learning_rate": 0.0001,
+      "loss": 1.3973,
+      "step": 7380
+    },
+    {
+      "epoch": 0.8478548044339785,
+      "grad_norm": 0.49197232723236084,
+      "learning_rate": 0.0001,
+      "loss": 1.7199,
+      "step": 7381
+    },
+    {
+      "epoch": 0.8479696743438057,
+      "grad_norm": 0.49532270431518555,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 7382
+    },
+    {
+      "epoch": 0.8480845442536328,
+      "grad_norm": 0.49762824177742004,
+      "learning_rate": 0.0001,
+      "loss": 1.6974,
+      "step": 7383
+    },
+    {
+      "epoch": 0.8481994141634599,
+      "grad_norm": 0.473664790391922,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 7384
+    },
+    {
+      "epoch": 0.848314284073287,
+      "grad_norm": 0.48429468274116516,
+      "learning_rate": 0.0001,
+      "loss": 1.4807,
+      "step": 7385
+    },
+    {
+      "epoch": 0.8484291539831141,
+      "grad_norm": 0.4701521396636963,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 7386
+    },
+    {
+      "epoch": 0.8485440238929413,
+      "grad_norm": 0.43333899974823,
+      "learning_rate": 0.0001,
+      "loss": 1.5189,
+      "step": 7387
+    },
+    {
+      "epoch": 0.8486588938027684,
+      "grad_norm": 0.47169727087020874,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 7388
+    },
+    {
+      "epoch": 0.8487737637125955,
+      "grad_norm": 0.4842875003814697,
+      "learning_rate": 0.0001,
+      "loss": 1.7324,
+      "step": 7389
+    },
+    {
+      "epoch": 0.8488886336224226,
+      "grad_norm": 0.49503180384635925,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 7390
+    },
+    {
+      "epoch": 0.8490035035322497,
+      "grad_norm": 0.44051480293273926,
+      "learning_rate": 0.0001,
+      "loss": 1.5385,
+      "step": 7391
+    },
+    {
+      "epoch": 0.8491183734420769,
+      "grad_norm": 0.4651011824607849,
+      "learning_rate": 0.0001,
+      "loss": 1.4514,
+      "step": 7392
+    },
+    {
+      "epoch": 0.849233243351904,
+      "grad_norm": 0.4789915978908539,
+      "learning_rate": 0.0001,
+      "loss": 1.6536,
+      "step": 7393
+    },
+    {
+      "epoch": 0.8493481132617311,
+      "grad_norm": 0.44147029519081116,
+      "learning_rate": 0.0001,
+      "loss": 1.6246,
+      "step": 7394
+    },
+    {
+      "epoch": 0.8494629831715582,
+      "grad_norm": 0.42696404457092285,
+      "learning_rate": 0.0001,
+      "loss": 1.3676,
+      "step": 7395
+    },
+    {
+      "epoch": 0.8495778530813853,
+      "grad_norm": 0.4485602080821991,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 7396
+    },
+    {
+      "epoch": 0.8496927229912125,
+      "grad_norm": 0.46463295817375183,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 7397
+    },
+    {
+      "epoch": 0.8498075929010396,
+      "grad_norm": 0.43345364928245544,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 7398
+    },
+    {
+      "epoch": 0.8499224628108667,
+      "grad_norm": 0.4817748963832855,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 7399
+    },
+    {
+      "epoch": 0.8500373327206938,
+      "grad_norm": 0.48344627022743225,
+      "learning_rate": 0.0001,
+      "loss": 1.434,
+      "step": 7400
+    },
+    {
+      "epoch": 0.8501522026305209,
+      "grad_norm": 0.4697633683681488,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 7401
+    },
+    {
+      "epoch": 0.8502670725403481,
+      "grad_norm": 0.4468742609024048,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 7402
+    },
+    {
+      "epoch": 0.8503819424501752,
+      "grad_norm": 0.4675341546535492,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 7403
+    },
+    {
+      "epoch": 0.8504968123600023,
+      "grad_norm": 0.43257632851600647,
+      "learning_rate": 0.0001,
+      "loss": 1.3002,
+      "step": 7404
+    },
+    {
+      "epoch": 0.8506116822698294,
+      "grad_norm": 0.4668475389480591,
+      "learning_rate": 0.0001,
+      "loss": 1.6644,
+      "step": 7405
+    },
+    {
+      "epoch": 0.8507265521796565,
+      "grad_norm": 0.4785885512828827,
+      "learning_rate": 0.0001,
+      "loss": 1.5659,
+      "step": 7406
+    },
+    {
+      "epoch": 0.8508414220894837,
+      "grad_norm": 0.43829622864723206,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 7407
+    },
+    {
+      "epoch": 0.8509562919993108,
+      "grad_norm": 0.5236424207687378,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 7408
+    },
+    {
+      "epoch": 0.8510711619091379,
+      "grad_norm": 0.45531997084617615,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 7409
+    },
+    {
+      "epoch": 0.851186031818965,
+      "grad_norm": 0.5072128176689148,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 7410
+    },
+    {
+      "epoch": 0.8513009017287921,
+      "grad_norm": 0.4630776643753052,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 7411
+    },
+    {
+      "epoch": 0.8514157716386193,
+      "grad_norm": 0.4900180697441101,
+      "learning_rate": 0.0001,
+      "loss": 1.7094,
+      "step": 7412
+    },
+    {
+      "epoch": 0.8515306415484464,
+      "grad_norm": 0.46314537525177,
+      "learning_rate": 0.0001,
+      "loss": 1.4107,
+      "step": 7413
+    },
+    {
+      "epoch": 0.8516455114582735,
+      "grad_norm": 0.5156530141830444,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 7414
+    },
+    {
+      "epoch": 0.8517603813681006,
+      "grad_norm": 0.5086135268211365,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 7415
+    },
+    {
+      "epoch": 0.8518752512779277,
+      "grad_norm": 0.48410969972610474,
+      "learning_rate": 0.0001,
+      "loss": 1.7056,
+      "step": 7416
+    },
+    {
+      "epoch": 0.8519901211877549,
+      "grad_norm": 0.5622872114181519,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 7417
+    },
+    {
+      "epoch": 0.852104991097582,
+      "grad_norm": 0.4555276036262512,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 7418
+    },
+    {
+      "epoch": 0.8522198610074091,
+      "grad_norm": 0.46399348974227905,
+      "learning_rate": 0.0001,
+      "loss": 1.6051,
+      "step": 7419
+    },
+    {
+      "epoch": 0.8523347309172362,
+      "grad_norm": 0.43579867482185364,
+      "learning_rate": 0.0001,
+      "loss": 1.3899,
+      "step": 7420
+    },
+    {
+      "epoch": 0.8524496008270633,
+      "grad_norm": 0.47155770659446716,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 7421
+    },
+    {
+      "epoch": 0.8525644707368905,
+      "grad_norm": 0.46017301082611084,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 7422
+    },
+    {
+      "epoch": 0.8526793406467176,
+      "grad_norm": 0.4408453404903412,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 7423
+    },
+    {
+      "epoch": 0.8527942105565447,
+      "grad_norm": 0.475716769695282,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 7424
+    },
+    {
+      "epoch": 0.8529090804663718,
+      "grad_norm": 0.4536953270435333,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 7425
+    },
+    {
+      "epoch": 0.8530239503761989,
+      "grad_norm": 0.476300448179245,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 7426
+    },
+    {
+      "epoch": 0.8531388202860261,
+      "grad_norm": 0.42572006583213806,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 7427
+    },
+    {
+      "epoch": 0.8532536901958532,
+      "grad_norm": 0.5320248007774353,
+      "learning_rate": 0.0001,
+      "loss": 1.6911,
+      "step": 7428
+    },
+    {
+      "epoch": 0.8533685601056803,
+      "grad_norm": 0.46820348501205444,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 7429
+    },
+    {
+      "epoch": 0.8534834300155074,
+      "grad_norm": 0.4351903796195984,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8535982999253345,
+      "grad_norm": 0.47440531849861145,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 7431
+    },
+    {
+      "epoch": 0.8537131698351617,
+      "grad_norm": 0.43947550654411316,
+      "learning_rate": 0.0001,
+      "loss": 1.5196,
+      "step": 7432
+    },
+    {
+      "epoch": 0.8538280397449888,
+      "grad_norm": 0.4965689182281494,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 7433
+    },
+    {
+      "epoch": 0.8539429096548159,
+      "grad_norm": 0.4858873188495636,
+      "learning_rate": 0.0001,
+      "loss": 1.8044,
+      "step": 7434
+    },
+    {
+      "epoch": 0.854057779564643,
+      "grad_norm": 0.4513735771179199,
+      "learning_rate": 0.0001,
+      "loss": 1.4199,
+      "step": 7435
+    },
+    {
+      "epoch": 0.8541726494744701,
+      "grad_norm": 0.4994846284389496,
+      "learning_rate": 0.0001,
+      "loss": 1.943,
+      "step": 7436
+    },
+    {
+      "epoch": 0.8542875193842973,
+      "grad_norm": 0.45225298404693604,
+      "learning_rate": 0.0001,
+      "loss": 1.6063,
+      "step": 7437
+    },
+    {
+      "epoch": 0.8544023892941244,
+      "grad_norm": 0.445266991853714,
+      "learning_rate": 0.0001,
+      "loss": 1.4866,
+      "step": 7438
+    },
+    {
+      "epoch": 0.8545172592039515,
+      "grad_norm": 0.4278365969657898,
+      "learning_rate": 0.0001,
+      "loss": 1.617,
+      "step": 7439
+    },
+    {
+      "epoch": 0.8546321291137786,
+      "grad_norm": 0.4820537865161896,
+      "learning_rate": 0.0001,
+      "loss": 1.7703,
+      "step": 7440
+    },
+    {
+      "epoch": 0.8547469990236057,
+      "grad_norm": 0.4873339533805847,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 7441
+    },
+    {
+      "epoch": 0.8548618689334329,
+      "grad_norm": 0.44816693663597107,
+      "learning_rate": 0.0001,
+      "loss": 1.6484,
+      "step": 7442
+    },
+    {
+      "epoch": 0.85497673884326,
+      "grad_norm": 0.47395050525665283,
+      "learning_rate": 0.0001,
+      "loss": 1.5569,
+      "step": 7443
+    },
+    {
+      "epoch": 0.8550916087530871,
+      "grad_norm": 0.42696264386177063,
+      "learning_rate": 0.0001,
+      "loss": 1.5543,
+      "step": 7444
+    },
+    {
+      "epoch": 0.8552064786629142,
+      "grad_norm": 0.5623756647109985,
+      "learning_rate": 0.0001,
+      "loss": 2.051,
+      "step": 7445
+    },
+    {
+      "epoch": 0.8553213485727413,
+      "grad_norm": 0.5125777721405029,
+      "learning_rate": 0.0001,
+      "loss": 1.7589,
+      "step": 7446
+    },
+    {
+      "epoch": 0.8554362184825685,
+      "grad_norm": 0.4917384684085846,
+      "learning_rate": 0.0001,
+      "loss": 1.8042,
+      "step": 7447
+    },
+    {
+      "epoch": 0.8555510883923956,
+      "grad_norm": 0.4971913695335388,
+      "learning_rate": 0.0001,
+      "loss": 1.6491,
+      "step": 7448
+    },
+    {
+      "epoch": 0.8556659583022227,
+      "grad_norm": 0.4413685202598572,
+      "learning_rate": 0.0001,
+      "loss": 1.4241,
+      "step": 7449
+    },
+    {
+      "epoch": 0.8557808282120498,
+      "grad_norm": 0.5352091789245605,
+      "learning_rate": 0.0001,
+      "loss": 1.9268,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8558956981218769,
+      "grad_norm": 0.5074852705001831,
+      "learning_rate": 0.0001,
+      "loss": 1.7119,
+      "step": 7451
+    },
+    {
+      "epoch": 0.8560105680317041,
+      "grad_norm": 0.45686012506484985,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 7452
+    },
+    {
+      "epoch": 0.8561254379415312,
+      "grad_norm": 0.48405084013938904,
+      "learning_rate": 0.0001,
+      "loss": 1.792,
+      "step": 7453
+    },
+    {
+      "epoch": 0.8562403078513583,
+      "grad_norm": 0.5095300674438477,
+      "learning_rate": 0.0001,
+      "loss": 1.7128,
+      "step": 7454
+    },
+    {
+      "epoch": 0.8563551777611854,
+      "grad_norm": 0.49315640330314636,
+      "learning_rate": 0.0001,
+      "loss": 1.4333,
+      "step": 7455
+    },
+    {
+      "epoch": 0.8564700476710125,
+      "grad_norm": 0.44458267092704773,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 7456
+    },
+    {
+      "epoch": 0.8565849175808397,
+      "grad_norm": 0.47778499126434326,
+      "learning_rate": 0.0001,
+      "loss": 1.7333,
+      "step": 7457
+    },
+    {
+      "epoch": 0.8566997874906668,
+      "grad_norm": 0.47323766350746155,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 7458
+    },
+    {
+      "epoch": 0.8568146574004939,
+      "grad_norm": 0.4731180965900421,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 7459
+    },
+    {
+      "epoch": 0.856929527310321,
+      "grad_norm": 0.4616163671016693,
+      "learning_rate": 0.0001,
+      "loss": 1.7287,
+      "step": 7460
+    },
+    {
+      "epoch": 0.8570443972201481,
+      "grad_norm": 0.4829119145870209,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 7461
+    },
+    {
+      "epoch": 0.8571592671299753,
+      "grad_norm": 0.5113245844841003,
+      "learning_rate": 0.0001,
+      "loss": 1.5072,
+      "step": 7462
+    },
+    {
+      "epoch": 0.8572741370398024,
+      "grad_norm": 0.45658257603645325,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 7463
+    },
+    {
+      "epoch": 0.8573890069496295,
+      "grad_norm": 0.44155845046043396,
+      "learning_rate": 0.0001,
+      "loss": 1.5001,
+      "step": 7464
+    },
+    {
+      "epoch": 0.8575038768594566,
+      "grad_norm": 0.46995145082473755,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 7465
+    },
+    {
+      "epoch": 0.8576187467692837,
+      "grad_norm": 0.43981465697288513,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 7466
+    },
+    {
+      "epoch": 0.8577336166791109,
+      "grad_norm": 0.4353700280189514,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 7467
+    },
+    {
+      "epoch": 0.857848486588938,
+      "grad_norm": 0.4613707959651947,
+      "learning_rate": 0.0001,
+      "loss": 1.6324,
+      "step": 7468
+    },
+    {
+      "epoch": 0.8579633564987651,
+      "grad_norm": 0.43150594830513,
+      "learning_rate": 0.0001,
+      "loss": 1.3483,
+      "step": 7469
+    },
+    {
+      "epoch": 0.8580782264085922,
+      "grad_norm": 0.49536994099617004,
+      "learning_rate": 0.0001,
+      "loss": 1.6696,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8581930963184193,
+      "grad_norm": 0.4285619258880615,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 7471
+    },
+    {
+      "epoch": 0.8583079662282466,
+      "grad_norm": 0.5003694891929626,
+      "learning_rate": 0.0001,
+      "loss": 1.5725,
+      "step": 7472
+    },
+    {
+      "epoch": 0.8584228361380737,
+      "grad_norm": 0.4515581429004669,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 7473
+    },
+    {
+      "epoch": 0.8585377060479008,
+      "grad_norm": 0.4796769320964813,
+      "learning_rate": 0.0001,
+      "loss": 1.4985,
+      "step": 7474
+    },
+    {
+      "epoch": 0.8586525759577279,
+      "grad_norm": 0.5017609596252441,
+      "learning_rate": 0.0001,
+      "loss": 1.7879,
+      "step": 7475
+    },
+    {
+      "epoch": 0.858767445867555,
+      "grad_norm": 0.4531811475753784,
+      "learning_rate": 0.0001,
+      "loss": 1.5423,
+      "step": 7476
+    },
+    {
+      "epoch": 0.8588823157773822,
+      "grad_norm": 0.4460836946964264,
+      "learning_rate": 0.0001,
+      "loss": 1.456,
+      "step": 7477
+    },
+    {
+      "epoch": 0.8589971856872093,
+      "grad_norm": 0.4407133162021637,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 7478
+    },
+    {
+      "epoch": 0.8591120555970364,
+      "grad_norm": 0.4690662622451782,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 7479
+    },
+    {
+      "epoch": 0.8592269255068635,
+      "grad_norm": 0.43073776364326477,
+      "learning_rate": 0.0001,
+      "loss": 1.3295,
+      "step": 7480
+    },
+    {
+      "epoch": 0.8593417954166906,
+      "grad_norm": 0.45200130343437195,
+      "learning_rate": 0.0001,
+      "loss": 1.435,
+      "step": 7481
+    },
+    {
+      "epoch": 0.8594566653265178,
+      "grad_norm": 0.45768219232559204,
+      "learning_rate": 0.0001,
+      "loss": 1.5326,
+      "step": 7482
+    },
+    {
+      "epoch": 0.8595715352363449,
+      "grad_norm": 0.4337681829929352,
+      "learning_rate": 0.0001,
+      "loss": 1.5591,
+      "step": 7483
+    },
+    {
+      "epoch": 0.859686405146172,
+      "grad_norm": 0.44868743419647217,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 7484
+    },
+    {
+      "epoch": 0.8598012750559991,
+      "grad_norm": 0.4528716206550598,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 7485
+    },
+    {
+      "epoch": 0.8599161449658262,
+      "grad_norm": 0.4922735095024109,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 7486
+    },
+    {
+      "epoch": 0.8600310148756534,
+      "grad_norm": 0.46525344252586365,
+      "learning_rate": 0.0001,
+      "loss": 1.5649,
+      "step": 7487
+    },
+    {
+      "epoch": 0.8601458847854805,
+      "grad_norm": 0.46007898449897766,
+      "learning_rate": 0.0001,
+      "loss": 1.6705,
+      "step": 7488
+    },
+    {
+      "epoch": 0.8602607546953076,
+      "grad_norm": 0.4645099341869354,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 7489
+    },
+    {
+      "epoch": 0.8603756246051347,
+      "grad_norm": 0.46048229932785034,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 7490
+    },
+    {
+      "epoch": 0.8604904945149618,
+      "grad_norm": 0.4628632366657257,
+      "learning_rate": 0.0001,
+      "loss": 1.6688,
+      "step": 7491
+    },
+    {
+      "epoch": 0.860605364424789,
+      "grad_norm": 0.5050228238105774,
+      "learning_rate": 0.0001,
+      "loss": 1.712,
+      "step": 7492
+    },
+    {
+      "epoch": 0.8607202343346161,
+      "grad_norm": 0.4507612884044647,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 7493
+    },
+    {
+      "epoch": 0.8608351042444432,
+      "grad_norm": 0.4406948983669281,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 7494
+    },
+    {
+      "epoch": 0.8609499741542703,
+      "grad_norm": 0.4973854124546051,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 7495
+    },
+    {
+      "epoch": 0.8610648440640974,
+      "grad_norm": 0.4488707482814789,
+      "learning_rate": 0.0001,
+      "loss": 1.4153,
+      "step": 7496
+    },
+    {
+      "epoch": 0.8611797139739246,
+      "grad_norm": 0.4815550744533539,
+      "learning_rate": 0.0001,
+      "loss": 1.6676,
+      "step": 7497
+    },
+    {
+      "epoch": 0.8612945838837517,
+      "grad_norm": 0.4687705338001251,
+      "learning_rate": 0.0001,
+      "loss": 1.3591,
+      "step": 7498
+    },
+    {
+      "epoch": 0.8614094537935788,
+      "grad_norm": 0.46637317538261414,
+      "learning_rate": 0.0001,
+      "loss": 1.6578,
+      "step": 7499
+    },
+    {
+      "epoch": 0.8615243237034059,
+      "grad_norm": 0.4513280987739563,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 7500
+    },
+    {
+      "epoch": 0.861639193613233,
+      "grad_norm": 0.4474887549877167,
+      "learning_rate": 0.0001,
+      "loss": 1.3628,
+      "step": 7501
+    },
+    {
+      "epoch": 0.8617540635230602,
+      "grad_norm": 0.4975503087043762,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 7502
+    },
+    {
+      "epoch": 0.8618689334328873,
+      "grad_norm": 0.5043979287147522,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 7503
+    },
+    {
+      "epoch": 0.8619838033427144,
+      "grad_norm": 0.4594765305519104,
+      "learning_rate": 0.0001,
+      "loss": 1.702,
+      "step": 7504
+    },
+    {
+      "epoch": 0.8620986732525415,
+      "grad_norm": 0.46996229887008667,
+      "learning_rate": 0.0001,
+      "loss": 1.714,
+      "step": 7505
+    },
+    {
+      "epoch": 0.8622135431623686,
+      "grad_norm": 0.45425885915756226,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 7506
+    },
+    {
+      "epoch": 0.8623284130721958,
+      "grad_norm": 0.48939937353134155,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 7507
+    },
+    {
+      "epoch": 0.8624432829820229,
+      "grad_norm": 0.44552651047706604,
+      "learning_rate": 0.0001,
+      "loss": 1.5571,
+      "step": 7508
+    },
+    {
+      "epoch": 0.86255815289185,
+      "grad_norm": 0.4959718883037567,
+      "learning_rate": 0.0001,
+      "loss": 1.6482,
+      "step": 7509
+    },
+    {
+      "epoch": 0.8626730228016771,
+      "grad_norm": 0.4864218533039093,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 7510
+    },
+    {
+      "epoch": 0.8627878927115042,
+      "grad_norm": 0.4416675865650177,
+      "learning_rate": 0.0001,
+      "loss": 1.686,
+      "step": 7511
+    },
+    {
+      "epoch": 0.8629027626213314,
+      "grad_norm": 0.46253499388694763,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 7512
+    },
+    {
+      "epoch": 0.8630176325311585,
+      "grad_norm": 0.47551655769348145,
+      "learning_rate": 0.0001,
+      "loss": 1.7636,
+      "step": 7513
+    },
+    {
+      "epoch": 0.8631325024409856,
+      "grad_norm": 0.48232796788215637,
+      "learning_rate": 0.0001,
+      "loss": 1.6967,
+      "step": 7514
+    },
+    {
+      "epoch": 0.8632473723508127,
+      "grad_norm": 0.4513819217681885,
+      "learning_rate": 0.0001,
+      "loss": 1.5229,
+      "step": 7515
+    },
+    {
+      "epoch": 0.8633622422606398,
+      "grad_norm": 0.4784489572048187,
+      "learning_rate": 0.0001,
+      "loss": 1.7298,
+      "step": 7516
+    },
+    {
+      "epoch": 0.863477112170467,
+      "grad_norm": 0.4839509427547455,
+      "learning_rate": 0.0001,
+      "loss": 1.8396,
+      "step": 7517
+    },
+    {
+      "epoch": 0.8635919820802941,
+      "grad_norm": 0.5013167858123779,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 7518
+    },
+    {
+      "epoch": 0.8637068519901212,
+      "grad_norm": 0.43714308738708496,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 7519
+    },
+    {
+      "epoch": 0.8638217218999483,
+      "grad_norm": 0.45084357261657715,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 7520
+    },
+    {
+      "epoch": 0.8639365918097754,
+      "grad_norm": 0.4939959943294525,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 7521
+    },
+    {
+      "epoch": 0.8640514617196026,
+      "grad_norm": 0.46642765402793884,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 7522
+    },
+    {
+      "epoch": 0.8641663316294297,
+      "grad_norm": 0.4534401595592499,
+      "learning_rate": 0.0001,
+      "loss": 1.582,
+      "step": 7523
+    },
+    {
+      "epoch": 0.8642812015392568,
+      "grad_norm": 0.43686237931251526,
+      "learning_rate": 0.0001,
+      "loss": 1.5726,
+      "step": 7524
+    },
+    {
+      "epoch": 0.8643960714490839,
+      "grad_norm": 0.4657284915447235,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 7525
+    },
+    {
+      "epoch": 0.864510941358911,
+      "grad_norm": 0.5079863667488098,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 7526
+    },
+    {
+      "epoch": 0.8646258112687382,
+      "grad_norm": 0.4670884311199188,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 7527
+    },
+    {
+      "epoch": 0.8647406811785653,
+      "grad_norm": 0.4581589102745056,
+      "learning_rate": 0.0001,
+      "loss": 1.3903,
+      "step": 7528
+    },
+    {
+      "epoch": 0.8648555510883924,
+      "grad_norm": 0.4376249313354492,
+      "learning_rate": 0.0001,
+      "loss": 1.5475,
+      "step": 7529
+    },
+    {
+      "epoch": 0.8649704209982195,
+      "grad_norm": 0.46135491132736206,
+      "learning_rate": 0.0001,
+      "loss": 1.5266,
+      "step": 7530
+    },
+    {
+      "epoch": 0.8650852909080466,
+      "grad_norm": 0.492291659116745,
+      "learning_rate": 0.0001,
+      "loss": 1.713,
+      "step": 7531
+    },
+    {
+      "epoch": 0.8652001608178738,
+      "grad_norm": 0.4598917067050934,
+      "learning_rate": 0.0001,
+      "loss": 1.3503,
+      "step": 7532
+    },
+    {
+      "epoch": 0.8653150307277009,
+      "grad_norm": 0.4626048505306244,
+      "learning_rate": 0.0001,
+      "loss": 1.4571,
+      "step": 7533
+    },
+    {
+      "epoch": 0.865429900637528,
+      "grad_norm": 0.49662452936172485,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 7534
+    },
+    {
+      "epoch": 0.8655447705473551,
+      "grad_norm": 0.4794904291629791,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 7535
+    },
+    {
+      "epoch": 0.8656596404571822,
+      "grad_norm": 0.4706816077232361,
+      "learning_rate": 0.0001,
+      "loss": 1.4572,
+      "step": 7536
+    },
+    {
+      "epoch": 0.8657745103670094,
+      "grad_norm": 0.4907233417034149,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 7537
+    },
+    {
+      "epoch": 0.8658893802768365,
+      "grad_norm": 0.46914729475975037,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 7538
+    },
+    {
+      "epoch": 0.8660042501866636,
+      "grad_norm": 0.47738802433013916,
+      "learning_rate": 0.0001,
+      "loss": 1.6708,
+      "step": 7539
+    },
+    {
+      "epoch": 0.8661191200964907,
+      "grad_norm": 0.47731783986091614,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 7540
+    },
+    {
+      "epoch": 0.8662339900063178,
+      "grad_norm": 0.4860245883464813,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 7541
+    },
+    {
+      "epoch": 0.866348859916145,
+      "grad_norm": 0.4608958065509796,
+      "learning_rate": 0.0001,
+      "loss": 1.4598,
+      "step": 7542
+    },
+    {
+      "epoch": 0.8664637298259721,
+      "grad_norm": 0.4907558262348175,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 7543
+    },
+    {
+      "epoch": 0.8665785997357992,
+      "grad_norm": 0.460859477519989,
+      "learning_rate": 0.0001,
+      "loss": 1.7288,
+      "step": 7544
+    },
+    {
+      "epoch": 0.8666934696456263,
+      "grad_norm": 0.5079466104507446,
+      "learning_rate": 0.0001,
+      "loss": 1.7155,
+      "step": 7545
+    },
+    {
+      "epoch": 0.8668083395554534,
+      "grad_norm": 0.4723812937736511,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 7546
+    },
+    {
+      "epoch": 0.8669232094652806,
+      "grad_norm": 0.457302063703537,
+      "learning_rate": 0.0001,
+      "loss": 1.5175,
+      "step": 7547
+    },
+    {
+      "epoch": 0.8670380793751077,
+      "grad_norm": 0.44479089975357056,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 7548
+    },
+    {
+      "epoch": 0.8671529492849348,
+      "grad_norm": 0.4218999147415161,
+      "learning_rate": 0.0001,
+      "loss": 1.5437,
+      "step": 7549
+    },
+    {
+      "epoch": 0.8672678191947619,
+      "grad_norm": 0.44676443934440613,
+      "learning_rate": 0.0001,
+      "loss": 1.6445,
+      "step": 7550
+    },
+    {
+      "epoch": 0.867382689104589,
+      "grad_norm": 0.4901319444179535,
+      "learning_rate": 0.0001,
+      "loss": 1.7671,
+      "step": 7551
+    },
+    {
+      "epoch": 0.8674975590144162,
+      "grad_norm": 0.4453258216381073,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 7552
+    },
+    {
+      "epoch": 0.8676124289242433,
+      "grad_norm": 0.4534637928009033,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 7553
+    },
+    {
+      "epoch": 0.8677272988340704,
+      "grad_norm": 0.46283942461013794,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 7554
+    },
+    {
+      "epoch": 0.8678421687438975,
+      "grad_norm": 0.5271000266075134,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 7555
+    },
+    {
+      "epoch": 0.8679570386537246,
+      "grad_norm": 0.4478399157524109,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 7556
+    },
+    {
+      "epoch": 0.8680719085635518,
+      "grad_norm": 0.4519982933998108,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 7557
+    },
+    {
+      "epoch": 0.8681867784733789,
+      "grad_norm": 0.49382635951042175,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 7558
+    },
+    {
+      "epoch": 0.868301648383206,
+      "grad_norm": 0.4385705888271332,
+      "learning_rate": 0.0001,
+      "loss": 1.4204,
+      "step": 7559
+    },
+    {
+      "epoch": 0.8684165182930331,
+      "grad_norm": 0.42191818356513977,
+      "learning_rate": 0.0001,
+      "loss": 1.3775,
+      "step": 7560
+    },
+    {
+      "epoch": 0.8685313882028602,
+      "grad_norm": 0.48729297518730164,
+      "learning_rate": 0.0001,
+      "loss": 1.5923,
+      "step": 7561
+    },
+    {
+      "epoch": 0.8686462581126874,
+      "grad_norm": 0.45936107635498047,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 7562
+    },
+    {
+      "epoch": 0.8687611280225145,
+      "grad_norm": 0.4694010317325592,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 7563
+    },
+    {
+      "epoch": 0.8688759979323416,
+      "grad_norm": 0.4324677288532257,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 7564
+    },
+    {
+      "epoch": 0.8689908678421687,
+      "grad_norm": 0.47414863109588623,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 7565
+    },
+    {
+      "epoch": 0.8691057377519958,
+      "grad_norm": 0.46548882126808167,
+      "learning_rate": 0.0001,
+      "loss": 1.3367,
+      "step": 7566
+    },
+    {
+      "epoch": 0.869220607661823,
+      "grad_norm": 0.44881516695022583,
+      "learning_rate": 0.0001,
+      "loss": 1.4306,
+      "step": 7567
+    },
+    {
+      "epoch": 0.8693354775716501,
+      "grad_norm": 0.475082129240036,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 7568
+    },
+    {
+      "epoch": 0.8694503474814772,
+      "grad_norm": 0.4312398135662079,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 7569
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.4381447434425354,
+      "learning_rate": 0.0001,
+      "loss": 1.4735,
+      "step": 7570
+    },
+    {
+      "epoch": 0.8696800873011314,
+      "grad_norm": 0.42871126532554626,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 7571
+    },
+    {
+      "epoch": 0.8697949572109586,
+      "grad_norm": 0.44468954205513,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 7572
+    },
+    {
+      "epoch": 0.8699098271207857,
+      "grad_norm": 0.49432846903800964,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 7573
+    },
+    {
+      "epoch": 0.8700246970306128,
+      "grad_norm": 0.4997519552707672,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 7574
+    },
+    {
+      "epoch": 0.8701395669404399,
+      "grad_norm": 0.47391921281814575,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 7575
+    },
+    {
+      "epoch": 0.870254436850267,
+      "grad_norm": 0.49615180492401123,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 7576
+    },
+    {
+      "epoch": 0.8703693067600942,
+      "grad_norm": 0.5302203297615051,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 7577
+    },
+    {
+      "epoch": 0.8704841766699213,
+      "grad_norm": 0.512974739074707,
+      "learning_rate": 0.0001,
+      "loss": 1.8474,
+      "step": 7578
+    },
+    {
+      "epoch": 0.8705990465797484,
+      "grad_norm": 0.5273780226707458,
+      "learning_rate": 0.0001,
+      "loss": 1.7045,
+      "step": 7579
+    },
+    {
+      "epoch": 0.8707139164895755,
+      "grad_norm": 0.45488354563713074,
+      "learning_rate": 0.0001,
+      "loss": 1.7054,
+      "step": 7580
+    },
+    {
+      "epoch": 0.8708287863994026,
+      "grad_norm": 0.4844851791858673,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 7581
+    },
+    {
+      "epoch": 0.8709436563092298,
+      "grad_norm": 0.47248736023902893,
+      "learning_rate": 0.0001,
+      "loss": 1.7882,
+      "step": 7582
+    },
+    {
+      "epoch": 0.8710585262190569,
+      "grad_norm": 0.4401882588863373,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 7583
+    },
+    {
+      "epoch": 0.871173396128884,
+      "grad_norm": 0.46107155084609985,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 7584
+    },
+    {
+      "epoch": 0.8712882660387111,
+      "grad_norm": 0.44512903690338135,
+      "learning_rate": 0.0001,
+      "loss": 1.4246,
+      "step": 7585
+    },
+    {
+      "epoch": 0.8714031359485382,
+      "grad_norm": 0.5119422078132629,
+      "learning_rate": 0.0001,
+      "loss": 1.6568,
+      "step": 7586
+    },
+    {
+      "epoch": 0.8715180058583654,
+      "grad_norm": 0.5448932647705078,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 7587
+    },
+    {
+      "epoch": 0.8716328757681925,
+      "grad_norm": 0.4820195138454437,
+      "learning_rate": 0.0001,
+      "loss": 1.5244,
+      "step": 7588
+    },
+    {
+      "epoch": 0.8717477456780196,
+      "grad_norm": 0.46764081716537476,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 7589
+    },
+    {
+      "epoch": 0.8718626155878467,
+      "grad_norm": 0.5043097734451294,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 7590
+    },
+    {
+      "epoch": 0.8719774854976738,
+      "grad_norm": 0.48596030473709106,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 7591
+    },
+    {
+      "epoch": 0.872092355407501,
+      "grad_norm": 0.4883681535720825,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 7592
+    },
+    {
+      "epoch": 0.8722072253173281,
+      "grad_norm": 0.4625040888786316,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 7593
+    },
+    {
+      "epoch": 0.8723220952271552,
+      "grad_norm": 0.4826609194278717,
+      "learning_rate": 0.0001,
+      "loss": 1.4658,
+      "step": 7594
+    },
+    {
+      "epoch": 0.8724369651369823,
+      "grad_norm": 0.4496923089027405,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 7595
+    },
+    {
+      "epoch": 0.8725518350468094,
+      "grad_norm": 0.5118072032928467,
+      "learning_rate": 0.0001,
+      "loss": 1.8564,
+      "step": 7596
+    },
+    {
+      "epoch": 0.8726667049566366,
+      "grad_norm": 0.4654580056667328,
+      "learning_rate": 0.0001,
+      "loss": 1.4611,
+      "step": 7597
+    },
+    {
+      "epoch": 0.8727815748664637,
+      "grad_norm": 0.42999371886253357,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 7598
+    },
+    {
+      "epoch": 0.8728964447762908,
+      "grad_norm": 0.5076425671577454,
+      "learning_rate": 0.0001,
+      "loss": 1.7994,
+      "step": 7599
+    },
+    {
+      "epoch": 0.8730113146861179,
+      "grad_norm": 0.4634290933609009,
+      "learning_rate": 0.0001,
+      "loss": 1.7437,
+      "step": 7600
+    },
+    {
+      "epoch": 0.873126184595945,
+      "grad_norm": 0.455967515707016,
+      "learning_rate": 0.0001,
+      "loss": 1.6326,
+      "step": 7601
+    },
+    {
+      "epoch": 0.8732410545057722,
+      "grad_norm": 0.5265469551086426,
+      "learning_rate": 0.0001,
+      "loss": 1.828,
+      "step": 7602
+    },
+    {
+      "epoch": 0.8733559244155993,
+      "grad_norm": 0.44739046692848206,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 7603
+    },
+    {
+      "epoch": 0.8734707943254264,
+      "grad_norm": 0.47437113523483276,
+      "learning_rate": 0.0001,
+      "loss": 1.7686,
+      "step": 7604
+    },
+    {
+      "epoch": 0.8735856642352535,
+      "grad_norm": 0.47178056836128235,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 7605
+    },
+    {
+      "epoch": 0.8737005341450806,
+      "grad_norm": 0.47476136684417725,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 7606
+    },
+    {
+      "epoch": 0.8738154040549078,
+      "grad_norm": 0.47085344791412354,
+      "learning_rate": 0.0001,
+      "loss": 1.6793,
+      "step": 7607
+    },
+    {
+      "epoch": 0.8739302739647349,
+      "grad_norm": 0.48475953936576843,
+      "learning_rate": 0.0001,
+      "loss": 1.7614,
+      "step": 7608
+    },
+    {
+      "epoch": 0.8740451438745621,
+      "grad_norm": 0.46257898211479187,
+      "learning_rate": 0.0001,
+      "loss": 1.5595,
+      "step": 7609
+    },
+    {
+      "epoch": 0.8741600137843892,
+      "grad_norm": 0.4603593051433563,
+      "learning_rate": 0.0001,
+      "loss": 1.6879,
+      "step": 7610
+    },
+    {
+      "epoch": 0.8742748836942164,
+      "grad_norm": 0.46927449107170105,
+      "learning_rate": 0.0001,
+      "loss": 1.7635,
+      "step": 7611
+    },
+    {
+      "epoch": 0.8743897536040435,
+      "grad_norm": 0.48620933294296265,
+      "learning_rate": 0.0001,
+      "loss": 1.6829,
+      "step": 7612
+    },
+    {
+      "epoch": 0.8745046235138706,
+      "grad_norm": 0.4888293445110321,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 7613
+    },
+    {
+      "epoch": 0.8746194934236977,
+      "grad_norm": 0.45221349596977234,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 7614
+    },
+    {
+      "epoch": 0.8747343633335248,
+      "grad_norm": 0.4384780824184418,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 7615
+    },
+    {
+      "epoch": 0.874849233243352,
+      "grad_norm": 0.463351309299469,
+      "learning_rate": 0.0001,
+      "loss": 1.2614,
+      "step": 7616
+    },
+    {
+      "epoch": 0.8749641031531791,
+      "grad_norm": 0.5050379037857056,
+      "learning_rate": 0.0001,
+      "loss": 1.7645,
+      "step": 7617
+    },
+    {
+      "epoch": 0.8750789730630062,
+      "grad_norm": 0.4900285005569458,
+      "learning_rate": 0.0001,
+      "loss": 1.668,
+      "step": 7618
+    },
+    {
+      "epoch": 0.8751938429728333,
+      "grad_norm": 0.4964447319507599,
+      "learning_rate": 0.0001,
+      "loss": 1.7461,
+      "step": 7619
+    },
+    {
+      "epoch": 0.8753087128826604,
+      "grad_norm": 0.4623139798641205,
+      "learning_rate": 0.0001,
+      "loss": 1.5311,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8754235827924876,
+      "grad_norm": 0.47225624322891235,
+      "learning_rate": 0.0001,
+      "loss": 1.6197,
+      "step": 7621
+    },
+    {
+      "epoch": 0.8755384527023147,
+      "grad_norm": 0.49223390221595764,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 7622
+    },
+    {
+      "epoch": 0.8756533226121418,
+      "grad_norm": 0.5095483064651489,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 7623
+    },
+    {
+      "epoch": 0.8757681925219689,
+      "grad_norm": 0.4903635084629059,
+      "learning_rate": 0.0001,
+      "loss": 1.7643,
+      "step": 7624
+    },
+    {
+      "epoch": 0.875883062431796,
+      "grad_norm": 0.4946781396865845,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 7625
+    },
+    {
+      "epoch": 0.8759979323416232,
+      "grad_norm": 0.4905765652656555,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 7626
+    },
+    {
+      "epoch": 0.8761128022514503,
+      "grad_norm": 0.505204975605011,
+      "learning_rate": 0.0001,
+      "loss": 1.7406,
+      "step": 7627
+    },
+    {
+      "epoch": 0.8762276721612774,
+      "grad_norm": 0.533305823802948,
+      "learning_rate": 0.0001,
+      "loss": 1.4833,
+      "step": 7628
+    },
+    {
+      "epoch": 0.8763425420711045,
+      "grad_norm": 0.47964945435523987,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 7629
+    },
+    {
+      "epoch": 0.8764574119809316,
+      "grad_norm": 0.4678000211715698,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8765722818907588,
+      "grad_norm": 0.47018638253211975,
+      "learning_rate": 0.0001,
+      "loss": 1.489,
+      "step": 7631
+    },
+    {
+      "epoch": 0.8766871518005859,
+      "grad_norm": 0.4635443091392517,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 7632
+    },
+    {
+      "epoch": 0.876802021710413,
+      "grad_norm": 0.46582552790641785,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 7633
+    },
+    {
+      "epoch": 0.8769168916202401,
+      "grad_norm": 0.47066211700439453,
+      "learning_rate": 0.0001,
+      "loss": 1.6757,
+      "step": 7634
+    },
+    {
+      "epoch": 0.8770317615300672,
+      "grad_norm": 0.569146990776062,
+      "learning_rate": 0.0001,
+      "loss": 1.7014,
+      "step": 7635
+    },
+    {
+      "epoch": 0.8771466314398944,
+      "grad_norm": 0.48655346035957336,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 7636
+    },
+    {
+      "epoch": 0.8772615013497215,
+      "grad_norm": 0.4453171491622925,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 7637
+    },
+    {
+      "epoch": 0.8773763712595486,
+      "grad_norm": 0.4952543079853058,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 7638
+    },
+    {
+      "epoch": 0.8774912411693757,
+      "grad_norm": 0.45405057072639465,
+      "learning_rate": 0.0001,
+      "loss": 1.5223,
+      "step": 7639
+    },
+    {
+      "epoch": 0.8776061110792028,
+      "grad_norm": 0.46041983366012573,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 7640
+    },
+    {
+      "epoch": 0.87772098098903,
+      "grad_norm": 0.4734715223312378,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 7641
+    },
+    {
+      "epoch": 0.8778358508988571,
+      "grad_norm": 0.4629979431629181,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 7642
+    },
+    {
+      "epoch": 0.8779507208086842,
+      "grad_norm": 0.47290873527526855,
+      "learning_rate": 0.0001,
+      "loss": 1.524,
+      "step": 7643
+    },
+    {
+      "epoch": 0.8780655907185113,
+      "grad_norm": 0.4647415578365326,
+      "learning_rate": 0.0001,
+      "loss": 1.5595,
+      "step": 7644
+    },
+    {
+      "epoch": 0.8781804606283384,
+      "grad_norm": 0.5042547583580017,
+      "learning_rate": 0.0001,
+      "loss": 1.7437,
+      "step": 7645
+    },
+    {
+      "epoch": 0.8782953305381656,
+      "grad_norm": 0.4914029836654663,
+      "learning_rate": 0.0001,
+      "loss": 1.8578,
+      "step": 7646
+    },
+    {
+      "epoch": 0.8784102004479927,
+      "grad_norm": 0.5158650875091553,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 7647
+    },
+    {
+      "epoch": 0.8785250703578198,
+      "grad_norm": 0.4820328950881958,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 7648
+    },
+    {
+      "epoch": 0.8786399402676469,
+      "grad_norm": 0.4517586827278137,
+      "learning_rate": 0.0001,
+      "loss": 1.4968,
+      "step": 7649
+    },
+    {
+      "epoch": 0.878754810177474,
+      "grad_norm": 0.45039236545562744,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 7650
+    },
+    {
+      "epoch": 0.8788696800873012,
+      "grad_norm": 0.4908886253833771,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 7651
+    },
+    {
+      "epoch": 0.8789845499971283,
+      "grad_norm": 0.4798978865146637,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 7652
+    },
+    {
+      "epoch": 0.8790994199069554,
+      "grad_norm": 0.4604727327823639,
+      "learning_rate": 0.0001,
+      "loss": 1.6227,
+      "step": 7653
+    },
+    {
+      "epoch": 0.8792142898167825,
+      "grad_norm": 0.5215774774551392,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 7654
+    },
+    {
+      "epoch": 0.8793291597266096,
+      "grad_norm": 0.5219174027442932,
+      "learning_rate": 0.0001,
+      "loss": 1.8248,
+      "step": 7655
+    },
+    {
+      "epoch": 0.8794440296364368,
+      "grad_norm": 0.46888914704322815,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 7656
+    },
+    {
+      "epoch": 0.8795588995462639,
+      "grad_norm": 0.49896708130836487,
+      "learning_rate": 0.0001,
+      "loss": 1.679,
+      "step": 7657
+    },
+    {
+      "epoch": 0.879673769456091,
+      "grad_norm": 0.4882703423500061,
+      "learning_rate": 0.0001,
+      "loss": 1.7798,
+      "step": 7658
+    },
+    {
+      "epoch": 0.8797886393659181,
+      "grad_norm": 0.4401630461215973,
+      "learning_rate": 0.0001,
+      "loss": 1.5311,
+      "step": 7659
+    },
+    {
+      "epoch": 0.8799035092757452,
+      "grad_norm": 0.46130862832069397,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8800183791855724,
+      "grad_norm": 0.4488151967525482,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 7661
+    },
+    {
+      "epoch": 0.8801332490953995,
+      "grad_norm": 0.47853875160217285,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 7662
+    },
+    {
+      "epoch": 0.8802481190052266,
+      "grad_norm": 0.4806554615497589,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 7663
+    },
+    {
+      "epoch": 0.8803629889150537,
+      "grad_norm": 0.44891056418418884,
+      "learning_rate": 0.0001,
+      "loss": 1.4271,
+      "step": 7664
+    },
+    {
+      "epoch": 0.8804778588248808,
+      "grad_norm": 0.4926340579986572,
+      "learning_rate": 0.0001,
+      "loss": 1.6632,
+      "step": 7665
+    },
+    {
+      "epoch": 0.880592728734708,
+      "grad_norm": 0.47855570912361145,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 7666
+    },
+    {
+      "epoch": 0.8807075986445351,
+      "grad_norm": 0.47924453020095825,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 7667
+    },
+    {
+      "epoch": 0.8808224685543622,
+      "grad_norm": 0.45178931951522827,
+      "learning_rate": 0.0001,
+      "loss": 1.3517,
+      "step": 7668
+    },
+    {
+      "epoch": 0.8809373384641893,
+      "grad_norm": 0.4857328534126282,
+      "learning_rate": 0.0001,
+      "loss": 1.5011,
+      "step": 7669
+    },
+    {
+      "epoch": 0.8810522083740164,
+      "grad_norm": 0.5388020277023315,
+      "learning_rate": 0.0001,
+      "loss": 1.7677,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8811670782838436,
+      "grad_norm": 0.4225468039512634,
+      "learning_rate": 0.0001,
+      "loss": 1.3599,
+      "step": 7671
+    },
+    {
+      "epoch": 0.8812819481936707,
+      "grad_norm": 0.4660169780254364,
+      "learning_rate": 0.0001,
+      "loss": 1.641,
+      "step": 7672
+    },
+    {
+      "epoch": 0.8813968181034978,
+      "grad_norm": 0.47213736176490784,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 7673
+    },
+    {
+      "epoch": 0.8815116880133249,
+      "grad_norm": 0.4642762541770935,
+      "learning_rate": 0.0001,
+      "loss": 1.3589,
+      "step": 7674
+    },
+    {
+      "epoch": 0.881626557923152,
+      "grad_norm": 0.47581663727760315,
+      "learning_rate": 0.0001,
+      "loss": 1.475,
+      "step": 7675
+    },
+    {
+      "epoch": 0.8817414278329792,
+      "grad_norm": 0.4813084900379181,
+      "learning_rate": 0.0001,
+      "loss": 1.426,
+      "step": 7676
+    },
+    {
+      "epoch": 0.8818562977428063,
+      "grad_norm": 0.46025902032852173,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 7677
+    },
+    {
+      "epoch": 0.8819711676526334,
+      "grad_norm": 0.5673764944076538,
+      "learning_rate": 0.0001,
+      "loss": 1.6262,
+      "step": 7678
+    },
+    {
+      "epoch": 0.8820860375624605,
+      "grad_norm": 0.4987919330596924,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 7679
+    },
+    {
+      "epoch": 0.8822009074722876,
+      "grad_norm": 0.4518562853336334,
+      "learning_rate": 0.0001,
+      "loss": 1.3551,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8823157773821148,
+      "grad_norm": 0.486644446849823,
+      "learning_rate": 0.0001,
+      "loss": 1.5322,
+      "step": 7681
+    },
+    {
+      "epoch": 0.8824306472919419,
+      "grad_norm": 0.5359245538711548,
+      "learning_rate": 0.0001,
+      "loss": 1.8432,
+      "step": 7682
+    },
+    {
+      "epoch": 0.882545517201769,
+      "grad_norm": 0.4623430371284485,
+      "learning_rate": 0.0001,
+      "loss": 1.5393,
+      "step": 7683
+    },
+    {
+      "epoch": 0.8826603871115961,
+      "grad_norm": 0.47618022561073303,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 7684
+    },
+    {
+      "epoch": 0.8827752570214232,
+      "grad_norm": 0.5313566327095032,
+      "learning_rate": 0.0001,
+      "loss": 1.8855,
+      "step": 7685
+    },
+    {
+      "epoch": 0.8828901269312504,
+      "grad_norm": 0.4783889055252075,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 7686
+    },
+    {
+      "epoch": 0.8830049968410775,
+      "grad_norm": 0.45966142416000366,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 7687
+    },
+    {
+      "epoch": 0.8831198667509046,
+      "grad_norm": 0.47145649790763855,
+      "learning_rate": 0.0001,
+      "loss": 1.7645,
+      "step": 7688
+    },
+    {
+      "epoch": 0.8832347366607317,
+      "grad_norm": 0.4668165445327759,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 7689
+    },
+    {
+      "epoch": 0.8833496065705588,
+      "grad_norm": 0.44585132598876953,
+      "learning_rate": 0.0001,
+      "loss": 1.3982,
+      "step": 7690
+    },
+    {
+      "epoch": 0.883464476480386,
+      "grad_norm": 0.49249133467674255,
+      "learning_rate": 0.0001,
+      "loss": 1.6614,
+      "step": 7691
+    },
+    {
+      "epoch": 0.8835793463902131,
+      "grad_norm": 0.4700791835784912,
+      "learning_rate": 0.0001,
+      "loss": 1.3881,
+      "step": 7692
+    },
+    {
+      "epoch": 0.8836942163000402,
+      "grad_norm": 0.44664838910102844,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 7693
+    },
+    {
+      "epoch": 0.8838090862098673,
+      "grad_norm": 0.4859354496002197,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 7694
+    },
+    {
+      "epoch": 0.8839239561196944,
+      "grad_norm": 0.4571913480758667,
+      "learning_rate": 0.0001,
+      "loss": 1.5796,
+      "step": 7695
+    },
+    {
+      "epoch": 0.8840388260295216,
+      "grad_norm": 0.4279733896255493,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 7696
+    },
+    {
+      "epoch": 0.8841536959393487,
+      "grad_norm": 0.4860191345214844,
+      "learning_rate": 0.0001,
+      "loss": 1.827,
+      "step": 7697
+    },
+    {
+      "epoch": 0.8842685658491758,
+      "grad_norm": 0.4540635049343109,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 7698
+    },
+    {
+      "epoch": 0.8843834357590029,
+      "grad_norm": 0.5169913172721863,
+      "learning_rate": 0.0001,
+      "loss": 1.6006,
+      "step": 7699
+    },
+    {
+      "epoch": 0.88449830566883,
+      "grad_norm": 0.4786296784877777,
+      "learning_rate": 0.0001,
+      "loss": 1.5426,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8846131755786572,
+      "grad_norm": 0.4788622558116913,
+      "learning_rate": 0.0001,
+      "loss": 1.4655,
+      "step": 7701
+    },
+    {
+      "epoch": 0.8847280454884843,
+      "grad_norm": 0.4708092212677002,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 7702
+    },
+    {
+      "epoch": 0.8848429153983114,
+      "grad_norm": 0.49218055605888367,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 7703
+    },
+    {
+      "epoch": 0.8849577853081385,
+      "grad_norm": 0.4458230137825012,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 7704
+    },
+    {
+      "epoch": 0.8850726552179656,
+      "grad_norm": 0.46966198086738586,
+      "learning_rate": 0.0001,
+      "loss": 1.6702,
+      "step": 7705
+    },
+    {
+      "epoch": 0.8851875251277928,
+      "grad_norm": 0.5042508244514465,
+      "learning_rate": 0.0001,
+      "loss": 1.6373,
+      "step": 7706
+    },
+    {
+      "epoch": 0.8853023950376199,
+      "grad_norm": 0.5178636312484741,
+      "learning_rate": 0.0001,
+      "loss": 1.8095,
+      "step": 7707
+    },
+    {
+      "epoch": 0.885417264947447,
+      "grad_norm": 0.5226283073425293,
+      "learning_rate": 0.0001,
+      "loss": 1.8296,
+      "step": 7708
+    },
+    {
+      "epoch": 0.8855321348572741,
+      "grad_norm": 0.4413732886314392,
+      "learning_rate": 0.0001,
+      "loss": 1.4009,
+      "step": 7709
+    },
+    {
+      "epoch": 0.8856470047671012,
+      "grad_norm": 0.44534966349601746,
+      "learning_rate": 0.0001,
+      "loss": 1.5705,
+      "step": 7710
+    },
+    {
+      "epoch": 0.8857618746769284,
+      "grad_norm": 0.49753180146217346,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 7711
+    },
+    {
+      "epoch": 0.8858767445867555,
+      "grad_norm": 0.4720037281513214,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 7712
+    },
+    {
+      "epoch": 0.8859916144965826,
+      "grad_norm": 0.5182157754898071,
+      "learning_rate": 0.0001,
+      "loss": 1.6232,
+      "step": 7713
+    },
+    {
+      "epoch": 0.8861064844064097,
+      "grad_norm": 0.47871580719947815,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 7714
+    },
+    {
+      "epoch": 0.8862213543162368,
+      "grad_norm": 0.508307158946991,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 7715
+    },
+    {
+      "epoch": 0.886336224226064,
+      "grad_norm": 0.47485652565956116,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 7716
+    },
+    {
+      "epoch": 0.8864510941358911,
+      "grad_norm": 0.5017088651657104,
+      "learning_rate": 0.0001,
+      "loss": 1.6983,
+      "step": 7717
+    },
+    {
+      "epoch": 0.8865659640457182,
+      "grad_norm": 0.46508511900901794,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 7718
+    },
+    {
+      "epoch": 0.8866808339555453,
+      "grad_norm": 0.4940198063850403,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 7719
+    },
+    {
+      "epoch": 0.8867957038653724,
+      "grad_norm": 0.508063554763794,
+      "learning_rate": 0.0001,
+      "loss": 1.7217,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8869105737751996,
+      "grad_norm": 0.5674510598182678,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 7721
+    },
+    {
+      "epoch": 0.8870254436850267,
+      "grad_norm": 0.47527050971984863,
+      "learning_rate": 0.0001,
+      "loss": 1.4227,
+      "step": 7722
+    },
+    {
+      "epoch": 0.8871403135948538,
+      "grad_norm": 0.45663002133369446,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 7723
+    },
+    {
+      "epoch": 0.8872551835046809,
+      "grad_norm": 0.4941721558570862,
+      "learning_rate": 0.0001,
+      "loss": 1.7747,
+      "step": 7724
+    },
+    {
+      "epoch": 0.887370053414508,
+      "grad_norm": 0.43632128834724426,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 7725
+    },
+    {
+      "epoch": 0.8874849233243352,
+      "grad_norm": 0.4669651389122009,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 7726
+    },
+    {
+      "epoch": 0.8875997932341623,
+      "grad_norm": 0.5105260014533997,
+      "learning_rate": 0.0001,
+      "loss": 1.7566,
+      "step": 7727
+    },
+    {
+      "epoch": 0.8877146631439894,
+      "grad_norm": 0.44809389114379883,
+      "learning_rate": 0.0001,
+      "loss": 1.4762,
+      "step": 7728
+    },
+    {
+      "epoch": 0.8878295330538165,
+      "grad_norm": 0.4762137234210968,
+      "learning_rate": 0.0001,
+      "loss": 1.7521,
+      "step": 7729
+    },
+    {
+      "epoch": 0.8879444029636436,
+      "grad_norm": 0.45027822256088257,
+      "learning_rate": 0.0001,
+      "loss": 1.4717,
+      "step": 7730
+    },
+    {
+      "epoch": 0.8880592728734708,
+      "grad_norm": 0.4824502170085907,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 7731
+    },
+    {
+      "epoch": 0.8881741427832979,
+      "grad_norm": 0.4795003831386566,
+      "learning_rate": 0.0001,
+      "loss": 1.5373,
+      "step": 7732
+    },
+    {
+      "epoch": 0.888289012693125,
+      "grad_norm": 0.5017417073249817,
+      "learning_rate": 0.0001,
+      "loss": 1.6815,
+      "step": 7733
+    },
+    {
+      "epoch": 0.8884038826029521,
+      "grad_norm": 0.44462695717811584,
+      "learning_rate": 0.0001,
+      "loss": 1.4145,
+      "step": 7734
+    },
+    {
+      "epoch": 0.8885187525127792,
+      "grad_norm": 0.4432191550731659,
+      "learning_rate": 0.0001,
+      "loss": 1.4622,
+      "step": 7735
+    },
+    {
+      "epoch": 0.8886336224226064,
+      "grad_norm": 0.4664211571216583,
+      "learning_rate": 0.0001,
+      "loss": 1.6906,
+      "step": 7736
+    },
+    {
+      "epoch": 0.8887484923324335,
+      "grad_norm": 0.4574565589427948,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 7737
+    },
+    {
+      "epoch": 0.8888633622422606,
+      "grad_norm": 0.44210490584373474,
+      "learning_rate": 0.0001,
+      "loss": 1.4454,
+      "step": 7738
+    },
+    {
+      "epoch": 0.8889782321520877,
+      "grad_norm": 0.4957959055900574,
+      "learning_rate": 0.0001,
+      "loss": 1.7905,
+      "step": 7739
+    },
+    {
+      "epoch": 0.8890931020619148,
+      "grad_norm": 0.43715575337409973,
+      "learning_rate": 0.0001,
+      "loss": 1.4055,
+      "step": 7740
+    },
+    {
+      "epoch": 0.889207971971742,
+      "grad_norm": 0.4835394024848938,
+      "learning_rate": 0.0001,
+      "loss": 1.7379,
+      "step": 7741
+    },
+    {
+      "epoch": 0.8893228418815691,
+      "grad_norm": 0.4787488281726837,
+      "learning_rate": 0.0001,
+      "loss": 1.6368,
+      "step": 7742
+    },
+    {
+      "epoch": 0.8894377117913962,
+      "grad_norm": 0.4879782199859619,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 7743
+    },
+    {
+      "epoch": 0.8895525817012233,
+      "grad_norm": 0.4805709719657898,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 7744
+    },
+    {
+      "epoch": 0.8896674516110504,
+      "grad_norm": 0.4606061577796936,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 7745
+    },
+    {
+      "epoch": 0.8897823215208777,
+      "grad_norm": 0.5064094662666321,
+      "learning_rate": 0.0001,
+      "loss": 1.322,
+      "step": 7746
+    },
+    {
+      "epoch": 0.8898971914307048,
+      "grad_norm": 0.45881563425064087,
+      "learning_rate": 0.0001,
+      "loss": 1.5164,
+      "step": 7747
+    },
+    {
+      "epoch": 0.8900120613405319,
+      "grad_norm": 0.48323944211006165,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 7748
+    },
+    {
+      "epoch": 0.890126931250359,
+      "grad_norm": 0.49248838424682617,
+      "learning_rate": 0.0001,
+      "loss": 1.5483,
+      "step": 7749
+    },
+    {
+      "epoch": 0.8902418011601861,
+      "grad_norm": 0.46828705072402954,
+      "learning_rate": 0.0001,
+      "loss": 1.422,
+      "step": 7750
+    },
+    {
+      "epoch": 0.8903566710700133,
+      "grad_norm": 0.45508071780204773,
+      "learning_rate": 0.0001,
+      "loss": 1.5604,
+      "step": 7751
+    },
+    {
+      "epoch": 0.8904715409798404,
+      "grad_norm": 0.4920658469200134,
+      "learning_rate": 0.0001,
+      "loss": 1.4792,
+      "step": 7752
+    },
+    {
+      "epoch": 0.8905864108896675,
+      "grad_norm": 0.46648716926574707,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 7753
+    },
+    {
+      "epoch": 0.8907012807994946,
+      "grad_norm": 0.4534609019756317,
+      "learning_rate": 0.0001,
+      "loss": 1.3492,
+      "step": 7754
+    },
+    {
+      "epoch": 0.8908161507093217,
+      "grad_norm": 0.5462976098060608,
+      "learning_rate": 0.0001,
+      "loss": 1.2098,
+      "step": 7755
+    },
+    {
+      "epoch": 0.8909310206191489,
+      "grad_norm": 0.4742743670940399,
+      "learning_rate": 0.0001,
+      "loss": 1.6341,
+      "step": 7756
+    },
+    {
+      "epoch": 0.891045890528976,
+      "grad_norm": 0.46202293038368225,
+      "learning_rate": 0.0001,
+      "loss": 1.5176,
+      "step": 7757
+    },
+    {
+      "epoch": 0.8911607604388031,
+      "grad_norm": 0.45886507630348206,
+      "learning_rate": 0.0001,
+      "loss": 1.4584,
+      "step": 7758
+    },
+    {
+      "epoch": 0.8912756303486302,
+      "grad_norm": 0.4991988241672516,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 7759
+    },
+    {
+      "epoch": 0.8913905002584573,
+      "grad_norm": 0.47904571890830994,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8915053701682845,
+      "grad_norm": 0.48273780941963196,
+      "learning_rate": 0.0001,
+      "loss": 1.6437,
+      "step": 7761
+    },
+    {
+      "epoch": 0.8916202400781116,
+      "grad_norm": 0.45888492465019226,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 7762
+    },
+    {
+      "epoch": 0.8917351099879387,
+      "grad_norm": 0.485304057598114,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 7763
+    },
+    {
+      "epoch": 0.8918499798977658,
+      "grad_norm": 0.49057838320732117,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 7764
+    },
+    {
+      "epoch": 0.8919648498075929,
+      "grad_norm": 0.503322958946228,
+      "learning_rate": 0.0001,
+      "loss": 1.4175,
+      "step": 7765
+    },
+    {
+      "epoch": 0.8920797197174201,
+      "grad_norm": 0.4755953550338745,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 7766
+    },
+    {
+      "epoch": 0.8921945896272472,
+      "grad_norm": 0.5407209396362305,
+      "learning_rate": 0.0001,
+      "loss": 1.8089,
+      "step": 7767
+    },
+    {
+      "epoch": 0.8923094595370743,
+      "grad_norm": 0.48187482357025146,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 7768
+    },
+    {
+      "epoch": 0.8924243294469014,
+      "grad_norm": 0.4826925992965698,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 7769
+    },
+    {
+      "epoch": 0.8925391993567285,
+      "grad_norm": 0.4857804775238037,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 7770
+    },
+    {
+      "epoch": 0.8926540692665557,
+      "grad_norm": 0.4642079472541809,
+      "learning_rate": 0.0001,
+      "loss": 1.4843,
+      "step": 7771
+    },
+    {
+      "epoch": 0.8927689391763828,
+      "grad_norm": 0.44618815183639526,
+      "learning_rate": 0.0001,
+      "loss": 1.3971,
+      "step": 7772
+    },
+    {
+      "epoch": 0.8928838090862099,
+      "grad_norm": 0.44455963373184204,
+      "learning_rate": 0.0001,
+      "loss": 1.3867,
+      "step": 7773
+    },
+    {
+      "epoch": 0.892998678996037,
+      "grad_norm": 0.46926748752593994,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 7774
+    },
+    {
+      "epoch": 0.8931135489058641,
+      "grad_norm": 0.5021077394485474,
+      "learning_rate": 0.0001,
+      "loss": 1.7976,
+      "step": 7775
+    },
+    {
+      "epoch": 0.8932284188156913,
+      "grad_norm": 0.4602797031402588,
+      "learning_rate": 0.0001,
+      "loss": 1.599,
+      "step": 7776
+    },
+    {
+      "epoch": 0.8933432887255184,
+      "grad_norm": 0.4697769582271576,
+      "learning_rate": 0.0001,
+      "loss": 1.4241,
+      "step": 7777
+    },
+    {
+      "epoch": 0.8934581586353455,
+      "grad_norm": 0.48710909485816956,
+      "learning_rate": 0.0001,
+      "loss": 1.4455,
+      "step": 7778
+    },
+    {
+      "epoch": 0.8935730285451726,
+      "grad_norm": 0.47241660952568054,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 7779
+    },
+    {
+      "epoch": 0.8936878984549997,
+      "grad_norm": 0.488765686750412,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 7780
+    },
+    {
+      "epoch": 0.8938027683648269,
+      "grad_norm": 0.45212939381599426,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 7781
+    },
+    {
+      "epoch": 0.893917638274654,
+      "grad_norm": 0.49566277861595154,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 7782
+    },
+    {
+      "epoch": 0.8940325081844811,
+      "grad_norm": 0.4758758544921875,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 7783
+    },
+    {
+      "epoch": 0.8941473780943082,
+      "grad_norm": 0.47846412658691406,
+      "learning_rate": 0.0001,
+      "loss": 1.409,
+      "step": 7784
+    },
+    {
+      "epoch": 0.8942622480041353,
+      "grad_norm": 0.465116947889328,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 7785
+    },
+    {
+      "epoch": 0.8943771179139625,
+      "grad_norm": 0.4536508321762085,
+      "learning_rate": 0.0001,
+      "loss": 1.7166,
+      "step": 7786
+    },
+    {
+      "epoch": 0.8944919878237896,
+      "grad_norm": 0.44384217262268066,
+      "learning_rate": 0.0001,
+      "loss": 1.5588,
+      "step": 7787
+    },
+    {
+      "epoch": 0.8946068577336167,
+      "grad_norm": 0.46297183632850647,
+      "learning_rate": 0.0001,
+      "loss": 1.4644,
+      "step": 7788
+    },
+    {
+      "epoch": 0.8947217276434438,
+      "grad_norm": 0.5043572783470154,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 7789
+    },
+    {
+      "epoch": 0.8948365975532709,
+      "grad_norm": 0.49215108156204224,
+      "learning_rate": 0.0001,
+      "loss": 1.7364,
+      "step": 7790
+    },
+    {
+      "epoch": 0.8949514674630981,
+      "grad_norm": 0.45499980449676514,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 7791
+    },
+    {
+      "epoch": 0.8950663373729252,
+      "grad_norm": 0.46621590852737427,
+      "learning_rate": 0.0001,
+      "loss": 1.6982,
+      "step": 7792
+    },
+    {
+      "epoch": 0.8951812072827523,
+      "grad_norm": 0.4747898578643799,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 7793
+    },
+    {
+      "epoch": 0.8952960771925794,
+      "grad_norm": 0.4650057256221771,
+      "learning_rate": 0.0001,
+      "loss": 1.4734,
+      "step": 7794
+    },
+    {
+      "epoch": 0.8954109471024065,
+      "grad_norm": 0.45124003291130066,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 7795
+    },
+    {
+      "epoch": 0.8955258170122337,
+      "grad_norm": 0.469590425491333,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 7796
+    },
+    {
+      "epoch": 0.8956406869220608,
+      "grad_norm": 0.4690435826778412,
+      "learning_rate": 0.0001,
+      "loss": 1.5853,
+      "step": 7797
+    },
+    {
+      "epoch": 0.8957555568318879,
+      "grad_norm": 0.48957929015159607,
+      "learning_rate": 0.0001,
+      "loss": 1.767,
+      "step": 7798
+    },
+    {
+      "epoch": 0.895870426741715,
+      "grad_norm": 0.459736168384552,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 7799
+    },
+    {
+      "epoch": 0.8959852966515421,
+      "grad_norm": 0.4722045063972473,
+      "learning_rate": 0.0001,
+      "loss": 1.8041,
+      "step": 7800
+    },
+    {
+      "epoch": 0.8961001665613693,
+      "grad_norm": 0.5619189143180847,
+      "learning_rate": 0.0001,
+      "loss": 1.6918,
+      "step": 7801
+    },
+    {
+      "epoch": 0.8962150364711964,
+      "grad_norm": 0.4744798243045807,
+      "learning_rate": 0.0001,
+      "loss": 1.7405,
+      "step": 7802
+    },
+    {
+      "epoch": 0.8963299063810235,
+      "grad_norm": 0.4761870801448822,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 7803
+    },
+    {
+      "epoch": 0.8964447762908506,
+      "grad_norm": 0.45697519183158875,
+      "learning_rate": 0.0001,
+      "loss": 1.5555,
+      "step": 7804
+    },
+    {
+      "epoch": 0.8965596462006777,
+      "grad_norm": 0.5395013689994812,
+      "learning_rate": 0.0001,
+      "loss": 1.8088,
+      "step": 7805
+    },
+    {
+      "epoch": 0.8966745161105049,
+      "grad_norm": 0.4655267000198364,
+      "learning_rate": 0.0001,
+      "loss": 1.5784,
+      "step": 7806
+    },
+    {
+      "epoch": 0.896789386020332,
+      "grad_norm": 0.4401511549949646,
+      "learning_rate": 0.0001,
+      "loss": 1.4877,
+      "step": 7807
+    },
+    {
+      "epoch": 0.8969042559301591,
+      "grad_norm": 0.5011132955551147,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 7808
+    },
+    {
+      "epoch": 0.8970191258399862,
+      "grad_norm": 0.4679315984249115,
+      "learning_rate": 0.0001,
+      "loss": 1.3788,
+      "step": 7809
+    },
+    {
+      "epoch": 0.8971339957498133,
+      "grad_norm": 0.487979918718338,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 7810
+    },
+    {
+      "epoch": 0.8972488656596405,
+      "grad_norm": 0.4689542353153229,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 7811
+    },
+    {
+      "epoch": 0.8973637355694676,
+      "grad_norm": 0.49013927578926086,
+      "learning_rate": 0.0001,
+      "loss": 1.8521,
+      "step": 7812
+    },
+    {
+      "epoch": 0.8974786054792947,
+      "grad_norm": 0.4543326497077942,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 7813
+    },
+    {
+      "epoch": 0.8975934753891218,
+      "grad_norm": 0.4831598103046417,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 7814
+    },
+    {
+      "epoch": 0.8977083452989489,
+      "grad_norm": 0.49325430393218994,
+      "learning_rate": 0.0001,
+      "loss": 1.6324,
+      "step": 7815
+    },
+    {
+      "epoch": 0.8978232152087761,
+      "grad_norm": 0.5205132365226746,
+      "learning_rate": 0.0001,
+      "loss": 1.5647,
+      "step": 7816
+    },
+    {
+      "epoch": 0.8979380851186032,
+      "grad_norm": 0.485248327255249,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 7817
+    },
+    {
+      "epoch": 0.8980529550284303,
+      "grad_norm": 0.467001736164093,
+      "learning_rate": 0.0001,
+      "loss": 1.5915,
+      "step": 7818
+    },
+    {
+      "epoch": 0.8981678249382574,
+      "grad_norm": 0.519828200340271,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 7819
+    },
+    {
+      "epoch": 0.8982826948480845,
+      "grad_norm": 0.4926901161670685,
+      "learning_rate": 0.0001,
+      "loss": 1.7966,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8983975647579117,
+      "grad_norm": 0.4958095848560333,
+      "learning_rate": 0.0001,
+      "loss": 1.6828,
+      "step": 7821
+    },
+    {
+      "epoch": 0.8985124346677388,
+      "grad_norm": 0.4742107093334198,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 7822
+    },
+    {
+      "epoch": 0.8986273045775659,
+      "grad_norm": 0.48758435249328613,
+      "learning_rate": 0.0001,
+      "loss": 1.7207,
+      "step": 7823
+    },
+    {
+      "epoch": 0.898742174487393,
+      "grad_norm": 0.48550042510032654,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 7824
+    },
+    {
+      "epoch": 0.8988570443972201,
+      "grad_norm": 0.5684822201728821,
+      "learning_rate": 0.0001,
+      "loss": 2.0053,
+      "step": 7825
+    },
+    {
+      "epoch": 0.8989719143070473,
+      "grad_norm": 0.4416176378726959,
+      "learning_rate": 0.0001,
+      "loss": 1.4682,
+      "step": 7826
+    },
+    {
+      "epoch": 0.8990867842168744,
+      "grad_norm": 0.44723203778266907,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 7827
+    },
+    {
+      "epoch": 0.8992016541267015,
+      "grad_norm": 0.4821613132953644,
+      "learning_rate": 0.0001,
+      "loss": 1.6573,
+      "step": 7828
+    },
+    {
+      "epoch": 0.8993165240365286,
+      "grad_norm": 0.4443539083003998,
+      "learning_rate": 0.0001,
+      "loss": 1.4682,
+      "step": 7829
+    },
+    {
+      "epoch": 0.8994313939463557,
+      "grad_norm": 0.5018233060836792,
+      "learning_rate": 0.0001,
+      "loss": 1.7302,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8995462638561829,
+      "grad_norm": 0.4475822448730469,
+      "learning_rate": 0.0001,
+      "loss": 1.5244,
+      "step": 7831
+    },
+    {
+      "epoch": 0.89966113376601,
+      "grad_norm": 0.4100770652294159,
+      "learning_rate": 0.0001,
+      "loss": 1.3041,
+      "step": 7832
+    },
+    {
+      "epoch": 0.8997760036758371,
+      "grad_norm": 0.5303254723548889,
+      "learning_rate": 0.0001,
+      "loss": 1.9058,
+      "step": 7833
+    },
+    {
+      "epoch": 0.8998908735856642,
+      "grad_norm": 0.4729917645454407,
+      "learning_rate": 0.0001,
+      "loss": 1.4515,
+      "step": 7834
+    },
+    {
+      "epoch": 0.9000057434954913,
+      "grad_norm": 0.45676717162132263,
+      "learning_rate": 0.0001,
+      "loss": 1.4337,
+      "step": 7835
+    },
+    {
+      "epoch": 0.9001206134053185,
+      "grad_norm": 0.4548957943916321,
+      "learning_rate": 0.0001,
+      "loss": 1.5653,
+      "step": 7836
+    },
+    {
+      "epoch": 0.9002354833151456,
+      "grad_norm": 0.46776890754699707,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 7837
+    },
+    {
+      "epoch": 0.9003503532249727,
+      "grad_norm": 0.8313576579093933,
+      "learning_rate": 0.0001,
+      "loss": 1.6672,
+      "step": 7838
+    },
+    {
+      "epoch": 0.9004652231347998,
+      "grad_norm": 0.45772501826286316,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 7839
+    },
+    {
+      "epoch": 0.9005800930446269,
+      "grad_norm": 0.4540563225746155,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 7840
+    },
+    {
+      "epoch": 0.9006949629544541,
+      "grad_norm": 0.43928417563438416,
+      "learning_rate": 0.0001,
+      "loss": 1.4581,
+      "step": 7841
+    },
+    {
+      "epoch": 0.9008098328642812,
+      "grad_norm": 0.46283161640167236,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 7842
+    },
+    {
+      "epoch": 0.9009247027741083,
+      "grad_norm": 0.49827006459236145,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 7843
+    },
+    {
+      "epoch": 0.9010395726839354,
+      "grad_norm": 0.48653677105903625,
+      "learning_rate": 0.0001,
+      "loss": 1.5447,
+      "step": 7844
+    },
+    {
+      "epoch": 0.9011544425937625,
+      "grad_norm": 0.5136744976043701,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 7845
+    },
+    {
+      "epoch": 0.9012693125035897,
+      "grad_norm": 0.4694797694683075,
+      "learning_rate": 0.0001,
+      "loss": 1.5367,
+      "step": 7846
+    },
+    {
+      "epoch": 0.9013841824134168,
+      "grad_norm": 0.47298017144203186,
+      "learning_rate": 0.0001,
+      "loss": 1.4799,
+      "step": 7847
+    },
+    {
+      "epoch": 0.9014990523232439,
+      "grad_norm": 0.46906399726867676,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 7848
+    },
+    {
+      "epoch": 0.901613922233071,
+      "grad_norm": 0.473000168800354,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 7849
+    },
+    {
+      "epoch": 0.9017287921428981,
+      "grad_norm": 0.462663471698761,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 7850
+    },
+    {
+      "epoch": 0.9018436620527253,
+      "grad_norm": 0.48268601298332214,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 7851
+    },
+    {
+      "epoch": 0.9019585319625524,
+      "grad_norm": 0.4828319251537323,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 7852
+    },
+    {
+      "epoch": 0.9020734018723795,
+      "grad_norm": 0.5251330733299255,
+      "learning_rate": 0.0001,
+      "loss": 1.8425,
+      "step": 7853
+    },
+    {
+      "epoch": 0.9021882717822066,
+      "grad_norm": 0.478097140789032,
+      "learning_rate": 0.0001,
+      "loss": 1.4669,
+      "step": 7854
+    },
+    {
+      "epoch": 0.9023031416920337,
+      "grad_norm": 0.4791843891143799,
+      "learning_rate": 0.0001,
+      "loss": 1.4991,
+      "step": 7855
+    },
+    {
+      "epoch": 0.9024180116018609,
+      "grad_norm": 0.4416709244251251,
+      "learning_rate": 0.0001,
+      "loss": 1.2615,
+      "step": 7856
+    },
+    {
+      "epoch": 0.902532881511688,
+      "grad_norm": 0.4547010362148285,
+      "learning_rate": 0.0001,
+      "loss": 1.6578,
+      "step": 7857
+    },
+    {
+      "epoch": 0.9026477514215151,
+      "grad_norm": 0.47191575169563293,
+      "learning_rate": 0.0001,
+      "loss": 1.6771,
+      "step": 7858
+    },
+    {
+      "epoch": 0.9027626213313422,
+      "grad_norm": 0.4907677173614502,
+      "learning_rate": 0.0001,
+      "loss": 1.6336,
+      "step": 7859
+    },
+    {
+      "epoch": 0.9028774912411693,
+      "grad_norm": 0.4836212396621704,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 7860
+    },
+    {
+      "epoch": 0.9029923611509965,
+      "grad_norm": 0.46854549646377563,
+      "learning_rate": 0.0001,
+      "loss": 1.7531,
+      "step": 7861
+    },
+    {
+      "epoch": 0.9031072310608236,
+      "grad_norm": 0.45791977643966675,
+      "learning_rate": 0.0001,
+      "loss": 1.6377,
+      "step": 7862
+    },
+    {
+      "epoch": 0.9032221009706507,
+      "grad_norm": 0.4522063732147217,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 7863
+    },
+    {
+      "epoch": 0.9033369708804778,
+      "grad_norm": 0.4975827932357788,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 7864
+    },
+    {
+      "epoch": 0.9034518407903049,
+      "grad_norm": 0.450082927942276,
+      "learning_rate": 0.0001,
+      "loss": 1.69,
+      "step": 7865
+    },
+    {
+      "epoch": 0.9035667107001321,
+      "grad_norm": 0.4922288656234741,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 7866
+    },
+    {
+      "epoch": 0.9036815806099592,
+      "grad_norm": 0.4563447833061218,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 7867
+    },
+    {
+      "epoch": 0.9037964505197863,
+      "grad_norm": 0.466033399105072,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 7868
+    },
+    {
+      "epoch": 0.9039113204296134,
+      "grad_norm": 0.48967087268829346,
+      "learning_rate": 0.0001,
+      "loss": 1.6251,
+      "step": 7869
+    },
+    {
+      "epoch": 0.9040261903394405,
+      "grad_norm": 0.4684128165245056,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 7870
+    },
+    {
+      "epoch": 0.9041410602492677,
+      "grad_norm": 0.5109823942184448,
+      "learning_rate": 0.0001,
+      "loss": 1.6527,
+      "step": 7871
+    },
+    {
+      "epoch": 0.9042559301590948,
+      "grad_norm": 0.46000662446022034,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 7872
+    },
+    {
+      "epoch": 0.9043708000689219,
+      "grad_norm": 0.5103839039802551,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 7873
+    },
+    {
+      "epoch": 0.904485669978749,
+      "grad_norm": 0.47217777371406555,
+      "learning_rate": 0.0001,
+      "loss": 1.6014,
+      "step": 7874
+    },
+    {
+      "epoch": 0.9046005398885761,
+      "grad_norm": 0.5040363073348999,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 7875
+    },
+    {
+      "epoch": 0.9047154097984033,
+      "grad_norm": 0.4671007990837097,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 7876
+    },
+    {
+      "epoch": 0.9048302797082304,
+      "grad_norm": 0.467616468667984,
+      "learning_rate": 0.0001,
+      "loss": 1.5478,
+      "step": 7877
+    },
+    {
+      "epoch": 0.9049451496180575,
+      "grad_norm": 0.4528937339782715,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 7878
+    },
+    {
+      "epoch": 0.9050600195278846,
+      "grad_norm": 0.4835106432437897,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 7879
+    },
+    {
+      "epoch": 0.9051748894377117,
+      "grad_norm": 0.4774007201194763,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 7880
+    },
+    {
+      "epoch": 0.9052897593475389,
+      "grad_norm": 0.4768272340297699,
+      "learning_rate": 0.0001,
+      "loss": 1.4522,
+      "step": 7881
+    },
+    {
+      "epoch": 0.905404629257366,
+      "grad_norm": 0.4725984036922455,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 7882
+    },
+    {
+      "epoch": 0.9055194991671932,
+      "grad_norm": 0.44116008281707764,
+      "learning_rate": 0.0001,
+      "loss": 1.3919,
+      "step": 7883
+    },
+    {
+      "epoch": 0.9056343690770203,
+      "grad_norm": 0.4748861789703369,
+      "learning_rate": 0.0001,
+      "loss": 1.6583,
+      "step": 7884
+    },
+    {
+      "epoch": 0.9057492389868474,
+      "grad_norm": 0.4855034053325653,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 7885
+    },
+    {
+      "epoch": 0.9058641088966746,
+      "grad_norm": 0.4543818533420563,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 7886
+    },
+    {
+      "epoch": 0.9059789788065017,
+      "grad_norm": 0.529973566532135,
+      "learning_rate": 0.0001,
+      "loss": 1.9031,
+      "step": 7887
+    },
+    {
+      "epoch": 0.9060938487163288,
+      "grad_norm": 0.445414662361145,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 7888
+    },
+    {
+      "epoch": 0.9062087186261559,
+      "grad_norm": 0.5017836689949036,
+      "learning_rate": 0.0001,
+      "loss": 1.4926,
+      "step": 7889
+    },
+    {
+      "epoch": 0.906323588535983,
+      "grad_norm": 0.4460631012916565,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 7890
+    },
+    {
+      "epoch": 0.9064384584458102,
+      "grad_norm": 0.4223814904689789,
+      "learning_rate": 0.0001,
+      "loss": 1.2901,
+      "step": 7891
+    },
+    {
+      "epoch": 0.9065533283556373,
+      "grad_norm": 0.4735008478164673,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 7892
+    },
+    {
+      "epoch": 0.9066681982654644,
+      "grad_norm": 0.4891226589679718,
+      "learning_rate": 0.0001,
+      "loss": 1.6429,
+      "step": 7893
+    },
+    {
+      "epoch": 0.9067830681752915,
+      "grad_norm": 0.49529507756233215,
+      "learning_rate": 0.0001,
+      "loss": 1.4392,
+      "step": 7894
+    },
+    {
+      "epoch": 0.9068979380851186,
+      "grad_norm": 0.4905736446380615,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 7895
+    },
+    {
+      "epoch": 0.9070128079949458,
+      "grad_norm": 0.53058922290802,
+      "learning_rate": 0.0001,
+      "loss": 1.4193,
+      "step": 7896
+    },
+    {
+      "epoch": 0.9071276779047729,
+      "grad_norm": 0.42824816703796387,
+      "learning_rate": 0.0001,
+      "loss": 1.4727,
+      "step": 7897
+    },
+    {
+      "epoch": 0.9072425478146,
+      "grad_norm": 0.4656597673892975,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 7898
+    },
+    {
+      "epoch": 0.9073574177244271,
+      "grad_norm": 0.5175133347511292,
+      "learning_rate": 0.0001,
+      "loss": 1.7236,
+      "step": 7899
+    },
+    {
+      "epoch": 0.9074722876342542,
+      "grad_norm": 0.4708557426929474,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 7900
+    },
+    {
+      "epoch": 0.9075871575440814,
+      "grad_norm": 0.49528563022613525,
+      "learning_rate": 0.0001,
+      "loss": 1.5486,
+      "step": 7901
+    },
+    {
+      "epoch": 0.9077020274539085,
+      "grad_norm": 0.45954427123069763,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 7902
+    },
+    {
+      "epoch": 0.9078168973637356,
+      "grad_norm": 0.48827245831489563,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 7903
+    },
+    {
+      "epoch": 0.9079317672735627,
+      "grad_norm": 0.4947773218154907,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 7904
+    },
+    {
+      "epoch": 0.9080466371833898,
+      "grad_norm": 0.4761560261249542,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 7905
+    },
+    {
+      "epoch": 0.908161507093217,
+      "grad_norm": 0.48969805240631104,
+      "learning_rate": 0.0001,
+      "loss": 1.5753,
+      "step": 7906
+    },
+    {
+      "epoch": 0.9082763770030441,
+      "grad_norm": 0.49315011501312256,
+      "learning_rate": 0.0001,
+      "loss": 1.6573,
+      "step": 7907
+    },
+    {
+      "epoch": 0.9083912469128712,
+      "grad_norm": 0.4530597925186157,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 7908
+    },
+    {
+      "epoch": 0.9085061168226983,
+      "grad_norm": 0.4942728877067566,
+      "learning_rate": 0.0001,
+      "loss": 1.7064,
+      "step": 7909
+    },
+    {
+      "epoch": 0.9086209867325254,
+      "grad_norm": 0.44348907470703125,
+      "learning_rate": 0.0001,
+      "loss": 1.506,
+      "step": 7910
+    },
+    {
+      "epoch": 0.9087358566423526,
+      "grad_norm": 0.4613859951496124,
+      "learning_rate": 0.0001,
+      "loss": 1.5753,
+      "step": 7911
+    },
+    {
+      "epoch": 0.9088507265521797,
+      "grad_norm": 0.46956050395965576,
+      "learning_rate": 0.0001,
+      "loss": 1.2878,
+      "step": 7912
+    },
+    {
+      "epoch": 0.9089655964620068,
+      "grad_norm": 0.5553674697875977,
+      "learning_rate": 0.0001,
+      "loss": 1.8513,
+      "step": 7913
+    },
+    {
+      "epoch": 0.9090804663718339,
+      "grad_norm": 0.48194214701652527,
+      "learning_rate": 0.0001,
+      "loss": 1.3442,
+      "step": 7914
+    },
+    {
+      "epoch": 0.909195336281661,
+      "grad_norm": 0.4378836750984192,
+      "learning_rate": 0.0001,
+      "loss": 1.4797,
+      "step": 7915
+    },
+    {
+      "epoch": 0.9093102061914882,
+      "grad_norm": 0.4518606662750244,
+      "learning_rate": 0.0001,
+      "loss": 1.5698,
+      "step": 7916
+    },
+    {
+      "epoch": 0.9094250761013153,
+      "grad_norm": 0.46091845631599426,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 7917
+    },
+    {
+      "epoch": 0.9095399460111424,
+      "grad_norm": 0.447496235370636,
+      "learning_rate": 0.0001,
+      "loss": 1.5605,
+      "step": 7918
+    },
+    {
+      "epoch": 0.9096548159209695,
+      "grad_norm": 0.5359869599342346,
+      "learning_rate": 0.0001,
+      "loss": 1.6687,
+      "step": 7919
+    },
+    {
+      "epoch": 0.9097696858307966,
+      "grad_norm": 0.5245752930641174,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 7920
+    },
+    {
+      "epoch": 0.9098845557406238,
+      "grad_norm": 0.5289234519004822,
+      "learning_rate": 0.0001,
+      "loss": 1.6148,
+      "step": 7921
+    },
+    {
+      "epoch": 0.9099994256504509,
+      "grad_norm": 0.49240848422050476,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 7922
+    },
+    {
+      "epoch": 0.910114295560278,
+      "grad_norm": 0.46257832646369934,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 7923
+    },
+    {
+      "epoch": 0.9102291654701051,
+      "grad_norm": 0.5155847072601318,
+      "learning_rate": 0.0001,
+      "loss": 1.6578,
+      "step": 7924
+    },
+    {
+      "epoch": 0.9103440353799322,
+      "grad_norm": 0.5221872925758362,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 7925
+    },
+    {
+      "epoch": 0.9104589052897594,
+      "grad_norm": 0.4366224408149719,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 7926
+    },
+    {
+      "epoch": 0.9105737751995865,
+      "grad_norm": 0.48819804191589355,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 7927
+    },
+    {
+      "epoch": 0.9106886451094136,
+      "grad_norm": 0.4560621678829193,
+      "learning_rate": 0.0001,
+      "loss": 1.4342,
+      "step": 7928
+    },
+    {
+      "epoch": 0.9108035150192407,
+      "grad_norm": 0.48248445987701416,
+      "learning_rate": 0.0001,
+      "loss": 1.7123,
+      "step": 7929
+    },
+    {
+      "epoch": 0.9109183849290678,
+      "grad_norm": 0.4500480890274048,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 7930
+    },
+    {
+      "epoch": 0.911033254838895,
+      "grad_norm": 0.52225661277771,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 7931
+    },
+    {
+      "epoch": 0.9111481247487221,
+      "grad_norm": 0.5065485239028931,
+      "learning_rate": 0.0001,
+      "loss": 1.3653,
+      "step": 7932
+    },
+    {
+      "epoch": 0.9112629946585492,
+      "grad_norm": 0.5204269886016846,
+      "learning_rate": 0.0001,
+      "loss": 1.7194,
+      "step": 7933
+    },
+    {
+      "epoch": 0.9113778645683763,
+      "grad_norm": 0.5047922134399414,
+      "learning_rate": 0.0001,
+      "loss": 1.6934,
+      "step": 7934
+    },
+    {
+      "epoch": 0.9114927344782034,
+      "grad_norm": 0.4803232252597809,
+      "learning_rate": 0.0001,
+      "loss": 1.7358,
+      "step": 7935
+    },
+    {
+      "epoch": 0.9116076043880306,
+      "grad_norm": 0.5010977983474731,
+      "learning_rate": 0.0001,
+      "loss": 1.6251,
+      "step": 7936
+    },
+    {
+      "epoch": 0.9117224742978577,
+      "grad_norm": 0.4615200459957123,
+      "learning_rate": 0.0001,
+      "loss": 1.5255,
+      "step": 7937
+    },
+    {
+      "epoch": 0.9118373442076848,
+      "grad_norm": 0.4917146861553192,
+      "learning_rate": 0.0001,
+      "loss": 1.5949,
+      "step": 7938
+    },
+    {
+      "epoch": 0.9119522141175119,
+      "grad_norm": 0.45587408542633057,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 7939
+    },
+    {
+      "epoch": 0.912067084027339,
+      "grad_norm": 0.4592360556125641,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 7940
+    },
+    {
+      "epoch": 0.9121819539371662,
+      "grad_norm": 0.49223262071609497,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 7941
+    },
+    {
+      "epoch": 0.9122968238469933,
+      "grad_norm": 0.46011883020401,
+      "learning_rate": 0.0001,
+      "loss": 1.3733,
+      "step": 7942
+    },
+    {
+      "epoch": 0.9124116937568204,
+      "grad_norm": 0.49059078097343445,
+      "learning_rate": 0.0001,
+      "loss": 1.7346,
+      "step": 7943
+    },
+    {
+      "epoch": 0.9125265636666475,
+      "grad_norm": 0.472307026386261,
+      "learning_rate": 0.0001,
+      "loss": 1.5103,
+      "step": 7944
+    },
+    {
+      "epoch": 0.9126414335764746,
+      "grad_norm": 0.48213303089141846,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 7945
+    },
+    {
+      "epoch": 0.9127563034863018,
+      "grad_norm": 0.5104955434799194,
+      "learning_rate": 0.0001,
+      "loss": 1.8161,
+      "step": 7946
+    },
+    {
+      "epoch": 0.9128711733961289,
+      "grad_norm": 0.4811389446258545,
+      "learning_rate": 0.0001,
+      "loss": 1.5841,
+      "step": 7947
+    },
+    {
+      "epoch": 0.912986043305956,
+      "grad_norm": 0.4906313419342041,
+      "learning_rate": 0.0001,
+      "loss": 1.7777,
+      "step": 7948
+    },
+    {
+      "epoch": 0.9131009132157831,
+      "grad_norm": 0.45326316356658936,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 7949
+    },
+    {
+      "epoch": 0.9132157831256102,
+      "grad_norm": 0.41644978523254395,
+      "learning_rate": 0.0001,
+      "loss": 1.306,
+      "step": 7950
+    },
+    {
+      "epoch": 0.9133306530354374,
+      "grad_norm": 0.432559996843338,
+      "learning_rate": 0.0001,
+      "loss": 1.3399,
+      "step": 7951
+    },
+    {
+      "epoch": 0.9134455229452645,
+      "grad_norm": 0.4607470631599426,
+      "learning_rate": 0.0001,
+      "loss": 1.6326,
+      "step": 7952
+    },
+    {
+      "epoch": 0.9135603928550916,
+      "grad_norm": 0.5131176710128784,
+      "learning_rate": 0.0001,
+      "loss": 1.7356,
+      "step": 7953
+    },
+    {
+      "epoch": 0.9136752627649187,
+      "grad_norm": 0.4790054261684418,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 7954
+    },
+    {
+      "epoch": 0.9137901326747458,
+      "grad_norm": 0.4506865441799164,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 7955
+    },
+    {
+      "epoch": 0.913905002584573,
+      "grad_norm": 0.44508710503578186,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 7956
+    },
+    {
+      "epoch": 0.9140198724944001,
+      "grad_norm": 0.43661433458328247,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 7957
+    },
+    {
+      "epoch": 0.9141347424042272,
+      "grad_norm": 0.456387460231781,
+      "learning_rate": 0.0001,
+      "loss": 1.6312,
+      "step": 7958
+    },
+    {
+      "epoch": 0.9142496123140543,
+      "grad_norm": 0.5062097311019897,
+      "learning_rate": 0.0001,
+      "loss": 1.7662,
+      "step": 7959
+    },
+    {
+      "epoch": 0.9143644822238814,
+      "grad_norm": 0.4519118070602417,
+      "learning_rate": 0.0001,
+      "loss": 1.5057,
+      "step": 7960
+    },
+    {
+      "epoch": 0.9144793521337086,
+      "grad_norm": 0.4753338098526001,
+      "learning_rate": 0.0001,
+      "loss": 1.5305,
+      "step": 7961
+    },
+    {
+      "epoch": 0.9145942220435357,
+      "grad_norm": 0.44463813304901123,
+      "learning_rate": 0.0001,
+      "loss": 1.4873,
+      "step": 7962
+    },
+    {
+      "epoch": 0.9147090919533628,
+      "grad_norm": 0.4657713770866394,
+      "learning_rate": 0.0001,
+      "loss": 1.5608,
+      "step": 7963
+    },
+    {
+      "epoch": 0.9148239618631899,
+      "grad_norm": 0.47514888644218445,
+      "learning_rate": 0.0001,
+      "loss": 1.4934,
+      "step": 7964
+    },
+    {
+      "epoch": 0.914938831773017,
+      "grad_norm": 0.5030425786972046,
+      "learning_rate": 0.0001,
+      "loss": 1.6392,
+      "step": 7965
+    },
+    {
+      "epoch": 0.9150537016828442,
+      "grad_norm": 0.49183422327041626,
+      "learning_rate": 0.0001,
+      "loss": 1.6318,
+      "step": 7966
+    },
+    {
+      "epoch": 0.9151685715926713,
+      "grad_norm": 0.4909827709197998,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 7967
+    },
+    {
+      "epoch": 0.9152834415024984,
+      "grad_norm": 0.5114017724990845,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 7968
+    },
+    {
+      "epoch": 0.9153983114123255,
+      "grad_norm": 0.5097264647483826,
+      "learning_rate": 0.0001,
+      "loss": 1.7322,
+      "step": 7969
+    },
+    {
+      "epoch": 0.9155131813221526,
+      "grad_norm": 0.4767652750015259,
+      "learning_rate": 0.0001,
+      "loss": 1.4866,
+      "step": 7970
+    },
+    {
+      "epoch": 0.9156280512319798,
+      "grad_norm": 0.4586257338523865,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 7971
+    },
+    {
+      "epoch": 0.9157429211418069,
+      "grad_norm": 0.4332074224948883,
+      "learning_rate": 0.0001,
+      "loss": 1.3948,
+      "step": 7972
+    },
+    {
+      "epoch": 0.915857791051634,
+      "grad_norm": 0.48241952061653137,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 7973
+    },
+    {
+      "epoch": 0.9159726609614611,
+      "grad_norm": 0.5874638557434082,
+      "learning_rate": 0.0001,
+      "loss": 1.895,
+      "step": 7974
+    },
+    {
+      "epoch": 0.9160875308712882,
+      "grad_norm": 0.4532858431339264,
+      "learning_rate": 0.0001,
+      "loss": 1.6298,
+      "step": 7975
+    },
+    {
+      "epoch": 0.9162024007811154,
+      "grad_norm": 0.47365015745162964,
+      "learning_rate": 0.0001,
+      "loss": 1.6582,
+      "step": 7976
+    },
+    {
+      "epoch": 0.9163172706909425,
+      "grad_norm": 0.5175490975379944,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 7977
+    },
+    {
+      "epoch": 0.9164321406007696,
+      "grad_norm": 0.47895553708076477,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 7978
+    },
+    {
+      "epoch": 0.9165470105105967,
+      "grad_norm": 0.4826950132846832,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 7979
+    },
+    {
+      "epoch": 0.9166618804204238,
+      "grad_norm": 0.4933220148086548,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 7980
+    },
+    {
+      "epoch": 0.916776750330251,
+      "grad_norm": 0.47146233916282654,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 7981
+    },
+    {
+      "epoch": 0.9168916202400781,
+      "grad_norm": 0.4537650942802429,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 7982
+    },
+    {
+      "epoch": 0.9170064901499052,
+      "grad_norm": 0.5133910179138184,
+      "learning_rate": 0.0001,
+      "loss": 1.75,
+      "step": 7983
+    },
+    {
+      "epoch": 0.9171213600597323,
+      "grad_norm": 0.4584057927131653,
+      "learning_rate": 0.0001,
+      "loss": 1.5925,
+      "step": 7984
+    },
+    {
+      "epoch": 0.9172362299695594,
+      "grad_norm": 0.4383944571018219,
+      "learning_rate": 0.0001,
+      "loss": 1.4504,
+      "step": 7985
+    },
+    {
+      "epoch": 0.9173510998793866,
+      "grad_norm": 0.482051819562912,
+      "learning_rate": 0.0001,
+      "loss": 1.4079,
+      "step": 7986
+    },
+    {
+      "epoch": 0.9174659697892137,
+      "grad_norm": 0.4862101078033447,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 7987
+    },
+    {
+      "epoch": 0.9175808396990408,
+      "grad_norm": 0.46888962388038635,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 7988
+    },
+    {
+      "epoch": 0.9176957096088679,
+      "grad_norm": 0.4773329198360443,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 7989
+    },
+    {
+      "epoch": 0.917810579518695,
+      "grad_norm": 0.5123506784439087,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 7990
+    },
+    {
+      "epoch": 0.9179254494285222,
+      "grad_norm": 0.4806014895439148,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 7991
+    },
+    {
+      "epoch": 0.9180403193383493,
+      "grad_norm": 0.45858410000801086,
+      "learning_rate": 0.0001,
+      "loss": 1.5983,
+      "step": 7992
+    },
+    {
+      "epoch": 0.9181551892481764,
+      "grad_norm": 0.5545932650566101,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 7993
+    },
+    {
+      "epoch": 0.9182700591580035,
+      "grad_norm": 0.4767701327800751,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 7994
+    },
+    {
+      "epoch": 0.9183849290678306,
+      "grad_norm": 0.5300646424293518,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 7995
+    },
+    {
+      "epoch": 0.9184997989776578,
+      "grad_norm": 0.4707062244415283,
+      "learning_rate": 0.0001,
+      "loss": 1.6684,
+      "step": 7996
+    },
+    {
+      "epoch": 0.9186146688874849,
+      "grad_norm": 0.47407713532447815,
+      "learning_rate": 0.0001,
+      "loss": 1.5561,
+      "step": 7997
+    },
+    {
+      "epoch": 0.918729538797312,
+      "grad_norm": 0.4721267521381378,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 7998
+    },
+    {
+      "epoch": 0.9188444087071391,
+      "grad_norm": 0.48417842388153076,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 7999
+    },
+    {
+      "epoch": 0.9189592786169662,
+      "grad_norm": 0.4994370639324188,
+      "learning_rate": 0.0001,
+      "loss": 1.786,
+      "step": 8000
+    },
+    {
+      "epoch": 0.9190741485267934,
+      "grad_norm": 0.5283456444740295,
+      "learning_rate": 0.0001,
+      "loss": 1.8096,
+      "step": 8001
+    },
+    {
+      "epoch": 0.9191890184366205,
+      "grad_norm": 0.4809178411960602,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 8002
+    },
+    {
+      "epoch": 0.9193038883464476,
+      "grad_norm": 0.5179066061973572,
+      "learning_rate": 0.0001,
+      "loss": 1.6071,
+      "step": 8003
+    },
+    {
+      "epoch": 0.9194187582562747,
+      "grad_norm": 0.5334687829017639,
+      "learning_rate": 0.0001,
+      "loss": 1.759,
+      "step": 8004
+    },
+    {
+      "epoch": 0.9195336281661018,
+      "grad_norm": 0.4858127236366272,
+      "learning_rate": 0.0001,
+      "loss": 1.6873,
+      "step": 8005
+    },
+    {
+      "epoch": 0.919648498075929,
+      "grad_norm": 0.47769853472709656,
+      "learning_rate": 0.0001,
+      "loss": 1.4748,
+      "step": 8006
+    },
+    {
+      "epoch": 0.9197633679857561,
+      "grad_norm": 0.5166919827461243,
+      "learning_rate": 0.0001,
+      "loss": 1.7652,
+      "step": 8007
+    },
+    {
+      "epoch": 0.9198782378955832,
+      "grad_norm": 0.45674726366996765,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 8008
+    },
+    {
+      "epoch": 0.9199931078054103,
+      "grad_norm": 0.45661020278930664,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 8009
+    },
+    {
+      "epoch": 0.9201079777152374,
+      "grad_norm": 0.5638238787651062,
+      "learning_rate": 0.0001,
+      "loss": 1.9338,
+      "step": 8010
+    },
+    {
+      "epoch": 0.9202228476250646,
+      "grad_norm": 0.4804632067680359,
+      "learning_rate": 0.0001,
+      "loss": 1.695,
+      "step": 8011
+    },
+    {
+      "epoch": 0.9203377175348917,
+      "grad_norm": 0.45596975088119507,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 8012
+    },
+    {
+      "epoch": 0.9204525874447188,
+      "grad_norm": 0.47516369819641113,
+      "learning_rate": 0.0001,
+      "loss": 1.6081,
+      "step": 8013
+    },
+    {
+      "epoch": 0.9205674573545459,
+      "grad_norm": 0.47999808192253113,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 8014
+    },
+    {
+      "epoch": 0.920682327264373,
+      "grad_norm": 0.4690406918525696,
+      "learning_rate": 0.0001,
+      "loss": 1.6159,
+      "step": 8015
+    },
+    {
+      "epoch": 0.9207971971742002,
+      "grad_norm": 0.4539242088794708,
+      "learning_rate": 0.0001,
+      "loss": 1.598,
+      "step": 8016
+    },
+    {
+      "epoch": 0.9209120670840273,
+      "grad_norm": 0.4828774333000183,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 8017
+    },
+    {
+      "epoch": 0.9210269369938544,
+      "grad_norm": 0.46138039231300354,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 8018
+    },
+    {
+      "epoch": 0.9211418069036815,
+      "grad_norm": 0.5137259364128113,
+      "learning_rate": 0.0001,
+      "loss": 1.7913,
+      "step": 8019
+    },
+    {
+      "epoch": 0.9212566768135086,
+      "grad_norm": 0.4621741771697998,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 8020
+    },
+    {
+      "epoch": 0.9213715467233359,
+      "grad_norm": 0.4695769250392914,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 8021
+    },
+    {
+      "epoch": 0.921486416633163,
+      "grad_norm": 0.4522899389266968,
+      "learning_rate": 0.0001,
+      "loss": 1.6889,
+      "step": 8022
+    },
+    {
+      "epoch": 0.9216012865429901,
+      "grad_norm": 0.4740762710571289,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 8023
+    },
+    {
+      "epoch": 0.9217161564528172,
+      "grad_norm": 0.4992024302482605,
+      "learning_rate": 0.0001,
+      "loss": 1.5142,
+      "step": 8024
+    },
+    {
+      "epoch": 0.9218310263626444,
+      "grad_norm": 0.4674678444862366,
+      "learning_rate": 0.0001,
+      "loss": 1.6359,
+      "step": 8025
+    },
+    {
+      "epoch": 0.9219458962724715,
+      "grad_norm": 0.5612513422966003,
+      "learning_rate": 0.0001,
+      "loss": 1.715,
+      "step": 8026
+    },
+    {
+      "epoch": 0.9220607661822986,
+      "grad_norm": 0.4955098330974579,
+      "learning_rate": 0.0001,
+      "loss": 1.6685,
+      "step": 8027
+    },
+    {
+      "epoch": 0.9221756360921257,
+      "grad_norm": 0.47574183344841003,
+      "learning_rate": 0.0001,
+      "loss": 1.5452,
+      "step": 8028
+    },
+    {
+      "epoch": 0.9222905060019528,
+      "grad_norm": 0.4642239212989807,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 8029
+    },
+    {
+      "epoch": 0.92240537591178,
+      "grad_norm": 0.473645955324173,
+      "learning_rate": 0.0001,
+      "loss": 1.6189,
+      "step": 8030
+    },
+    {
+      "epoch": 0.9225202458216071,
+      "grad_norm": 0.4713420569896698,
+      "learning_rate": 0.0001,
+      "loss": 1.6377,
+      "step": 8031
+    },
+    {
+      "epoch": 0.9226351157314342,
+      "grad_norm": 0.44111305475234985,
+      "learning_rate": 0.0001,
+      "loss": 1.4193,
+      "step": 8032
+    },
+    {
+      "epoch": 0.9227499856412613,
+      "grad_norm": 0.4585154354572296,
+      "learning_rate": 0.0001,
+      "loss": 1.5419,
+      "step": 8033
+    },
+    {
+      "epoch": 0.9228648555510884,
+      "grad_norm": 0.5225629210472107,
+      "learning_rate": 0.0001,
+      "loss": 1.8615,
+      "step": 8034
+    },
+    {
+      "epoch": 0.9229797254609156,
+      "grad_norm": 0.43736404180526733,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 8035
+    },
+    {
+      "epoch": 0.9230945953707427,
+      "grad_norm": 0.4762333333492279,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 8036
+    },
+    {
+      "epoch": 0.9232094652805698,
+      "grad_norm": 0.47877663373947144,
+      "learning_rate": 0.0001,
+      "loss": 1.7119,
+      "step": 8037
+    },
+    {
+      "epoch": 0.9233243351903969,
+      "grad_norm": 0.47204989194869995,
+      "learning_rate": 0.0001,
+      "loss": 1.5251,
+      "step": 8038
+    },
+    {
+      "epoch": 0.923439205100224,
+      "grad_norm": 0.4415188133716583,
+      "learning_rate": 0.0001,
+      "loss": 1.6022,
+      "step": 8039
+    },
+    {
+      "epoch": 0.9235540750100512,
+      "grad_norm": 0.4474656879901886,
+      "learning_rate": 0.0001,
+      "loss": 1.4882,
+      "step": 8040
+    },
+    {
+      "epoch": 0.9236689449198783,
+      "grad_norm": 0.4732295274734497,
+      "learning_rate": 0.0001,
+      "loss": 1.3283,
+      "step": 8041
+    },
+    {
+      "epoch": 0.9237838148297054,
+      "grad_norm": 0.5105409026145935,
+      "learning_rate": 0.0001,
+      "loss": 1.7868,
+      "step": 8042
+    },
+    {
+      "epoch": 0.9238986847395325,
+      "grad_norm": 0.48805803060531616,
+      "learning_rate": 0.0001,
+      "loss": 1.3698,
+      "step": 8043
+    },
+    {
+      "epoch": 0.9240135546493596,
+      "grad_norm": 0.5286892652511597,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 8044
+    },
+    {
+      "epoch": 0.9241284245591868,
+      "grad_norm": 0.5003154873847961,
+      "learning_rate": 0.0001,
+      "loss": 1.5898,
+      "step": 8045
+    },
+    {
+      "epoch": 0.9242432944690139,
+      "grad_norm": 0.503452718257904,
+      "learning_rate": 0.0001,
+      "loss": 1.2122,
+      "step": 8046
+    },
+    {
+      "epoch": 0.924358164378841,
+      "grad_norm": 0.46282196044921875,
+      "learning_rate": 0.0001,
+      "loss": 1.4744,
+      "step": 8047
+    },
+    {
+      "epoch": 0.9244730342886681,
+      "grad_norm": 0.48527947068214417,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 8048
+    },
+    {
+      "epoch": 0.9245879041984952,
+      "grad_norm": 0.5114087462425232,
+      "learning_rate": 0.0001,
+      "loss": 1.8104,
+      "step": 8049
+    },
+    {
+      "epoch": 0.9247027741083224,
+      "grad_norm": 0.5129145979881287,
+      "learning_rate": 0.0001,
+      "loss": 1.7043,
+      "step": 8050
+    },
+    {
+      "epoch": 0.9248176440181495,
+      "grad_norm": 0.4763910174369812,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 8051
+    },
+    {
+      "epoch": 0.9249325139279766,
+      "grad_norm": 0.47218838334083557,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 8052
+    },
+    {
+      "epoch": 0.9250473838378037,
+      "grad_norm": 0.48110297322273254,
+      "learning_rate": 0.0001,
+      "loss": 1.4001,
+      "step": 8053
+    },
+    {
+      "epoch": 0.9251622537476308,
+      "grad_norm": 0.4787732660770416,
+      "learning_rate": 0.0001,
+      "loss": 1.6857,
+      "step": 8054
+    },
+    {
+      "epoch": 0.925277123657458,
+      "grad_norm": 0.48069849610328674,
+      "learning_rate": 0.0001,
+      "loss": 1.7946,
+      "step": 8055
+    },
+    {
+      "epoch": 0.9253919935672851,
+      "grad_norm": 0.47537073493003845,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 8056
+    },
+    {
+      "epoch": 0.9255068634771122,
+      "grad_norm": 0.5298829078674316,
+      "learning_rate": 0.0001,
+      "loss": 1.76,
+      "step": 8057
+    },
+    {
+      "epoch": 0.9256217333869393,
+      "grad_norm": 0.5032019019126892,
+      "learning_rate": 0.0001,
+      "loss": 1.8279,
+      "step": 8058
+    },
+    {
+      "epoch": 0.9257366032967664,
+      "grad_norm": 0.45885175466537476,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 8059
+    },
+    {
+      "epoch": 0.9258514732065936,
+      "grad_norm": 0.5285278558731079,
+      "learning_rate": 0.0001,
+      "loss": 1.8198,
+      "step": 8060
+    },
+    {
+      "epoch": 0.9259663431164207,
+      "grad_norm": 0.4382495880126953,
+      "learning_rate": 0.0001,
+      "loss": 1.3852,
+      "step": 8061
+    },
+    {
+      "epoch": 0.9260812130262478,
+      "grad_norm": 0.47816163301467896,
+      "learning_rate": 0.0001,
+      "loss": 1.7725,
+      "step": 8062
+    },
+    {
+      "epoch": 0.9261960829360749,
+      "grad_norm": 0.4544413685798645,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 8063
+    },
+    {
+      "epoch": 0.926310952845902,
+      "grad_norm": 0.46362125873565674,
+      "learning_rate": 0.0001,
+      "loss": 1.583,
+      "step": 8064
+    },
+    {
+      "epoch": 0.9264258227557292,
+      "grad_norm": 0.4957447052001953,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 8065
+    },
+    {
+      "epoch": 0.9265406926655563,
+      "grad_norm": 0.4854178726673126,
+      "learning_rate": 0.0001,
+      "loss": 1.7782,
+      "step": 8066
+    },
+    {
+      "epoch": 0.9266555625753834,
+      "grad_norm": 0.4702478051185608,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 8067
+    },
+    {
+      "epoch": 0.9267704324852105,
+      "grad_norm": 0.5195414423942566,
+      "learning_rate": 0.0001,
+      "loss": 1.8032,
+      "step": 8068
+    },
+    {
+      "epoch": 0.9268853023950376,
+      "grad_norm": 0.45996007323265076,
+      "learning_rate": 0.0001,
+      "loss": 1.3316,
+      "step": 8069
+    },
+    {
+      "epoch": 0.9270001723048648,
+      "grad_norm": 0.482707679271698,
+      "learning_rate": 0.0001,
+      "loss": 1.7089,
+      "step": 8070
+    },
+    {
+      "epoch": 0.9271150422146919,
+      "grad_norm": 0.4624692499637604,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 8071
+    },
+    {
+      "epoch": 0.927229912124519,
+      "grad_norm": 0.46416664123535156,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 8072
+    },
+    {
+      "epoch": 0.9273447820343461,
+      "grad_norm": 0.4962540566921234,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 8073
+    },
+    {
+      "epoch": 0.9274596519441732,
+      "grad_norm": 0.4407590627670288,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 8074
+    },
+    {
+      "epoch": 0.9275745218540004,
+      "grad_norm": 0.4883404076099396,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 8075
+    },
+    {
+      "epoch": 0.9276893917638275,
+      "grad_norm": 0.46453699469566345,
+      "learning_rate": 0.0001,
+      "loss": 1.4894,
+      "step": 8076
+    },
+    {
+      "epoch": 0.9278042616736546,
+      "grad_norm": 0.45241397619247437,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 8077
+    },
+    {
+      "epoch": 0.9279191315834817,
+      "grad_norm": 0.48242220282554626,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 8078
+    },
+    {
+      "epoch": 0.9280340014933088,
+      "grad_norm": 0.546755313873291,
+      "learning_rate": 0.0001,
+      "loss": 1.9116,
+      "step": 8079
+    },
+    {
+      "epoch": 0.928148871403136,
+      "grad_norm": 0.4613960087299347,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 8080
+    },
+    {
+      "epoch": 0.9282637413129631,
+      "grad_norm": 0.47544580698013306,
+      "learning_rate": 0.0001,
+      "loss": 1.5371,
+      "step": 8081
+    },
+    {
+      "epoch": 0.9283786112227902,
+      "grad_norm": 0.5122746825218201,
+      "learning_rate": 0.0001,
+      "loss": 1.686,
+      "step": 8082
+    },
+    {
+      "epoch": 0.9284934811326173,
+      "grad_norm": 0.47522497177124023,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 8083
+    },
+    {
+      "epoch": 0.9286083510424444,
+      "grad_norm": 0.46367111802101135,
+      "learning_rate": 0.0001,
+      "loss": 1.4283,
+      "step": 8084
+    },
+    {
+      "epoch": 0.9287232209522716,
+      "grad_norm": 0.5942444205284119,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 8085
+    },
+    {
+      "epoch": 0.9288380908620987,
+      "grad_norm": 0.46138957142829895,
+      "learning_rate": 0.0001,
+      "loss": 1.7417,
+      "step": 8086
+    },
+    {
+      "epoch": 0.9289529607719258,
+      "grad_norm": 0.5159793496131897,
+      "learning_rate": 0.0001,
+      "loss": 1.757,
+      "step": 8087
+    },
+    {
+      "epoch": 0.9290678306817529,
+      "grad_norm": 0.48674991726875305,
+      "learning_rate": 0.0001,
+      "loss": 1.617,
+      "step": 8088
+    },
+    {
+      "epoch": 0.92918270059158,
+      "grad_norm": 0.5087023973464966,
+      "learning_rate": 0.0001,
+      "loss": 1.8192,
+      "step": 8089
+    },
+    {
+      "epoch": 0.9292975705014072,
+      "grad_norm": 0.5132076740264893,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 8090
+    },
+    {
+      "epoch": 0.9294124404112343,
+      "grad_norm": 0.4683862030506134,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 8091
+    },
+    {
+      "epoch": 0.9295273103210614,
+      "grad_norm": 0.5067598819732666,
+      "learning_rate": 0.0001,
+      "loss": 1.6037,
+      "step": 8092
+    },
+    {
+      "epoch": 0.9296421802308885,
+      "grad_norm": 0.4644726514816284,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 8093
+    },
+    {
+      "epoch": 0.9297570501407156,
+      "grad_norm": 0.44829845428466797,
+      "learning_rate": 0.0001,
+      "loss": 1.43,
+      "step": 8094
+    },
+    {
+      "epoch": 0.9298719200505428,
+      "grad_norm": 0.5093761682510376,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 8095
+    },
+    {
+      "epoch": 0.9299867899603699,
+      "grad_norm": 0.4932906925678253,
+      "learning_rate": 0.0001,
+      "loss": 1.7515,
+      "step": 8096
+    },
+    {
+      "epoch": 0.930101659870197,
+      "grad_norm": 0.4766685366630554,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 8097
+    },
+    {
+      "epoch": 0.9302165297800241,
+      "grad_norm": 0.46401628851890564,
+      "learning_rate": 0.0001,
+      "loss": 1.5306,
+      "step": 8098
+    },
+    {
+      "epoch": 0.9303313996898512,
+      "grad_norm": 0.45654991269111633,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 8099
+    },
+    {
+      "epoch": 0.9304462695996784,
+      "grad_norm": 0.46206626296043396,
+      "learning_rate": 0.0001,
+      "loss": 1.3484,
+      "step": 8100
+    },
+    {
+      "epoch": 0.9305611395095055,
+      "grad_norm": 0.4578942060470581,
+      "learning_rate": 0.0001,
+      "loss": 1.3981,
+      "step": 8101
+    },
+    {
+      "epoch": 0.9306760094193326,
+      "grad_norm": 0.4928850531578064,
+      "learning_rate": 0.0001,
+      "loss": 1.5062,
+      "step": 8102
+    },
+    {
+      "epoch": 0.9307908793291597,
+      "grad_norm": 0.46304240822792053,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 8103
+    },
+    {
+      "epoch": 0.9309057492389868,
+      "grad_norm": 0.43924397230148315,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 8104
+    },
+    {
+      "epoch": 0.931020619148814,
+      "grad_norm": 0.5079434514045715,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 8105
+    },
+    {
+      "epoch": 0.9311354890586411,
+      "grad_norm": 0.4720153510570526,
+      "learning_rate": 0.0001,
+      "loss": 1.7547,
+      "step": 8106
+    },
+    {
+      "epoch": 0.9312503589684682,
+      "grad_norm": 0.4631122648715973,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 8107
+    },
+    {
+      "epoch": 0.9313652288782953,
+      "grad_norm": 0.4567249119281769,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 8108
+    },
+    {
+      "epoch": 0.9314800987881224,
+      "grad_norm": 0.49746426939964294,
+      "learning_rate": 0.0001,
+      "loss": 1.6227,
+      "step": 8109
+    },
+    {
+      "epoch": 0.9315949686979496,
+      "grad_norm": 0.47081831097602844,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 8110
+    },
+    {
+      "epoch": 0.9317098386077767,
+      "grad_norm": 0.5293380618095398,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 8111
+    },
+    {
+      "epoch": 0.9318247085176038,
+      "grad_norm": 0.47562193870544434,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 8112
+    },
+    {
+      "epoch": 0.9319395784274309,
+      "grad_norm": 0.5127376317977905,
+      "learning_rate": 0.0001,
+      "loss": 1.6752,
+      "step": 8113
+    },
+    {
+      "epoch": 0.932054448337258,
+      "grad_norm": 0.47655048966407776,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 8114
+    },
+    {
+      "epoch": 0.9321693182470852,
+      "grad_norm": 0.4715765416622162,
+      "learning_rate": 0.0001,
+      "loss": 1.6537,
+      "step": 8115
+    },
+    {
+      "epoch": 0.9322841881569123,
+      "grad_norm": 0.4677276313304901,
+      "learning_rate": 0.0001,
+      "loss": 1.5214,
+      "step": 8116
+    },
+    {
+      "epoch": 0.9323990580667394,
+      "grad_norm": 0.4286912977695465,
+      "learning_rate": 0.0001,
+      "loss": 1.4507,
+      "step": 8117
+    },
+    {
+      "epoch": 0.9325139279765665,
+      "grad_norm": 0.4614866375923157,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 8118
+    },
+    {
+      "epoch": 0.9326287978863936,
+      "grad_norm": 0.44791609048843384,
+      "learning_rate": 0.0001,
+      "loss": 1.5154,
+      "step": 8119
+    },
+    {
+      "epoch": 0.9327436677962208,
+      "grad_norm": 0.49002981185913086,
+      "learning_rate": 0.0001,
+      "loss": 1.7432,
+      "step": 8120
+    },
+    {
+      "epoch": 0.9328585377060479,
+      "grad_norm": 0.4595596194267273,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 8121
+    },
+    {
+      "epoch": 0.932973407615875,
+      "grad_norm": 0.47388172149658203,
+      "learning_rate": 0.0001,
+      "loss": 1.4072,
+      "step": 8122
+    },
+    {
+      "epoch": 0.9330882775257021,
+      "grad_norm": 0.5274661183357239,
+      "learning_rate": 0.0001,
+      "loss": 1.7443,
+      "step": 8123
+    },
+    {
+      "epoch": 0.9332031474355292,
+      "grad_norm": 0.5174868702888489,
+      "learning_rate": 0.0001,
+      "loss": 1.7705,
+      "step": 8124
+    },
+    {
+      "epoch": 0.9333180173453564,
+      "grad_norm": 0.5129168629646301,
+      "learning_rate": 0.0001,
+      "loss": 1.6475,
+      "step": 8125
+    },
+    {
+      "epoch": 0.9334328872551835,
+      "grad_norm": 0.4941151440143585,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 8126
+    },
+    {
+      "epoch": 0.9335477571650106,
+      "grad_norm": 0.4885686933994293,
+      "learning_rate": 0.0001,
+      "loss": 1.7374,
+      "step": 8127
+    },
+    {
+      "epoch": 0.9336626270748377,
+      "grad_norm": 0.47207513451576233,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 8128
+    },
+    {
+      "epoch": 0.9337774969846648,
+      "grad_norm": 0.5630714893341064,
+      "learning_rate": 0.0001,
+      "loss": 2.0647,
+      "step": 8129
+    },
+    {
+      "epoch": 0.933892366894492,
+      "grad_norm": 0.47757160663604736,
+      "learning_rate": 0.0001,
+      "loss": 1.6114,
+      "step": 8130
+    },
+    {
+      "epoch": 0.9340072368043191,
+      "grad_norm": 0.4331253170967102,
+      "learning_rate": 0.0001,
+      "loss": 1.4655,
+      "step": 8131
+    },
+    {
+      "epoch": 0.9341221067141462,
+      "grad_norm": 0.48484155535697937,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 8132
+    },
+    {
+      "epoch": 0.9342369766239733,
+      "grad_norm": 0.47229698300361633,
+      "learning_rate": 0.0001,
+      "loss": 1.7429,
+      "step": 8133
+    },
+    {
+      "epoch": 0.9343518465338004,
+      "grad_norm": 0.45240116119384766,
+      "learning_rate": 0.0001,
+      "loss": 1.3286,
+      "step": 8134
+    },
+    {
+      "epoch": 0.9344667164436276,
+      "grad_norm": 0.4636879563331604,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 8135
+    },
+    {
+      "epoch": 0.9345815863534547,
+      "grad_norm": 0.4439462721347809,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 8136
+    },
+    {
+      "epoch": 0.9346964562632818,
+      "grad_norm": 0.46923813223838806,
+      "learning_rate": 0.0001,
+      "loss": 1.3393,
+      "step": 8137
+    },
+    {
+      "epoch": 0.9348113261731089,
+      "grad_norm": 0.47300711274147034,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 8138
+    },
+    {
+      "epoch": 0.934926196082936,
+      "grad_norm": 0.5104739665985107,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 8139
+    },
+    {
+      "epoch": 0.9350410659927632,
+      "grad_norm": 0.4453759491443634,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 8140
+    },
+    {
+      "epoch": 0.9351559359025903,
+      "grad_norm": 0.46000081300735474,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 8141
+    },
+    {
+      "epoch": 0.9352708058124174,
+      "grad_norm": 0.4699609577655792,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 8142
+    },
+    {
+      "epoch": 0.9353856757222445,
+      "grad_norm": 0.4497356414794922,
+      "learning_rate": 0.0001,
+      "loss": 1.6447,
+      "step": 8143
+    },
+    {
+      "epoch": 0.9355005456320716,
+      "grad_norm": 0.4631606638431549,
+      "learning_rate": 0.0001,
+      "loss": 1.7243,
+      "step": 8144
+    },
+    {
+      "epoch": 0.9356154155418988,
+      "grad_norm": 0.4655974805355072,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 8145
+    },
+    {
+      "epoch": 0.9357302854517259,
+      "grad_norm": 0.464070588350296,
+      "learning_rate": 0.0001,
+      "loss": 1.5003,
+      "step": 8146
+    },
+    {
+      "epoch": 0.935845155361553,
+      "grad_norm": 0.5652233362197876,
+      "learning_rate": 0.0001,
+      "loss": 1.5475,
+      "step": 8147
+    },
+    {
+      "epoch": 0.9359600252713801,
+      "grad_norm": 0.47151249647140503,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 8148
+    },
+    {
+      "epoch": 0.9360748951812072,
+      "grad_norm": 0.4593266248703003,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 8149
+    },
+    {
+      "epoch": 0.9361897650910344,
+      "grad_norm": 0.491393119096756,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 8150
+    },
+    {
+      "epoch": 0.9363046350008615,
+      "grad_norm": 0.5595842003822327,
+      "learning_rate": 0.0001,
+      "loss": 1.4467,
+      "step": 8151
+    },
+    {
+      "epoch": 0.9364195049106886,
+      "grad_norm": 0.46376144886016846,
+      "learning_rate": 0.0001,
+      "loss": 1.5592,
+      "step": 8152
+    },
+    {
+      "epoch": 0.9365343748205157,
+      "grad_norm": 0.4688231348991394,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 8153
+    },
+    {
+      "epoch": 0.9366492447303428,
+      "grad_norm": 0.45528239011764526,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 8154
+    },
+    {
+      "epoch": 0.93676411464017,
+      "grad_norm": 0.48385414481163025,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 8155
+    },
+    {
+      "epoch": 0.9368789845499971,
+      "grad_norm": 0.4678395688533783,
+      "learning_rate": 0.0001,
+      "loss": 1.5949,
+      "step": 8156
+    },
+    {
+      "epoch": 0.9369938544598242,
+      "grad_norm": 0.5076170563697815,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 8157
+    },
+    {
+      "epoch": 0.9371087243696514,
+      "grad_norm": 0.4750152826309204,
+      "learning_rate": 0.0001,
+      "loss": 1.4804,
+      "step": 8158
+    },
+    {
+      "epoch": 0.9372235942794785,
+      "grad_norm": 0.4693584740161896,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 8159
+    },
+    {
+      "epoch": 0.9373384641893057,
+      "grad_norm": 0.45886656641960144,
+      "learning_rate": 0.0001,
+      "loss": 1.7251,
+      "step": 8160
+    },
+    {
+      "epoch": 0.9374533340991328,
+      "grad_norm": 0.554453432559967,
+      "learning_rate": 0.0001,
+      "loss": 1.9527,
+      "step": 8161
+    },
+    {
+      "epoch": 0.9375682040089599,
+      "grad_norm": 0.46930959820747375,
+      "learning_rate": 0.0001,
+      "loss": 1.4046,
+      "step": 8162
+    },
+    {
+      "epoch": 0.937683073918787,
+      "grad_norm": 0.44557827711105347,
+      "learning_rate": 0.0001,
+      "loss": 1.4906,
+      "step": 8163
+    },
+    {
+      "epoch": 0.9377979438286141,
+      "grad_norm": 0.4786651134490967,
+      "learning_rate": 0.0001,
+      "loss": 1.7044,
+      "step": 8164
+    },
+    {
+      "epoch": 0.9379128137384413,
+      "grad_norm": 0.47346630692481995,
+      "learning_rate": 0.0001,
+      "loss": 1.384,
+      "step": 8165
+    },
+    {
+      "epoch": 0.9380276836482684,
+      "grad_norm": 0.4756239056587219,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 8166
+    },
+    {
+      "epoch": 0.9381425535580955,
+      "grad_norm": 0.4796939194202423,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 8167
+    },
+    {
+      "epoch": 0.9382574234679226,
+      "grad_norm": 0.4924411475658417,
+      "learning_rate": 0.0001,
+      "loss": 1.5232,
+      "step": 8168
+    },
+    {
+      "epoch": 0.9383722933777497,
+      "grad_norm": 0.4914194941520691,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 8169
+    },
+    {
+      "epoch": 0.9384871632875769,
+      "grad_norm": 0.46854540705680847,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 8170
+    },
+    {
+      "epoch": 0.938602033197404,
+      "grad_norm": 0.46316492557525635,
+      "learning_rate": 0.0001,
+      "loss": 1.523,
+      "step": 8171
+    },
+    {
+      "epoch": 0.9387169031072311,
+      "grad_norm": 0.4873157739639282,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 8172
+    },
+    {
+      "epoch": 0.9388317730170582,
+      "grad_norm": 0.4700479805469513,
+      "learning_rate": 0.0001,
+      "loss": 1.4223,
+      "step": 8173
+    },
+    {
+      "epoch": 0.9389466429268853,
+      "grad_norm": 0.5385305285453796,
+      "learning_rate": 0.0001,
+      "loss": 1.821,
+      "step": 8174
+    },
+    {
+      "epoch": 0.9390615128367125,
+      "grad_norm": 0.48141562938690186,
+      "learning_rate": 0.0001,
+      "loss": 1.4552,
+      "step": 8175
+    },
+    {
+      "epoch": 0.9391763827465396,
+      "grad_norm": 0.4608069360256195,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 8176
+    },
+    {
+      "epoch": 0.9392912526563667,
+      "grad_norm": 0.4807290732860565,
+      "learning_rate": 0.0001,
+      "loss": 1.4567,
+      "step": 8177
+    },
+    {
+      "epoch": 0.9394061225661938,
+      "grad_norm": 0.5841659307479858,
+      "learning_rate": 0.0001,
+      "loss": 1.8741,
+      "step": 8178
+    },
+    {
+      "epoch": 0.9395209924760209,
+      "grad_norm": 0.4911169111728668,
+      "learning_rate": 0.0001,
+      "loss": 1.6685,
+      "step": 8179
+    },
+    {
+      "epoch": 0.9396358623858481,
+      "grad_norm": 0.4507903754711151,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 8180
+    },
+    {
+      "epoch": 0.9397507322956752,
+      "grad_norm": 0.4615688621997833,
+      "learning_rate": 0.0001,
+      "loss": 1.6094,
+      "step": 8181
+    },
+    {
+      "epoch": 0.9398656022055023,
+      "grad_norm": 0.5235313773155212,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 8182
+    },
+    {
+      "epoch": 0.9399804721153294,
+      "grad_norm": 0.4749388098716736,
+      "learning_rate": 0.0001,
+      "loss": 1.4103,
+      "step": 8183
+    },
+    {
+      "epoch": 0.9400953420251565,
+      "grad_norm": 0.49427589774131775,
+      "learning_rate": 0.0001,
+      "loss": 1.7691,
+      "step": 8184
+    },
+    {
+      "epoch": 0.9402102119349837,
+      "grad_norm": 0.4776000678539276,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 8185
+    },
+    {
+      "epoch": 0.9403250818448108,
+      "grad_norm": 0.47408178448677063,
+      "learning_rate": 0.0001,
+      "loss": 1.4462,
+      "step": 8186
+    },
+    {
+      "epoch": 0.9404399517546379,
+      "grad_norm": 0.5133025646209717,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 8187
+    },
+    {
+      "epoch": 0.940554821664465,
+      "grad_norm": 0.45842912793159485,
+      "learning_rate": 0.0001,
+      "loss": 1.6081,
+      "step": 8188
+    },
+    {
+      "epoch": 0.9406696915742921,
+      "grad_norm": 0.4748247563838959,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 8189
+    },
+    {
+      "epoch": 0.9407845614841193,
+      "grad_norm": 0.46958670020103455,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 8190
+    },
+    {
+      "epoch": 0.9408994313939464,
+      "grad_norm": 0.48225289583206177,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 8191
+    },
+    {
+      "epoch": 0.9410143013037735,
+      "grad_norm": 0.5528295636177063,
+      "learning_rate": 0.0001,
+      "loss": 1.6373,
+      "step": 8192
+    },
+    {
+      "epoch": 0.9411291712136006,
+      "grad_norm": 0.5050978064537048,
+      "learning_rate": 0.0001,
+      "loss": 1.702,
+      "step": 8193
+    },
+    {
+      "epoch": 0.9412440411234277,
+      "grad_norm": 0.4624803066253662,
+      "learning_rate": 0.0001,
+      "loss": 1.6395,
+      "step": 8194
+    },
+    {
+      "epoch": 0.9413589110332549,
+      "grad_norm": 0.5099507570266724,
+      "learning_rate": 0.0001,
+      "loss": 1.6849,
+      "step": 8195
+    },
+    {
+      "epoch": 0.941473780943082,
+      "grad_norm": 0.4514469504356384,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 8196
+    },
+    {
+      "epoch": 0.9415886508529091,
+      "grad_norm": 0.5006936192512512,
+      "learning_rate": 0.0001,
+      "loss": 1.5678,
+      "step": 8197
+    },
+    {
+      "epoch": 0.9417035207627362,
+      "grad_norm": 0.5389548540115356,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 8198
+    },
+    {
+      "epoch": 0.9418183906725633,
+      "grad_norm": 0.46252140402793884,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 8199
+    },
+    {
+      "epoch": 0.9419332605823905,
+      "grad_norm": 0.46771207451820374,
+      "learning_rate": 0.0001,
+      "loss": 1.3591,
+      "step": 8200
+    },
+    {
+      "epoch": 0.9420481304922176,
+      "grad_norm": 0.492191880941391,
+      "learning_rate": 0.0001,
+      "loss": 1.4178,
+      "step": 8201
+    },
+    {
+      "epoch": 0.9421630004020447,
+      "grad_norm": 0.4770967960357666,
+      "learning_rate": 0.0001,
+      "loss": 1.7547,
+      "step": 8202
+    },
+    {
+      "epoch": 0.9422778703118718,
+      "grad_norm": 0.5047717094421387,
+      "learning_rate": 0.0001,
+      "loss": 1.6103,
+      "step": 8203
+    },
+    {
+      "epoch": 0.9423927402216989,
+      "grad_norm": 0.47742435336112976,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 8204
+    },
+    {
+      "epoch": 0.9425076101315261,
+      "grad_norm": 0.5914652943611145,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 8205
+    },
+    {
+      "epoch": 0.9426224800413532,
+      "grad_norm": 0.4261223077774048,
+      "learning_rate": 0.0001,
+      "loss": 1.3505,
+      "step": 8206
+    },
+    {
+      "epoch": 0.9427373499511803,
+      "grad_norm": 0.47556623816490173,
+      "learning_rate": 0.0001,
+      "loss": 1.4745,
+      "step": 8207
+    },
+    {
+      "epoch": 0.9428522198610074,
+      "grad_norm": 0.4878052771091461,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 8208
+    },
+    {
+      "epoch": 0.9429670897708345,
+      "grad_norm": 0.44221365451812744,
+      "learning_rate": 0.0001,
+      "loss": 1.3707,
+      "step": 8209
+    },
+    {
+      "epoch": 0.9430819596806617,
+      "grad_norm": 0.47154805064201355,
+      "learning_rate": 0.0001,
+      "loss": 1.5895,
+      "step": 8210
+    },
+    {
+      "epoch": 0.9431968295904888,
+      "grad_norm": 0.5039512515068054,
+      "learning_rate": 0.0001,
+      "loss": 1.7088,
+      "step": 8211
+    },
+    {
+      "epoch": 0.9433116995003159,
+      "grad_norm": 0.45005252957344055,
+      "learning_rate": 0.0001,
+      "loss": 1.4901,
+      "step": 8212
+    },
+    {
+      "epoch": 0.943426569410143,
+      "grad_norm": 0.47104600071907043,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 8213
+    },
+    {
+      "epoch": 0.9435414393199701,
+      "grad_norm": 0.4666808843612671,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 8214
+    },
+    {
+      "epoch": 0.9436563092297973,
+      "grad_norm": 0.510503351688385,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 8215
+    },
+    {
+      "epoch": 0.9437711791396244,
+      "grad_norm": 0.4818345606327057,
+      "learning_rate": 0.0001,
+      "loss": 1.4745,
+      "step": 8216
+    },
+    {
+      "epoch": 0.9438860490494515,
+      "grad_norm": 0.4899197518825531,
+      "learning_rate": 0.0001,
+      "loss": 1.5966,
+      "step": 8217
+    },
+    {
+      "epoch": 0.9440009189592786,
+      "grad_norm": 0.5139290690422058,
+      "learning_rate": 0.0001,
+      "loss": 1.406,
+      "step": 8218
+    },
+    {
+      "epoch": 0.9441157888691057,
+      "grad_norm": 0.4770418107509613,
+      "learning_rate": 0.0001,
+      "loss": 1.4606,
+      "step": 8219
+    },
+    {
+      "epoch": 0.9442306587789329,
+      "grad_norm": 0.617544949054718,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 8220
+    },
+    {
+      "epoch": 0.94434552868876,
+      "grad_norm": 0.49356672167778015,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 8221
+    },
+    {
+      "epoch": 0.9444603985985871,
+      "grad_norm": 0.4679422974586487,
+      "learning_rate": 0.0001,
+      "loss": 1.5843,
+      "step": 8222
+    },
+    {
+      "epoch": 0.9445752685084142,
+      "grad_norm": 0.5478677749633789,
+      "learning_rate": 0.0001,
+      "loss": 1.8256,
+      "step": 8223
+    },
+    {
+      "epoch": 0.9446901384182413,
+      "grad_norm": 0.4296529293060303,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 8224
+    },
+    {
+      "epoch": 0.9448050083280685,
+      "grad_norm": 0.4777142405509949,
+      "learning_rate": 0.0001,
+      "loss": 1.4079,
+      "step": 8225
+    },
+    {
+      "epoch": 0.9449198782378956,
+      "grad_norm": 0.5253008604049683,
+      "learning_rate": 0.0001,
+      "loss": 1.6611,
+      "step": 8226
+    },
+    {
+      "epoch": 0.9450347481477227,
+      "grad_norm": 0.4972079396247864,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 8227
+    },
+    {
+      "epoch": 0.9451496180575498,
+      "grad_norm": 0.4560241997241974,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 8228
+    },
+    {
+      "epoch": 0.9452644879673769,
+      "grad_norm": 0.4682483673095703,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 8229
+    },
+    {
+      "epoch": 0.9453793578772041,
+      "grad_norm": 0.4736382067203522,
+      "learning_rate": 0.0001,
+      "loss": 1.3865,
+      "step": 8230
+    },
+    {
+      "epoch": 0.9454942277870312,
+      "grad_norm": 0.4748673737049103,
+      "learning_rate": 0.0001,
+      "loss": 1.427,
+      "step": 8231
+    },
+    {
+      "epoch": 0.9456090976968583,
+      "grad_norm": 0.45922762155532837,
+      "learning_rate": 0.0001,
+      "loss": 1.3151,
+      "step": 8232
+    },
+    {
+      "epoch": 0.9457239676066854,
+      "grad_norm": 0.47256186604499817,
+      "learning_rate": 0.0001,
+      "loss": 1.4456,
+      "step": 8233
+    },
+    {
+      "epoch": 0.9458388375165125,
+      "grad_norm": 0.46213608980178833,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 8234
+    },
+    {
+      "epoch": 0.9459537074263397,
+      "grad_norm": 0.4915864169597626,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 8235
+    },
+    {
+      "epoch": 0.9460685773361668,
+      "grad_norm": 0.4443720579147339,
+      "learning_rate": 0.0001,
+      "loss": 1.5858,
+      "step": 8236
+    },
+    {
+      "epoch": 0.9461834472459939,
+      "grad_norm": 0.4873688519001007,
+      "learning_rate": 0.0001,
+      "loss": 1.7889,
+      "step": 8237
+    },
+    {
+      "epoch": 0.946298317155821,
+      "grad_norm": 0.49472931027412415,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 8238
+    },
+    {
+      "epoch": 0.9464131870656481,
+      "grad_norm": 0.4404892921447754,
+      "learning_rate": 0.0001,
+      "loss": 1.3307,
+      "step": 8239
+    },
+    {
+      "epoch": 0.9465280569754753,
+      "grad_norm": 0.48074471950531006,
+      "learning_rate": 0.0001,
+      "loss": 1.6203,
+      "step": 8240
+    },
+    {
+      "epoch": 0.9466429268853024,
+      "grad_norm": 0.5083596110343933,
+      "learning_rate": 0.0001,
+      "loss": 1.8559,
+      "step": 8241
+    },
+    {
+      "epoch": 0.9467577967951295,
+      "grad_norm": 0.4485703110694885,
+      "learning_rate": 0.0001,
+      "loss": 1.5608,
+      "step": 8242
+    },
+    {
+      "epoch": 0.9468726667049566,
+      "grad_norm": 0.48620954155921936,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 8243
+    },
+    {
+      "epoch": 0.9469875366147837,
+      "grad_norm": 0.4664001166820526,
+      "learning_rate": 0.0001,
+      "loss": 1.5831,
+      "step": 8244
+    },
+    {
+      "epoch": 0.9471024065246109,
+      "grad_norm": 0.4595847427845001,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 8245
+    },
+    {
+      "epoch": 0.947217276434438,
+      "grad_norm": 0.5062698721885681,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 8246
+    },
+    {
+      "epoch": 0.9473321463442651,
+      "grad_norm": 0.44905591011047363,
+      "learning_rate": 0.0001,
+      "loss": 1.5338,
+      "step": 8247
+    },
+    {
+      "epoch": 0.9474470162540922,
+      "grad_norm": 0.47497665882110596,
+      "learning_rate": 0.0001,
+      "loss": 1.4712,
+      "step": 8248
+    },
+    {
+      "epoch": 0.9475618861639193,
+      "grad_norm": 0.48770657181739807,
+      "learning_rate": 0.0001,
+      "loss": 1.43,
+      "step": 8249
+    },
+    {
+      "epoch": 0.9476767560737465,
+      "grad_norm": 0.4931548833847046,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 8250
+    },
+    {
+      "epoch": 0.9477916259835736,
+      "grad_norm": 0.47381991147994995,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 8251
+    },
+    {
+      "epoch": 0.9479064958934007,
+      "grad_norm": 0.5158901810646057,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 8252
+    },
+    {
+      "epoch": 0.9480213658032278,
+      "grad_norm": 0.48369431495666504,
+      "learning_rate": 0.0001,
+      "loss": 1.6776,
+      "step": 8253
+    },
+    {
+      "epoch": 0.9481362357130549,
+      "grad_norm": 0.5041664838790894,
+      "learning_rate": 0.0001,
+      "loss": 1.6957,
+      "step": 8254
+    },
+    {
+      "epoch": 0.9482511056228821,
+      "grad_norm": 0.46258240938186646,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 8255
+    },
+    {
+      "epoch": 0.9483659755327092,
+      "grad_norm": 0.4723878800868988,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 8256
+    },
+    {
+      "epoch": 0.9484808454425363,
+      "grad_norm": 0.45941516757011414,
+      "learning_rate": 0.0001,
+      "loss": 1.6875,
+      "step": 8257
+    },
+    {
+      "epoch": 0.9485957153523634,
+      "grad_norm": 0.4574301540851593,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 8258
+    },
+    {
+      "epoch": 0.9487105852621905,
+      "grad_norm": 0.4727140963077545,
+      "learning_rate": 0.0001,
+      "loss": 1.6921,
+      "step": 8259
+    },
+    {
+      "epoch": 0.9488254551720177,
+      "grad_norm": 0.46204841136932373,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 8260
+    },
+    {
+      "epoch": 0.9489403250818448,
+      "grad_norm": 0.48473307490348816,
+      "learning_rate": 0.0001,
+      "loss": 1.3974,
+      "step": 8261
+    },
+    {
+      "epoch": 0.9490551949916719,
+      "grad_norm": 0.4569971561431885,
+      "learning_rate": 0.0001,
+      "loss": 1.7005,
+      "step": 8262
+    },
+    {
+      "epoch": 0.949170064901499,
+      "grad_norm": 0.4910833239555359,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 8263
+    },
+    {
+      "epoch": 0.9492849348113261,
+      "grad_norm": 0.469138503074646,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 8264
+    },
+    {
+      "epoch": 0.9493998047211533,
+      "grad_norm": 0.4757539927959442,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 8265
+    },
+    {
+      "epoch": 0.9495146746309804,
+      "grad_norm": 0.4612511992454529,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 8266
+    },
+    {
+      "epoch": 0.9496295445408075,
+      "grad_norm": 0.5166419744491577,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 8267
+    },
+    {
+      "epoch": 0.9497444144506346,
+      "grad_norm": 0.451444149017334,
+      "learning_rate": 0.0001,
+      "loss": 1.5371,
+      "step": 8268
+    },
+    {
+      "epoch": 0.9498592843604617,
+      "grad_norm": 0.4882412850856781,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 8269
+    },
+    {
+      "epoch": 0.9499741542702889,
+      "grad_norm": 0.5102916359901428,
+      "learning_rate": 0.0001,
+      "loss": 1.6798,
+      "step": 8270
+    },
+    {
+      "epoch": 0.950089024180116,
+      "grad_norm": 0.46115705370903015,
+      "learning_rate": 0.0001,
+      "loss": 1.2963,
+      "step": 8271
+    },
+    {
+      "epoch": 0.9502038940899431,
+      "grad_norm": 0.5228103399276733,
+      "learning_rate": 0.0001,
+      "loss": 1.4118,
+      "step": 8272
+    },
+    {
+      "epoch": 0.9503187639997702,
+      "grad_norm": 0.5342441201210022,
+      "learning_rate": 0.0001,
+      "loss": 1.7615,
+      "step": 8273
+    },
+    {
+      "epoch": 0.9504336339095973,
+      "grad_norm": 0.4900018274784088,
+      "learning_rate": 0.0001,
+      "loss": 1.3293,
+      "step": 8274
+    },
+    {
+      "epoch": 0.9505485038194245,
+      "grad_norm": 0.5222525000572205,
+      "learning_rate": 0.0001,
+      "loss": 1.6752,
+      "step": 8275
+    },
+    {
+      "epoch": 0.9506633737292516,
+      "grad_norm": 0.5265666842460632,
+      "learning_rate": 0.0001,
+      "loss": 1.7776,
+      "step": 8276
+    },
+    {
+      "epoch": 0.9507782436390787,
+      "grad_norm": 0.4796416759490967,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 8277
+    },
+    {
+      "epoch": 0.9508931135489058,
+      "grad_norm": 0.4717714488506317,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 8278
+    },
+    {
+      "epoch": 0.9510079834587329,
+      "grad_norm": 0.5052562952041626,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 8279
+    },
+    {
+      "epoch": 0.9511228533685601,
+      "grad_norm": 0.45067209005355835,
+      "learning_rate": 0.0001,
+      "loss": 1.5466,
+      "step": 8280
+    },
+    {
+      "epoch": 0.9512377232783872,
+      "grad_norm": 0.527546226978302,
+      "learning_rate": 0.0001,
+      "loss": 1.4335,
+      "step": 8281
+    },
+    {
+      "epoch": 0.9513525931882143,
+      "grad_norm": 0.46170440316200256,
+      "learning_rate": 0.0001,
+      "loss": 1.3969,
+      "step": 8282
+    },
+    {
+      "epoch": 0.9514674630980414,
+      "grad_norm": 0.5153254270553589,
+      "learning_rate": 0.0001,
+      "loss": 1.6541,
+      "step": 8283
+    },
+    {
+      "epoch": 0.9515823330078685,
+      "grad_norm": 0.48110654950141907,
+      "learning_rate": 0.0001,
+      "loss": 1.5888,
+      "step": 8284
+    },
+    {
+      "epoch": 0.9516972029176957,
+      "grad_norm": 0.5225505828857422,
+      "learning_rate": 0.0001,
+      "loss": 1.7728,
+      "step": 8285
+    },
+    {
+      "epoch": 0.9518120728275228,
+      "grad_norm": 0.4549621343612671,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 8286
+    },
+    {
+      "epoch": 0.9519269427373499,
+      "grad_norm": 0.5324844121932983,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 8287
+    },
+    {
+      "epoch": 0.952041812647177,
+      "grad_norm": 0.4963549077510834,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 8288
+    },
+    {
+      "epoch": 0.9521566825570041,
+      "grad_norm": 0.460531085729599,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 8289
+    },
+    {
+      "epoch": 0.9522715524668313,
+      "grad_norm": 0.4704626798629761,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 8290
+    },
+    {
+      "epoch": 0.9523864223766584,
+      "grad_norm": 0.4891350269317627,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 8291
+    },
+    {
+      "epoch": 0.9525012922864855,
+      "grad_norm": 0.4812767207622528,
+      "learning_rate": 0.0001,
+      "loss": 1.6853,
+      "step": 8292
+    },
+    {
+      "epoch": 0.9526161621963126,
+      "grad_norm": 0.47794899344444275,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 8293
+    },
+    {
+      "epoch": 0.9527310321061397,
+      "grad_norm": 0.49231424927711487,
+      "learning_rate": 0.0001,
+      "loss": 1.4125,
+      "step": 8294
+    },
+    {
+      "epoch": 0.952845902015967,
+      "grad_norm": 0.47585782408714294,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 8295
+    },
+    {
+      "epoch": 0.9529607719257941,
+      "grad_norm": 0.5035213828086853,
+      "learning_rate": 0.0001,
+      "loss": 1.6766,
+      "step": 8296
+    },
+    {
+      "epoch": 0.9530756418356212,
+      "grad_norm": 0.5115920305252075,
+      "learning_rate": 0.0001,
+      "loss": 1.4955,
+      "step": 8297
+    },
+    {
+      "epoch": 0.9531905117454483,
+      "grad_norm": 0.5224997401237488,
+      "learning_rate": 0.0001,
+      "loss": 1.6931,
+      "step": 8298
+    },
+    {
+      "epoch": 0.9533053816552755,
+      "grad_norm": 0.5272790789604187,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 8299
+    },
+    {
+      "epoch": 0.9534202515651026,
+      "grad_norm": 0.516810953617096,
+      "learning_rate": 0.0001,
+      "loss": 1.7744,
+      "step": 8300
+    },
+    {
+      "epoch": 0.9535351214749297,
+      "grad_norm": 0.46272823214530945,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 8301
+    },
+    {
+      "epoch": 0.9536499913847568,
+      "grad_norm": 0.5371290445327759,
+      "learning_rate": 0.0001,
+      "loss": 1.5101,
+      "step": 8302
+    },
+    {
+      "epoch": 0.9537648612945839,
+      "grad_norm": 0.5088568329811096,
+      "learning_rate": 0.0001,
+      "loss": 1.7429,
+      "step": 8303
+    },
+    {
+      "epoch": 0.953879731204411,
+      "grad_norm": 0.4637158215045929,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 8304
+    },
+    {
+      "epoch": 0.9539946011142382,
+      "grad_norm": 0.44325053691864014,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 8305
+    },
+    {
+      "epoch": 0.9541094710240653,
+      "grad_norm": 0.4635404944419861,
+      "learning_rate": 0.0001,
+      "loss": 1.4581,
+      "step": 8306
+    },
+    {
+      "epoch": 0.9542243409338924,
+      "grad_norm": 0.5078271627426147,
+      "learning_rate": 0.0001,
+      "loss": 1.7131,
+      "step": 8307
+    },
+    {
+      "epoch": 0.9543392108437195,
+      "grad_norm": 0.5453723669052124,
+      "learning_rate": 0.0001,
+      "loss": 1.6502,
+      "step": 8308
+    },
+    {
+      "epoch": 0.9544540807535467,
+      "grad_norm": 0.4599123001098633,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 8309
+    },
+    {
+      "epoch": 0.9545689506633738,
+      "grad_norm": 0.5290499329566956,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 8310
+    },
+    {
+      "epoch": 0.9546838205732009,
+      "grad_norm": 0.44906917214393616,
+      "learning_rate": 0.0001,
+      "loss": 1.5134,
+      "step": 8311
+    },
+    {
+      "epoch": 0.954798690483028,
+      "grad_norm": 0.44247668981552124,
+      "learning_rate": 0.0001,
+      "loss": 1.4854,
+      "step": 8312
+    },
+    {
+      "epoch": 0.9549135603928551,
+      "grad_norm": 0.49097177386283875,
+      "learning_rate": 0.0001,
+      "loss": 1.3212,
+      "step": 8313
+    },
+    {
+      "epoch": 0.9550284303026823,
+      "grad_norm": 0.47003665566444397,
+      "learning_rate": 0.0001,
+      "loss": 1.5748,
+      "step": 8314
+    },
+    {
+      "epoch": 0.9551433002125094,
+      "grad_norm": 0.4841088056564331,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 8315
+    },
+    {
+      "epoch": 0.9552581701223365,
+      "grad_norm": 0.47243985533714294,
+      "learning_rate": 0.0001,
+      "loss": 1.4881,
+      "step": 8316
+    },
+    {
+      "epoch": 0.9553730400321636,
+      "grad_norm": 0.47702085971832275,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 8317
+    },
+    {
+      "epoch": 0.9554879099419907,
+      "grad_norm": 0.45515045523643494,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 8318
+    },
+    {
+      "epoch": 0.9556027798518179,
+      "grad_norm": 0.5219390392303467,
+      "learning_rate": 0.0001,
+      "loss": 1.7652,
+      "step": 8319
+    },
+    {
+      "epoch": 0.955717649761645,
+      "grad_norm": 0.5401485562324524,
+      "learning_rate": 0.0001,
+      "loss": 1.8357,
+      "step": 8320
+    },
+    {
+      "epoch": 0.9558325196714721,
+      "grad_norm": 0.4558376669883728,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 8321
+    },
+    {
+      "epoch": 0.9559473895812992,
+      "grad_norm": 0.4861636459827423,
+      "learning_rate": 0.0001,
+      "loss": 1.6804,
+      "step": 8322
+    },
+    {
+      "epoch": 0.9560622594911263,
+      "grad_norm": 0.4567401707172394,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 8323
+    },
+    {
+      "epoch": 0.9561771294009535,
+      "grad_norm": 0.5179786086082458,
+      "learning_rate": 0.0001,
+      "loss": 1.8311,
+      "step": 8324
+    },
+    {
+      "epoch": 0.9562919993107806,
+      "grad_norm": 0.45889440178871155,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 8325
+    },
+    {
+      "epoch": 0.9564068692206077,
+      "grad_norm": 0.4716584086418152,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 8326
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 0.43330761790275574,
+      "learning_rate": 0.0001,
+      "loss": 1.2869,
+      "step": 8327
+    },
+    {
+      "epoch": 0.9566366090402619,
+      "grad_norm": 0.4397355914115906,
+      "learning_rate": 0.0001,
+      "loss": 1.5012,
+      "step": 8328
+    },
+    {
+      "epoch": 0.956751478950089,
+      "grad_norm": 0.4772343933582306,
+      "learning_rate": 0.0001,
+      "loss": 1.3369,
+      "step": 8329
+    },
+    {
+      "epoch": 0.9568663488599162,
+      "grad_norm": 0.5061171054840088,
+      "learning_rate": 0.0001,
+      "loss": 1.7989,
+      "step": 8330
+    },
+    {
+      "epoch": 0.9569812187697433,
+      "grad_norm": 0.46907833218574524,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 8331
+    },
+    {
+      "epoch": 0.9570960886795704,
+      "grad_norm": 0.5048026442527771,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 8332
+    },
+    {
+      "epoch": 0.9572109585893975,
+      "grad_norm": 0.4772169888019562,
+      "learning_rate": 0.0001,
+      "loss": 1.4803,
+      "step": 8333
+    },
+    {
+      "epoch": 0.9573258284992247,
+      "grad_norm": 0.47201335430145264,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 8334
+    },
+    {
+      "epoch": 0.9574406984090518,
+      "grad_norm": 0.49312329292297363,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 8335
+    },
+    {
+      "epoch": 0.9575555683188789,
+      "grad_norm": 0.4100249111652374,
+      "learning_rate": 0.0001,
+      "loss": 1.3143,
+      "step": 8336
+    },
+    {
+      "epoch": 0.957670438228706,
+      "grad_norm": 0.5461440086364746,
+      "learning_rate": 0.0001,
+      "loss": 1.7893,
+      "step": 8337
+    },
+    {
+      "epoch": 0.9577853081385331,
+      "grad_norm": 0.47124072909355164,
+      "learning_rate": 0.0001,
+      "loss": 1.3658,
+      "step": 8338
+    },
+    {
+      "epoch": 0.9579001780483603,
+      "grad_norm": 0.4608875811100006,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 8339
+    },
+    {
+      "epoch": 0.9580150479581874,
+      "grad_norm": 0.49988433718681335,
+      "learning_rate": 0.0001,
+      "loss": 1.6149,
+      "step": 8340
+    },
+    {
+      "epoch": 0.9581299178680145,
+      "grad_norm": 0.47735628485679626,
+      "learning_rate": 0.0001,
+      "loss": 1.5887,
+      "step": 8341
+    },
+    {
+      "epoch": 0.9582447877778416,
+      "grad_norm": 0.5070236921310425,
+      "learning_rate": 0.0001,
+      "loss": 1.4978,
+      "step": 8342
+    },
+    {
+      "epoch": 0.9583596576876687,
+      "grad_norm": 0.5401067137718201,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 8343
+    },
+    {
+      "epoch": 0.9584745275974959,
+      "grad_norm": 0.45689040422439575,
+      "learning_rate": 0.0001,
+      "loss": 1.6352,
+      "step": 8344
+    },
+    {
+      "epoch": 0.958589397507323,
+      "grad_norm": 0.4751141667366028,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 8345
+    },
+    {
+      "epoch": 0.9587042674171501,
+      "grad_norm": 0.5322357416152954,
+      "learning_rate": 0.0001,
+      "loss": 1.7816,
+      "step": 8346
+    },
+    {
+      "epoch": 0.9588191373269772,
+      "grad_norm": 0.472920298576355,
+      "learning_rate": 0.0001,
+      "loss": 1.4187,
+      "step": 8347
+    },
+    {
+      "epoch": 0.9589340072368043,
+      "grad_norm": 0.5017462372779846,
+      "learning_rate": 0.0001,
+      "loss": 1.7655,
+      "step": 8348
+    },
+    {
+      "epoch": 0.9590488771466315,
+      "grad_norm": 0.4727727770805359,
+      "learning_rate": 0.0001,
+      "loss": 1.5264,
+      "step": 8349
+    },
+    {
+      "epoch": 0.9591637470564586,
+      "grad_norm": 0.45175325870513916,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 8350
+    },
+    {
+      "epoch": 0.9592786169662857,
+      "grad_norm": 0.500636100769043,
+      "learning_rate": 0.0001,
+      "loss": 1.7486,
+      "step": 8351
+    },
+    {
+      "epoch": 0.9593934868761128,
+      "grad_norm": 0.44014236330986023,
+      "learning_rate": 0.0001,
+      "loss": 1.3911,
+      "step": 8352
+    },
+    {
+      "epoch": 0.9595083567859399,
+      "grad_norm": 0.4677266478538513,
+      "learning_rate": 0.0001,
+      "loss": 1.8267,
+      "step": 8353
+    },
+    {
+      "epoch": 0.959623226695767,
+      "grad_norm": 0.4543503522872925,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 8354
+    },
+    {
+      "epoch": 0.9597380966055942,
+      "grad_norm": 0.48893970251083374,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 8355
+    },
+    {
+      "epoch": 0.9598529665154213,
+      "grad_norm": 0.4677874445915222,
+      "learning_rate": 0.0001,
+      "loss": 1.6743,
+      "step": 8356
+    },
+    {
+      "epoch": 0.9599678364252484,
+      "grad_norm": 0.46414315700531006,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 8357
+    },
+    {
+      "epoch": 0.9600827063350755,
+      "grad_norm": 0.46315401792526245,
+      "learning_rate": 0.0001,
+      "loss": 1.5053,
+      "step": 8358
+    },
+    {
+      "epoch": 0.9601975762449027,
+      "grad_norm": 0.4876864552497864,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 8359
+    },
+    {
+      "epoch": 0.9603124461547298,
+      "grad_norm": 0.4578522741794586,
+      "learning_rate": 0.0001,
+      "loss": 1.5981,
+      "step": 8360
+    },
+    {
+      "epoch": 0.9604273160645569,
+      "grad_norm": 0.48835068941116333,
+      "learning_rate": 0.0001,
+      "loss": 1.7534,
+      "step": 8361
+    },
+    {
+      "epoch": 0.960542185974384,
+      "grad_norm": 0.4757782518863678,
+      "learning_rate": 0.0001,
+      "loss": 1.4454,
+      "step": 8362
+    },
+    {
+      "epoch": 0.9606570558842111,
+      "grad_norm": 0.45514336228370667,
+      "learning_rate": 0.0001,
+      "loss": 1.476,
+      "step": 8363
+    },
+    {
+      "epoch": 0.9607719257940383,
+      "grad_norm": 0.4917604625225067,
+      "learning_rate": 0.0001,
+      "loss": 1.6986,
+      "step": 8364
+    },
+    {
+      "epoch": 0.9608867957038654,
+      "grad_norm": 0.4693104326725006,
+      "learning_rate": 0.0001,
+      "loss": 1.4697,
+      "step": 8365
+    },
+    {
+      "epoch": 0.9610016656136925,
+      "grad_norm": 0.4452751874923706,
+      "learning_rate": 0.0001,
+      "loss": 1.5288,
+      "step": 8366
+    },
+    {
+      "epoch": 0.9611165355235196,
+      "grad_norm": 0.4996024966239929,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 8367
+    },
+    {
+      "epoch": 0.9612314054333467,
+      "grad_norm": 0.4728527069091797,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 8368
+    },
+    {
+      "epoch": 0.9613462753431739,
+      "grad_norm": 0.5003929734230042,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 8369
+    },
+    {
+      "epoch": 0.961461145253001,
+      "grad_norm": 0.49857136607170105,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 8370
+    },
+    {
+      "epoch": 0.9615760151628281,
+      "grad_norm": 0.49390268325805664,
+      "learning_rate": 0.0001,
+      "loss": 1.5569,
+      "step": 8371
+    },
+    {
+      "epoch": 0.9616908850726552,
+      "grad_norm": 0.5034048557281494,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 8372
+    },
+    {
+      "epoch": 0.9618057549824823,
+      "grad_norm": 0.45734164118766785,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 8373
+    },
+    {
+      "epoch": 0.9619206248923095,
+      "grad_norm": 0.47848743200302124,
+      "learning_rate": 0.0001,
+      "loss": 1.6597,
+      "step": 8374
+    },
+    {
+      "epoch": 0.9620354948021366,
+      "grad_norm": 0.4872235357761383,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 8375
+    },
+    {
+      "epoch": 0.9621503647119637,
+      "grad_norm": 0.491102933883667,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 8376
+    },
+    {
+      "epoch": 0.9622652346217908,
+      "grad_norm": 0.4527552127838135,
+      "learning_rate": 0.0001,
+      "loss": 1.3692,
+      "step": 8377
+    },
+    {
+      "epoch": 0.9623801045316179,
+      "grad_norm": 0.4430198073387146,
+      "learning_rate": 0.0001,
+      "loss": 1.5954,
+      "step": 8378
+    },
+    {
+      "epoch": 0.962494974441445,
+      "grad_norm": 0.4682437479496002,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 8379
+    },
+    {
+      "epoch": 0.9626098443512722,
+      "grad_norm": 0.4858921766281128,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 8380
+    },
+    {
+      "epoch": 0.9627247142610993,
+      "grad_norm": 0.4844062626361847,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 8381
+    },
+    {
+      "epoch": 0.9628395841709264,
+      "grad_norm": 0.5199929475784302,
+      "learning_rate": 0.0001,
+      "loss": 1.7057,
+      "step": 8382
+    },
+    {
+      "epoch": 0.9629544540807535,
+      "grad_norm": 0.49841341376304626,
+      "learning_rate": 0.0001,
+      "loss": 1.5898,
+      "step": 8383
+    },
+    {
+      "epoch": 0.9630693239905807,
+      "grad_norm": 0.5616093873977661,
+      "learning_rate": 0.0001,
+      "loss": 1.9157,
+      "step": 8384
+    },
+    {
+      "epoch": 0.9631841939004078,
+      "grad_norm": 0.45158112049102783,
+      "learning_rate": 0.0001,
+      "loss": 1.5281,
+      "step": 8385
+    },
+    {
+      "epoch": 0.9632990638102349,
+      "grad_norm": 0.5046762228012085,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 8386
+    },
+    {
+      "epoch": 0.963413933720062,
+      "grad_norm": 0.5029253959655762,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 8387
+    },
+    {
+      "epoch": 0.9635288036298891,
+      "grad_norm": 0.48361513018608093,
+      "learning_rate": 0.0001,
+      "loss": 1.7588,
+      "step": 8388
+    },
+    {
+      "epoch": 0.9636436735397163,
+      "grad_norm": 0.4966965317726135,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 8389
+    },
+    {
+      "epoch": 0.9637585434495434,
+      "grad_norm": 0.4773005545139313,
+      "learning_rate": 0.0001,
+      "loss": 1.6763,
+      "step": 8390
+    },
+    {
+      "epoch": 0.9638734133593705,
+      "grad_norm": 0.4541994035243988,
+      "learning_rate": 0.0001,
+      "loss": 1.5686,
+      "step": 8391
+    },
+    {
+      "epoch": 0.9639882832691976,
+      "grad_norm": 0.43786925077438354,
+      "learning_rate": 0.0001,
+      "loss": 1.3373,
+      "step": 8392
+    },
+    {
+      "epoch": 0.9641031531790247,
+      "grad_norm": 0.5206244587898254,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 8393
+    },
+    {
+      "epoch": 0.9642180230888518,
+      "grad_norm": 0.5168501138687134,
+      "learning_rate": 0.0001,
+      "loss": 1.7587,
+      "step": 8394
+    },
+    {
+      "epoch": 0.964332892998679,
+      "grad_norm": 0.48464956879615784,
+      "learning_rate": 0.0001,
+      "loss": 1.7185,
+      "step": 8395
+    },
+    {
+      "epoch": 0.9644477629085061,
+      "grad_norm": 0.4584466814994812,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 8396
+    },
+    {
+      "epoch": 0.9645626328183332,
+      "grad_norm": 0.49749574065208435,
+      "learning_rate": 0.0001,
+      "loss": 1.785,
+      "step": 8397
+    },
+    {
+      "epoch": 0.9646775027281603,
+      "grad_norm": 0.4844261407852173,
+      "learning_rate": 0.0001,
+      "loss": 1.4182,
+      "step": 8398
+    },
+    {
+      "epoch": 0.9647923726379874,
+      "grad_norm": 0.4621402323246002,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 8399
+    },
+    {
+      "epoch": 0.9649072425478146,
+      "grad_norm": 0.4868880808353424,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 8400
+    },
+    {
+      "epoch": 0.9650221124576417,
+      "grad_norm": 0.4497327506542206,
+      "learning_rate": 0.0001,
+      "loss": 1.5056,
+      "step": 8401
+    },
+    {
+      "epoch": 0.9651369823674688,
+      "grad_norm": 0.5474899411201477,
+      "learning_rate": 0.0001,
+      "loss": 1.8234,
+      "step": 8402
+    },
+    {
+      "epoch": 0.9652518522772959,
+      "grad_norm": 0.46715471148490906,
+      "learning_rate": 0.0001,
+      "loss": 1.3831,
+      "step": 8403
+    },
+    {
+      "epoch": 0.965366722187123,
+      "grad_norm": 0.5017038583755493,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 8404
+    },
+    {
+      "epoch": 0.9654815920969502,
+      "grad_norm": 0.5196341872215271,
+      "learning_rate": 0.0001,
+      "loss": 1.8523,
+      "step": 8405
+    },
+    {
+      "epoch": 0.9655964620067773,
+      "grad_norm": 0.4290624260902405,
+      "learning_rate": 0.0001,
+      "loss": 1.3917,
+      "step": 8406
+    },
+    {
+      "epoch": 0.9657113319166044,
+      "grad_norm": 0.5225144028663635,
+      "learning_rate": 0.0001,
+      "loss": 1.7594,
+      "step": 8407
+    },
+    {
+      "epoch": 0.9658262018264315,
+      "grad_norm": 0.44787734746932983,
+      "learning_rate": 0.0001,
+      "loss": 1.5222,
+      "step": 8408
+    },
+    {
+      "epoch": 0.9659410717362586,
+      "grad_norm": 0.4654848575592041,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 8409
+    },
+    {
+      "epoch": 0.9660559416460858,
+      "grad_norm": 0.49902090430259705,
+      "learning_rate": 0.0001,
+      "loss": 1.7194,
+      "step": 8410
+    },
+    {
+      "epoch": 0.9661708115559129,
+      "grad_norm": 0.5180195569992065,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 8411
+    },
+    {
+      "epoch": 0.96628568146574,
+      "grad_norm": 0.4926517903804779,
+      "learning_rate": 0.0001,
+      "loss": 1.6377,
+      "step": 8412
+    },
+    {
+      "epoch": 0.9664005513755671,
+      "grad_norm": 0.4600144028663635,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 8413
+    },
+    {
+      "epoch": 0.9665154212853942,
+      "grad_norm": 0.4665214717388153,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 8414
+    },
+    {
+      "epoch": 0.9666302911952214,
+      "grad_norm": 0.47527435421943665,
+      "learning_rate": 0.0001,
+      "loss": 1.5603,
+      "step": 8415
+    },
+    {
+      "epoch": 0.9667451611050485,
+      "grad_norm": 0.4811752140522003,
+      "learning_rate": 0.0001,
+      "loss": 1.4612,
+      "step": 8416
+    },
+    {
+      "epoch": 0.9668600310148756,
+      "grad_norm": 0.4516500234603882,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 8417
+    },
+    {
+      "epoch": 0.9669749009247027,
+      "grad_norm": 0.5186557769775391,
+      "learning_rate": 0.0001,
+      "loss": 1.7303,
+      "step": 8418
+    },
+    {
+      "epoch": 0.9670897708345298,
+      "grad_norm": 0.4529307782649994,
+      "learning_rate": 0.0001,
+      "loss": 1.4181,
+      "step": 8419
+    },
+    {
+      "epoch": 0.967204640744357,
+      "grad_norm": 0.4979994297027588,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 8420
+    },
+    {
+      "epoch": 0.9673195106541841,
+      "grad_norm": 0.4783253073692322,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 8421
+    },
+    {
+      "epoch": 0.9674343805640112,
+      "grad_norm": 0.4694981276988983,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 8422
+    },
+    {
+      "epoch": 0.9675492504738383,
+      "grad_norm": 0.4916779398918152,
+      "learning_rate": 0.0001,
+      "loss": 1.7472,
+      "step": 8423
+    },
+    {
+      "epoch": 0.9676641203836654,
+      "grad_norm": 0.42479321360588074,
+      "learning_rate": 0.0001,
+      "loss": 1.3967,
+      "step": 8424
+    },
+    {
+      "epoch": 0.9677789902934926,
+      "grad_norm": 0.5579563975334167,
+      "learning_rate": 0.0001,
+      "loss": 1.8852,
+      "step": 8425
+    },
+    {
+      "epoch": 0.9678938602033197,
+      "grad_norm": 0.5487393736839294,
+      "learning_rate": 0.0001,
+      "loss": 1.7019,
+      "step": 8426
+    },
+    {
+      "epoch": 0.9680087301131468,
+      "grad_norm": 0.46243569254875183,
+      "learning_rate": 0.0001,
+      "loss": 1.5185,
+      "step": 8427
+    },
+    {
+      "epoch": 0.9681236000229739,
+      "grad_norm": 0.468492716550827,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 8428
+    },
+    {
+      "epoch": 0.968238469932801,
+      "grad_norm": 0.4968058466911316,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 8429
+    },
+    {
+      "epoch": 0.9683533398426282,
+      "grad_norm": 0.44949015974998474,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 8430
+    },
+    {
+      "epoch": 0.9684682097524553,
+      "grad_norm": 0.46042609214782715,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 8431
+    },
+    {
+      "epoch": 0.9685830796622825,
+      "grad_norm": 0.48226192593574524,
+      "learning_rate": 0.0001,
+      "loss": 1.6218,
+      "step": 8432
+    },
+    {
+      "epoch": 0.9686979495721096,
+      "grad_norm": 0.4611875116825104,
+      "learning_rate": 0.0001,
+      "loss": 1.4546,
+      "step": 8433
+    },
+    {
+      "epoch": 0.9688128194819368,
+      "grad_norm": 0.49678662419319153,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 8434
+    },
+    {
+      "epoch": 0.9689276893917639,
+      "grad_norm": 0.4989873170852661,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 8435
+    },
+    {
+      "epoch": 0.969042559301591,
+      "grad_norm": 0.49776941537857056,
+      "learning_rate": 0.0001,
+      "loss": 1.6435,
+      "step": 8436
+    },
+    {
+      "epoch": 0.9691574292114181,
+      "grad_norm": 0.4603618085384369,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 8437
+    },
+    {
+      "epoch": 0.9692722991212452,
+      "grad_norm": 0.503463625907898,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 8438
+    },
+    {
+      "epoch": 0.9693871690310724,
+      "grad_norm": 0.520193874835968,
+      "learning_rate": 0.0001,
+      "loss": 1.7394,
+      "step": 8439
+    },
+    {
+      "epoch": 0.9695020389408995,
+      "grad_norm": 0.4809243679046631,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 8440
+    },
+    {
+      "epoch": 0.9696169088507266,
+      "grad_norm": 0.49442121386528015,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 8441
+    },
+    {
+      "epoch": 0.9697317787605537,
+      "grad_norm": 0.43582677841186523,
+      "learning_rate": 0.0001,
+      "loss": 1.5189,
+      "step": 8442
+    },
+    {
+      "epoch": 0.9698466486703808,
+      "grad_norm": 0.4740058481693268,
+      "learning_rate": 0.0001,
+      "loss": 1.6358,
+      "step": 8443
+    },
+    {
+      "epoch": 0.969961518580208,
+      "grad_norm": 0.4624651372432709,
+      "learning_rate": 0.0001,
+      "loss": 1.6075,
+      "step": 8444
+    },
+    {
+      "epoch": 0.9700763884900351,
+      "grad_norm": 0.5144423842430115,
+      "learning_rate": 0.0001,
+      "loss": 1.7943,
+      "step": 8445
+    },
+    {
+      "epoch": 0.9701912583998622,
+      "grad_norm": 0.4762701392173767,
+      "learning_rate": 0.0001,
+      "loss": 1.4843,
+      "step": 8446
+    },
+    {
+      "epoch": 0.9703061283096893,
+      "grad_norm": 0.46587637066841125,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 8447
+    },
+    {
+      "epoch": 0.9704209982195164,
+      "grad_norm": 0.516172468662262,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 8448
+    },
+    {
+      "epoch": 0.9705358681293436,
+      "grad_norm": 0.4806175231933594,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 8449
+    },
+    {
+      "epoch": 0.9706507380391707,
+      "grad_norm": 0.4877374768257141,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 8450
+    },
+    {
+      "epoch": 0.9707656079489978,
+      "grad_norm": 0.5047240257263184,
+      "learning_rate": 0.0001,
+      "loss": 1.7531,
+      "step": 8451
+    },
+    {
+      "epoch": 0.9708804778588249,
+      "grad_norm": 0.4985336363315582,
+      "learning_rate": 0.0001,
+      "loss": 1.8455,
+      "step": 8452
+    },
+    {
+      "epoch": 0.970995347768652,
+      "grad_norm": 0.5148325562477112,
+      "learning_rate": 0.0001,
+      "loss": 1.6977,
+      "step": 8453
+    },
+    {
+      "epoch": 0.9711102176784792,
+      "grad_norm": 0.4783836901187897,
+      "learning_rate": 0.0001,
+      "loss": 1.7402,
+      "step": 8454
+    },
+    {
+      "epoch": 0.9712250875883063,
+      "grad_norm": 0.5134006142616272,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 8455
+    },
+    {
+      "epoch": 0.9713399574981334,
+      "grad_norm": 0.4870409071445465,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 8456
+    },
+    {
+      "epoch": 0.9714548274079605,
+      "grad_norm": 0.47401508688926697,
+      "learning_rate": 0.0001,
+      "loss": 1.7521,
+      "step": 8457
+    },
+    {
+      "epoch": 0.9715696973177876,
+      "grad_norm": 0.46967458724975586,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 8458
+    },
+    {
+      "epoch": 0.9716845672276148,
+      "grad_norm": 0.5170222520828247,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 8459
+    },
+    {
+      "epoch": 0.9717994371374419,
+      "grad_norm": 0.503593921661377,
+      "learning_rate": 0.0001,
+      "loss": 1.5592,
+      "step": 8460
+    },
+    {
+      "epoch": 0.971914307047269,
+      "grad_norm": 0.47926294803619385,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 8461
+    },
+    {
+      "epoch": 0.9720291769570961,
+      "grad_norm": 0.48252803087234497,
+      "learning_rate": 0.0001,
+      "loss": 1.7432,
+      "step": 8462
+    },
+    {
+      "epoch": 0.9721440468669232,
+      "grad_norm": 0.5037051439285278,
+      "learning_rate": 0.0001,
+      "loss": 1.7139,
+      "step": 8463
+    },
+    {
+      "epoch": 0.9722589167767504,
+      "grad_norm": 0.46094417572021484,
+      "learning_rate": 0.0001,
+      "loss": 1.406,
+      "step": 8464
+    },
+    {
+      "epoch": 0.9723737866865775,
+      "grad_norm": 0.5147666335105896,
+      "learning_rate": 0.0001,
+      "loss": 1.6715,
+      "step": 8465
+    },
+    {
+      "epoch": 0.9724886565964046,
+      "grad_norm": 0.4747392237186432,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 8466
+    },
+    {
+      "epoch": 0.9726035265062317,
+      "grad_norm": 0.5164050459861755,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 8467
+    },
+    {
+      "epoch": 0.9727183964160588,
+      "grad_norm": 0.46309229731559753,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 8468
+    },
+    {
+      "epoch": 0.972833266325886,
+      "grad_norm": 0.4927702248096466,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 8469
+    },
+    {
+      "epoch": 0.9729481362357131,
+      "grad_norm": 0.5026853084564209,
+      "learning_rate": 0.0001,
+      "loss": 1.5314,
+      "step": 8470
+    },
+    {
+      "epoch": 0.9730630061455402,
+      "grad_norm": 0.5030810832977295,
+      "learning_rate": 0.0001,
+      "loss": 1.7262,
+      "step": 8471
+    },
+    {
+      "epoch": 0.9731778760553673,
+      "grad_norm": 0.4909353256225586,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 8472
+    },
+    {
+      "epoch": 0.9732927459651944,
+      "grad_norm": 0.4842395484447479,
+      "learning_rate": 0.0001,
+      "loss": 1.6103,
+      "step": 8473
+    },
+    {
+      "epoch": 0.9734076158750216,
+      "grad_norm": 0.44747745990753174,
+      "learning_rate": 0.0001,
+      "loss": 1.4004,
+      "step": 8474
+    },
+    {
+      "epoch": 0.9735224857848487,
+      "grad_norm": 0.4720572233200073,
+      "learning_rate": 0.0001,
+      "loss": 1.4371,
+      "step": 8475
+    },
+    {
+      "epoch": 0.9736373556946758,
+      "grad_norm": 0.5061703324317932,
+      "learning_rate": 0.0001,
+      "loss": 1.5303,
+      "step": 8476
+    },
+    {
+      "epoch": 0.9737522256045029,
+      "grad_norm": 0.46352124214172363,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 8477
+    },
+    {
+      "epoch": 0.97386709551433,
+      "grad_norm": 0.4936666488647461,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 8478
+    },
+    {
+      "epoch": 0.9739819654241572,
+      "grad_norm": 0.4917965829372406,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 8479
+    },
+    {
+      "epoch": 0.9740968353339843,
+      "grad_norm": 0.5103937387466431,
+      "learning_rate": 0.0001,
+      "loss": 1.7674,
+      "step": 8480
+    },
+    {
+      "epoch": 0.9742117052438114,
+      "grad_norm": 0.44612032175064087,
+      "learning_rate": 0.0001,
+      "loss": 1.3708,
+      "step": 8481
+    },
+    {
+      "epoch": 0.9743265751536385,
+      "grad_norm": 0.48888033628463745,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 8482
+    },
+    {
+      "epoch": 0.9744414450634656,
+      "grad_norm": 0.49829840660095215,
+      "learning_rate": 0.0001,
+      "loss": 1.7156,
+      "step": 8483
+    },
+    {
+      "epoch": 0.9745563149732928,
+      "grad_norm": 0.4825059473514557,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 8484
+    },
+    {
+      "epoch": 0.9746711848831199,
+      "grad_norm": 0.44610595703125,
+      "learning_rate": 0.0001,
+      "loss": 1.3814,
+      "step": 8485
+    },
+    {
+      "epoch": 0.974786054792947,
+      "grad_norm": 0.5563763976097107,
+      "learning_rate": 0.0001,
+      "loss": 1.7003,
+      "step": 8486
+    },
+    {
+      "epoch": 0.9749009247027741,
+      "grad_norm": 0.4703843295574188,
+      "learning_rate": 0.0001,
+      "loss": 1.4677,
+      "step": 8487
+    },
+    {
+      "epoch": 0.9750157946126012,
+      "grad_norm": 0.524733304977417,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 8488
+    },
+    {
+      "epoch": 0.9751306645224284,
+      "grad_norm": 0.5146064758300781,
+      "learning_rate": 0.0001,
+      "loss": 1.8297,
+      "step": 8489
+    },
+    {
+      "epoch": 0.9752455344322555,
+      "grad_norm": 0.432230681180954,
+      "learning_rate": 0.0001,
+      "loss": 1.5052,
+      "step": 8490
+    },
+    {
+      "epoch": 0.9753604043420826,
+      "grad_norm": 0.4458877742290497,
+      "learning_rate": 0.0001,
+      "loss": 1.5611,
+      "step": 8491
+    },
+    {
+      "epoch": 0.9754752742519097,
+      "grad_norm": 0.474324107170105,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 8492
+    },
+    {
+      "epoch": 0.9755901441617368,
+      "grad_norm": 0.4817248582839966,
+      "learning_rate": 0.0001,
+      "loss": 1.4067,
+      "step": 8493
+    },
+    {
+      "epoch": 0.975705014071564,
+      "grad_norm": 0.46746066212654114,
+      "learning_rate": 0.0001,
+      "loss": 1.5686,
+      "step": 8494
+    },
+    {
+      "epoch": 0.9758198839813911,
+      "grad_norm": 0.4995744824409485,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 8495
+    },
+    {
+      "epoch": 0.9759347538912182,
+      "grad_norm": 0.4750157594680786,
+      "learning_rate": 0.0001,
+      "loss": 1.4601,
+      "step": 8496
+    },
+    {
+      "epoch": 0.9760496238010453,
+      "grad_norm": 0.474306583404541,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 8497
+    },
+    {
+      "epoch": 0.9761644937108724,
+      "grad_norm": 0.494228720664978,
+      "learning_rate": 0.0001,
+      "loss": 1.4225,
+      "step": 8498
+    },
+    {
+      "epoch": 0.9762793636206996,
+      "grad_norm": 0.4527890384197235,
+      "learning_rate": 0.0001,
+      "loss": 1.5073,
+      "step": 8499
+    },
+    {
+      "epoch": 0.9763942335305267,
+      "grad_norm": 0.4580267071723938,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 8500
+    },
+    {
+      "epoch": 0.9765091034403538,
+      "grad_norm": 0.5154402852058411,
+      "learning_rate": 0.0001,
+      "loss": 1.6121,
+      "step": 8501
+    },
+    {
+      "epoch": 0.9766239733501809,
+      "grad_norm": 0.4666356146335602,
+      "learning_rate": 0.0001,
+      "loss": 1.6026,
+      "step": 8502
+    },
+    {
+      "epoch": 0.976738843260008,
+      "grad_norm": 0.4806033670902252,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 8503
+    },
+    {
+      "epoch": 0.9768537131698352,
+      "grad_norm": 0.49173009395599365,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 8504
+    },
+    {
+      "epoch": 0.9769685830796623,
+      "grad_norm": 0.528270959854126,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 8505
+    },
+    {
+      "epoch": 0.9770834529894894,
+      "grad_norm": 0.4734746217727661,
+      "learning_rate": 0.0001,
+      "loss": 1.7124,
+      "step": 8506
+    },
+    {
+      "epoch": 0.9771983228993165,
+      "grad_norm": 0.44804245233535767,
+      "learning_rate": 0.0001,
+      "loss": 1.4199,
+      "step": 8507
+    },
+    {
+      "epoch": 0.9773131928091436,
+      "grad_norm": 0.4997158944606781,
+      "learning_rate": 0.0001,
+      "loss": 1.5853,
+      "step": 8508
+    },
+    {
+      "epoch": 0.9774280627189708,
+      "grad_norm": 0.5285566449165344,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 8509
+    },
+    {
+      "epoch": 0.9775429326287979,
+      "grad_norm": 0.5019384026527405,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 8510
+    },
+    {
+      "epoch": 0.977657802538625,
+      "grad_norm": 0.49325329065322876,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 8511
+    },
+    {
+      "epoch": 0.9777726724484521,
+      "grad_norm": 0.439513236284256,
+      "learning_rate": 0.0001,
+      "loss": 1.3171,
+      "step": 8512
+    },
+    {
+      "epoch": 0.9778875423582792,
+      "grad_norm": 0.4921404719352722,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 8513
+    },
+    {
+      "epoch": 0.9780024122681064,
+      "grad_norm": 0.48236212134361267,
+      "learning_rate": 0.0001,
+      "loss": 1.3529,
+      "step": 8514
+    },
+    {
+      "epoch": 0.9781172821779335,
+      "grad_norm": 0.437320739030838,
+      "learning_rate": 0.0001,
+      "loss": 1.345,
+      "step": 8515
+    },
+    {
+      "epoch": 0.9782321520877606,
+      "grad_norm": 0.48497122526168823,
+      "learning_rate": 0.0001,
+      "loss": 1.2823,
+      "step": 8516
+    },
+    {
+      "epoch": 0.9783470219975877,
+      "grad_norm": 0.48262646794319153,
+      "learning_rate": 0.0001,
+      "loss": 1.6981,
+      "step": 8517
+    },
+    {
+      "epoch": 0.9784618919074148,
+      "grad_norm": 0.48424267768859863,
+      "learning_rate": 0.0001,
+      "loss": 1.5371,
+      "step": 8518
+    },
+    {
+      "epoch": 0.978576761817242,
+      "grad_norm": 0.48815423250198364,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 8519
+    },
+    {
+      "epoch": 0.9786916317270691,
+      "grad_norm": 0.48007214069366455,
+      "learning_rate": 0.0001,
+      "loss": 1.696,
+      "step": 8520
+    },
+    {
+      "epoch": 0.9788065016368962,
+      "grad_norm": 0.501133918762207,
+      "learning_rate": 0.0001,
+      "loss": 1.6893,
+      "step": 8521
+    },
+    {
+      "epoch": 0.9789213715467233,
+      "grad_norm": 0.48664984107017517,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 8522
+    },
+    {
+      "epoch": 0.9790362414565504,
+      "grad_norm": 0.47685080766677856,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 8523
+    },
+    {
+      "epoch": 0.9791511113663776,
+      "grad_norm": 0.458021342754364,
+      "learning_rate": 0.0001,
+      "loss": 1.3507,
+      "step": 8524
+    },
+    {
+      "epoch": 0.9792659812762047,
+      "grad_norm": 0.48686715960502625,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 8525
+    },
+    {
+      "epoch": 0.9793808511860318,
+      "grad_norm": 0.4461478590965271,
+      "learning_rate": 0.0001,
+      "loss": 1.4896,
+      "step": 8526
+    },
+    {
+      "epoch": 0.9794957210958589,
+      "grad_norm": 0.47489675879478455,
+      "learning_rate": 0.0001,
+      "loss": 1.691,
+      "step": 8527
+    },
+    {
+      "epoch": 0.979610591005686,
+      "grad_norm": 0.4845513105392456,
+      "learning_rate": 0.0001,
+      "loss": 1.5867,
+      "step": 8528
+    },
+    {
+      "epoch": 0.9797254609155132,
+      "grad_norm": 0.4592495858669281,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 8529
+    },
+    {
+      "epoch": 0.9798403308253403,
+      "grad_norm": 0.4863679111003876,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 8530
+    },
+    {
+      "epoch": 0.9799552007351674,
+      "grad_norm": 0.47153112292289734,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 8531
+    },
+    {
+      "epoch": 0.9800700706449945,
+      "grad_norm": 0.4806155562400818,
+      "learning_rate": 0.0001,
+      "loss": 1.3655,
+      "step": 8532
+    },
+    {
+      "epoch": 0.9801849405548216,
+      "grad_norm": 0.5179005265235901,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 8533
+    },
+    {
+      "epoch": 0.9802998104646488,
+      "grad_norm": 0.5735377073287964,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 8534
+    },
+    {
+      "epoch": 0.9804146803744759,
+      "grad_norm": 0.47526344656944275,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 8535
+    },
+    {
+      "epoch": 0.980529550284303,
+      "grad_norm": 0.4913672208786011,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 8536
+    },
+    {
+      "epoch": 0.9806444201941301,
+      "grad_norm": 0.5402122735977173,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 8537
+    },
+    {
+      "epoch": 0.9807592901039572,
+      "grad_norm": 0.4758612811565399,
+      "learning_rate": 0.0001,
+      "loss": 1.6862,
+      "step": 8538
+    },
+    {
+      "epoch": 0.9808741600137844,
+      "grad_norm": 0.46299320459365845,
+      "learning_rate": 0.0001,
+      "loss": 1.5098,
+      "step": 8539
+    },
+    {
+      "epoch": 0.9809890299236115,
+      "grad_norm": 0.48221555352211,
+      "learning_rate": 0.0001,
+      "loss": 1.5331,
+      "step": 8540
+    },
+    {
+      "epoch": 0.9811038998334386,
+      "grad_norm": 0.47534555196762085,
+      "learning_rate": 0.0001,
+      "loss": 1.6351,
+      "step": 8541
+    },
+    {
+      "epoch": 0.9812187697432657,
+      "grad_norm": 0.44301724433898926,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 8542
+    },
+    {
+      "epoch": 0.9813336396530928,
+      "grad_norm": 0.5158907771110535,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 8543
+    },
+    {
+      "epoch": 0.98144850956292,
+      "grad_norm": 0.5194290280342102,
+      "learning_rate": 0.0001,
+      "loss": 1.4948,
+      "step": 8544
+    },
+    {
+      "epoch": 0.9815633794727471,
+      "grad_norm": 0.5136659741401672,
+      "learning_rate": 0.0001,
+      "loss": 1.6842,
+      "step": 8545
+    },
+    {
+      "epoch": 0.9816782493825742,
+      "grad_norm": 0.5167720913887024,
+      "learning_rate": 0.0001,
+      "loss": 1.844,
+      "step": 8546
+    },
+    {
+      "epoch": 0.9817931192924013,
+      "grad_norm": 0.45328572392463684,
+      "learning_rate": 0.0001,
+      "loss": 1.513,
+      "step": 8547
+    },
+    {
+      "epoch": 0.9819079892022284,
+      "grad_norm": 0.4788568317890167,
+      "learning_rate": 0.0001,
+      "loss": 1.6795,
+      "step": 8548
+    },
+    {
+      "epoch": 0.9820228591120556,
+      "grad_norm": 0.4906943440437317,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 8549
+    },
+    {
+      "epoch": 0.9821377290218827,
+      "grad_norm": 0.4383232295513153,
+      "learning_rate": 0.0001,
+      "loss": 1.3974,
+      "step": 8550
+    },
+    {
+      "epoch": 0.9822525989317098,
+      "grad_norm": 0.480825275182724,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 8551
+    },
+    {
+      "epoch": 0.9823674688415369,
+      "grad_norm": 0.47409677505493164,
+      "learning_rate": 0.0001,
+      "loss": 1.5389,
+      "step": 8552
+    },
+    {
+      "epoch": 0.982482338751364,
+      "grad_norm": 0.4691976010799408,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 8553
+    },
+    {
+      "epoch": 0.9825972086611912,
+      "grad_norm": 0.48811331391334534,
+      "learning_rate": 0.0001,
+      "loss": 1.3919,
+      "step": 8554
+    },
+    {
+      "epoch": 0.9827120785710183,
+      "grad_norm": 0.5102879405021667,
+      "learning_rate": 0.0001,
+      "loss": 1.7138,
+      "step": 8555
+    },
+    {
+      "epoch": 0.9828269484808454,
+      "grad_norm": 0.4797075092792511,
+      "learning_rate": 0.0001,
+      "loss": 1.4157,
+      "step": 8556
+    },
+    {
+      "epoch": 0.9829418183906725,
+      "grad_norm": 0.4744819104671478,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 8557
+    },
+    {
+      "epoch": 0.9830566883004996,
+      "grad_norm": 0.5046994090080261,
+      "learning_rate": 0.0001,
+      "loss": 1.5037,
+      "step": 8558
+    },
+    {
+      "epoch": 0.9831715582103268,
+      "grad_norm": 0.49901989102363586,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 8559
+    },
+    {
+      "epoch": 0.9832864281201539,
+      "grad_norm": 0.49144455790519714,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 8560
+    },
+    {
+      "epoch": 0.983401298029981,
+      "grad_norm": 0.47116243839263916,
+      "learning_rate": 0.0001,
+      "loss": 1.5554,
+      "step": 8561
+    },
+    {
+      "epoch": 0.9835161679398081,
+      "grad_norm": 0.5342057347297668,
+      "learning_rate": 0.0001,
+      "loss": 1.7248,
+      "step": 8562
+    },
+    {
+      "epoch": 0.9836310378496352,
+      "grad_norm": 0.4303724467754364,
+      "learning_rate": 0.0001,
+      "loss": 1.3374,
+      "step": 8563
+    },
+    {
+      "epoch": 0.9837459077594624,
+      "grad_norm": 0.5251704454421997,
+      "learning_rate": 0.0001,
+      "loss": 1.572,
+      "step": 8564
+    },
+    {
+      "epoch": 0.9838607776692895,
+      "grad_norm": 0.4820622503757477,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 8565
+    },
+    {
+      "epoch": 0.9839756475791166,
+      "grad_norm": 0.4786358177661896,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 8566
+    },
+    {
+      "epoch": 0.9840905174889437,
+      "grad_norm": 0.508850634098053,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 8567
+    },
+    {
+      "epoch": 0.9842053873987708,
+      "grad_norm": 0.5014883875846863,
+      "learning_rate": 0.0001,
+      "loss": 1.7386,
+      "step": 8568
+    },
+    {
+      "epoch": 0.9843202573085981,
+      "grad_norm": 0.48899659514427185,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 8569
+    },
+    {
+      "epoch": 0.9844351272184252,
+      "grad_norm": 0.5107044577598572,
+      "learning_rate": 0.0001,
+      "loss": 1.8143,
+      "step": 8570
+    },
+    {
+      "epoch": 0.9845499971282523,
+      "grad_norm": 0.5331456065177917,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 8571
+    },
+    {
+      "epoch": 0.9846648670380794,
+      "grad_norm": 0.46554502844810486,
+      "learning_rate": 0.0001,
+      "loss": 1.4281,
+      "step": 8572
+    },
+    {
+      "epoch": 0.9847797369479065,
+      "grad_norm": 0.45582613348960876,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 8573
+    },
+    {
+      "epoch": 0.9848946068577337,
+      "grad_norm": 0.5047944188117981,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 8574
+    },
+    {
+      "epoch": 0.9850094767675608,
+      "grad_norm": 0.4860698878765106,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 8575
+    },
+    {
+      "epoch": 0.9851243466773879,
+      "grad_norm": 0.5398543477058411,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 8576
+    },
+    {
+      "epoch": 0.985239216587215,
+      "grad_norm": 0.4853983521461487,
+      "learning_rate": 0.0001,
+      "loss": 1.4137,
+      "step": 8577
+    },
+    {
+      "epoch": 0.9853540864970421,
+      "grad_norm": 0.47747766971588135,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 8578
+    },
+    {
+      "epoch": 0.9854689564068693,
+      "grad_norm": 0.5553230047225952,
+      "learning_rate": 0.0001,
+      "loss": 1.9805,
+      "step": 8579
+    },
+    {
+      "epoch": 0.9855838263166964,
+      "grad_norm": 0.5310031771659851,
+      "learning_rate": 0.0001,
+      "loss": 1.5902,
+      "step": 8580
+    },
+    {
+      "epoch": 0.9856986962265235,
+      "grad_norm": 0.46402546763420105,
+      "learning_rate": 0.0001,
+      "loss": 1.3974,
+      "step": 8581
+    },
+    {
+      "epoch": 0.9858135661363506,
+      "grad_norm": 0.4859161674976349,
+      "learning_rate": 0.0001,
+      "loss": 1.7247,
+      "step": 8582
+    },
+    {
+      "epoch": 0.9859284360461777,
+      "grad_norm": 0.47045204043388367,
+      "learning_rate": 0.0001,
+      "loss": 1.3744,
+      "step": 8583
+    },
+    {
+      "epoch": 0.9860433059560049,
+      "grad_norm": 0.525871992111206,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 8584
+    },
+    {
+      "epoch": 0.986158175865832,
+      "grad_norm": 0.47140932083129883,
+      "learning_rate": 0.0001,
+      "loss": 1.5403,
+      "step": 8585
+    },
+    {
+      "epoch": 0.9862730457756591,
+      "grad_norm": 0.4621630311012268,
+      "learning_rate": 0.0001,
+      "loss": 1.4708,
+      "step": 8586
+    },
+    {
+      "epoch": 0.9863879156854862,
+      "grad_norm": 0.4546113610267639,
+      "learning_rate": 0.0001,
+      "loss": 1.5826,
+      "step": 8587
+    },
+    {
+      "epoch": 0.9865027855953133,
+      "grad_norm": 0.47744020819664,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 8588
+    },
+    {
+      "epoch": 0.9866176555051405,
+      "grad_norm": 0.45344898104667664,
+      "learning_rate": 0.0001,
+      "loss": 1.3902,
+      "step": 8589
+    },
+    {
+      "epoch": 0.9867325254149676,
+      "grad_norm": 0.5098491907119751,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 8590
+    },
+    {
+      "epoch": 0.9868473953247947,
+      "grad_norm": 0.5069968700408936,
+      "learning_rate": 0.0001,
+      "loss": 1.4687,
+      "step": 8591
+    },
+    {
+      "epoch": 0.9869622652346218,
+      "grad_norm": 0.45141881704330444,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 8592
+    },
+    {
+      "epoch": 0.987077135144449,
+      "grad_norm": 0.49101027846336365,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 8593
+    },
+    {
+      "epoch": 0.9871920050542761,
+      "grad_norm": 0.4652627408504486,
+      "learning_rate": 0.0001,
+      "loss": 1.6014,
+      "step": 8594
+    },
+    {
+      "epoch": 0.9873068749641032,
+      "grad_norm": 0.4782879650592804,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 8595
+    },
+    {
+      "epoch": 0.9874217448739303,
+      "grad_norm": 0.4874049425125122,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 8596
+    },
+    {
+      "epoch": 0.9875366147837574,
+      "grad_norm": 0.5173846483230591,
+      "learning_rate": 0.0001,
+      "loss": 1.6132,
+      "step": 8597
+    },
+    {
+      "epoch": 0.9876514846935845,
+      "grad_norm": 0.520719051361084,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 8598
+    },
+    {
+      "epoch": 0.9877663546034117,
+      "grad_norm": 0.5008309483528137,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 8599
+    },
+    {
+      "epoch": 0.9878812245132388,
+      "grad_norm": 0.491559237241745,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 8600
+    },
+    {
+      "epoch": 0.9879960944230659,
+      "grad_norm": 0.5189545750617981,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 8601
+    },
+    {
+      "epoch": 0.988110964332893,
+      "grad_norm": 0.4896295964717865,
+      "learning_rate": 0.0001,
+      "loss": 1.4129,
+      "step": 8602
+    },
+    {
+      "epoch": 0.9882258342427201,
+      "grad_norm": 0.47650331258773804,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 8603
+    },
+    {
+      "epoch": 0.9883407041525473,
+      "grad_norm": 0.46054142713546753,
+      "learning_rate": 0.0001,
+      "loss": 1.4276,
+      "step": 8604
+    },
+    {
+      "epoch": 0.9884555740623744,
+      "grad_norm": 0.5123742818832397,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 8605
+    },
+    {
+      "epoch": 0.9885704439722015,
+      "grad_norm": 0.48375701904296875,
+      "learning_rate": 0.0001,
+      "loss": 1.3923,
+      "step": 8606
+    },
+    {
+      "epoch": 0.9886853138820286,
+      "grad_norm": 0.5176060795783997,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 8607
+    },
+    {
+      "epoch": 0.9888001837918557,
+      "grad_norm": 0.5452961921691895,
+      "learning_rate": 0.0001,
+      "loss": 1.7097,
+      "step": 8608
+    },
+    {
+      "epoch": 0.9889150537016829,
+      "grad_norm": 0.5161251425743103,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 8609
+    },
+    {
+      "epoch": 0.98902992361151,
+      "grad_norm": 0.47460034489631653,
+      "learning_rate": 0.0001,
+      "loss": 1.6462,
+      "step": 8610
+    },
+    {
+      "epoch": 0.9891447935213371,
+      "grad_norm": 0.46245843172073364,
+      "learning_rate": 0.0001,
+      "loss": 1.2765,
+      "step": 8611
+    },
+    {
+      "epoch": 0.9892596634311642,
+      "grad_norm": 0.46704110503196716,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 8612
+    },
+    {
+      "epoch": 0.9893745333409913,
+      "grad_norm": 0.585635781288147,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 8613
+    },
+    {
+      "epoch": 0.9894894032508185,
+      "grad_norm": 0.4674919843673706,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 8614
+    },
+    {
+      "epoch": 0.9896042731606456,
+      "grad_norm": 0.4823983609676361,
+      "learning_rate": 0.0001,
+      "loss": 1.5698,
+      "step": 8615
+    },
+    {
+      "epoch": 0.9897191430704727,
+      "grad_norm": 0.5025752186775208,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 8616
+    },
+    {
+      "epoch": 0.9898340129802998,
+      "grad_norm": 0.5351320505142212,
+      "learning_rate": 0.0001,
+      "loss": 1.7404,
+      "step": 8617
+    },
+    {
+      "epoch": 0.989948882890127,
+      "grad_norm": 0.4923611283302307,
+      "learning_rate": 0.0001,
+      "loss": 1.4485,
+      "step": 8618
+    },
+    {
+      "epoch": 0.9900637527999541,
+      "grad_norm": 0.4848472476005554,
+      "learning_rate": 0.0001,
+      "loss": 1.5072,
+      "step": 8619
+    },
+    {
+      "epoch": 0.9901786227097812,
+      "grad_norm": 0.5281382203102112,
+      "learning_rate": 0.0001,
+      "loss": 1.6906,
+      "step": 8620
+    },
+    {
+      "epoch": 0.9902934926196083,
+      "grad_norm": 0.4649568498134613,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 8621
+    },
+    {
+      "epoch": 0.9904083625294354,
+      "grad_norm": 0.4958634674549103,
+      "learning_rate": 0.0001,
+      "loss": 1.6043,
+      "step": 8622
+    },
+    {
+      "epoch": 0.9905232324392625,
+      "grad_norm": 0.501117467880249,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 8623
+    },
+    {
+      "epoch": 0.9906381023490897,
+      "grad_norm": 0.4949086308479309,
+      "learning_rate": 0.0001,
+      "loss": 1.7568,
+      "step": 8624
+    },
+    {
+      "epoch": 0.9907529722589168,
+      "grad_norm": 0.46853989362716675,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 8625
+    },
+    {
+      "epoch": 0.9908678421687439,
+      "grad_norm": 0.47322335839271545,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 8626
+    },
+    {
+      "epoch": 0.990982712078571,
+      "grad_norm": 0.5503798127174377,
+      "learning_rate": 0.0001,
+      "loss": 1.7705,
+      "step": 8627
+    },
+    {
+      "epoch": 0.9910975819883981,
+      "grad_norm": 0.48998063802719116,
+      "learning_rate": 0.0001,
+      "loss": 1.7953,
+      "step": 8628
+    },
+    {
+      "epoch": 0.9912124518982253,
+      "grad_norm": 0.4801502525806427,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 8629
+    },
+    {
+      "epoch": 0.9913273218080524,
+      "grad_norm": 0.5271478295326233,
+      "learning_rate": 0.0001,
+      "loss": 1.618,
+      "step": 8630
+    },
+    {
+      "epoch": 0.9914421917178795,
+      "grad_norm": 0.501660943031311,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 8631
+    },
+    {
+      "epoch": 0.9915570616277066,
+      "grad_norm": 0.47359898686408997,
+      "learning_rate": 0.0001,
+      "loss": 1.5764,
+      "step": 8632
+    },
+    {
+      "epoch": 0.9916719315375337,
+      "grad_norm": 0.49344560503959656,
+      "learning_rate": 0.0001,
+      "loss": 1.3197,
+      "step": 8633
+    },
+    {
+      "epoch": 0.9917868014473609,
+      "grad_norm": 0.4800715446472168,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 8634
+    },
+    {
+      "epoch": 0.991901671357188,
+      "grad_norm": 0.49181854724884033,
+      "learning_rate": 0.0001,
+      "loss": 1.6261,
+      "step": 8635
+    },
+    {
+      "epoch": 0.9920165412670151,
+      "grad_norm": 0.4867425560951233,
+      "learning_rate": 0.0001,
+      "loss": 1.5423,
+      "step": 8636
+    },
+    {
+      "epoch": 0.9921314111768422,
+      "grad_norm": 0.5553606152534485,
+      "learning_rate": 0.0001,
+      "loss": 1.8589,
+      "step": 8637
+    },
+    {
+      "epoch": 0.9922462810866693,
+      "grad_norm": 0.4396592080593109,
+      "learning_rate": 0.0001,
+      "loss": 1.3746,
+      "step": 8638
+    },
+    {
+      "epoch": 0.9923611509964965,
+      "grad_norm": 0.5199365019798279,
+      "learning_rate": 0.0001,
+      "loss": 1.7913,
+      "step": 8639
+    },
+    {
+      "epoch": 0.9924760209063236,
+      "grad_norm": 0.5626462697982788,
+      "learning_rate": 0.0001,
+      "loss": 1.583,
+      "step": 8640
+    },
+    {
+      "epoch": 0.9925908908161507,
+      "grad_norm": 0.4940207600593567,
+      "learning_rate": 0.0001,
+      "loss": 1.3618,
+      "step": 8641
+    },
+    {
+      "epoch": 0.9927057607259778,
+      "grad_norm": 0.5407283902168274,
+      "learning_rate": 0.0001,
+      "loss": 1.5611,
+      "step": 8642
+    },
+    {
+      "epoch": 0.992820630635805,
+      "grad_norm": 0.5184419751167297,
+      "learning_rate": 0.0001,
+      "loss": 1.7046,
+      "step": 8643
+    },
+    {
+      "epoch": 0.9929355005456321,
+      "grad_norm": 0.4674243628978729,
+      "learning_rate": 0.0001,
+      "loss": 1.4962,
+      "step": 8644
+    },
+    {
+      "epoch": 0.9930503704554592,
+      "grad_norm": 0.519338846206665,
+      "learning_rate": 0.0001,
+      "loss": 1.731,
+      "step": 8645
+    },
+    {
+      "epoch": 0.9931652403652863,
+      "grad_norm": 0.43663740158081055,
+      "learning_rate": 0.0001,
+      "loss": 1.34,
+      "step": 8646
+    },
+    {
+      "epoch": 0.9932801102751134,
+      "grad_norm": 0.4806329309940338,
+      "learning_rate": 0.0001,
+      "loss": 1.322,
+      "step": 8647
+    },
+    {
+      "epoch": 0.9933949801849405,
+      "grad_norm": 0.509090006351471,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 8648
+    },
+    {
+      "epoch": 0.9935098500947677,
+      "grad_norm": 0.4767511487007141,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 8649
+    },
+    {
+      "epoch": 0.9936247200045948,
+      "grad_norm": 0.4973703920841217,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 8650
+    },
+    {
+      "epoch": 0.9937395899144219,
+      "grad_norm": 0.4734041392803192,
+      "learning_rate": 0.0001,
+      "loss": 1.7253,
+      "step": 8651
+    },
+    {
+      "epoch": 0.993854459824249,
+      "grad_norm": 0.44982102513313293,
+      "learning_rate": 0.0001,
+      "loss": 1.5608,
+      "step": 8652
+    },
+    {
+      "epoch": 0.9939693297340761,
+      "grad_norm": 0.4307469129562378,
+      "learning_rate": 0.0001,
+      "loss": 1.3253,
+      "step": 8653
+    },
+    {
+      "epoch": 0.9940841996439033,
+      "grad_norm": 0.5234149694442749,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 8654
+    },
+    {
+      "epoch": 0.9941990695537304,
+      "grad_norm": 0.4377621114253998,
+      "learning_rate": 0.0001,
+      "loss": 1.1823,
+      "step": 8655
+    },
+    {
+      "epoch": 0.9943139394635575,
+      "grad_norm": 0.4438837766647339,
+      "learning_rate": 0.0001,
+      "loss": 1.415,
+      "step": 8656
+    },
+    {
+      "epoch": 0.9944288093733846,
+      "grad_norm": 0.4784393012523651,
+      "learning_rate": 0.0001,
+      "loss": 1.499,
+      "step": 8657
+    },
+    {
+      "epoch": 0.9945436792832117,
+      "grad_norm": 0.468770831823349,
+      "learning_rate": 0.0001,
+      "loss": 1.5585,
+      "step": 8658
+    },
+    {
+      "epoch": 0.9946585491930389,
+      "grad_norm": 0.4884924292564392,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 8659
+    },
+    {
+      "epoch": 0.994773419102866,
+      "grad_norm": 0.5758072733879089,
+      "learning_rate": 0.0001,
+      "loss": 1.6505,
+      "step": 8660
+    },
+    {
+      "epoch": 0.9948882890126931,
+      "grad_norm": 0.5178779363632202,
+      "learning_rate": 0.0001,
+      "loss": 1.7349,
+      "step": 8661
+    },
+    {
+      "epoch": 0.9950031589225202,
+      "grad_norm": 0.49969327449798584,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 8662
+    },
+    {
+      "epoch": 0.9951180288323473,
+      "grad_norm": 0.4617246985435486,
+      "learning_rate": 0.0001,
+      "loss": 1.467,
+      "step": 8663
+    },
+    {
+      "epoch": 0.9952328987421745,
+      "grad_norm": 0.4882105886936188,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 8664
+    },
+    {
+      "epoch": 0.9953477686520016,
+      "grad_norm": 0.47335708141326904,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 8665
+    },
+    {
+      "epoch": 0.9954626385618287,
+      "grad_norm": 0.5231687426567078,
+      "learning_rate": 0.0001,
+      "loss": 1.3034,
+      "step": 8666
+    },
+    {
+      "epoch": 0.9955775084716558,
+      "grad_norm": 0.48819267749786377,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 8667
+    },
+    {
+      "epoch": 0.995692378381483,
+      "grad_norm": 0.4395897686481476,
+      "learning_rate": 0.0001,
+      "loss": 1.349,
+      "step": 8668
+    },
+    {
+      "epoch": 0.9958072482913101,
+      "grad_norm": 0.44600117206573486,
+      "learning_rate": 0.0001,
+      "loss": 1.2248,
+      "step": 8669
+    },
+    {
+      "epoch": 0.9959221182011372,
+      "grad_norm": 0.5204833745956421,
+      "learning_rate": 0.0001,
+      "loss": 1.8804,
+      "step": 8670
+    },
+    {
+      "epoch": 0.9960369881109643,
+      "grad_norm": 0.5219905972480774,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 8671
+    },
+    {
+      "epoch": 0.9961518580207914,
+      "grad_norm": 0.5194002389907837,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 8672
+    },
+    {
+      "epoch": 0.9962667279306185,
+      "grad_norm": 0.4883606433868408,
+      "learning_rate": 0.0001,
+      "loss": 1.5022,
+      "step": 8673
+    },
+    {
+      "epoch": 0.9963815978404457,
+      "grad_norm": 0.5156763792037964,
+      "learning_rate": 0.0001,
+      "loss": 1.7083,
+      "step": 8674
+    },
+    {
+      "epoch": 0.9964964677502728,
+      "grad_norm": 0.4738801121711731,
+      "learning_rate": 0.0001,
+      "loss": 1.4915,
+      "step": 8675
+    },
+    {
+      "epoch": 0.9966113376600999,
+      "grad_norm": 0.4632324278354645,
+      "learning_rate": 0.0001,
+      "loss": 1.702,
+      "step": 8676
+    },
+    {
+      "epoch": 0.996726207569927,
+      "grad_norm": 0.4966660439968109,
+      "learning_rate": 0.0001,
+      "loss": 1.6701,
+      "step": 8677
+    },
+    {
+      "epoch": 0.9968410774797541,
+      "grad_norm": 0.48212072253227234,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 8678
+    },
+    {
+      "epoch": 0.9969559473895813,
+      "grad_norm": 0.498323529958725,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 8679
+    },
+    {
+      "epoch": 0.9970708172994084,
+      "grad_norm": 0.5070053935050964,
+      "learning_rate": 0.0001,
+      "loss": 1.8766,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9971856872092355,
+      "grad_norm": 0.4994966387748718,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 8681
+    },
+    {
+      "epoch": 0.9973005571190626,
+      "grad_norm": 0.5102283954620361,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 8682
+    },
+    {
+      "epoch": 0.9974154270288897,
+      "grad_norm": 0.5465260148048401,
+      "learning_rate": 0.0001,
+      "loss": 1.7164,
+      "step": 8683
+    },
+    {
+      "epoch": 0.9975302969387169,
+      "grad_norm": 0.46429330110549927,
+      "learning_rate": 0.0001,
+      "loss": 1.6374,
+      "step": 8684
+    },
+    {
+      "epoch": 0.997645166848544,
+      "grad_norm": 0.4978303611278534,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 8685
+    },
+    {
+      "epoch": 0.9977600367583711,
+      "grad_norm": 0.4677812457084656,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 8686
+    },
+    {
+      "epoch": 0.9978749066681982,
+      "grad_norm": 0.49235597252845764,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 8687
+    },
+    {
+      "epoch": 0.9979897765780253,
+      "grad_norm": 0.48482832312583923,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 8688
+    },
+    {
+      "epoch": 0.9981046464878525,
+      "grad_norm": 0.5265589356422424,
+      "learning_rate": 0.0001,
+      "loss": 1.7991,
+      "step": 8689
+    },
+    {
+      "epoch": 0.9982195163976796,
+      "grad_norm": 0.47693178057670593,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9983343863075067,
+      "grad_norm": 0.4894190728664398,
+      "learning_rate": 0.0001,
+      "loss": 1.6614,
+      "step": 8691
+    },
+    {
+      "epoch": 0.9984492562173338,
+      "grad_norm": 0.46358221769332886,
+      "learning_rate": 0.0001,
+      "loss": 1.5936,
+      "step": 8692
+    },
+    {
+      "epoch": 0.998564126127161,
+      "grad_norm": 0.45858120918273926,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 8693
+    },
+    {
+      "epoch": 0.9986789960369881,
+      "grad_norm": 0.45518410205841064,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 8694
+    },
+    {
+      "epoch": 0.9987938659468152,
+      "grad_norm": 0.47189682722091675,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 8695
+    },
+    {
+      "epoch": 0.9989087358566423,
+      "grad_norm": 0.5133928060531616,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 8696
+    },
+    {
+      "epoch": 0.9990236057664694,
+      "grad_norm": 0.5051494240760803,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 8697
+    },
+    {
+      "epoch": 0.9991384756762965,
+      "grad_norm": 0.5052590370178223,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 8698
+    },
+    {
+      "epoch": 0.9992533455861237,
+      "grad_norm": 0.4864301085472107,
+      "learning_rate": 0.0001,
+      "loss": 1.4405,
+      "step": 8699
+    },
+    {
+      "epoch": 0.9993682154959508,
+      "grad_norm": 0.49039098620414734,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 8700
+    },
+    {
+      "epoch": 0.9994830854057779,
+      "grad_norm": 0.5003101229667664,
+      "learning_rate": 0.0001,
+      "loss": 1.729,
+      "step": 8701
+    },
+    {
+      "epoch": 0.999597955315605,
+      "grad_norm": 0.4698851704597473,
+      "learning_rate": 0.0001,
+      "loss": 1.5738,
+      "step": 8702
+    },
+    {
+      "epoch": 0.9997128252254321,
+      "grad_norm": 0.5079613327980042,
+      "learning_rate": 0.0001,
+      "loss": 1.6768,
+      "step": 8703
+    },
+    {
+      "epoch": 0.9998276951352593,
+      "grad_norm": 0.5211082100868225,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 8704
+    },
+    {
+      "epoch": 0.9999425650450864,
+      "grad_norm": 0.45943862199783325,
+      "learning_rate": 0.0001,
+      "loss": 1.5753,
+      "step": 8705
+    },
+    {
+      "epoch": 1.0000574349549136,
+      "grad_norm": 0.45664501190185547,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 8706
+    },
+    {
+      "epoch": 1.0001723048647406,
+      "grad_norm": 0.4673193097114563,
+      "learning_rate": 0.0001,
+      "loss": 1.6225,
+      "step": 8707
+    },
+    {
+      "epoch": 1.0002871747745679,
+      "grad_norm": 0.46662116050720215,
+      "learning_rate": 0.0001,
+      "loss": 1.3963,
+      "step": 8708
+    },
+    {
+      "epoch": 1.0004020446843949,
+      "grad_norm": 0.543308675289154,
+      "learning_rate": 0.0001,
+      "loss": 1.2983,
+      "step": 8709
+    },
+    {
+      "epoch": 1.000516914594222,
+      "grad_norm": 0.49053019285202026,
+      "learning_rate": 0.0001,
+      "loss": 1.417,
+      "step": 8710
+    },
+    {
+      "epoch": 1.000631784504049,
+      "grad_norm": 0.504199743270874,
+      "learning_rate": 0.0001,
+      "loss": 1.4342,
+      "step": 8711
+    },
+    {
+      "epoch": 1.0007466544138763,
+      "grad_norm": 0.5157631635665894,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 8712
+    },
+    {
+      "epoch": 1.0008615243237033,
+      "grad_norm": 0.5185043215751648,
+      "learning_rate": 0.0001,
+      "loss": 1.401,
+      "step": 8713
+    },
+    {
+      "epoch": 1.0009763942335306,
+      "grad_norm": 0.5186070203781128,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 8714
+    },
+    {
+      "epoch": 1.0010912641433576,
+      "grad_norm": 0.5666443705558777,
+      "learning_rate": 0.0001,
+      "loss": 1.579,
+      "step": 8715
+    },
+    {
+      "epoch": 1.0012061340531848,
+      "grad_norm": 0.6023586988449097,
+      "learning_rate": 0.0001,
+      "loss": 1.4872,
+      "step": 8716
+    },
+    {
+      "epoch": 1.0013210039630118,
+      "grad_norm": 0.504416823387146,
+      "learning_rate": 0.0001,
+      "loss": 1.4496,
+      "step": 8717
+    },
+    {
+      "epoch": 1.001435873872839,
+      "grad_norm": 0.563830554485321,
+      "learning_rate": 0.0001,
+      "loss": 1.2972,
+      "step": 8718
+    },
+    {
+      "epoch": 1.001550743782666,
+      "grad_norm": 0.5681375861167908,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 8719
+    },
+    {
+      "epoch": 1.0016656136924933,
+      "grad_norm": 0.5679755210876465,
+      "learning_rate": 0.0001,
+      "loss": 1.4425,
+      "step": 8720
+    },
+    {
+      "epoch": 1.0017804836023203,
+      "grad_norm": 0.5443891286849976,
+      "learning_rate": 0.0001,
+      "loss": 1.6377,
+      "step": 8721
+    },
+    {
+      "epoch": 1.0018953535121475,
+      "grad_norm": 0.5426947474479675,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 8722
+    },
+    {
+      "epoch": 1.0020102234219745,
+      "grad_norm": 0.5179216861724854,
+      "learning_rate": 0.0001,
+      "loss": 1.463,
+      "step": 8723
+    },
+    {
+      "epoch": 1.0021250933318018,
+      "grad_norm": 0.5044360160827637,
+      "learning_rate": 0.0001,
+      "loss": 1.4617,
+      "step": 8724
+    },
+    {
+      "epoch": 1.0022399632416288,
+      "grad_norm": 0.5553032159805298,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 8725
+    },
+    {
+      "epoch": 1.002354833151456,
+      "grad_norm": 0.5114759206771851,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 8726
+    },
+    {
+      "epoch": 1.002469703061283,
+      "grad_norm": 0.5439996719360352,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 8727
+    },
+    {
+      "epoch": 1.0025845729711103,
+      "grad_norm": 0.5239347815513611,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 8728
+    },
+    {
+      "epoch": 1.0026994428809373,
+      "grad_norm": 0.49250611662864685,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 8729
+    },
+    {
+      "epoch": 1.0028143127907645,
+      "grad_norm": 0.5004076361656189,
+      "learning_rate": 0.0001,
+      "loss": 1.435,
+      "step": 8730
+    },
+    {
+      "epoch": 1.0029291827005915,
+      "grad_norm": 0.519535481929779,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 8731
+    },
+    {
+      "epoch": 1.0030440526104187,
+      "grad_norm": 0.536308228969574,
+      "learning_rate": 0.0001,
+      "loss": 1.5055,
+      "step": 8732
+    },
+    {
+      "epoch": 1.0031589225202457,
+      "grad_norm": 0.5679396986961365,
+      "learning_rate": 0.0001,
+      "loss": 1.8589,
+      "step": 8733
+    },
+    {
+      "epoch": 1.003273792430073,
+      "grad_norm": 0.5036899447441101,
+      "learning_rate": 0.0001,
+      "loss": 1.4946,
+      "step": 8734
+    },
+    {
+      "epoch": 1.0033886623399,
+      "grad_norm": 0.4876912236213684,
+      "learning_rate": 0.0001,
+      "loss": 1.4041,
+      "step": 8735
+    },
+    {
+      "epoch": 1.0035035322497272,
+      "grad_norm": 0.4986098110675812,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 8736
+    },
+    {
+      "epoch": 1.0036184021595542,
+      "grad_norm": 0.4937874674797058,
+      "learning_rate": 0.0001,
+      "loss": 1.4384,
+      "step": 8737
+    },
+    {
+      "epoch": 1.0037332720693815,
+      "grad_norm": 0.5576295852661133,
+      "learning_rate": 0.0001,
+      "loss": 1.4278,
+      "step": 8738
+    },
+    {
+      "epoch": 1.0038481419792085,
+      "grad_norm": 0.5405202507972717,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 8739
+    },
+    {
+      "epoch": 1.0039630118890357,
+      "grad_norm": 0.5165756344795227,
+      "learning_rate": 0.0001,
+      "loss": 1.4411,
+      "step": 8740
+    },
+    {
+      "epoch": 1.0040778817988627,
+      "grad_norm": 0.4816116988658905,
+      "learning_rate": 0.0001,
+      "loss": 1.2506,
+      "step": 8741
+    },
+    {
+      "epoch": 1.00419275170869,
+      "grad_norm": 0.5027560591697693,
+      "learning_rate": 0.0001,
+      "loss": 1.4022,
+      "step": 8742
+    },
+    {
+      "epoch": 1.004307621618517,
+      "grad_norm": 0.5379549264907837,
+      "learning_rate": 0.0001,
+      "loss": 1.6524,
+      "step": 8743
+    },
+    {
+      "epoch": 1.0044224915283442,
+      "grad_norm": 0.6135854721069336,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 8744
+    },
+    {
+      "epoch": 1.0045373614381712,
+      "grad_norm": 0.4991096556186676,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 8745
+    },
+    {
+      "epoch": 1.0046522313479984,
+      "grad_norm": 0.5324615836143494,
+      "learning_rate": 0.0001,
+      "loss": 1.3607,
+      "step": 8746
+    },
+    {
+      "epoch": 1.0047671012578254,
+      "grad_norm": 0.5299088358879089,
+      "learning_rate": 0.0001,
+      "loss": 1.4187,
+      "step": 8747
+    },
+    {
+      "epoch": 1.0048819711676527,
+      "grad_norm": 0.4894624650478363,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 8748
+    },
+    {
+      "epoch": 1.0049968410774797,
+      "grad_norm": 0.5205270648002625,
+      "learning_rate": 0.0001,
+      "loss": 1.4387,
+      "step": 8749
+    },
+    {
+      "epoch": 1.005111710987307,
+      "grad_norm": 0.4865090250968933,
+      "learning_rate": 0.0001,
+      "loss": 1.478,
+      "step": 8750
+    },
+    {
+      "epoch": 1.005226580897134,
+      "grad_norm": 0.5512129664421082,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 8751
+    },
+    {
+      "epoch": 1.0053414508069611,
+      "grad_norm": 0.484320729970932,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 8752
+    },
+    {
+      "epoch": 1.0054563207167881,
+      "grad_norm": 0.5077235698699951,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 8753
+    },
+    {
+      "epoch": 1.0055711906266154,
+      "grad_norm": 0.5155119895935059,
+      "learning_rate": 0.0001,
+      "loss": 1.4048,
+      "step": 8754
+    },
+    {
+      "epoch": 1.0056860605364424,
+      "grad_norm": 0.54921954870224,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 8755
+    },
+    {
+      "epoch": 1.0058009304462696,
+      "grad_norm": 0.4844646751880646,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 8756
+    },
+    {
+      "epoch": 1.0059158003560966,
+      "grad_norm": 0.47289541363716125,
+      "learning_rate": 0.0001,
+      "loss": 1.3917,
+      "step": 8757
+    },
+    {
+      "epoch": 1.0060306702659239,
+      "grad_norm": 0.5090060234069824,
+      "learning_rate": 0.0001,
+      "loss": 1.4632,
+      "step": 8758
+    },
+    {
+      "epoch": 1.0061455401757509,
+      "grad_norm": 0.5446418523788452,
+      "learning_rate": 0.0001,
+      "loss": 1.4641,
+      "step": 8759
+    },
+    {
+      "epoch": 1.006260410085578,
+      "grad_norm": 0.4822421669960022,
+      "learning_rate": 0.0001,
+      "loss": 1.3481,
+      "step": 8760
+    },
+    {
+      "epoch": 1.006375279995405,
+      "grad_norm": 0.5040774345397949,
+      "learning_rate": 0.0001,
+      "loss": 1.5635,
+      "step": 8761
+    },
+    {
+      "epoch": 1.0064901499052323,
+      "grad_norm": 0.5044432282447815,
+      "learning_rate": 0.0001,
+      "loss": 1.3254,
+      "step": 8762
+    },
+    {
+      "epoch": 1.0066050198150593,
+      "grad_norm": 0.5607876181602478,
+      "learning_rate": 0.0001,
+      "loss": 1.5483,
+      "step": 8763
+    },
+    {
+      "epoch": 1.0067198897248866,
+      "grad_norm": 0.5420013666152954,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 8764
+    },
+    {
+      "epoch": 1.0068347596347136,
+      "grad_norm": 0.5551548004150391,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 8765
+    },
+    {
+      "epoch": 1.0069496295445408,
+      "grad_norm": 0.48607149720191956,
+      "learning_rate": 0.0001,
+      "loss": 1.4455,
+      "step": 8766
+    },
+    {
+      "epoch": 1.0070644994543678,
+      "grad_norm": 0.5381131172180176,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 8767
+    },
+    {
+      "epoch": 1.007179369364195,
+      "grad_norm": 0.5239553451538086,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 8768
+    },
+    {
+      "epoch": 1.007294239274022,
+      "grad_norm": 0.5344405174255371,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 8769
+    },
+    {
+      "epoch": 1.0074091091838493,
+      "grad_norm": 0.4784035384654999,
+      "learning_rate": 0.0001,
+      "loss": 1.3902,
+      "step": 8770
+    },
+    {
+      "epoch": 1.0075239790936763,
+      "grad_norm": 0.49824607372283936,
+      "learning_rate": 0.0001,
+      "loss": 1.4876,
+      "step": 8771
+    },
+    {
+      "epoch": 1.0076388490035035,
+      "grad_norm": 0.4996454119682312,
+      "learning_rate": 0.0001,
+      "loss": 1.4361,
+      "step": 8772
+    },
+    {
+      "epoch": 1.0077537189133305,
+      "grad_norm": 0.5669626593589783,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 8773
+    },
+    {
+      "epoch": 1.0078685888231578,
+      "grad_norm": 0.5511415600776672,
+      "learning_rate": 0.0001,
+      "loss": 1.4873,
+      "step": 8774
+    },
+    {
+      "epoch": 1.0079834587329848,
+      "grad_norm": 0.5880183577537537,
+      "learning_rate": 0.0001,
+      "loss": 1.6994,
+      "step": 8775
+    },
+    {
+      "epoch": 1.008098328642812,
+      "grad_norm": 0.47547656297683716,
+      "learning_rate": 0.0001,
+      "loss": 1.3952,
+      "step": 8776
+    },
+    {
+      "epoch": 1.0082131985526392,
+      "grad_norm": 0.49340614676475525,
+      "learning_rate": 0.0001,
+      "loss": 1.4364,
+      "step": 8777
+    },
+    {
+      "epoch": 1.0083280684624663,
+      "grad_norm": 0.5298573970794678,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 8778
+    },
+    {
+      "epoch": 1.0084429383722935,
+      "grad_norm": 0.5538324117660522,
+      "learning_rate": 0.0001,
+      "loss": 1.6803,
+      "step": 8779
+    },
+    {
+      "epoch": 1.0085578082821205,
+      "grad_norm": 0.5182627439498901,
+      "learning_rate": 0.0001,
+      "loss": 1.5341,
+      "step": 8780
+    },
+    {
+      "epoch": 1.0086726781919477,
+      "grad_norm": 0.516372561454773,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 8781
+    },
+    {
+      "epoch": 1.0087875481017747,
+      "grad_norm": 0.47818052768707275,
+      "learning_rate": 0.0001,
+      "loss": 1.4983,
+      "step": 8782
+    },
+    {
+      "epoch": 1.008902418011602,
+      "grad_norm": 0.5226326584815979,
+      "learning_rate": 0.0001,
+      "loss": 1.3928,
+      "step": 8783
+    },
+    {
+      "epoch": 1.009017287921429,
+      "grad_norm": 0.4940192401409149,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 8784
+    },
+    {
+      "epoch": 1.0091321578312562,
+      "grad_norm": 0.5674400329589844,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 8785
+    },
+    {
+      "epoch": 1.0092470277410832,
+      "grad_norm": 0.4928450286388397,
+      "learning_rate": 0.0001,
+      "loss": 1.4894,
+      "step": 8786
+    },
+    {
+      "epoch": 1.0093618976509104,
+      "grad_norm": 0.48630955815315247,
+      "learning_rate": 0.0001,
+      "loss": 1.3444,
+      "step": 8787
+    },
+    {
+      "epoch": 1.0094767675607375,
+      "grad_norm": 0.4933887720108032,
+      "learning_rate": 0.0001,
+      "loss": 1.414,
+      "step": 8788
+    },
+    {
+      "epoch": 1.0095916374705647,
+      "grad_norm": 0.587792158126831,
+      "learning_rate": 0.0001,
+      "loss": 1.3733,
+      "step": 8789
+    },
+    {
+      "epoch": 1.0097065073803917,
+      "grad_norm": 0.5260916352272034,
+      "learning_rate": 0.0001,
+      "loss": 1.3751,
+      "step": 8790
+    },
+    {
+      "epoch": 1.009821377290219,
+      "grad_norm": 0.55689936876297,
+      "learning_rate": 0.0001,
+      "loss": 1.501,
+      "step": 8791
+    },
+    {
+      "epoch": 1.009936247200046,
+      "grad_norm": 0.5287038087844849,
+      "learning_rate": 0.0001,
+      "loss": 1.3097,
+      "step": 8792
+    },
+    {
+      "epoch": 1.0100511171098732,
+      "grad_norm": 0.521734356880188,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 8793
+    },
+    {
+      "epoch": 1.0101659870197002,
+      "grad_norm": 0.5244892239570618,
+      "learning_rate": 0.0001,
+      "loss": 1.3727,
+      "step": 8794
+    },
+    {
+      "epoch": 1.0102808569295274,
+      "grad_norm": 0.519815981388092,
+      "learning_rate": 0.0001,
+      "loss": 1.3018,
+      "step": 8795
+    },
+    {
+      "epoch": 1.0103957268393544,
+      "grad_norm": 0.5823325514793396,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 8796
+    },
+    {
+      "epoch": 1.0105105967491816,
+      "grad_norm": 0.51022869348526,
+      "learning_rate": 0.0001,
+      "loss": 1.2856,
+      "step": 8797
+    },
+    {
+      "epoch": 1.0106254666590087,
+      "grad_norm": 0.530922532081604,
+      "learning_rate": 0.0001,
+      "loss": 1.7368,
+      "step": 8798
+    },
+    {
+      "epoch": 1.0107403365688359,
+      "grad_norm": 0.48666539788246155,
+      "learning_rate": 0.0001,
+      "loss": 1.3965,
+      "step": 8799
+    },
+    {
+      "epoch": 1.010855206478663,
+      "grad_norm": 0.5583677291870117,
+      "learning_rate": 0.0001,
+      "loss": 1.499,
+      "step": 8800
+    },
+    {
+      "epoch": 1.0109700763884901,
+      "grad_norm": 0.5303579568862915,
+      "learning_rate": 0.0001,
+      "loss": 1.5556,
+      "step": 8801
+    },
+    {
+      "epoch": 1.0110849462983171,
+      "grad_norm": 0.5240969657897949,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 8802
+    },
+    {
+      "epoch": 1.0111998162081444,
+      "grad_norm": 0.5126948356628418,
+      "learning_rate": 0.0001,
+      "loss": 1.4357,
+      "step": 8803
+    },
+    {
+      "epoch": 1.0113146861179714,
+      "grad_norm": 0.5634292960166931,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 8804
+    },
+    {
+      "epoch": 1.0114295560277986,
+      "grad_norm": 0.5226359963417053,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 8805
+    },
+    {
+      "epoch": 1.0115444259376256,
+      "grad_norm": 0.4996303617954254,
+      "learning_rate": 0.0001,
+      "loss": 1.3072,
+      "step": 8806
+    },
+    {
+      "epoch": 1.0116592958474528,
+      "grad_norm": 0.49595144391059875,
+      "learning_rate": 0.0001,
+      "loss": 1.5657,
+      "step": 8807
+    },
+    {
+      "epoch": 1.0117741657572799,
+      "grad_norm": 0.5020740628242493,
+      "learning_rate": 0.0001,
+      "loss": 1.3781,
+      "step": 8808
+    },
+    {
+      "epoch": 1.011889035667107,
+      "grad_norm": 0.49597975611686707,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 8809
+    },
+    {
+      "epoch": 1.012003905576934,
+      "grad_norm": 0.5139532685279846,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 8810
+    },
+    {
+      "epoch": 1.0121187754867613,
+      "grad_norm": 0.4895060658454895,
+      "learning_rate": 0.0001,
+      "loss": 1.3705,
+      "step": 8811
+    },
+    {
+      "epoch": 1.0122336453965883,
+      "grad_norm": 0.47956687211990356,
+      "learning_rate": 0.0001,
+      "loss": 1.3134,
+      "step": 8812
+    },
+    {
+      "epoch": 1.0123485153064156,
+      "grad_norm": 0.4845159649848938,
+      "learning_rate": 0.0001,
+      "loss": 1.3541,
+      "step": 8813
+    },
+    {
+      "epoch": 1.0124633852162426,
+      "grad_norm": 0.5296355485916138,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 8814
+    },
+    {
+      "epoch": 1.0125782551260698,
+      "grad_norm": 0.5246522426605225,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 8815
+    },
+    {
+      "epoch": 1.0126931250358968,
+      "grad_norm": 0.5685099959373474,
+      "learning_rate": 0.0001,
+      "loss": 1.4409,
+      "step": 8816
+    },
+    {
+      "epoch": 1.012807994945724,
+      "grad_norm": 0.47896215319633484,
+      "learning_rate": 0.0001,
+      "loss": 1.4345,
+      "step": 8817
+    },
+    {
+      "epoch": 1.012922864855551,
+      "grad_norm": 0.487551748752594,
+      "learning_rate": 0.0001,
+      "loss": 1.1566,
+      "step": 8818
+    },
+    {
+      "epoch": 1.0130377347653783,
+      "grad_norm": 0.5409836173057556,
+      "learning_rate": 0.0001,
+      "loss": 1.6797,
+      "step": 8819
+    },
+    {
+      "epoch": 1.0131526046752053,
+      "grad_norm": 0.5281198620796204,
+      "learning_rate": 0.0001,
+      "loss": 1.5419,
+      "step": 8820
+    },
+    {
+      "epoch": 1.0132674745850325,
+      "grad_norm": 0.5546361804008484,
+      "learning_rate": 0.0001,
+      "loss": 1.5202,
+      "step": 8821
+    },
+    {
+      "epoch": 1.0133823444948595,
+      "grad_norm": 0.5375561118125916,
+      "learning_rate": 0.0001,
+      "loss": 1.4132,
+      "step": 8822
+    },
+    {
+      "epoch": 1.0134972144046868,
+      "grad_norm": 0.5230275988578796,
+      "learning_rate": 0.0001,
+      "loss": 1.242,
+      "step": 8823
+    },
+    {
+      "epoch": 1.0136120843145138,
+      "grad_norm": 0.5615382194519043,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 8824
+    },
+    {
+      "epoch": 1.013726954224341,
+      "grad_norm": 0.49817347526550293,
+      "learning_rate": 0.0001,
+      "loss": 1.2932,
+      "step": 8825
+    },
+    {
+      "epoch": 1.013841824134168,
+      "grad_norm": 0.5396572947502136,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 8826
+    },
+    {
+      "epoch": 1.0139566940439952,
+      "grad_norm": 0.5678941607475281,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 8827
+    },
+    {
+      "epoch": 1.0140715639538223,
+      "grad_norm": 0.5782578587532043,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 8828
+    },
+    {
+      "epoch": 1.0141864338636495,
+      "grad_norm": 0.502582311630249,
+      "learning_rate": 0.0001,
+      "loss": 1.614,
+      "step": 8829
+    },
+    {
+      "epoch": 1.0143013037734765,
+      "grad_norm": 0.6554863452911377,
+      "learning_rate": 0.0001,
+      "loss": 1.4321,
+      "step": 8830
+    },
+    {
+      "epoch": 1.0144161736833037,
+      "grad_norm": 0.5245055556297302,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 8831
+    },
+    {
+      "epoch": 1.0145310435931307,
+      "grad_norm": 0.4950437843799591,
+      "learning_rate": 0.0001,
+      "loss": 1.404,
+      "step": 8832
+    },
+    {
+      "epoch": 1.014645913502958,
+      "grad_norm": 0.48556363582611084,
+      "learning_rate": 0.0001,
+      "loss": 1.3682,
+      "step": 8833
+    },
+    {
+      "epoch": 1.014760783412785,
+      "grad_norm": 0.5288364291191101,
+      "learning_rate": 0.0001,
+      "loss": 1.3036,
+      "step": 8834
+    },
+    {
+      "epoch": 1.0148756533226122,
+      "grad_norm": 0.5161809325218201,
+      "learning_rate": 0.0001,
+      "loss": 1.4239,
+      "step": 8835
+    },
+    {
+      "epoch": 1.0149905232324392,
+      "grad_norm": 0.6864306330680847,
+      "learning_rate": 0.0001,
+      "loss": 1.4739,
+      "step": 8836
+    },
+    {
+      "epoch": 1.0151053931422664,
+      "grad_norm": 0.5075218081474304,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 8837
+    },
+    {
+      "epoch": 1.0152202630520935,
+      "grad_norm": 0.5431599617004395,
+      "learning_rate": 0.0001,
+      "loss": 1.522,
+      "step": 8838
+    },
+    {
+      "epoch": 1.0153351329619207,
+      "grad_norm": 0.49773699045181274,
+      "learning_rate": 0.0001,
+      "loss": 1.4085,
+      "step": 8839
+    },
+    {
+      "epoch": 1.0154500028717477,
+      "grad_norm": 0.5445903539657593,
+      "learning_rate": 0.0001,
+      "loss": 1.6883,
+      "step": 8840
+    },
+    {
+      "epoch": 1.015564872781575,
+      "grad_norm": 0.5013106465339661,
+      "learning_rate": 0.0001,
+      "loss": 1.489,
+      "step": 8841
+    },
+    {
+      "epoch": 1.015679742691402,
+      "grad_norm": 0.48422935605049133,
+      "learning_rate": 0.0001,
+      "loss": 1.4696,
+      "step": 8842
+    },
+    {
+      "epoch": 1.0157946126012292,
+      "grad_norm": 0.507689893245697,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 8843
+    },
+    {
+      "epoch": 1.0159094825110562,
+      "grad_norm": 0.48591098189353943,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 8844
+    },
+    {
+      "epoch": 1.0160243524208834,
+      "grad_norm": 0.5136837363243103,
+      "learning_rate": 0.0001,
+      "loss": 1.3958,
+      "step": 8845
+    },
+    {
+      "epoch": 1.0161392223307104,
+      "grad_norm": 0.5529723167419434,
+      "learning_rate": 0.0001,
+      "loss": 1.6141,
+      "step": 8846
+    },
+    {
+      "epoch": 1.0162540922405376,
+      "grad_norm": 0.5329626202583313,
+      "learning_rate": 0.0001,
+      "loss": 1.5841,
+      "step": 8847
+    },
+    {
+      "epoch": 1.0163689621503647,
+      "grad_norm": 0.5492246747016907,
+      "learning_rate": 0.0001,
+      "loss": 1.322,
+      "step": 8848
+    },
+    {
+      "epoch": 1.0164838320601919,
+      "grad_norm": 0.5101956725120544,
+      "learning_rate": 0.0001,
+      "loss": 1.4037,
+      "step": 8849
+    },
+    {
+      "epoch": 1.016598701970019,
+      "grad_norm": 0.5277436971664429,
+      "learning_rate": 0.0001,
+      "loss": 1.5163,
+      "step": 8850
+    },
+    {
+      "epoch": 1.0167135718798461,
+      "grad_norm": 0.5436544418334961,
+      "learning_rate": 0.0001,
+      "loss": 1.4159,
+      "step": 8851
+    },
+    {
+      "epoch": 1.0168284417896731,
+      "grad_norm": 0.5090106725692749,
+      "learning_rate": 0.0001,
+      "loss": 1.4402,
+      "step": 8852
+    },
+    {
+      "epoch": 1.0169433116995004,
+      "grad_norm": 0.5208089351654053,
+      "learning_rate": 0.0001,
+      "loss": 1.3969,
+      "step": 8853
+    },
+    {
+      "epoch": 1.0170581816093274,
+      "grad_norm": 0.506205141544342,
+      "learning_rate": 0.0001,
+      "loss": 1.2677,
+      "step": 8854
+    },
+    {
+      "epoch": 1.0171730515191546,
+      "grad_norm": 0.4834211468696594,
+      "learning_rate": 0.0001,
+      "loss": 1.4207,
+      "step": 8855
+    },
+    {
+      "epoch": 1.0172879214289816,
+      "grad_norm": 0.4916054606437683,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 8856
+    },
+    {
+      "epoch": 1.0174027913388088,
+      "grad_norm": 0.5485023856163025,
+      "learning_rate": 0.0001,
+      "loss": 1.4735,
+      "step": 8857
+    },
+    {
+      "epoch": 1.0175176612486359,
+      "grad_norm": 0.5115214586257935,
+      "learning_rate": 0.0001,
+      "loss": 1.4413,
+      "step": 8858
+    },
+    {
+      "epoch": 1.017632531158463,
+      "grad_norm": 0.526879608631134,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 8859
+    },
+    {
+      "epoch": 1.01774740106829,
+      "grad_norm": 0.547588050365448,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 8860
+    },
+    {
+      "epoch": 1.0178622709781173,
+      "grad_norm": 0.4805614948272705,
+      "learning_rate": 0.0001,
+      "loss": 1.2985,
+      "step": 8861
+    },
+    {
+      "epoch": 1.0179771408879443,
+      "grad_norm": 0.5069999098777771,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 8862
+    },
+    {
+      "epoch": 1.0180920107977716,
+      "grad_norm": 0.4949933588504791,
+      "learning_rate": 0.0001,
+      "loss": 1.483,
+      "step": 8863
+    },
+    {
+      "epoch": 1.0182068807075986,
+      "grad_norm": 0.4933357238769531,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 8864
+    },
+    {
+      "epoch": 1.0183217506174258,
+      "grad_norm": 0.5064468383789062,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 8865
+    },
+    {
+      "epoch": 1.0184366205272528,
+      "grad_norm": 0.5541262030601501,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 8866
+    },
+    {
+      "epoch": 1.01855149043708,
+      "grad_norm": 0.5688692331314087,
+      "learning_rate": 0.0001,
+      "loss": 1.565,
+      "step": 8867
+    },
+    {
+      "epoch": 1.018666360346907,
+      "grad_norm": 0.5120310187339783,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 8868
+    },
+    {
+      "epoch": 1.0187812302567343,
+      "grad_norm": 0.5274046063423157,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 8869
+    },
+    {
+      "epoch": 1.0188961001665613,
+      "grad_norm": 0.5022189021110535,
+      "learning_rate": 0.0001,
+      "loss": 1.4302,
+      "step": 8870
+    },
+    {
+      "epoch": 1.0190109700763885,
+      "grad_norm": 0.5581625699996948,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 8871
+    },
+    {
+      "epoch": 1.0191258399862155,
+      "grad_norm": 0.535330593585968,
+      "learning_rate": 0.0001,
+      "loss": 1.4295,
+      "step": 8872
+    },
+    {
+      "epoch": 1.0192407098960428,
+      "grad_norm": 0.4544619023799896,
+      "learning_rate": 0.0001,
+      "loss": 1.3384,
+      "step": 8873
+    },
+    {
+      "epoch": 1.0193555798058698,
+      "grad_norm": 0.6101300716400146,
+      "learning_rate": 0.0001,
+      "loss": 1.3963,
+      "step": 8874
+    },
+    {
+      "epoch": 1.019470449715697,
+      "grad_norm": 0.5120131969451904,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 8875
+    },
+    {
+      "epoch": 1.019585319625524,
+      "grad_norm": 0.49319207668304443,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 8876
+    },
+    {
+      "epoch": 1.0197001895353512,
+      "grad_norm": 0.5753055810928345,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 8877
+    },
+    {
+      "epoch": 1.0198150594451783,
+      "grad_norm": 0.5088034868240356,
+      "learning_rate": 0.0001,
+      "loss": 1.3886,
+      "step": 8878
+    },
+    {
+      "epoch": 1.0199299293550055,
+      "grad_norm": 0.5577898621559143,
+      "learning_rate": 0.0001,
+      "loss": 1.4367,
+      "step": 8879
+    },
+    {
+      "epoch": 1.0200447992648325,
+      "grad_norm": 0.5172377228736877,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 8880
+    },
+    {
+      "epoch": 1.0201596691746597,
+      "grad_norm": 0.49840047955513,
+      "learning_rate": 0.0001,
+      "loss": 1.3737,
+      "step": 8881
+    },
+    {
+      "epoch": 1.0202745390844867,
+      "grad_norm": 0.5156157612800598,
+      "learning_rate": 0.0001,
+      "loss": 1.4591,
+      "step": 8882
+    },
+    {
+      "epoch": 1.020389408994314,
+      "grad_norm": 0.4850468337535858,
+      "learning_rate": 0.0001,
+      "loss": 1.4474,
+      "step": 8883
+    },
+    {
+      "epoch": 1.020504278904141,
+      "grad_norm": 0.5439092516899109,
+      "learning_rate": 0.0001,
+      "loss": 1.3706,
+      "step": 8884
+    },
+    {
+      "epoch": 1.0206191488139682,
+      "grad_norm": 0.5357840657234192,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 8885
+    },
+    {
+      "epoch": 1.0207340187237952,
+      "grad_norm": 0.5319818258285522,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 8886
+    },
+    {
+      "epoch": 1.0208488886336224,
+      "grad_norm": 0.52201247215271,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 8887
+    },
+    {
+      "epoch": 1.0209637585434495,
+      "grad_norm": 0.4789310097694397,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 8888
+    },
+    {
+      "epoch": 1.0210786284532767,
+      "grad_norm": 0.5354008078575134,
+      "learning_rate": 0.0001,
+      "loss": 1.6706,
+      "step": 8889
+    },
+    {
+      "epoch": 1.0211934983631037,
+      "grad_norm": 0.5723029971122742,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 8890
+    },
+    {
+      "epoch": 1.021308368272931,
+      "grad_norm": 0.4998638927936554,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 8891
+    },
+    {
+      "epoch": 1.021423238182758,
+      "grad_norm": 0.5805894732475281,
+      "learning_rate": 0.0001,
+      "loss": 1.7066,
+      "step": 8892
+    },
+    {
+      "epoch": 1.0215381080925852,
+      "grad_norm": 0.6122562885284424,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 8893
+    },
+    {
+      "epoch": 1.0216529780024122,
+      "grad_norm": 0.5063309073448181,
+      "learning_rate": 0.0001,
+      "loss": 1.4165,
+      "step": 8894
+    },
+    {
+      "epoch": 1.0217678479122394,
+      "grad_norm": 0.5316261053085327,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 8895
+    },
+    {
+      "epoch": 1.0218827178220664,
+      "grad_norm": 0.5456799268722534,
+      "learning_rate": 0.0001,
+      "loss": 1.3952,
+      "step": 8896
+    },
+    {
+      "epoch": 1.0219975877318936,
+      "grad_norm": 0.5277631878852844,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 8897
+    },
+    {
+      "epoch": 1.0221124576417207,
+      "grad_norm": 0.5488672852516174,
+      "learning_rate": 0.0001,
+      "loss": 1.4201,
+      "step": 8898
+    },
+    {
+      "epoch": 1.0222273275515479,
+      "grad_norm": 0.5567140579223633,
+      "learning_rate": 0.0001,
+      "loss": 1.5074,
+      "step": 8899
+    },
+    {
+      "epoch": 1.022342197461375,
+      "grad_norm": 0.5117017030715942,
+      "learning_rate": 0.0001,
+      "loss": 1.5041,
+      "step": 8900
+    },
+    {
+      "epoch": 1.0224570673712021,
+      "grad_norm": 0.569999098777771,
+      "learning_rate": 0.0001,
+      "loss": 1.4689,
+      "step": 8901
+    },
+    {
+      "epoch": 1.0225719372810291,
+      "grad_norm": 0.5409071445465088,
+      "learning_rate": 0.0001,
+      "loss": 1.3848,
+      "step": 8902
+    },
+    {
+      "epoch": 1.0226868071908564,
+      "grad_norm": 0.5431987047195435,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 8903
+    },
+    {
+      "epoch": 1.0228016771006834,
+      "grad_norm": 0.4773072302341461,
+      "learning_rate": 0.0001,
+      "loss": 1.3969,
+      "step": 8904
+    },
+    {
+      "epoch": 1.0229165470105106,
+      "grad_norm": 0.48939380049705505,
+      "learning_rate": 0.0001,
+      "loss": 1.2671,
+      "step": 8905
+    },
+    {
+      "epoch": 1.0230314169203376,
+      "grad_norm": 0.5301325917243958,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 8906
+    },
+    {
+      "epoch": 1.0231462868301648,
+      "grad_norm": 0.5332849621772766,
+      "learning_rate": 0.0001,
+      "loss": 1.4009,
+      "step": 8907
+    },
+    {
+      "epoch": 1.0232611567399919,
+      "grad_norm": 0.5592193603515625,
+      "learning_rate": 0.0001,
+      "loss": 1.4986,
+      "step": 8908
+    },
+    {
+      "epoch": 1.023376026649819,
+      "grad_norm": 0.5041161179542542,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 8909
+    },
+    {
+      "epoch": 1.0234908965596463,
+      "grad_norm": 0.5014781951904297,
+      "learning_rate": 0.0001,
+      "loss": 1.424,
+      "step": 8910
+    },
+    {
+      "epoch": 1.0236057664694733,
+      "grad_norm": 0.49788087606430054,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 8911
+    },
+    {
+      "epoch": 1.0237206363793003,
+      "grad_norm": 0.5221337080001831,
+      "learning_rate": 0.0001,
+      "loss": 1.3539,
+      "step": 8912
+    },
+    {
+      "epoch": 1.0238355062891276,
+      "grad_norm": 0.5878422856330872,
+      "learning_rate": 0.0001,
+      "loss": 1.5483,
+      "step": 8913
+    },
+    {
+      "epoch": 1.0239503761989548,
+      "grad_norm": 0.5616274476051331,
+      "learning_rate": 0.0001,
+      "loss": 1.4821,
+      "step": 8914
+    },
+    {
+      "epoch": 1.0240652461087818,
+      "grad_norm": 0.4898712635040283,
+      "learning_rate": 0.0001,
+      "loss": 1.4507,
+      "step": 8915
+    },
+    {
+      "epoch": 1.024180116018609,
+      "grad_norm": 0.5116493105888367,
+      "learning_rate": 0.0001,
+      "loss": 1.4732,
+      "step": 8916
+    },
+    {
+      "epoch": 1.024294985928436,
+      "grad_norm": 0.5652375221252441,
+      "learning_rate": 0.0001,
+      "loss": 1.7441,
+      "step": 8917
+    },
+    {
+      "epoch": 1.0244098558382633,
+      "grad_norm": 0.525242805480957,
+      "learning_rate": 0.0001,
+      "loss": 1.3921,
+      "step": 8918
+    },
+    {
+      "epoch": 1.0245247257480903,
+      "grad_norm": 0.5136474370956421,
+      "learning_rate": 0.0001,
+      "loss": 1.4856,
+      "step": 8919
+    },
+    {
+      "epoch": 1.0246395956579175,
+      "grad_norm": 0.5519577860832214,
+      "learning_rate": 0.0001,
+      "loss": 1.4674,
+      "step": 8920
+    },
+    {
+      "epoch": 1.0247544655677445,
+      "grad_norm": 0.5397735834121704,
+      "learning_rate": 0.0001,
+      "loss": 1.4107,
+      "step": 8921
+    },
+    {
+      "epoch": 1.0248693354775718,
+      "grad_norm": 0.530933141708374,
+      "learning_rate": 0.0001,
+      "loss": 1.399,
+      "step": 8922
+    },
+    {
+      "epoch": 1.0249842053873988,
+      "grad_norm": 0.5000669956207275,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 8923
+    },
+    {
+      "epoch": 1.025099075297226,
+      "grad_norm": 0.5061250329017639,
+      "learning_rate": 0.0001,
+      "loss": 1.2578,
+      "step": 8924
+    },
+    {
+      "epoch": 1.025213945207053,
+      "grad_norm": 0.5225130915641785,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 8925
+    },
+    {
+      "epoch": 1.0253288151168802,
+      "grad_norm": 0.567387044429779,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 8926
+    },
+    {
+      "epoch": 1.0254436850267072,
+      "grad_norm": 0.6644077897071838,
+      "learning_rate": 0.0001,
+      "loss": 1.3388,
+      "step": 8927
+    },
+    {
+      "epoch": 1.0255585549365345,
+      "grad_norm": 0.5568031668663025,
+      "learning_rate": 0.0001,
+      "loss": 1.5316,
+      "step": 8928
+    },
+    {
+      "epoch": 1.0256734248463615,
+      "grad_norm": 0.45952072739601135,
+      "learning_rate": 0.0001,
+      "loss": 1.2552,
+      "step": 8929
+    },
+    {
+      "epoch": 1.0257882947561887,
+      "grad_norm": 0.4860810935497284,
+      "learning_rate": 0.0001,
+      "loss": 1.2508,
+      "step": 8930
+    },
+    {
+      "epoch": 1.0259031646660157,
+      "grad_norm": 0.5749561190605164,
+      "learning_rate": 0.0001,
+      "loss": 1.4491,
+      "step": 8931
+    },
+    {
+      "epoch": 1.026018034575843,
+      "grad_norm": 0.5022095441818237,
+      "learning_rate": 0.0001,
+      "loss": 1.2589,
+      "step": 8932
+    },
+    {
+      "epoch": 1.02613290448567,
+      "grad_norm": 0.4758133888244629,
+      "learning_rate": 0.0001,
+      "loss": 1.3553,
+      "step": 8933
+    },
+    {
+      "epoch": 1.0262477743954972,
+      "grad_norm": 0.5137754678726196,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 8934
+    },
+    {
+      "epoch": 1.0263626443053242,
+      "grad_norm": 0.517978847026825,
+      "learning_rate": 0.0001,
+      "loss": 1.2764,
+      "step": 8935
+    },
+    {
+      "epoch": 1.0264775142151514,
+      "grad_norm": 0.527491569519043,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 8936
+    },
+    {
+      "epoch": 1.0265923841249784,
+      "grad_norm": 0.5537298321723938,
+      "learning_rate": 0.0001,
+      "loss": 1.4757,
+      "step": 8937
+    },
+    {
+      "epoch": 1.0267072540348057,
+      "grad_norm": 0.49527880549430847,
+      "learning_rate": 0.0001,
+      "loss": 1.4042,
+      "step": 8938
+    },
+    {
+      "epoch": 1.0268221239446327,
+      "grad_norm": 0.5914473533630371,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 8939
+    },
+    {
+      "epoch": 1.02693699385446,
+      "grad_norm": 0.5270557403564453,
+      "learning_rate": 0.0001,
+      "loss": 1.5776,
+      "step": 8940
+    },
+    {
+      "epoch": 1.027051863764287,
+      "grad_norm": 0.4984856843948364,
+      "learning_rate": 0.0001,
+      "loss": 1.4058,
+      "step": 8941
+    },
+    {
+      "epoch": 1.0271667336741142,
+      "grad_norm": 0.5142775177955627,
+      "learning_rate": 0.0001,
+      "loss": 1.4986,
+      "step": 8942
+    },
+    {
+      "epoch": 1.0272816035839412,
+      "grad_norm": 0.5666712522506714,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 8943
+    },
+    {
+      "epoch": 1.0273964734937684,
+      "grad_norm": 0.515430212020874,
+      "learning_rate": 0.0001,
+      "loss": 1.5156,
+      "step": 8944
+    },
+    {
+      "epoch": 1.0275113434035954,
+      "grad_norm": 0.4515966475009918,
+      "learning_rate": 0.0001,
+      "loss": 1.0998,
+      "step": 8945
+    },
+    {
+      "epoch": 1.0276262133134226,
+      "grad_norm": 0.5609683394432068,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 8946
+    },
+    {
+      "epoch": 1.0277410832232496,
+      "grad_norm": 0.616452157497406,
+      "learning_rate": 0.0001,
+      "loss": 1.4311,
+      "step": 8947
+    },
+    {
+      "epoch": 1.0278559531330769,
+      "grad_norm": 0.5427160859107971,
+      "learning_rate": 0.0001,
+      "loss": 1.6402,
+      "step": 8948
+    },
+    {
+      "epoch": 1.0279708230429039,
+      "grad_norm": 0.5293609499931335,
+      "learning_rate": 0.0001,
+      "loss": 1.3271,
+      "step": 8949
+    },
+    {
+      "epoch": 1.0280856929527311,
+      "grad_norm": 0.5510618686676025,
+      "learning_rate": 0.0001,
+      "loss": 1.4305,
+      "step": 8950
+    },
+    {
+      "epoch": 1.0282005628625581,
+      "grad_norm": 0.5403707027435303,
+      "learning_rate": 0.0001,
+      "loss": 1.5898,
+      "step": 8951
+    },
+    {
+      "epoch": 1.0283154327723854,
+      "grad_norm": 0.5419697761535645,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 8952
+    },
+    {
+      "epoch": 1.0284303026822124,
+      "grad_norm": 0.5279735326766968,
+      "learning_rate": 0.0001,
+      "loss": 1.3857,
+      "step": 8953
+    },
+    {
+      "epoch": 1.0285451725920396,
+      "grad_norm": 0.48474159836769104,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 8954
+    },
+    {
+      "epoch": 1.0286600425018666,
+      "grad_norm": 0.5220118761062622,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 8955
+    },
+    {
+      "epoch": 1.0287749124116938,
+      "grad_norm": 0.5131214261054993,
+      "learning_rate": 0.0001,
+      "loss": 1.449,
+      "step": 8956
+    },
+    {
+      "epoch": 1.0288897823215208,
+      "grad_norm": 0.5555469989776611,
+      "learning_rate": 0.0001,
+      "loss": 1.4271,
+      "step": 8957
+    },
+    {
+      "epoch": 1.029004652231348,
+      "grad_norm": 0.5716768503189087,
+      "learning_rate": 0.0001,
+      "loss": 1.7778,
+      "step": 8958
+    },
+    {
+      "epoch": 1.029119522141175,
+      "grad_norm": 0.5097254514694214,
+      "learning_rate": 0.0001,
+      "loss": 1.5314,
+      "step": 8959
+    },
+    {
+      "epoch": 1.0292343920510023,
+      "grad_norm": 0.5094704031944275,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 8960
+    },
+    {
+      "epoch": 1.0293492619608293,
+      "grad_norm": 0.49823570251464844,
+      "learning_rate": 0.0001,
+      "loss": 1.4139,
+      "step": 8961
+    },
+    {
+      "epoch": 1.0294641318706566,
+      "grad_norm": 0.5155112147331238,
+      "learning_rate": 0.0001,
+      "loss": 1.4046,
+      "step": 8962
+    },
+    {
+      "epoch": 1.0295790017804836,
+      "grad_norm": 0.4962422549724579,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 8963
+    },
+    {
+      "epoch": 1.0296938716903108,
+      "grad_norm": 0.530081570148468,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 8964
+    },
+    {
+      "epoch": 1.0298087416001378,
+      "grad_norm": 0.5400624871253967,
+      "learning_rate": 0.0001,
+      "loss": 1.5078,
+      "step": 8965
+    },
+    {
+      "epoch": 1.029923611509965,
+      "grad_norm": 0.57288658618927,
+      "learning_rate": 0.0001,
+      "loss": 1.2249,
+      "step": 8966
+    },
+    {
+      "epoch": 1.030038481419792,
+      "grad_norm": 0.5239868760108948,
+      "learning_rate": 0.0001,
+      "loss": 1.3921,
+      "step": 8967
+    },
+    {
+      "epoch": 1.0301533513296193,
+      "grad_norm": 0.5611276030540466,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 8968
+    },
+    {
+      "epoch": 1.0302682212394463,
+      "grad_norm": 0.5283910036087036,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 8969
+    },
+    {
+      "epoch": 1.0303830911492735,
+      "grad_norm": 0.5439105033874512,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 8970
+    },
+    {
+      "epoch": 1.0304979610591005,
+      "grad_norm": 0.5273348093032837,
+      "learning_rate": 0.0001,
+      "loss": 1.683,
+      "step": 8971
+    },
+    {
+      "epoch": 1.0306128309689278,
+      "grad_norm": 0.5008038282394409,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 8972
+    },
+    {
+      "epoch": 1.0307277008787548,
+      "grad_norm": 0.5743789672851562,
+      "learning_rate": 0.0001,
+      "loss": 1.4544,
+      "step": 8973
+    },
+    {
+      "epoch": 1.030842570788582,
+      "grad_norm": 0.5148991346359253,
+      "learning_rate": 0.0001,
+      "loss": 1.3179,
+      "step": 8974
+    },
+    {
+      "epoch": 1.030957440698409,
+      "grad_norm": 0.5285124182701111,
+      "learning_rate": 0.0001,
+      "loss": 1.5723,
+      "step": 8975
+    },
+    {
+      "epoch": 1.0310723106082362,
+      "grad_norm": 0.5112460851669312,
+      "learning_rate": 0.0001,
+      "loss": 1.486,
+      "step": 8976
+    },
+    {
+      "epoch": 1.0311871805180632,
+      "grad_norm": 0.5338481068611145,
+      "learning_rate": 0.0001,
+      "loss": 1.3635,
+      "step": 8977
+    },
+    {
+      "epoch": 1.0313020504278905,
+      "grad_norm": 0.5114549994468689,
+      "learning_rate": 0.0001,
+      "loss": 1.1468,
+      "step": 8978
+    },
+    {
+      "epoch": 1.0314169203377175,
+      "grad_norm": 0.5220407247543335,
+      "learning_rate": 0.0001,
+      "loss": 1.5003,
+      "step": 8979
+    },
+    {
+      "epoch": 1.0315317902475447,
+      "grad_norm": 0.47584792971611023,
+      "learning_rate": 0.0001,
+      "loss": 1.4724,
+      "step": 8980
+    },
+    {
+      "epoch": 1.0316466601573717,
+      "grad_norm": 0.5218107104301453,
+      "learning_rate": 0.0001,
+      "loss": 1.4703,
+      "step": 8981
+    },
+    {
+      "epoch": 1.031761530067199,
+      "grad_norm": 0.5085425972938538,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 8982
+    },
+    {
+      "epoch": 1.031876399977026,
+      "grad_norm": 0.5143735408782959,
+      "learning_rate": 0.0001,
+      "loss": 1.2407,
+      "step": 8983
+    },
+    {
+      "epoch": 1.0319912698868532,
+      "grad_norm": 0.47457659244537354,
+      "learning_rate": 0.0001,
+      "loss": 1.3498,
+      "step": 8984
+    },
+    {
+      "epoch": 1.0321061397966802,
+      "grad_norm": 0.5300225615501404,
+      "learning_rate": 0.0001,
+      "loss": 1.4482,
+      "step": 8985
+    },
+    {
+      "epoch": 1.0322210097065074,
+      "grad_norm": 0.5350293517112732,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 8986
+    },
+    {
+      "epoch": 1.0323358796163344,
+      "grad_norm": 0.5425633788108826,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 8987
+    },
+    {
+      "epoch": 1.0324507495261617,
+      "grad_norm": 0.5491638779640198,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 8988
+    },
+    {
+      "epoch": 1.0325656194359887,
+      "grad_norm": 0.47783395648002625,
+      "learning_rate": 0.0001,
+      "loss": 1.3218,
+      "step": 8989
+    },
+    {
+      "epoch": 1.032680489345816,
+      "grad_norm": 0.57433021068573,
+      "learning_rate": 0.0001,
+      "loss": 1.5094,
+      "step": 8990
+    },
+    {
+      "epoch": 1.032795359255643,
+      "grad_norm": 0.5770861506462097,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 8991
+    },
+    {
+      "epoch": 1.0329102291654702,
+      "grad_norm": 0.5487734079360962,
+      "learning_rate": 0.0001,
+      "loss": 1.6067,
+      "step": 8992
+    },
+    {
+      "epoch": 1.0330250990752972,
+      "grad_norm": 0.5157055854797363,
+      "learning_rate": 0.0001,
+      "loss": 1.4258,
+      "step": 8993
+    },
+    {
+      "epoch": 1.0331399689851244,
+      "grad_norm": 0.5037803649902344,
+      "learning_rate": 0.0001,
+      "loss": 1.3896,
+      "step": 8994
+    },
+    {
+      "epoch": 1.0332548388949514,
+      "grad_norm": 0.5406895875930786,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 8995
+    },
+    {
+      "epoch": 1.0333697088047786,
+      "grad_norm": 0.5066888332366943,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 8996
+    },
+    {
+      "epoch": 1.0334845787146056,
+      "grad_norm": 0.48853886127471924,
+      "learning_rate": 0.0001,
+      "loss": 1.5338,
+      "step": 8997
+    },
+    {
+      "epoch": 1.0335994486244329,
+      "grad_norm": 0.48481160402297974,
+      "learning_rate": 0.0001,
+      "loss": 1.406,
+      "step": 8998
+    },
+    {
+      "epoch": 1.0337143185342599,
+      "grad_norm": 0.4979383945465088,
+      "learning_rate": 0.0001,
+      "loss": 1.3085,
+      "step": 8999
+    },
+    {
+      "epoch": 1.033829188444087,
+      "grad_norm": 0.5427919030189514,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 9000
+    },
+    {
+      "epoch": 1.0339440583539141,
+      "grad_norm": 0.5497532486915588,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 9001
+    },
+    {
+      "epoch": 1.0340589282637414,
+      "grad_norm": 0.5383864045143127,
+      "learning_rate": 0.0001,
+      "loss": 1.263,
+      "step": 9002
+    },
+    {
+      "epoch": 1.0341737981735684,
+      "grad_norm": 0.5036965012550354,
+      "learning_rate": 0.0001,
+      "loss": 1.4651,
+      "step": 9003
+    },
+    {
+      "epoch": 1.0342886680833956,
+      "grad_norm": 0.555041491985321,
+      "learning_rate": 0.0001,
+      "loss": 1.5879,
+      "step": 9004
+    },
+    {
+      "epoch": 1.0344035379932226,
+      "grad_norm": 0.5549968481063843,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 9005
+    },
+    {
+      "epoch": 1.0345184079030498,
+      "grad_norm": 0.579571008682251,
+      "learning_rate": 0.0001,
+      "loss": 1.458,
+      "step": 9006
+    },
+    {
+      "epoch": 1.0346332778128768,
+      "grad_norm": 0.5244538187980652,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 9007
+    },
+    {
+      "epoch": 1.034748147722704,
+      "grad_norm": 0.544011116027832,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 9008
+    },
+    {
+      "epoch": 1.034863017632531,
+      "grad_norm": 0.5533955097198486,
+      "learning_rate": 0.0001,
+      "loss": 1.6356,
+      "step": 9009
+    },
+    {
+      "epoch": 1.0349778875423583,
+      "grad_norm": 0.5618937611579895,
+      "learning_rate": 0.0001,
+      "loss": 1.6885,
+      "step": 9010
+    },
+    {
+      "epoch": 1.0350927574521853,
+      "grad_norm": 0.5388464331626892,
+      "learning_rate": 0.0001,
+      "loss": 1.4549,
+      "step": 9011
+    },
+    {
+      "epoch": 1.0352076273620126,
+      "grad_norm": 0.5120463967323303,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 9012
+    },
+    {
+      "epoch": 1.0353224972718396,
+      "grad_norm": 0.5093674659729004,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 9013
+    },
+    {
+      "epoch": 1.0354373671816668,
+      "grad_norm": 0.4842431843280792,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 9014
+    },
+    {
+      "epoch": 1.0355522370914938,
+      "grad_norm": 0.5035402774810791,
+      "learning_rate": 0.0001,
+      "loss": 1.443,
+      "step": 9015
+    },
+    {
+      "epoch": 1.035667107001321,
+      "grad_norm": 0.5231695771217346,
+      "learning_rate": 0.0001,
+      "loss": 1.4145,
+      "step": 9016
+    },
+    {
+      "epoch": 1.035781976911148,
+      "grad_norm": 0.5481805801391602,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 9017
+    },
+    {
+      "epoch": 1.0358968468209753,
+      "grad_norm": 0.510319173336029,
+      "learning_rate": 0.0001,
+      "loss": 1.4618,
+      "step": 9018
+    },
+    {
+      "epoch": 1.0360117167308023,
+      "grad_norm": 0.5354859828948975,
+      "learning_rate": 0.0001,
+      "loss": 1.3844,
+      "step": 9019
+    },
+    {
+      "epoch": 1.0361265866406295,
+      "grad_norm": 0.4900701642036438,
+      "learning_rate": 0.0001,
+      "loss": 1.3788,
+      "step": 9020
+    },
+    {
+      "epoch": 1.0362414565504565,
+      "grad_norm": 0.5398866534233093,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 9021
+    },
+    {
+      "epoch": 1.0363563264602837,
+      "grad_norm": 0.5214775204658508,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 9022
+    },
+    {
+      "epoch": 1.0364711963701108,
+      "grad_norm": 0.4958128035068512,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 9023
+    },
+    {
+      "epoch": 1.036586066279938,
+      "grad_norm": 0.5190637111663818,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 9024
+    },
+    {
+      "epoch": 1.036700936189765,
+      "grad_norm": 0.568577766418457,
+      "learning_rate": 0.0001,
+      "loss": 1.5747,
+      "step": 9025
+    },
+    {
+      "epoch": 1.0368158060995922,
+      "grad_norm": 0.5356246829032898,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 9026
+    },
+    {
+      "epoch": 1.0369306760094192,
+      "grad_norm": 0.5341411232948303,
+      "learning_rate": 0.0001,
+      "loss": 1.5258,
+      "step": 9027
+    },
+    {
+      "epoch": 1.0370455459192465,
+      "grad_norm": 0.5894232392311096,
+      "learning_rate": 0.0001,
+      "loss": 1.7847,
+      "step": 9028
+    },
+    {
+      "epoch": 1.0371604158290735,
+      "grad_norm": 0.4876749515533447,
+      "learning_rate": 0.0001,
+      "loss": 1.4103,
+      "step": 9029
+    },
+    {
+      "epoch": 1.0372752857389007,
+      "grad_norm": 0.5098420977592468,
+      "learning_rate": 0.0001,
+      "loss": 1.59,
+      "step": 9030
+    },
+    {
+      "epoch": 1.0373901556487277,
+      "grad_norm": 0.5374711751937866,
+      "learning_rate": 0.0001,
+      "loss": 1.406,
+      "step": 9031
+    },
+    {
+      "epoch": 1.037505025558555,
+      "grad_norm": 0.5379756689071655,
+      "learning_rate": 0.0001,
+      "loss": 1.4881,
+      "step": 9032
+    },
+    {
+      "epoch": 1.037619895468382,
+      "grad_norm": 0.5455212593078613,
+      "learning_rate": 0.0001,
+      "loss": 1.4375,
+      "step": 9033
+    },
+    {
+      "epoch": 1.0377347653782092,
+      "grad_norm": 0.5194243788719177,
+      "learning_rate": 0.0001,
+      "loss": 1.4984,
+      "step": 9034
+    },
+    {
+      "epoch": 1.0378496352880362,
+      "grad_norm": 0.5264743566513062,
+      "learning_rate": 0.0001,
+      "loss": 1.3457,
+      "step": 9035
+    },
+    {
+      "epoch": 1.0379645051978634,
+      "grad_norm": 0.5833313465118408,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 9036
+    },
+    {
+      "epoch": 1.0380793751076904,
+      "grad_norm": 0.5064496994018555,
+      "learning_rate": 0.0001,
+      "loss": 1.3317,
+      "step": 9037
+    },
+    {
+      "epoch": 1.0381942450175177,
+      "grad_norm": 0.5563880205154419,
+      "learning_rate": 0.0001,
+      "loss": 1.5072,
+      "step": 9038
+    },
+    {
+      "epoch": 1.0383091149273447,
+      "grad_norm": 0.5034350156784058,
+      "learning_rate": 0.0001,
+      "loss": 1.5021,
+      "step": 9039
+    },
+    {
+      "epoch": 1.038423984837172,
+      "grad_norm": 0.5645545125007629,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 9040
+    },
+    {
+      "epoch": 1.038538854746999,
+      "grad_norm": 0.5010533928871155,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 9041
+    },
+    {
+      "epoch": 1.0386537246568261,
+      "grad_norm": 0.5693106651306152,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 9042
+    },
+    {
+      "epoch": 1.0387685945666532,
+      "grad_norm": 0.5178468227386475,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 9043
+    },
+    {
+      "epoch": 1.0388834644764804,
+      "grad_norm": 0.49142345786094666,
+      "learning_rate": 0.0001,
+      "loss": 1.3921,
+      "step": 9044
+    },
+    {
+      "epoch": 1.0389983343863074,
+      "grad_norm": 0.512225329875946,
+      "learning_rate": 0.0001,
+      "loss": 1.6559,
+      "step": 9045
+    },
+    {
+      "epoch": 1.0391132042961346,
+      "grad_norm": 0.500827968120575,
+      "learning_rate": 0.0001,
+      "loss": 1.444,
+      "step": 9046
+    },
+    {
+      "epoch": 1.0392280742059619,
+      "grad_norm": 0.6273430585861206,
+      "learning_rate": 0.0001,
+      "loss": 1.6012,
+      "step": 9047
+    },
+    {
+      "epoch": 1.0393429441157889,
+      "grad_norm": 0.610727071762085,
+      "learning_rate": 0.0001,
+      "loss": 1.6243,
+      "step": 9048
+    },
+    {
+      "epoch": 1.0394578140256159,
+      "grad_norm": 0.4880732297897339,
+      "learning_rate": 0.0001,
+      "loss": 1.3747,
+      "step": 9049
+    },
+    {
+      "epoch": 1.039572683935443,
+      "grad_norm": 0.5448058843612671,
+      "learning_rate": 0.0001,
+      "loss": 1.422,
+      "step": 9050
+    },
+    {
+      "epoch": 1.0396875538452703,
+      "grad_norm": 0.5537201762199402,
+      "learning_rate": 0.0001,
+      "loss": 1.4965,
+      "step": 9051
+    },
+    {
+      "epoch": 1.0398024237550973,
+      "grad_norm": 0.518325686454773,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 9052
+    },
+    {
+      "epoch": 1.0399172936649246,
+      "grad_norm": 0.49593982100486755,
+      "learning_rate": 0.0001,
+      "loss": 1.3273,
+      "step": 9053
+    },
+    {
+      "epoch": 1.0400321635747516,
+      "grad_norm": 0.5287415385246277,
+      "learning_rate": 0.0001,
+      "loss": 1.5204,
+      "step": 9054
+    },
+    {
+      "epoch": 1.0401470334845788,
+      "grad_norm": 0.48811841011047363,
+      "learning_rate": 0.0001,
+      "loss": 1.3151,
+      "step": 9055
+    },
+    {
+      "epoch": 1.0402619033944058,
+      "grad_norm": 0.557843804359436,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 9056
+    },
+    {
+      "epoch": 1.040376773304233,
+      "grad_norm": 0.5163830518722534,
+      "learning_rate": 0.0001,
+      "loss": 1.5059,
+      "step": 9057
+    },
+    {
+      "epoch": 1.04049164321406,
+      "grad_norm": 0.5076297521591187,
+      "learning_rate": 0.0001,
+      "loss": 1.3535,
+      "step": 9058
+    },
+    {
+      "epoch": 1.0406065131238873,
+      "grad_norm": 0.5205563902854919,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 9059
+    },
+    {
+      "epoch": 1.0407213830337143,
+      "grad_norm": 0.5657637119293213,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 9060
+    },
+    {
+      "epoch": 1.0408362529435415,
+      "grad_norm": 0.5690082907676697,
+      "learning_rate": 0.0001,
+      "loss": 1.4616,
+      "step": 9061
+    },
+    {
+      "epoch": 1.0409511228533685,
+      "grad_norm": 0.5683187246322632,
+      "learning_rate": 0.0001,
+      "loss": 1.3052,
+      "step": 9062
+    },
+    {
+      "epoch": 1.0410659927631958,
+      "grad_norm": 0.5876998901367188,
+      "learning_rate": 0.0001,
+      "loss": 1.5554,
+      "step": 9063
+    },
+    {
+      "epoch": 1.0411808626730228,
+      "grad_norm": 0.5582718849182129,
+      "learning_rate": 0.0001,
+      "loss": 1.3038,
+      "step": 9064
+    },
+    {
+      "epoch": 1.04129573258285,
+      "grad_norm": 0.4977602958679199,
+      "learning_rate": 0.0001,
+      "loss": 1.4045,
+      "step": 9065
+    },
+    {
+      "epoch": 1.041410602492677,
+      "grad_norm": 0.5169378519058228,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 9066
+    },
+    {
+      "epoch": 1.0415254724025043,
+      "grad_norm": 0.5189085006713867,
+      "learning_rate": 0.0001,
+      "loss": 1.4188,
+      "step": 9067
+    },
+    {
+      "epoch": 1.0416403423123313,
+      "grad_norm": 0.49691158533096313,
+      "learning_rate": 0.0001,
+      "loss": 1.2885,
+      "step": 9068
+    },
+    {
+      "epoch": 1.0417552122221585,
+      "grad_norm": 0.5555067658424377,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 9069
+    },
+    {
+      "epoch": 1.0418700821319855,
+      "grad_norm": 0.5489440560340881,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 9070
+    },
+    {
+      "epoch": 1.0419849520418127,
+      "grad_norm": 0.5622000098228455,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 9071
+    },
+    {
+      "epoch": 1.0420998219516397,
+      "grad_norm": 0.546467661857605,
+      "learning_rate": 0.0001,
+      "loss": 1.2848,
+      "step": 9072
+    },
+    {
+      "epoch": 1.042214691861467,
+      "grad_norm": 0.496011883020401,
+      "learning_rate": 0.0001,
+      "loss": 1.2943,
+      "step": 9073
+    },
+    {
+      "epoch": 1.042329561771294,
+      "grad_norm": 0.500908613204956,
+      "learning_rate": 0.0001,
+      "loss": 1.3917,
+      "step": 9074
+    },
+    {
+      "epoch": 1.0424444316811212,
+      "grad_norm": 0.554749071598053,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 9075
+    },
+    {
+      "epoch": 1.0425593015909482,
+      "grad_norm": 0.5059463977813721,
+      "learning_rate": 0.0001,
+      "loss": 1.4333,
+      "step": 9076
+    },
+    {
+      "epoch": 1.0426741715007755,
+      "grad_norm": 0.5207396745681763,
+      "learning_rate": 0.0001,
+      "loss": 1.387,
+      "step": 9077
+    },
+    {
+      "epoch": 1.0427890414106025,
+      "grad_norm": 0.48321661353111267,
+      "learning_rate": 0.0001,
+      "loss": 1.4378,
+      "step": 9078
+    },
+    {
+      "epoch": 1.0429039113204297,
+      "grad_norm": 0.5234763622283936,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 9079
+    },
+    {
+      "epoch": 1.0430187812302567,
+      "grad_norm": 0.5880540609359741,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 9080
+    },
+    {
+      "epoch": 1.043133651140084,
+      "grad_norm": 0.548507809638977,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 9081
+    },
+    {
+      "epoch": 1.043248521049911,
+      "grad_norm": 0.545957088470459,
+      "learning_rate": 0.0001,
+      "loss": 1.5419,
+      "step": 9082
+    },
+    {
+      "epoch": 1.0433633909597382,
+      "grad_norm": 0.5886163115501404,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 9083
+    },
+    {
+      "epoch": 1.0434782608695652,
+      "grad_norm": 0.5363225340843201,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 9084
+    },
+    {
+      "epoch": 1.0435931307793924,
+      "grad_norm": 0.5168259739875793,
+      "learning_rate": 0.0001,
+      "loss": 1.3905,
+      "step": 9085
+    },
+    {
+      "epoch": 1.0437080006892194,
+      "grad_norm": 0.5514031052589417,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 9086
+    },
+    {
+      "epoch": 1.0438228705990467,
+      "grad_norm": 0.5472466945648193,
+      "learning_rate": 0.0001,
+      "loss": 1.382,
+      "step": 9087
+    },
+    {
+      "epoch": 1.0439377405088737,
+      "grad_norm": 0.5088561773300171,
+      "learning_rate": 0.0001,
+      "loss": 1.4711,
+      "step": 9088
+    },
+    {
+      "epoch": 1.044052610418701,
+      "grad_norm": 0.535428524017334,
+      "learning_rate": 0.0001,
+      "loss": 1.4639,
+      "step": 9089
+    },
+    {
+      "epoch": 1.044167480328528,
+      "grad_norm": 0.4890363812446594,
+      "learning_rate": 0.0001,
+      "loss": 1.3678,
+      "step": 9090
+    },
+    {
+      "epoch": 1.0442823502383551,
+      "grad_norm": 0.5183672308921814,
+      "learning_rate": 0.0001,
+      "loss": 1.4534,
+      "step": 9091
+    },
+    {
+      "epoch": 1.0443972201481821,
+      "grad_norm": 0.5594877004623413,
+      "learning_rate": 0.0001,
+      "loss": 1.6638,
+      "step": 9092
+    },
+    {
+      "epoch": 1.0445120900580094,
+      "grad_norm": 0.5082018375396729,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 9093
+    },
+    {
+      "epoch": 1.0446269599678364,
+      "grad_norm": 0.5126281380653381,
+      "learning_rate": 0.0001,
+      "loss": 1.3822,
+      "step": 9094
+    },
+    {
+      "epoch": 1.0447418298776636,
+      "grad_norm": 0.5137059092521667,
+      "learning_rate": 0.0001,
+      "loss": 1.4601,
+      "step": 9095
+    },
+    {
+      "epoch": 1.0448566997874906,
+      "grad_norm": 0.491517573595047,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 9096
+    },
+    {
+      "epoch": 1.0449715696973179,
+      "grad_norm": 0.5336720943450928,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 9097
+    },
+    {
+      "epoch": 1.0450864396071449,
+      "grad_norm": 0.5771209001541138,
+      "learning_rate": 0.0001,
+      "loss": 1.6016,
+      "step": 9098
+    },
+    {
+      "epoch": 1.045201309516972,
+      "grad_norm": 0.5206882357597351,
+      "learning_rate": 0.0001,
+      "loss": 1.5159,
+      "step": 9099
+    },
+    {
+      "epoch": 1.045316179426799,
+      "grad_norm": 0.5490942597389221,
+      "learning_rate": 0.0001,
+      "loss": 1.6517,
+      "step": 9100
+    },
+    {
+      "epoch": 1.0454310493366263,
+      "grad_norm": 0.5432296991348267,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 9101
+    },
+    {
+      "epoch": 1.0455459192464533,
+      "grad_norm": 0.5428939461708069,
+      "learning_rate": 0.0001,
+      "loss": 1.3368,
+      "step": 9102
+    },
+    {
+      "epoch": 1.0456607891562806,
+      "grad_norm": 0.4910656213760376,
+      "learning_rate": 0.0001,
+      "loss": 1.2355,
+      "step": 9103
+    },
+    {
+      "epoch": 1.0457756590661076,
+      "grad_norm": 0.5123563408851624,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 9104
+    },
+    {
+      "epoch": 1.0458905289759348,
+      "grad_norm": 0.5455781817436218,
+      "learning_rate": 0.0001,
+      "loss": 1.4858,
+      "step": 9105
+    },
+    {
+      "epoch": 1.0460053988857618,
+      "grad_norm": 0.5809556245803833,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 9106
+    },
+    {
+      "epoch": 1.046120268795589,
+      "grad_norm": 0.5579778552055359,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 9107
+    },
+    {
+      "epoch": 1.046235138705416,
+      "grad_norm": 0.5217399597167969,
+      "learning_rate": 0.0001,
+      "loss": 1.4473,
+      "step": 9108
+    },
+    {
+      "epoch": 1.0463500086152433,
+      "grad_norm": 0.5513988137245178,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 9109
+    },
+    {
+      "epoch": 1.0464648785250703,
+      "grad_norm": 0.5201188921928406,
+      "learning_rate": 0.0001,
+      "loss": 1.3266,
+      "step": 9110
+    },
+    {
+      "epoch": 1.0465797484348975,
+      "grad_norm": 0.5076174736022949,
+      "learning_rate": 0.0001,
+      "loss": 1.4519,
+      "step": 9111
+    },
+    {
+      "epoch": 1.0466946183447245,
+      "grad_norm": 0.543914258480072,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 9112
+    },
+    {
+      "epoch": 1.0468094882545518,
+      "grad_norm": 0.5245856642723083,
+      "learning_rate": 0.0001,
+      "loss": 1.587,
+      "step": 9113
+    },
+    {
+      "epoch": 1.0469243581643788,
+      "grad_norm": 0.5603649616241455,
+      "learning_rate": 0.0001,
+      "loss": 1.462,
+      "step": 9114
+    },
+    {
+      "epoch": 1.047039228074206,
+      "grad_norm": 0.5366967916488647,
+      "learning_rate": 0.0001,
+      "loss": 1.7216,
+      "step": 9115
+    },
+    {
+      "epoch": 1.047154097984033,
+      "grad_norm": 0.6067121624946594,
+      "learning_rate": 0.0001,
+      "loss": 1.6592,
+      "step": 9116
+    },
+    {
+      "epoch": 1.0472689678938603,
+      "grad_norm": 0.5686327219009399,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 9117
+    },
+    {
+      "epoch": 1.0473838378036873,
+      "grad_norm": 0.5154968500137329,
+      "learning_rate": 0.0001,
+      "loss": 1.2529,
+      "step": 9118
+    },
+    {
+      "epoch": 1.0474987077135145,
+      "grad_norm": 0.49103155732154846,
+      "learning_rate": 0.0001,
+      "loss": 1.3669,
+      "step": 9119
+    },
+    {
+      "epoch": 1.0476135776233415,
+      "grad_norm": 0.5241073966026306,
+      "learning_rate": 0.0001,
+      "loss": 1.4359,
+      "step": 9120
+    },
+    {
+      "epoch": 1.0477284475331687,
+      "grad_norm": 0.5242918133735657,
+      "learning_rate": 0.0001,
+      "loss": 1.4032,
+      "step": 9121
+    },
+    {
+      "epoch": 1.0478433174429957,
+      "grad_norm": 0.500344455242157,
+      "learning_rate": 0.0001,
+      "loss": 1.3488,
+      "step": 9122
+    },
+    {
+      "epoch": 1.047958187352823,
+      "grad_norm": 0.5362123250961304,
+      "learning_rate": 0.0001,
+      "loss": 1.3767,
+      "step": 9123
+    },
+    {
+      "epoch": 1.04807305726265,
+      "grad_norm": 0.5414429903030396,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 9124
+    },
+    {
+      "epoch": 1.0481879271724772,
+      "grad_norm": 0.5353612899780273,
+      "learning_rate": 0.0001,
+      "loss": 1.5028,
+      "step": 9125
+    },
+    {
+      "epoch": 1.0483027970823042,
+      "grad_norm": 0.5600303411483765,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 9126
+    },
+    {
+      "epoch": 1.0484176669921315,
+      "grad_norm": 0.5072823166847229,
+      "learning_rate": 0.0001,
+      "loss": 1.4421,
+      "step": 9127
+    },
+    {
+      "epoch": 1.0485325369019585,
+      "grad_norm": 0.5189083218574524,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 9128
+    },
+    {
+      "epoch": 1.0486474068117857,
+      "grad_norm": 0.5282391309738159,
+      "learning_rate": 0.0001,
+      "loss": 1.4874,
+      "step": 9129
+    },
+    {
+      "epoch": 1.0487622767216127,
+      "grad_norm": 0.5408895015716553,
+      "learning_rate": 0.0001,
+      "loss": 1.4552,
+      "step": 9130
+    },
+    {
+      "epoch": 1.04887714663144,
+      "grad_norm": 0.5891426205635071,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 9131
+    },
+    {
+      "epoch": 1.048992016541267,
+      "grad_norm": 0.5693600177764893,
+      "learning_rate": 0.0001,
+      "loss": 1.4055,
+      "step": 9132
+    },
+    {
+      "epoch": 1.0491068864510942,
+      "grad_norm": 0.534770667552948,
+      "learning_rate": 0.0001,
+      "loss": 1.4429,
+      "step": 9133
+    },
+    {
+      "epoch": 1.0492217563609212,
+      "grad_norm": 0.5606382489204407,
+      "learning_rate": 0.0001,
+      "loss": 1.5205,
+      "step": 9134
+    },
+    {
+      "epoch": 1.0493366262707484,
+      "grad_norm": 0.5300989747047424,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 9135
+    },
+    {
+      "epoch": 1.0494514961805754,
+      "grad_norm": 0.56827712059021,
+      "learning_rate": 0.0001,
+      "loss": 1.6758,
+      "step": 9136
+    },
+    {
+      "epoch": 1.0495663660904027,
+      "grad_norm": 0.5045410394668579,
+      "learning_rate": 0.0001,
+      "loss": 1.277,
+      "step": 9137
+    },
+    {
+      "epoch": 1.0496812360002297,
+      "grad_norm": 0.5462251901626587,
+      "learning_rate": 0.0001,
+      "loss": 1.3595,
+      "step": 9138
+    },
+    {
+      "epoch": 1.049796105910057,
+      "grad_norm": 0.5165095925331116,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 9139
+    },
+    {
+      "epoch": 1.049910975819884,
+      "grad_norm": 0.4886758327484131,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 9140
+    },
+    {
+      "epoch": 1.0500258457297111,
+      "grad_norm": 0.5420746207237244,
+      "learning_rate": 0.0001,
+      "loss": 1.2517,
+      "step": 9141
+    },
+    {
+      "epoch": 1.0501407156395381,
+      "grad_norm": 0.5480278730392456,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 9142
+    },
+    {
+      "epoch": 1.0502555855493654,
+      "grad_norm": 0.5709562301635742,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 9143
+    },
+    {
+      "epoch": 1.0503704554591924,
+      "grad_norm": 0.5308666229248047,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 9144
+    },
+    {
+      "epoch": 1.0504853253690196,
+      "grad_norm": 0.5184751749038696,
+      "learning_rate": 0.0001,
+      "loss": 1.4136,
+      "step": 9145
+    },
+    {
+      "epoch": 1.0506001952788466,
+      "grad_norm": 0.5020022392272949,
+      "learning_rate": 0.0001,
+      "loss": 1.4594,
+      "step": 9146
+    },
+    {
+      "epoch": 1.0507150651886739,
+      "grad_norm": 0.531156063079834,
+      "learning_rate": 0.0001,
+      "loss": 1.4978,
+      "step": 9147
+    },
+    {
+      "epoch": 1.0508299350985009,
+      "grad_norm": 0.5110874772071838,
+      "learning_rate": 0.0001,
+      "loss": 1.3658,
+      "step": 9148
+    },
+    {
+      "epoch": 1.050944805008328,
+      "grad_norm": 0.6354340314865112,
+      "learning_rate": 0.0001,
+      "loss": 1.511,
+      "step": 9149
+    },
+    {
+      "epoch": 1.051059674918155,
+      "grad_norm": 0.5681877136230469,
+      "learning_rate": 0.0001,
+      "loss": 1.382,
+      "step": 9150
+    },
+    {
+      "epoch": 1.0511745448279823,
+      "grad_norm": 0.4902039170265198,
+      "learning_rate": 0.0001,
+      "loss": 1.3208,
+      "step": 9151
+    },
+    {
+      "epoch": 1.0512894147378093,
+      "grad_norm": 0.5818510055541992,
+      "learning_rate": 0.0001,
+      "loss": 1.4707,
+      "step": 9152
+    },
+    {
+      "epoch": 1.0514042846476366,
+      "grad_norm": 0.5937033891677856,
+      "learning_rate": 0.0001,
+      "loss": 1.5029,
+      "step": 9153
+    },
+    {
+      "epoch": 1.0515191545574636,
+      "grad_norm": 0.5286517143249512,
+      "learning_rate": 0.0001,
+      "loss": 1.4477,
+      "step": 9154
+    },
+    {
+      "epoch": 1.0516340244672908,
+      "grad_norm": 0.5208611488342285,
+      "learning_rate": 0.0001,
+      "loss": 1.3308,
+      "step": 9155
+    },
+    {
+      "epoch": 1.0517488943771178,
+      "grad_norm": 0.5702365636825562,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 9156
+    },
+    {
+      "epoch": 1.051863764286945,
+      "grad_norm": 0.551328182220459,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 9157
+    },
+    {
+      "epoch": 1.051978634196772,
+      "grad_norm": 0.5106463432312012,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 9158
+    },
+    {
+      "epoch": 1.0520935041065993,
+      "grad_norm": 0.5503958463668823,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 9159
+    },
+    {
+      "epoch": 1.0522083740164263,
+      "grad_norm": 0.5540024042129517,
+      "learning_rate": 0.0001,
+      "loss": 1.3085,
+      "step": 9160
+    },
+    {
+      "epoch": 1.0523232439262535,
+      "grad_norm": 0.5419454574584961,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 9161
+    },
+    {
+      "epoch": 1.0524381138360805,
+      "grad_norm": 0.5251337885856628,
+      "learning_rate": 0.0001,
+      "loss": 1.3589,
+      "step": 9162
+    },
+    {
+      "epoch": 1.0525529837459078,
+      "grad_norm": 0.553993284702301,
+      "learning_rate": 0.0001,
+      "loss": 1.5545,
+      "step": 9163
+    },
+    {
+      "epoch": 1.0526678536557348,
+      "grad_norm": 0.5526601672172546,
+      "learning_rate": 0.0001,
+      "loss": 1.3881,
+      "step": 9164
+    },
+    {
+      "epoch": 1.052782723565562,
+      "grad_norm": 0.5293091535568237,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 9165
+    },
+    {
+      "epoch": 1.052897593475389,
+      "grad_norm": 0.5309327840805054,
+      "learning_rate": 0.0001,
+      "loss": 1.2998,
+      "step": 9166
+    },
+    {
+      "epoch": 1.0530124633852163,
+      "grad_norm": 0.5376285314559937,
+      "learning_rate": 0.0001,
+      "loss": 1.5854,
+      "step": 9167
+    },
+    {
+      "epoch": 1.0531273332950433,
+      "grad_norm": 0.5321171283721924,
+      "learning_rate": 0.0001,
+      "loss": 1.4715,
+      "step": 9168
+    },
+    {
+      "epoch": 1.0532422032048705,
+      "grad_norm": 0.5454419851303101,
+      "learning_rate": 0.0001,
+      "loss": 1.565,
+      "step": 9169
+    },
+    {
+      "epoch": 1.0533570731146975,
+      "grad_norm": 0.5634780526161194,
+      "learning_rate": 0.0001,
+      "loss": 1.2175,
+      "step": 9170
+    },
+    {
+      "epoch": 1.0534719430245247,
+      "grad_norm": 0.6618204712867737,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 9171
+    },
+    {
+      "epoch": 1.0535868129343517,
+      "grad_norm": 0.5510662198066711,
+      "learning_rate": 0.0001,
+      "loss": 1.5464,
+      "step": 9172
+    },
+    {
+      "epoch": 1.053701682844179,
+      "grad_norm": 0.5555408596992493,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 9173
+    },
+    {
+      "epoch": 1.053816552754006,
+      "grad_norm": 0.5211841464042664,
+      "learning_rate": 0.0001,
+      "loss": 1.3639,
+      "step": 9174
+    },
+    {
+      "epoch": 1.0539314226638332,
+      "grad_norm": 0.5432996153831482,
+      "learning_rate": 0.0001,
+      "loss": 1.6417,
+      "step": 9175
+    },
+    {
+      "epoch": 1.0540462925736602,
+      "grad_norm": 0.5261744260787964,
+      "learning_rate": 0.0001,
+      "loss": 1.4003,
+      "step": 9176
+    },
+    {
+      "epoch": 1.0541611624834875,
+      "grad_norm": 0.5317165851593018,
+      "learning_rate": 0.0001,
+      "loss": 1.3629,
+      "step": 9177
+    },
+    {
+      "epoch": 1.0542760323933145,
+      "grad_norm": 0.5396149754524231,
+      "learning_rate": 0.0001,
+      "loss": 1.4707,
+      "step": 9178
+    },
+    {
+      "epoch": 1.0543909023031417,
+      "grad_norm": 0.5117886066436768,
+      "learning_rate": 0.0001,
+      "loss": 1.2257,
+      "step": 9179
+    },
+    {
+      "epoch": 1.0545057722129687,
+      "grad_norm": 0.5107698440551758,
+      "learning_rate": 0.0001,
+      "loss": 1.5142,
+      "step": 9180
+    },
+    {
+      "epoch": 1.054620642122796,
+      "grad_norm": 0.5465172529220581,
+      "learning_rate": 0.0001,
+      "loss": 1.4611,
+      "step": 9181
+    },
+    {
+      "epoch": 1.054735512032623,
+      "grad_norm": 0.5007506012916565,
+      "learning_rate": 0.0001,
+      "loss": 1.3207,
+      "step": 9182
+    },
+    {
+      "epoch": 1.0548503819424502,
+      "grad_norm": 0.5821637511253357,
+      "learning_rate": 0.0001,
+      "loss": 1.5268,
+      "step": 9183
+    },
+    {
+      "epoch": 1.0549652518522774,
+      "grad_norm": 0.5477774739265442,
+      "learning_rate": 0.0001,
+      "loss": 1.1992,
+      "step": 9184
+    },
+    {
+      "epoch": 1.0550801217621044,
+      "grad_norm": 0.5345506072044373,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 9185
+    },
+    {
+      "epoch": 1.0551949916719314,
+      "grad_norm": 0.4949226379394531,
+      "learning_rate": 0.0001,
+      "loss": 1.4013,
+      "step": 9186
+    },
+    {
+      "epoch": 1.0553098615817587,
+      "grad_norm": 0.543696403503418,
+      "learning_rate": 0.0001,
+      "loss": 1.6937,
+      "step": 9187
+    },
+    {
+      "epoch": 1.0554247314915859,
+      "grad_norm": 0.5482311248779297,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 9188
+    },
+    {
+      "epoch": 1.055539601401413,
+      "grad_norm": 0.500032901763916,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 9189
+    },
+    {
+      "epoch": 1.0556544713112401,
+      "grad_norm": 0.5224020481109619,
+      "learning_rate": 0.0001,
+      "loss": 1.4317,
+      "step": 9190
+    },
+    {
+      "epoch": 1.0557693412210671,
+      "grad_norm": 0.5447744131088257,
+      "learning_rate": 0.0001,
+      "loss": 1.3627,
+      "step": 9191
+    },
+    {
+      "epoch": 1.0558842111308944,
+      "grad_norm": 0.551868200302124,
+      "learning_rate": 0.0001,
+      "loss": 1.5095,
+      "step": 9192
+    },
+    {
+      "epoch": 1.0559990810407214,
+      "grad_norm": 0.5407156944274902,
+      "learning_rate": 0.0001,
+      "loss": 1.6561,
+      "step": 9193
+    },
+    {
+      "epoch": 1.0561139509505486,
+      "grad_norm": 0.5364171862602234,
+      "learning_rate": 0.0001,
+      "loss": 1.5194,
+      "step": 9194
+    },
+    {
+      "epoch": 1.0562288208603756,
+      "grad_norm": 0.5270219445228577,
+      "learning_rate": 0.0001,
+      "loss": 1.3945,
+      "step": 9195
+    },
+    {
+      "epoch": 1.0563436907702028,
+      "grad_norm": 0.5508518218994141,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 9196
+    },
+    {
+      "epoch": 1.0564585606800299,
+      "grad_norm": 0.5767947435379028,
+      "learning_rate": 0.0001,
+      "loss": 1.3796,
+      "step": 9197
+    },
+    {
+      "epoch": 1.056573430589857,
+      "grad_norm": 0.516804575920105,
+      "learning_rate": 0.0001,
+      "loss": 1.3958,
+      "step": 9198
+    },
+    {
+      "epoch": 1.056688300499684,
+      "grad_norm": 0.5120447278022766,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 9199
+    },
+    {
+      "epoch": 1.0568031704095113,
+      "grad_norm": 0.5243872404098511,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 9200
+    },
+    {
+      "epoch": 1.0569180403193383,
+      "grad_norm": 0.5689402222633362,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 9201
+    },
+    {
+      "epoch": 1.0570329102291656,
+      "grad_norm": 0.5817130208015442,
+      "learning_rate": 0.0001,
+      "loss": 1.4481,
+      "step": 9202
+    },
+    {
+      "epoch": 1.0571477801389926,
+      "grad_norm": 0.5574333667755127,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 9203
+    },
+    {
+      "epoch": 1.0572626500488198,
+      "grad_norm": 0.5260115265846252,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 9204
+    },
+    {
+      "epoch": 1.0573775199586468,
+      "grad_norm": 0.5010888576507568,
+      "learning_rate": 0.0001,
+      "loss": 1.46,
+      "step": 9205
+    },
+    {
+      "epoch": 1.057492389868474,
+      "grad_norm": 0.5643918514251709,
+      "learning_rate": 0.0001,
+      "loss": 1.335,
+      "step": 9206
+    },
+    {
+      "epoch": 1.057607259778301,
+      "grad_norm": 0.5270254015922546,
+      "learning_rate": 0.0001,
+      "loss": 1.2919,
+      "step": 9207
+    },
+    {
+      "epoch": 1.0577221296881283,
+      "grad_norm": 0.5090634226799011,
+      "learning_rate": 0.0001,
+      "loss": 1.4459,
+      "step": 9208
+    },
+    {
+      "epoch": 1.0578369995979553,
+      "grad_norm": 0.5546037554740906,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 9209
+    },
+    {
+      "epoch": 1.0579518695077825,
+      "grad_norm": 0.5387788414955139,
+      "learning_rate": 0.0001,
+      "loss": 1.5268,
+      "step": 9210
+    },
+    {
+      "epoch": 1.0580667394176095,
+      "grad_norm": 0.5352441668510437,
+      "learning_rate": 0.0001,
+      "loss": 1.5164,
+      "step": 9211
+    },
+    {
+      "epoch": 1.0581816093274368,
+      "grad_norm": 0.5658225417137146,
+      "learning_rate": 0.0001,
+      "loss": 1.472,
+      "step": 9212
+    },
+    {
+      "epoch": 1.0582964792372638,
+      "grad_norm": 0.541875958442688,
+      "learning_rate": 0.0001,
+      "loss": 1.0823,
+      "step": 9213
+    },
+    {
+      "epoch": 1.058411349147091,
+      "grad_norm": 0.5292019248008728,
+      "learning_rate": 0.0001,
+      "loss": 1.4035,
+      "step": 9214
+    },
+    {
+      "epoch": 1.058526219056918,
+      "grad_norm": 0.5192826986312866,
+      "learning_rate": 0.0001,
+      "loss": 1.2651,
+      "step": 9215
+    },
+    {
+      "epoch": 1.0586410889667452,
+      "grad_norm": 0.5812119245529175,
+      "learning_rate": 0.0001,
+      "loss": 1.3717,
+      "step": 9216
+    },
+    {
+      "epoch": 1.0587559588765723,
+      "grad_norm": 0.6301900744438171,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 9217
+    },
+    {
+      "epoch": 1.0588708287863995,
+      "grad_norm": 0.5290337800979614,
+      "learning_rate": 0.0001,
+      "loss": 1.3963,
+      "step": 9218
+    },
+    {
+      "epoch": 1.0589856986962265,
+      "grad_norm": 0.6721698641777039,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 9219
+    },
+    {
+      "epoch": 1.0591005686060537,
+      "grad_norm": 0.6142293810844421,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 9220
+    },
+    {
+      "epoch": 1.0592154385158807,
+      "grad_norm": 0.5450447797775269,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 9221
+    },
+    {
+      "epoch": 1.059330308425708,
+      "grad_norm": 0.5755720138549805,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 9222
+    },
+    {
+      "epoch": 1.059445178335535,
+      "grad_norm": 0.6007996201515198,
+      "learning_rate": 0.0001,
+      "loss": 1.7402,
+      "step": 9223
+    },
+    {
+      "epoch": 1.0595600482453622,
+      "grad_norm": 0.5676028728485107,
+      "learning_rate": 0.0001,
+      "loss": 1.4648,
+      "step": 9224
+    },
+    {
+      "epoch": 1.0596749181551892,
+      "grad_norm": 0.5260946750640869,
+      "learning_rate": 0.0001,
+      "loss": 1.5316,
+      "step": 9225
+    },
+    {
+      "epoch": 1.0597897880650164,
+      "grad_norm": 0.5359583497047424,
+      "learning_rate": 0.0001,
+      "loss": 1.3536,
+      "step": 9226
+    },
+    {
+      "epoch": 1.0599046579748435,
+      "grad_norm": 0.5493340492248535,
+      "learning_rate": 0.0001,
+      "loss": 1.6997,
+      "step": 9227
+    },
+    {
+      "epoch": 1.0600195278846707,
+      "grad_norm": 0.5583151578903198,
+      "learning_rate": 0.0001,
+      "loss": 1.5545,
+      "step": 9228
+    },
+    {
+      "epoch": 1.0601343977944977,
+      "grad_norm": 0.5902903079986572,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 9229
+    },
+    {
+      "epoch": 1.060249267704325,
+      "grad_norm": 0.5786603689193726,
+      "learning_rate": 0.0001,
+      "loss": 1.3942,
+      "step": 9230
+    },
+    {
+      "epoch": 1.060364137614152,
+      "grad_norm": 0.5523187518119812,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 9231
+    },
+    {
+      "epoch": 1.0604790075239792,
+      "grad_norm": 0.5302920341491699,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 9232
+    },
+    {
+      "epoch": 1.0605938774338062,
+      "grad_norm": 0.5174841284751892,
+      "learning_rate": 0.0001,
+      "loss": 1.4171,
+      "step": 9233
+    },
+    {
+      "epoch": 1.0607087473436334,
+      "grad_norm": 0.5988504886627197,
+      "learning_rate": 0.0001,
+      "loss": 1.6997,
+      "step": 9234
+    },
+    {
+      "epoch": 1.0608236172534604,
+      "grad_norm": 0.5470107197761536,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 9235
+    },
+    {
+      "epoch": 1.0609384871632876,
+      "grad_norm": 0.5042878985404968,
+      "learning_rate": 0.0001,
+      "loss": 1.4093,
+      "step": 9236
+    },
+    {
+      "epoch": 1.0610533570731147,
+      "grad_norm": 0.5748067498207092,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 9237
+    },
+    {
+      "epoch": 1.0611682269829419,
+      "grad_norm": 0.5632085204124451,
+      "learning_rate": 0.0001,
+      "loss": 1.6162,
+      "step": 9238
+    },
+    {
+      "epoch": 1.061283096892769,
+      "grad_norm": 0.5615068674087524,
+      "learning_rate": 0.0001,
+      "loss": 1.5225,
+      "step": 9239
+    },
+    {
+      "epoch": 1.0613979668025961,
+      "grad_norm": 0.5865307450294495,
+      "learning_rate": 0.0001,
+      "loss": 1.3333,
+      "step": 9240
+    },
+    {
+      "epoch": 1.0615128367124231,
+      "grad_norm": 0.5403387546539307,
+      "learning_rate": 0.0001,
+      "loss": 1.3786,
+      "step": 9241
+    },
+    {
+      "epoch": 1.0616277066222504,
+      "grad_norm": 0.5474202036857605,
+      "learning_rate": 0.0001,
+      "loss": 1.2461,
+      "step": 9242
+    },
+    {
+      "epoch": 1.0617425765320774,
+      "grad_norm": 0.5543806552886963,
+      "learning_rate": 0.0001,
+      "loss": 1.4252,
+      "step": 9243
+    },
+    {
+      "epoch": 1.0618574464419046,
+      "grad_norm": 0.5339075326919556,
+      "learning_rate": 0.0001,
+      "loss": 1.3884,
+      "step": 9244
+    },
+    {
+      "epoch": 1.0619723163517316,
+      "grad_norm": 0.49664506316185,
+      "learning_rate": 0.0001,
+      "loss": 1.376,
+      "step": 9245
+    },
+    {
+      "epoch": 1.0620871862615588,
+      "grad_norm": 0.5427022576332092,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 9246
+    },
+    {
+      "epoch": 1.0622020561713859,
+      "grad_norm": 0.5456292629241943,
+      "learning_rate": 0.0001,
+      "loss": 1.6116,
+      "step": 9247
+    },
+    {
+      "epoch": 1.062316926081213,
+      "grad_norm": 0.533386766910553,
+      "learning_rate": 0.0001,
+      "loss": 1.4291,
+      "step": 9248
+    },
+    {
+      "epoch": 1.06243179599104,
+      "grad_norm": 0.5511561632156372,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 9249
+    },
+    {
+      "epoch": 1.0625466659008673,
+      "grad_norm": 0.5116701126098633,
+      "learning_rate": 0.0001,
+      "loss": 1.367,
+      "step": 9250
+    },
+    {
+      "epoch": 1.0626615358106943,
+      "grad_norm": 0.5287088751792908,
+      "learning_rate": 0.0001,
+      "loss": 1.5248,
+      "step": 9251
+    },
+    {
+      "epoch": 1.0627764057205216,
+      "grad_norm": 0.5381811261177063,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 9252
+    },
+    {
+      "epoch": 1.0628912756303486,
+      "grad_norm": 0.5121027231216431,
+      "learning_rate": 0.0001,
+      "loss": 1.3487,
+      "step": 9253
+    },
+    {
+      "epoch": 1.0630061455401758,
+      "grad_norm": 0.49183768033981323,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 9254
+    },
+    {
+      "epoch": 1.0631210154500028,
+      "grad_norm": 0.5818896293640137,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 9255
+    },
+    {
+      "epoch": 1.06323588535983,
+      "grad_norm": 0.5403487682342529,
+      "learning_rate": 0.0001,
+      "loss": 1.0887,
+      "step": 9256
+    },
+    {
+      "epoch": 1.063350755269657,
+      "grad_norm": 0.5411499738693237,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 9257
+    },
+    {
+      "epoch": 1.0634656251794843,
+      "grad_norm": 0.5758698582649231,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 9258
+    },
+    {
+      "epoch": 1.0635804950893113,
+      "grad_norm": 0.5343381762504578,
+      "learning_rate": 0.0001,
+      "loss": 1.5186,
+      "step": 9259
+    },
+    {
+      "epoch": 1.0636953649991385,
+      "grad_norm": 0.5107303261756897,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 9260
+    },
+    {
+      "epoch": 1.0638102349089655,
+      "grad_norm": 0.5439027547836304,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 9261
+    },
+    {
+      "epoch": 1.0639251048187928,
+      "grad_norm": 0.5605586767196655,
+      "learning_rate": 0.0001,
+      "loss": 1.6073,
+      "step": 9262
+    },
+    {
+      "epoch": 1.0640399747286198,
+      "grad_norm": 0.5504233837127686,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 9263
+    },
+    {
+      "epoch": 1.064154844638447,
+      "grad_norm": 0.5588496923446655,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 9264
+    },
+    {
+      "epoch": 1.064269714548274,
+      "grad_norm": 0.5234324932098389,
+      "learning_rate": 0.0001,
+      "loss": 1.4739,
+      "step": 9265
+    },
+    {
+      "epoch": 1.0643845844581012,
+      "grad_norm": 0.5559567809104919,
+      "learning_rate": 0.0001,
+      "loss": 1.4358,
+      "step": 9266
+    },
+    {
+      "epoch": 1.0644994543679283,
+      "grad_norm": 0.5315650105476379,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 9267
+    },
+    {
+      "epoch": 1.0646143242777555,
+      "grad_norm": 0.586476743221283,
+      "learning_rate": 0.0001,
+      "loss": 1.693,
+      "step": 9268
+    },
+    {
+      "epoch": 1.0647291941875825,
+      "grad_norm": 0.545502245426178,
+      "learning_rate": 0.0001,
+      "loss": 1.5206,
+      "step": 9269
+    },
+    {
+      "epoch": 1.0648440640974097,
+      "grad_norm": 0.4984760582447052,
+      "learning_rate": 0.0001,
+      "loss": 1.3982,
+      "step": 9270
+    },
+    {
+      "epoch": 1.0649589340072367,
+      "grad_norm": 0.5218265056610107,
+      "learning_rate": 0.0001,
+      "loss": 1.4516,
+      "step": 9271
+    },
+    {
+      "epoch": 1.065073803917064,
+      "grad_norm": 0.5094042420387268,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 9272
+    },
+    {
+      "epoch": 1.065188673826891,
+      "grad_norm": 0.5259179472923279,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 9273
+    },
+    {
+      "epoch": 1.0653035437367182,
+      "grad_norm": 0.4852963387966156,
+      "learning_rate": 0.0001,
+      "loss": 1.2565,
+      "step": 9274
+    },
+    {
+      "epoch": 1.0654184136465452,
+      "grad_norm": 0.5180572867393494,
+      "learning_rate": 0.0001,
+      "loss": 1.4978,
+      "step": 9275
+    },
+    {
+      "epoch": 1.0655332835563724,
+      "grad_norm": 0.5113215446472168,
+      "learning_rate": 0.0001,
+      "loss": 1.5048,
+      "step": 9276
+    },
+    {
+      "epoch": 1.0656481534661995,
+      "grad_norm": 0.5396996140480042,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 9277
+    },
+    {
+      "epoch": 1.0657630233760267,
+      "grad_norm": 0.5498672723770142,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 9278
+    },
+    {
+      "epoch": 1.0658778932858537,
+      "grad_norm": 0.5394907593727112,
+      "learning_rate": 0.0001,
+      "loss": 1.4526,
+      "step": 9279
+    },
+    {
+      "epoch": 1.065992763195681,
+      "grad_norm": 0.5096206665039062,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 9280
+    },
+    {
+      "epoch": 1.066107633105508,
+      "grad_norm": 0.49931639432907104,
+      "learning_rate": 0.0001,
+      "loss": 1.3194,
+      "step": 9281
+    },
+    {
+      "epoch": 1.0662225030153352,
+      "grad_norm": 0.5749096274375916,
+      "learning_rate": 0.0001,
+      "loss": 1.634,
+      "step": 9282
+    },
+    {
+      "epoch": 1.0663373729251622,
+      "grad_norm": 0.5646559000015259,
+      "learning_rate": 0.0001,
+      "loss": 1.5262,
+      "step": 9283
+    },
+    {
+      "epoch": 1.0664522428349894,
+      "grad_norm": 0.5303424000740051,
+      "learning_rate": 0.0001,
+      "loss": 1.2186,
+      "step": 9284
+    },
+    {
+      "epoch": 1.0665671127448164,
+      "grad_norm": 0.5843081474304199,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 9285
+    },
+    {
+      "epoch": 1.0666819826546436,
+      "grad_norm": 0.506814181804657,
+      "learning_rate": 0.0001,
+      "loss": 1.2877,
+      "step": 9286
+    },
+    {
+      "epoch": 1.0667968525644707,
+      "grad_norm": 0.5348950028419495,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 9287
+    },
+    {
+      "epoch": 1.0669117224742979,
+      "grad_norm": 0.5017212629318237,
+      "learning_rate": 0.0001,
+      "loss": 1.4226,
+      "step": 9288
+    },
+    {
+      "epoch": 1.067026592384125,
+      "grad_norm": 0.5333235859870911,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 9289
+    },
+    {
+      "epoch": 1.0671414622939521,
+      "grad_norm": 0.4682038426399231,
+      "learning_rate": 0.0001,
+      "loss": 1.3008,
+      "step": 9290
+    },
+    {
+      "epoch": 1.0672563322037791,
+      "grad_norm": 0.4889301061630249,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 9291
+    },
+    {
+      "epoch": 1.0673712021136064,
+      "grad_norm": 0.52678382396698,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 9292
+    },
+    {
+      "epoch": 1.0674860720234334,
+      "grad_norm": 0.5729466676712036,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 9293
+    },
+    {
+      "epoch": 1.0676009419332606,
+      "grad_norm": 0.5632121562957764,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 9294
+    },
+    {
+      "epoch": 1.0677158118430876,
+      "grad_norm": 0.5657988786697388,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 9295
+    },
+    {
+      "epoch": 1.0678306817529148,
+      "grad_norm": 0.5650860071182251,
+      "learning_rate": 0.0001,
+      "loss": 1.4778,
+      "step": 9296
+    },
+    {
+      "epoch": 1.0679455516627419,
+      "grad_norm": 0.5312690138816833,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 9297
+    },
+    {
+      "epoch": 1.068060421572569,
+      "grad_norm": 0.5419080853462219,
+      "learning_rate": 0.0001,
+      "loss": 1.544,
+      "step": 9298
+    },
+    {
+      "epoch": 1.068175291482396,
+      "grad_norm": 0.5316713452339172,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 9299
+    },
+    {
+      "epoch": 1.0682901613922233,
+      "grad_norm": 0.5229997634887695,
+      "learning_rate": 0.0001,
+      "loss": 1.333,
+      "step": 9300
+    },
+    {
+      "epoch": 1.0684050313020503,
+      "grad_norm": 0.5338220596313477,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 9301
+    },
+    {
+      "epoch": 1.0685199012118776,
+      "grad_norm": 0.5003429651260376,
+      "learning_rate": 0.0001,
+      "loss": 1.3834,
+      "step": 9302
+    },
+    {
+      "epoch": 1.0686347711217046,
+      "grad_norm": 0.5164094567298889,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 9303
+    },
+    {
+      "epoch": 1.0687496410315318,
+      "grad_norm": 0.5205984115600586,
+      "learning_rate": 0.0001,
+      "loss": 1.2881,
+      "step": 9304
+    },
+    {
+      "epoch": 1.0688645109413588,
+      "grad_norm": 0.5387089252471924,
+      "learning_rate": 0.0001,
+      "loss": 1.4241,
+      "step": 9305
+    },
+    {
+      "epoch": 1.068979380851186,
+      "grad_norm": 0.5411062836647034,
+      "learning_rate": 0.0001,
+      "loss": 1.4488,
+      "step": 9306
+    },
+    {
+      "epoch": 1.069094250761013,
+      "grad_norm": 0.522082507610321,
+      "learning_rate": 0.0001,
+      "loss": 1.48,
+      "step": 9307
+    },
+    {
+      "epoch": 1.0692091206708403,
+      "grad_norm": 0.5525926351547241,
+      "learning_rate": 0.0001,
+      "loss": 1.359,
+      "step": 9308
+    },
+    {
+      "epoch": 1.0693239905806673,
+      "grad_norm": 0.6042261719703674,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 9309
+    },
+    {
+      "epoch": 1.0694388604904945,
+      "grad_norm": 0.5462369918823242,
+      "learning_rate": 0.0001,
+      "loss": 1.713,
+      "step": 9310
+    },
+    {
+      "epoch": 1.0695537304003215,
+      "grad_norm": 0.5386967658996582,
+      "learning_rate": 0.0001,
+      "loss": 1.4737,
+      "step": 9311
+    },
+    {
+      "epoch": 1.0696686003101488,
+      "grad_norm": 0.5753105282783508,
+      "learning_rate": 0.0001,
+      "loss": 1.7007,
+      "step": 9312
+    },
+    {
+      "epoch": 1.0697834702199758,
+      "grad_norm": 0.509254515171051,
+      "learning_rate": 0.0001,
+      "loss": 1.3826,
+      "step": 9313
+    },
+    {
+      "epoch": 1.069898340129803,
+      "grad_norm": 0.5281634330749512,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 9314
+    },
+    {
+      "epoch": 1.07001321003963,
+      "grad_norm": 0.5292205214500427,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 9315
+    },
+    {
+      "epoch": 1.0701280799494572,
+      "grad_norm": 0.522649347782135,
+      "learning_rate": 0.0001,
+      "loss": 1.4171,
+      "step": 9316
+    },
+    {
+      "epoch": 1.0702429498592845,
+      "grad_norm": 0.5478872656822205,
+      "learning_rate": 0.0001,
+      "loss": 1.5693,
+      "step": 9317
+    },
+    {
+      "epoch": 1.0703578197691115,
+      "grad_norm": 0.5441164374351501,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 9318
+    },
+    {
+      "epoch": 1.0704726896789385,
+      "grad_norm": 0.5701571702957153,
+      "learning_rate": 0.0001,
+      "loss": 1.4287,
+      "step": 9319
+    },
+    {
+      "epoch": 1.0705875595887657,
+      "grad_norm": 0.5586724281311035,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 9320
+    },
+    {
+      "epoch": 1.070702429498593,
+      "grad_norm": 0.5121877789497375,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 9321
+    },
+    {
+      "epoch": 1.07081729940842,
+      "grad_norm": 0.5557401776313782,
+      "learning_rate": 0.0001,
+      "loss": 1.4617,
+      "step": 9322
+    },
+    {
+      "epoch": 1.070932169318247,
+      "grad_norm": 0.5157942175865173,
+      "learning_rate": 0.0001,
+      "loss": 1.5935,
+      "step": 9323
+    },
+    {
+      "epoch": 1.0710470392280742,
+      "grad_norm": 0.5756588578224182,
+      "learning_rate": 0.0001,
+      "loss": 1.5785,
+      "step": 9324
+    },
+    {
+      "epoch": 1.0711619091379014,
+      "grad_norm": 0.5151377320289612,
+      "learning_rate": 0.0001,
+      "loss": 1.4551,
+      "step": 9325
+    },
+    {
+      "epoch": 1.0712767790477284,
+      "grad_norm": 0.5527158975601196,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 9326
+    },
+    {
+      "epoch": 1.0713916489575555,
+      "grad_norm": 0.517097532749176,
+      "learning_rate": 0.0001,
+      "loss": 1.2917,
+      "step": 9327
+    },
+    {
+      "epoch": 1.0715065188673827,
+      "grad_norm": 0.5368314385414124,
+      "learning_rate": 0.0001,
+      "loss": 1.4936,
+      "step": 9328
+    },
+    {
+      "epoch": 1.07162138877721,
+      "grad_norm": 0.5495009422302246,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 9329
+    },
+    {
+      "epoch": 1.071736258687037,
+      "grad_norm": 0.5558755993843079,
+      "learning_rate": 0.0001,
+      "loss": 1.4963,
+      "step": 9330
+    },
+    {
+      "epoch": 1.0718511285968642,
+      "grad_norm": 0.5218509435653687,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 9331
+    },
+    {
+      "epoch": 1.0719659985066912,
+      "grad_norm": 0.5873258709907532,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 9332
+    },
+    {
+      "epoch": 1.0720808684165184,
+      "grad_norm": 0.542597234249115,
+      "learning_rate": 0.0001,
+      "loss": 1.4964,
+      "step": 9333
+    },
+    {
+      "epoch": 1.0721957383263454,
+      "grad_norm": 0.5635952949523926,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 9334
+    },
+    {
+      "epoch": 1.0723106082361726,
+      "grad_norm": 0.5015137791633606,
+      "learning_rate": 0.0001,
+      "loss": 1.409,
+      "step": 9335
+    },
+    {
+      "epoch": 1.0724254781459996,
+      "grad_norm": 0.5085453987121582,
+      "learning_rate": 0.0001,
+      "loss": 1.4133,
+      "step": 9336
+    },
+    {
+      "epoch": 1.0725403480558269,
+      "grad_norm": 0.5499141216278076,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 9337
+    },
+    {
+      "epoch": 1.0726552179656539,
+      "grad_norm": 0.5491046905517578,
+      "learning_rate": 0.0001,
+      "loss": 1.389,
+      "step": 9338
+    },
+    {
+      "epoch": 1.0727700878754811,
+      "grad_norm": 0.58219975233078,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 9339
+    },
+    {
+      "epoch": 1.0728849577853081,
+      "grad_norm": 0.5443148612976074,
+      "learning_rate": 0.0001,
+      "loss": 1.4332,
+      "step": 9340
+    },
+    {
+      "epoch": 1.0729998276951354,
+      "grad_norm": 0.5684245824813843,
+      "learning_rate": 0.0001,
+      "loss": 1.5124,
+      "step": 9341
+    },
+    {
+      "epoch": 1.0731146976049624,
+      "grad_norm": 0.6064300537109375,
+      "learning_rate": 0.0001,
+      "loss": 1.5588,
+      "step": 9342
+    },
+    {
+      "epoch": 1.0732295675147896,
+      "grad_norm": 0.5258216857910156,
+      "learning_rate": 0.0001,
+      "loss": 1.3385,
+      "step": 9343
+    },
+    {
+      "epoch": 1.0733444374246166,
+      "grad_norm": 0.5391144752502441,
+      "learning_rate": 0.0001,
+      "loss": 1.4488,
+      "step": 9344
+    },
+    {
+      "epoch": 1.0734593073344438,
+      "grad_norm": 0.5189342498779297,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 9345
+    },
+    {
+      "epoch": 1.0735741772442708,
+      "grad_norm": 0.527388870716095,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 9346
+    },
+    {
+      "epoch": 1.073689047154098,
+      "grad_norm": 0.569824755191803,
+      "learning_rate": 0.0001,
+      "loss": 1.2466,
+      "step": 9347
+    },
+    {
+      "epoch": 1.073803917063925,
+      "grad_norm": 0.5429185628890991,
+      "learning_rate": 0.0001,
+      "loss": 1.4456,
+      "step": 9348
+    },
+    {
+      "epoch": 1.0739187869737523,
+      "grad_norm": 0.5612589120864868,
+      "learning_rate": 0.0001,
+      "loss": 1.2836,
+      "step": 9349
+    },
+    {
+      "epoch": 1.0740336568835793,
+      "grad_norm": 0.562879204750061,
+      "learning_rate": 0.0001,
+      "loss": 1.2671,
+      "step": 9350
+    },
+    {
+      "epoch": 1.0741485267934066,
+      "grad_norm": 0.6129858493804932,
+      "learning_rate": 0.0001,
+      "loss": 1.6126,
+      "step": 9351
+    },
+    {
+      "epoch": 1.0742633967032336,
+      "grad_norm": 0.5683223009109497,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 9352
+    },
+    {
+      "epoch": 1.0743782666130608,
+      "grad_norm": 0.549032986164093,
+      "learning_rate": 0.0001,
+      "loss": 1.3925,
+      "step": 9353
+    },
+    {
+      "epoch": 1.0744931365228878,
+      "grad_norm": 0.5353028774261475,
+      "learning_rate": 0.0001,
+      "loss": 1.4364,
+      "step": 9354
+    },
+    {
+      "epoch": 1.074608006432715,
+      "grad_norm": 0.5511956810951233,
+      "learning_rate": 0.0001,
+      "loss": 1.6354,
+      "step": 9355
+    },
+    {
+      "epoch": 1.074722876342542,
+      "grad_norm": 0.5705670118331909,
+      "learning_rate": 0.0001,
+      "loss": 1.3258,
+      "step": 9356
+    },
+    {
+      "epoch": 1.0748377462523693,
+      "grad_norm": 0.5503584742546082,
+      "learning_rate": 0.0001,
+      "loss": 1.3599,
+      "step": 9357
+    },
+    {
+      "epoch": 1.0749526161621963,
+      "grad_norm": 0.48784855008125305,
+      "learning_rate": 0.0001,
+      "loss": 1.3298,
+      "step": 9358
+    },
+    {
+      "epoch": 1.0750674860720235,
+      "grad_norm": 0.5487546324729919,
+      "learning_rate": 0.0001,
+      "loss": 1.3159,
+      "step": 9359
+    },
+    {
+      "epoch": 1.0751823559818505,
+      "grad_norm": 0.6427622437477112,
+      "learning_rate": 0.0001,
+      "loss": 1.4016,
+      "step": 9360
+    },
+    {
+      "epoch": 1.0752972258916778,
+      "grad_norm": 0.5325155258178711,
+      "learning_rate": 0.0001,
+      "loss": 1.4784,
+      "step": 9361
+    },
+    {
+      "epoch": 1.0754120958015048,
+      "grad_norm": 0.5536829829216003,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 9362
+    },
+    {
+      "epoch": 1.075526965711332,
+      "grad_norm": 0.49843937158584595,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 9363
+    },
+    {
+      "epoch": 1.075641835621159,
+      "grad_norm": 0.5196369290351868,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 9364
+    },
+    {
+      "epoch": 1.0757567055309862,
+      "grad_norm": 0.5939671993255615,
+      "learning_rate": 0.0001,
+      "loss": 1.532,
+      "step": 9365
+    },
+    {
+      "epoch": 1.0758715754408132,
+      "grad_norm": 0.5654057860374451,
+      "learning_rate": 0.0001,
+      "loss": 1.7005,
+      "step": 9366
+    },
+    {
+      "epoch": 1.0759864453506405,
+      "grad_norm": 0.5311039686203003,
+      "learning_rate": 0.0001,
+      "loss": 1.4949,
+      "step": 9367
+    },
+    {
+      "epoch": 1.0761013152604675,
+      "grad_norm": 0.5418568849563599,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 9368
+    },
+    {
+      "epoch": 1.0762161851702947,
+      "grad_norm": 0.5042740106582642,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 9369
+    },
+    {
+      "epoch": 1.0763310550801217,
+      "grad_norm": 0.5391519665718079,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 9370
+    },
+    {
+      "epoch": 1.076445924989949,
+      "grad_norm": 0.48869240283966064,
+      "learning_rate": 0.0001,
+      "loss": 1.3108,
+      "step": 9371
+    },
+    {
+      "epoch": 1.076560794899776,
+      "grad_norm": 0.5286645889282227,
+      "learning_rate": 0.0001,
+      "loss": 1.6169,
+      "step": 9372
+    },
+    {
+      "epoch": 1.0766756648096032,
+      "grad_norm": 0.5066417455673218,
+      "learning_rate": 0.0001,
+      "loss": 1.2543,
+      "step": 9373
+    },
+    {
+      "epoch": 1.0767905347194302,
+      "grad_norm": 0.5383975505828857,
+      "learning_rate": 0.0001,
+      "loss": 1.2786,
+      "step": 9374
+    },
+    {
+      "epoch": 1.0769054046292574,
+      "grad_norm": 0.5614655017852783,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 9375
+    },
+    {
+      "epoch": 1.0770202745390844,
+      "grad_norm": 0.5506321787834167,
+      "learning_rate": 0.0001,
+      "loss": 1.4668,
+      "step": 9376
+    },
+    {
+      "epoch": 1.0771351444489117,
+      "grad_norm": 0.6028522253036499,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 9377
+    },
+    {
+      "epoch": 1.0772500143587387,
+      "grad_norm": 0.535199761390686,
+      "learning_rate": 0.0001,
+      "loss": 1.5098,
+      "step": 9378
+    },
+    {
+      "epoch": 1.077364884268566,
+      "grad_norm": 0.5362054705619812,
+      "learning_rate": 0.0001,
+      "loss": 1.6251,
+      "step": 9379
+    },
+    {
+      "epoch": 1.077479754178393,
+      "grad_norm": 0.5681738257408142,
+      "learning_rate": 0.0001,
+      "loss": 1.6243,
+      "step": 9380
+    },
+    {
+      "epoch": 1.0775946240882202,
+      "grad_norm": 0.5353713631629944,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 9381
+    },
+    {
+      "epoch": 1.0777094939980472,
+      "grad_norm": 0.5305677652359009,
+      "learning_rate": 0.0001,
+      "loss": 1.5472,
+      "step": 9382
+    },
+    {
+      "epoch": 1.0778243639078744,
+      "grad_norm": 0.5264517068862915,
+      "learning_rate": 0.0001,
+      "loss": 1.3933,
+      "step": 9383
+    },
+    {
+      "epoch": 1.0779392338177014,
+      "grad_norm": 0.5325207114219666,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 9384
+    },
+    {
+      "epoch": 1.0780541037275286,
+      "grad_norm": 0.5129626989364624,
+      "learning_rate": 0.0001,
+      "loss": 1.392,
+      "step": 9385
+    },
+    {
+      "epoch": 1.0781689736373556,
+      "grad_norm": 0.5592479109764099,
+      "learning_rate": 0.0001,
+      "loss": 1.4948,
+      "step": 9386
+    },
+    {
+      "epoch": 1.0782838435471829,
+      "grad_norm": 0.5640060901641846,
+      "learning_rate": 0.0001,
+      "loss": 1.4874,
+      "step": 9387
+    },
+    {
+      "epoch": 1.0783987134570099,
+      "grad_norm": 0.5315003991127014,
+      "learning_rate": 0.0001,
+      "loss": 1.4705,
+      "step": 9388
+    },
+    {
+      "epoch": 1.0785135833668371,
+      "grad_norm": 0.5562155842781067,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 9389
+    },
+    {
+      "epoch": 1.0786284532766641,
+      "grad_norm": 0.5401049256324768,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 9390
+    },
+    {
+      "epoch": 1.0787433231864914,
+      "grad_norm": 0.5349462032318115,
+      "learning_rate": 0.0001,
+      "loss": 1.471,
+      "step": 9391
+    },
+    {
+      "epoch": 1.0788581930963184,
+      "grad_norm": 0.543636679649353,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 9392
+    },
+    {
+      "epoch": 1.0789730630061456,
+      "grad_norm": 0.5344265103340149,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 9393
+    },
+    {
+      "epoch": 1.0790879329159726,
+      "grad_norm": 0.5137317180633545,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 9394
+    },
+    {
+      "epoch": 1.0792028028257998,
+      "grad_norm": 0.5402754545211792,
+      "learning_rate": 0.0001,
+      "loss": 1.422,
+      "step": 9395
+    },
+    {
+      "epoch": 1.0793176727356268,
+      "grad_norm": 0.6155654191970825,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 9396
+    },
+    {
+      "epoch": 1.079432542645454,
+      "grad_norm": 0.5445329546928406,
+      "learning_rate": 0.0001,
+      "loss": 1.3084,
+      "step": 9397
+    },
+    {
+      "epoch": 1.079547412555281,
+      "grad_norm": 0.5331481099128723,
+      "learning_rate": 0.0001,
+      "loss": 1.2886,
+      "step": 9398
+    },
+    {
+      "epoch": 1.0796622824651083,
+      "grad_norm": 0.4990472197532654,
+      "learning_rate": 0.0001,
+      "loss": 1.4427,
+      "step": 9399
+    },
+    {
+      "epoch": 1.0797771523749353,
+      "grad_norm": 0.5740399956703186,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 9400
+    },
+    {
+      "epoch": 1.0798920222847626,
+      "grad_norm": 0.5508021712303162,
+      "learning_rate": 0.0001,
+      "loss": 1.725,
+      "step": 9401
+    },
+    {
+      "epoch": 1.0800068921945896,
+      "grad_norm": 0.5360609889030457,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 9402
+    },
+    {
+      "epoch": 1.0801217621044168,
+      "grad_norm": 0.4839753806591034,
+      "learning_rate": 0.0001,
+      "loss": 1.3208,
+      "step": 9403
+    },
+    {
+      "epoch": 1.0802366320142438,
+      "grad_norm": 0.5315743684768677,
+      "learning_rate": 0.0001,
+      "loss": 1.4254,
+      "step": 9404
+    },
+    {
+      "epoch": 1.080351501924071,
+      "grad_norm": 0.62214195728302,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 9405
+    },
+    {
+      "epoch": 1.080466371833898,
+      "grad_norm": 0.5492658615112305,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 9406
+    },
+    {
+      "epoch": 1.0805812417437253,
+      "grad_norm": 0.5915753841400146,
+      "learning_rate": 0.0001,
+      "loss": 1.6729,
+      "step": 9407
+    },
+    {
+      "epoch": 1.0806961116535523,
+      "grad_norm": 0.49104854464530945,
+      "learning_rate": 0.0001,
+      "loss": 1.284,
+      "step": 9408
+    },
+    {
+      "epoch": 1.0808109815633795,
+      "grad_norm": 0.5419100522994995,
+      "learning_rate": 0.0001,
+      "loss": 1.3962,
+      "step": 9409
+    },
+    {
+      "epoch": 1.0809258514732065,
+      "grad_norm": 0.5406461358070374,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 9410
+    },
+    {
+      "epoch": 1.0810407213830338,
+      "grad_norm": 0.5958341360092163,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 9411
+    },
+    {
+      "epoch": 1.0811555912928608,
+      "grad_norm": 0.5315396189689636,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 9412
+    },
+    {
+      "epoch": 1.081270461202688,
+      "grad_norm": 0.5059750080108643,
+      "learning_rate": 0.0001,
+      "loss": 1.3242,
+      "step": 9413
+    },
+    {
+      "epoch": 1.081385331112515,
+      "grad_norm": 0.5587995052337646,
+      "learning_rate": 0.0001,
+      "loss": 1.446,
+      "step": 9414
+    },
+    {
+      "epoch": 1.0815002010223422,
+      "grad_norm": 0.5836725831031799,
+      "learning_rate": 0.0001,
+      "loss": 1.6805,
+      "step": 9415
+    },
+    {
+      "epoch": 1.0816150709321692,
+      "grad_norm": 0.5699766278266907,
+      "learning_rate": 0.0001,
+      "loss": 1.3352,
+      "step": 9416
+    },
+    {
+      "epoch": 1.0817299408419965,
+      "grad_norm": 0.5126357078552246,
+      "learning_rate": 0.0001,
+      "loss": 1.3888,
+      "step": 9417
+    },
+    {
+      "epoch": 1.0818448107518235,
+      "grad_norm": 0.5475831627845764,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 9418
+    },
+    {
+      "epoch": 1.0819596806616507,
+      "grad_norm": 0.5603591203689575,
+      "learning_rate": 0.0001,
+      "loss": 1.5595,
+      "step": 9419
+    },
+    {
+      "epoch": 1.0820745505714777,
+      "grad_norm": 0.5825121998786926,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 9420
+    },
+    {
+      "epoch": 1.082189420481305,
+      "grad_norm": 0.5266685485839844,
+      "learning_rate": 0.0001,
+      "loss": 1.5018,
+      "step": 9421
+    },
+    {
+      "epoch": 1.082304290391132,
+      "grad_norm": 0.5489354133605957,
+      "learning_rate": 0.0001,
+      "loss": 1.3732,
+      "step": 9422
+    },
+    {
+      "epoch": 1.0824191603009592,
+      "grad_norm": 0.5415682792663574,
+      "learning_rate": 0.0001,
+      "loss": 1.4901,
+      "step": 9423
+    },
+    {
+      "epoch": 1.0825340302107862,
+      "grad_norm": 0.5336417555809021,
+      "learning_rate": 0.0001,
+      "loss": 1.4311,
+      "step": 9424
+    },
+    {
+      "epoch": 1.0826489001206134,
+      "grad_norm": 0.5154435038566589,
+      "learning_rate": 0.0001,
+      "loss": 1.4462,
+      "step": 9425
+    },
+    {
+      "epoch": 1.0827637700304404,
+      "grad_norm": 0.5345861911773682,
+      "learning_rate": 0.0001,
+      "loss": 1.3614,
+      "step": 9426
+    },
+    {
+      "epoch": 1.0828786399402677,
+      "grad_norm": 0.5687355399131775,
+      "learning_rate": 0.0001,
+      "loss": 1.4912,
+      "step": 9427
+    },
+    {
+      "epoch": 1.0829935098500947,
+      "grad_norm": 0.508594810962677,
+      "learning_rate": 0.0001,
+      "loss": 1.471,
+      "step": 9428
+    },
+    {
+      "epoch": 1.083108379759922,
+      "grad_norm": 0.5026642084121704,
+      "learning_rate": 0.0001,
+      "loss": 1.3456,
+      "step": 9429
+    },
+    {
+      "epoch": 1.083223249669749,
+      "grad_norm": 0.5498985648155212,
+      "learning_rate": 0.0001,
+      "loss": 1.3439,
+      "step": 9430
+    },
+    {
+      "epoch": 1.0833381195795762,
+      "grad_norm": 0.5488834977149963,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 9431
+    },
+    {
+      "epoch": 1.0834529894894032,
+      "grad_norm": 0.5577558279037476,
+      "learning_rate": 0.0001,
+      "loss": 1.4781,
+      "step": 9432
+    },
+    {
+      "epoch": 1.0835678593992304,
+      "grad_norm": 0.5541431307792664,
+      "learning_rate": 0.0001,
+      "loss": 1.5649,
+      "step": 9433
+    },
+    {
+      "epoch": 1.0836827293090574,
+      "grad_norm": 0.6540974378585815,
+      "learning_rate": 0.0001,
+      "loss": 1.7835,
+      "step": 9434
+    },
+    {
+      "epoch": 1.0837975992188846,
+      "grad_norm": 0.6181141138076782,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 9435
+    },
+    {
+      "epoch": 1.0839124691287116,
+      "grad_norm": 0.5547922849655151,
+      "learning_rate": 0.0001,
+      "loss": 1.4657,
+      "step": 9436
+    },
+    {
+      "epoch": 1.0840273390385389,
+      "grad_norm": 0.5842359066009521,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 9437
+    },
+    {
+      "epoch": 1.0841422089483659,
+      "grad_norm": 0.5383824110031128,
+      "learning_rate": 0.0001,
+      "loss": 1.5272,
+      "step": 9438
+    },
+    {
+      "epoch": 1.0842570788581931,
+      "grad_norm": 0.5450701713562012,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 9439
+    },
+    {
+      "epoch": 1.0843719487680201,
+      "grad_norm": 0.5500561594963074,
+      "learning_rate": 0.0001,
+      "loss": 1.5371,
+      "step": 9440
+    },
+    {
+      "epoch": 1.0844868186778474,
+      "grad_norm": 0.6040080785751343,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 9441
+    },
+    {
+      "epoch": 1.0846016885876744,
+      "grad_norm": 0.5557257533073425,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 9442
+    },
+    {
+      "epoch": 1.0847165584975016,
+      "grad_norm": 0.5492468476295471,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 9443
+    },
+    {
+      "epoch": 1.0848314284073286,
+      "grad_norm": 0.5362855195999146,
+      "learning_rate": 0.0001,
+      "loss": 1.3116,
+      "step": 9444
+    },
+    {
+      "epoch": 1.0849462983171558,
+      "grad_norm": 0.491189569234848,
+      "learning_rate": 0.0001,
+      "loss": 1.1433,
+      "step": 9445
+    },
+    {
+      "epoch": 1.0850611682269828,
+      "grad_norm": 0.561486005783081,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 9446
+    },
+    {
+      "epoch": 1.08517603813681,
+      "grad_norm": 0.5247159600257874,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 9447
+    },
+    {
+      "epoch": 1.085290908046637,
+      "grad_norm": 0.5002375841140747,
+      "learning_rate": 0.0001,
+      "loss": 1.3614,
+      "step": 9448
+    },
+    {
+      "epoch": 1.0854057779564643,
+      "grad_norm": 0.5415583252906799,
+      "learning_rate": 0.0001,
+      "loss": 1.3459,
+      "step": 9449
+    },
+    {
+      "epoch": 1.0855206478662913,
+      "grad_norm": 0.5931771993637085,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 9450
+    },
+    {
+      "epoch": 1.0856355177761186,
+      "grad_norm": 0.6163171529769897,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 9451
+    },
+    {
+      "epoch": 1.0857503876859456,
+      "grad_norm": 0.5527577996253967,
+      "learning_rate": 0.0001,
+      "loss": 1.3209,
+      "step": 9452
+    },
+    {
+      "epoch": 1.0858652575957728,
+      "grad_norm": 0.5498605966567993,
+      "learning_rate": 0.0001,
+      "loss": 1.3325,
+      "step": 9453
+    },
+    {
+      "epoch": 1.0859801275056,
+      "grad_norm": 0.5718228220939636,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 9454
+    },
+    {
+      "epoch": 1.086094997415427,
+      "grad_norm": 0.537427544593811,
+      "learning_rate": 0.0001,
+      "loss": 1.3966,
+      "step": 9455
+    },
+    {
+      "epoch": 1.086209867325254,
+      "grad_norm": 0.5483804941177368,
+      "learning_rate": 0.0001,
+      "loss": 1.2163,
+      "step": 9456
+    },
+    {
+      "epoch": 1.0863247372350813,
+      "grad_norm": 0.5748531222343445,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 9457
+    },
+    {
+      "epoch": 1.0864396071449085,
+      "grad_norm": 0.5303325653076172,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 9458
+    },
+    {
+      "epoch": 1.0865544770547355,
+      "grad_norm": 0.5701431632041931,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 9459
+    },
+    {
+      "epoch": 1.0866693469645625,
+      "grad_norm": 0.5297312140464783,
+      "learning_rate": 0.0001,
+      "loss": 1.49,
+      "step": 9460
+    },
+    {
+      "epoch": 1.0867842168743898,
+      "grad_norm": 0.5745691061019897,
+      "learning_rate": 0.0001,
+      "loss": 1.6774,
+      "step": 9461
+    },
+    {
+      "epoch": 1.086899086784217,
+      "grad_norm": 0.5294743180274963,
+      "learning_rate": 0.0001,
+      "loss": 1.3204,
+      "step": 9462
+    },
+    {
+      "epoch": 1.087013956694044,
+      "grad_norm": 0.5992701649665833,
+      "learning_rate": 0.0001,
+      "loss": 1.3512,
+      "step": 9463
+    },
+    {
+      "epoch": 1.087128826603871,
+      "grad_norm": 0.5554730892181396,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 9464
+    },
+    {
+      "epoch": 1.0872436965136982,
+      "grad_norm": 0.555023193359375,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 9465
+    },
+    {
+      "epoch": 1.0873585664235255,
+      "grad_norm": 0.5271770358085632,
+      "learning_rate": 0.0001,
+      "loss": 1.2415,
+      "step": 9466
+    },
+    {
+      "epoch": 1.0874734363333525,
+      "grad_norm": 0.520576536655426,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 9467
+    },
+    {
+      "epoch": 1.0875883062431795,
+      "grad_norm": 0.5427173972129822,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 9468
+    },
+    {
+      "epoch": 1.0877031761530067,
+      "grad_norm": 0.5307012796401978,
+      "learning_rate": 0.0001,
+      "loss": 1.3187,
+      "step": 9469
+    },
+    {
+      "epoch": 1.087818046062834,
+      "grad_norm": 0.49438974261283875,
+      "learning_rate": 0.0001,
+      "loss": 1.3237,
+      "step": 9470
+    },
+    {
+      "epoch": 1.087932915972661,
+      "grad_norm": 0.5442484617233276,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 9471
+    },
+    {
+      "epoch": 1.0880477858824882,
+      "grad_norm": 0.5300331711769104,
+      "learning_rate": 0.0001,
+      "loss": 1.6,
+      "step": 9472
+    },
+    {
+      "epoch": 1.0881626557923152,
+      "grad_norm": 0.5837119817733765,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 9473
+    },
+    {
+      "epoch": 1.0882775257021424,
+      "grad_norm": 0.5622014999389648,
+      "learning_rate": 0.0001,
+      "loss": 1.5176,
+      "step": 9474
+    },
+    {
+      "epoch": 1.0883923956119694,
+      "grad_norm": 0.5535881519317627,
+      "learning_rate": 0.0001,
+      "loss": 1.4017,
+      "step": 9475
+    },
+    {
+      "epoch": 1.0885072655217967,
+      "grad_norm": 0.5968254804611206,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 9476
+    },
+    {
+      "epoch": 1.0886221354316237,
+      "grad_norm": 0.5204367637634277,
+      "learning_rate": 0.0001,
+      "loss": 1.4491,
+      "step": 9477
+    },
+    {
+      "epoch": 1.088737005341451,
+      "grad_norm": 0.5856305956840515,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 9478
+    },
+    {
+      "epoch": 1.088851875251278,
+      "grad_norm": 0.5582146048545837,
+      "learning_rate": 0.0001,
+      "loss": 1.5893,
+      "step": 9479
+    },
+    {
+      "epoch": 1.0889667451611051,
+      "grad_norm": 0.5371931195259094,
+      "learning_rate": 0.0001,
+      "loss": 1.4195,
+      "step": 9480
+    },
+    {
+      "epoch": 1.0890816150709322,
+      "grad_norm": 0.5727693438529968,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 9481
+    },
+    {
+      "epoch": 1.0891964849807594,
+      "grad_norm": 0.5168283581733704,
+      "learning_rate": 0.0001,
+      "loss": 1.4037,
+      "step": 9482
+    },
+    {
+      "epoch": 1.0893113548905864,
+      "grad_norm": 0.5147133469581604,
+      "learning_rate": 0.0001,
+      "loss": 1.3784,
+      "step": 9483
+    },
+    {
+      "epoch": 1.0894262248004136,
+      "grad_norm": 0.5450608730316162,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 9484
+    },
+    {
+      "epoch": 1.0895410947102406,
+      "grad_norm": 0.5697365403175354,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 9485
+    },
+    {
+      "epoch": 1.0896559646200679,
+      "grad_norm": 0.5053684711456299,
+      "learning_rate": 0.0001,
+      "loss": 1.3655,
+      "step": 9486
+    },
+    {
+      "epoch": 1.0897708345298949,
+      "grad_norm": 0.5777620673179626,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 9487
+    },
+    {
+      "epoch": 1.089885704439722,
+      "grad_norm": 0.5835351943969727,
+      "learning_rate": 0.0001,
+      "loss": 1.3695,
+      "step": 9488
+    },
+    {
+      "epoch": 1.0900005743495491,
+      "grad_norm": 0.5509904623031616,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 9489
+    },
+    {
+      "epoch": 1.0901154442593763,
+      "grad_norm": 0.5363668203353882,
+      "learning_rate": 0.0001,
+      "loss": 1.321,
+      "step": 9490
+    },
+    {
+      "epoch": 1.0902303141692034,
+      "grad_norm": 0.5192736983299255,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 9491
+    },
+    {
+      "epoch": 1.0903451840790306,
+      "grad_norm": 0.5779258608818054,
+      "learning_rate": 0.0001,
+      "loss": 1.4376,
+      "step": 9492
+    },
+    {
+      "epoch": 1.0904600539888576,
+      "grad_norm": 0.5055257678031921,
+      "learning_rate": 0.0001,
+      "loss": 1.3177,
+      "step": 9493
+    },
+    {
+      "epoch": 1.0905749238986848,
+      "grad_norm": 0.5672529935836792,
+      "learning_rate": 0.0001,
+      "loss": 1.5611,
+      "step": 9494
+    },
+    {
+      "epoch": 1.0906897938085118,
+      "grad_norm": 0.5167456865310669,
+      "learning_rate": 0.0001,
+      "loss": 1.4562,
+      "step": 9495
+    },
+    {
+      "epoch": 1.090804663718339,
+      "grad_norm": 0.522533118724823,
+      "learning_rate": 0.0001,
+      "loss": 1.4801,
+      "step": 9496
+    },
+    {
+      "epoch": 1.090919533628166,
+      "grad_norm": 0.5050001740455627,
+      "learning_rate": 0.0001,
+      "loss": 1.5072,
+      "step": 9497
+    },
+    {
+      "epoch": 1.0910344035379933,
+      "grad_norm": 0.5358273983001709,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 9498
+    },
+    {
+      "epoch": 1.0911492734478203,
+      "grad_norm": 0.5162444114685059,
+      "learning_rate": 0.0001,
+      "loss": 1.4414,
+      "step": 9499
+    },
+    {
+      "epoch": 1.0912641433576475,
+      "grad_norm": 0.533893346786499,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 9500
+    },
+    {
+      "epoch": 1.0913790132674746,
+      "grad_norm": 0.5578415989875793,
+      "learning_rate": 0.0001,
+      "loss": 1.5303,
+      "step": 9501
+    },
+    {
+      "epoch": 1.0914938831773018,
+      "grad_norm": 0.5179681777954102,
+      "learning_rate": 0.0001,
+      "loss": 1.4064,
+      "step": 9502
+    },
+    {
+      "epoch": 1.0916087530871288,
+      "grad_norm": 0.5305806994438171,
+      "learning_rate": 0.0001,
+      "loss": 1.5176,
+      "step": 9503
+    },
+    {
+      "epoch": 1.091723622996956,
+      "grad_norm": 0.5723759531974792,
+      "learning_rate": 0.0001,
+      "loss": 1.1344,
+      "step": 9504
+    },
+    {
+      "epoch": 1.091838492906783,
+      "grad_norm": 0.520270049571991,
+      "learning_rate": 0.0001,
+      "loss": 1.3782,
+      "step": 9505
+    },
+    {
+      "epoch": 1.0919533628166103,
+      "grad_norm": 0.5493273138999939,
+      "learning_rate": 0.0001,
+      "loss": 1.4887,
+      "step": 9506
+    },
+    {
+      "epoch": 1.0920682327264373,
+      "grad_norm": 0.5690314769744873,
+      "learning_rate": 0.0001,
+      "loss": 1.49,
+      "step": 9507
+    },
+    {
+      "epoch": 1.0921831026362645,
+      "grad_norm": 0.5376046895980835,
+      "learning_rate": 0.0001,
+      "loss": 1.4984,
+      "step": 9508
+    },
+    {
+      "epoch": 1.0922979725460915,
+      "grad_norm": 0.5700758695602417,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 9509
+    },
+    {
+      "epoch": 1.0924128424559187,
+      "grad_norm": 0.5583578944206238,
+      "learning_rate": 0.0001,
+      "loss": 1.4936,
+      "step": 9510
+    },
+    {
+      "epoch": 1.0925277123657458,
+      "grad_norm": 0.5892508625984192,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 9511
+    },
+    {
+      "epoch": 1.092642582275573,
+      "grad_norm": 0.5539071559906006,
+      "learning_rate": 0.0001,
+      "loss": 1.4654,
+      "step": 9512
+    },
+    {
+      "epoch": 1.0927574521854,
+      "grad_norm": 0.5687013268470764,
+      "learning_rate": 0.0001,
+      "loss": 1.4186,
+      "step": 9513
+    },
+    {
+      "epoch": 1.0928723220952272,
+      "grad_norm": 0.612593412399292,
+      "learning_rate": 0.0001,
+      "loss": 1.6754,
+      "step": 9514
+    },
+    {
+      "epoch": 1.0929871920050542,
+      "grad_norm": 0.5612267851829529,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 9515
+    },
+    {
+      "epoch": 1.0931020619148815,
+      "grad_norm": 0.5222997665405273,
+      "learning_rate": 0.0001,
+      "loss": 1.3799,
+      "step": 9516
+    },
+    {
+      "epoch": 1.0932169318247085,
+      "grad_norm": 0.5479604005813599,
+      "learning_rate": 0.0001,
+      "loss": 1.5003,
+      "step": 9517
+    },
+    {
+      "epoch": 1.0933318017345357,
+      "grad_norm": 0.5224593281745911,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 9518
+    },
+    {
+      "epoch": 1.0934466716443627,
+      "grad_norm": 0.524739682674408,
+      "learning_rate": 0.0001,
+      "loss": 1.3937,
+      "step": 9519
+    },
+    {
+      "epoch": 1.09356154155419,
+      "grad_norm": 0.5201284885406494,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 9520
+    },
+    {
+      "epoch": 1.093676411464017,
+      "grad_norm": 0.5348097681999207,
+      "learning_rate": 0.0001,
+      "loss": 1.3621,
+      "step": 9521
+    },
+    {
+      "epoch": 1.0937912813738442,
+      "grad_norm": 0.5572234988212585,
+      "learning_rate": 0.0001,
+      "loss": 1.3629,
+      "step": 9522
+    },
+    {
+      "epoch": 1.0939061512836712,
+      "grad_norm": 0.5672652721405029,
+      "learning_rate": 0.0001,
+      "loss": 1.6203,
+      "step": 9523
+    },
+    {
+      "epoch": 1.0940210211934984,
+      "grad_norm": 0.5444334149360657,
+      "learning_rate": 0.0001,
+      "loss": 1.7165,
+      "step": 9524
+    },
+    {
+      "epoch": 1.0941358911033254,
+      "grad_norm": 0.54752516746521,
+      "learning_rate": 0.0001,
+      "loss": 1.4252,
+      "step": 9525
+    },
+    {
+      "epoch": 1.0942507610131527,
+      "grad_norm": 0.49809858202934265,
+      "learning_rate": 0.0001,
+      "loss": 1.387,
+      "step": 9526
+    },
+    {
+      "epoch": 1.0943656309229797,
+      "grad_norm": 0.512008786201477,
+      "learning_rate": 0.0001,
+      "loss": 1.295,
+      "step": 9527
+    },
+    {
+      "epoch": 1.094480500832807,
+      "grad_norm": 0.5522312521934509,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 9528
+    },
+    {
+      "epoch": 1.094595370742634,
+      "grad_norm": 0.5582612752914429,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 9529
+    },
+    {
+      "epoch": 1.0947102406524611,
+      "grad_norm": 0.48772454261779785,
+      "learning_rate": 0.0001,
+      "loss": 1.3134,
+      "step": 9530
+    },
+    {
+      "epoch": 1.0948251105622882,
+      "grad_norm": 0.49887439608573914,
+      "learning_rate": 0.0001,
+      "loss": 1.2972,
+      "step": 9531
+    },
+    {
+      "epoch": 1.0949399804721154,
+      "grad_norm": 0.5389269590377808,
+      "learning_rate": 0.0001,
+      "loss": 1.5936,
+      "step": 9532
+    },
+    {
+      "epoch": 1.0950548503819424,
+      "grad_norm": 0.5234885215759277,
+      "learning_rate": 0.0001,
+      "loss": 1.3984,
+      "step": 9533
+    },
+    {
+      "epoch": 1.0951697202917696,
+      "grad_norm": 0.5532096028327942,
+      "learning_rate": 0.0001,
+      "loss": 1.7103,
+      "step": 9534
+    },
+    {
+      "epoch": 1.0952845902015966,
+      "grad_norm": 0.5728646516799927,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 9535
+    },
+    {
+      "epoch": 1.0953994601114239,
+      "grad_norm": 0.5275560021400452,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 9536
+    },
+    {
+      "epoch": 1.0955143300212509,
+      "grad_norm": 0.5168836116790771,
+      "learning_rate": 0.0001,
+      "loss": 1.4736,
+      "step": 9537
+    },
+    {
+      "epoch": 1.095629199931078,
+      "grad_norm": 0.5980050563812256,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 9538
+    },
+    {
+      "epoch": 1.095744069840905,
+      "grad_norm": 0.553145706653595,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 9539
+    },
+    {
+      "epoch": 1.0958589397507323,
+      "grad_norm": 0.5515928864479065,
+      "learning_rate": 0.0001,
+      "loss": 1.3338,
+      "step": 9540
+    },
+    {
+      "epoch": 1.0959738096605594,
+      "grad_norm": 0.5219181776046753,
+      "learning_rate": 0.0001,
+      "loss": 1.3579,
+      "step": 9541
+    },
+    {
+      "epoch": 1.0960886795703866,
+      "grad_norm": 0.4816209077835083,
+      "learning_rate": 0.0001,
+      "loss": 1.3447,
+      "step": 9542
+    },
+    {
+      "epoch": 1.0962035494802136,
+      "grad_norm": 0.5502380132675171,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 9543
+    },
+    {
+      "epoch": 1.0963184193900408,
+      "grad_norm": 0.5180455446243286,
+      "learning_rate": 0.0001,
+      "loss": 1.4077,
+      "step": 9544
+    },
+    {
+      "epoch": 1.0964332892998678,
+      "grad_norm": 0.524360716342926,
+      "learning_rate": 0.0001,
+      "loss": 1.3433,
+      "step": 9545
+    },
+    {
+      "epoch": 1.096548159209695,
+      "grad_norm": 0.5326607823371887,
+      "learning_rate": 0.0001,
+      "loss": 1.4864,
+      "step": 9546
+    },
+    {
+      "epoch": 1.096663029119522,
+      "grad_norm": 0.578113853931427,
+      "learning_rate": 0.0001,
+      "loss": 1.3848,
+      "step": 9547
+    },
+    {
+      "epoch": 1.0967778990293493,
+      "grad_norm": 0.5228033065795898,
+      "learning_rate": 0.0001,
+      "loss": 1.4384,
+      "step": 9548
+    },
+    {
+      "epoch": 1.0968927689391763,
+      "grad_norm": 0.5391709804534912,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 9549
+    },
+    {
+      "epoch": 1.0970076388490035,
+      "grad_norm": 0.5661507248878479,
+      "learning_rate": 0.0001,
+      "loss": 1.5187,
+      "step": 9550
+    },
+    {
+      "epoch": 1.0971225087588306,
+      "grad_norm": 0.5772484540939331,
+      "learning_rate": 0.0001,
+      "loss": 1.526,
+      "step": 9551
+    },
+    {
+      "epoch": 1.0972373786686578,
+      "grad_norm": 0.5205997824668884,
+      "learning_rate": 0.0001,
+      "loss": 1.2596,
+      "step": 9552
+    },
+    {
+      "epoch": 1.0973522485784848,
+      "grad_norm": 0.4969817101955414,
+      "learning_rate": 0.0001,
+      "loss": 1.3083,
+      "step": 9553
+    },
+    {
+      "epoch": 1.097467118488312,
+      "grad_norm": 0.5085188746452332,
+      "learning_rate": 0.0001,
+      "loss": 1.4074,
+      "step": 9554
+    },
+    {
+      "epoch": 1.097581988398139,
+      "grad_norm": 0.5870746970176697,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 9555
+    },
+    {
+      "epoch": 1.0976968583079663,
+      "grad_norm": 0.5884862542152405,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 9556
+    },
+    {
+      "epoch": 1.0978117282177933,
+      "grad_norm": 0.6046068668365479,
+      "learning_rate": 0.0001,
+      "loss": 1.5751,
+      "step": 9557
+    },
+    {
+      "epoch": 1.0979265981276205,
+      "grad_norm": 0.5737109780311584,
+      "learning_rate": 0.0001,
+      "loss": 1.5331,
+      "step": 9558
+    },
+    {
+      "epoch": 1.0980414680374475,
+      "grad_norm": 0.5459645390510559,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 9559
+    },
+    {
+      "epoch": 1.0981563379472747,
+      "grad_norm": 0.5659840106964111,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 9560
+    },
+    {
+      "epoch": 1.0982712078571017,
+      "grad_norm": 0.6050505042076111,
+      "learning_rate": 0.0001,
+      "loss": 1.4345,
+      "step": 9561
+    },
+    {
+      "epoch": 1.098386077766929,
+      "grad_norm": 0.5667203068733215,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 9562
+    },
+    {
+      "epoch": 1.098500947676756,
+      "grad_norm": 0.5636523962020874,
+      "learning_rate": 0.0001,
+      "loss": 1.6369,
+      "step": 9563
+    },
+    {
+      "epoch": 1.0986158175865832,
+      "grad_norm": 0.6931208372116089,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 9564
+    },
+    {
+      "epoch": 1.0987306874964102,
+      "grad_norm": 0.6440662145614624,
+      "learning_rate": 0.0001,
+      "loss": 1.7034,
+      "step": 9565
+    },
+    {
+      "epoch": 1.0988455574062375,
+      "grad_norm": 0.5347530245780945,
+      "learning_rate": 0.0001,
+      "loss": 1.4818,
+      "step": 9566
+    },
+    {
+      "epoch": 1.0989604273160645,
+      "grad_norm": 0.504833996295929,
+      "learning_rate": 0.0001,
+      "loss": 1.3958,
+      "step": 9567
+    },
+    {
+      "epoch": 1.0990752972258917,
+      "grad_norm": 0.5506554245948792,
+      "learning_rate": 0.0001,
+      "loss": 1.3691,
+      "step": 9568
+    },
+    {
+      "epoch": 1.0991901671357187,
+      "grad_norm": 0.573026180267334,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 9569
+    },
+    {
+      "epoch": 1.099305037045546,
+      "grad_norm": 0.5038986206054688,
+      "learning_rate": 0.0001,
+      "loss": 1.4027,
+      "step": 9570
+    },
+    {
+      "epoch": 1.099419906955373,
+      "grad_norm": 0.5019423961639404,
+      "learning_rate": 0.0001,
+      "loss": 1.239,
+      "step": 9571
+    },
+    {
+      "epoch": 1.0995347768652002,
+      "grad_norm": 0.5748392343521118,
+      "learning_rate": 0.0001,
+      "loss": 1.346,
+      "step": 9572
+    },
+    {
+      "epoch": 1.0996496467750272,
+      "grad_norm": 0.5261626243591309,
+      "learning_rate": 0.0001,
+      "loss": 1.3476,
+      "step": 9573
+    },
+    {
+      "epoch": 1.0997645166848544,
+      "grad_norm": 0.5816366672515869,
+      "learning_rate": 0.0001,
+      "loss": 1.6714,
+      "step": 9574
+    },
+    {
+      "epoch": 1.0998793865946814,
+      "grad_norm": 0.5417282581329346,
+      "learning_rate": 0.0001,
+      "loss": 1.3579,
+      "step": 9575
+    },
+    {
+      "epoch": 1.0999942565045087,
+      "grad_norm": 0.5732975602149963,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 9576
+    },
+    {
+      "epoch": 1.1001091264143357,
+      "grad_norm": 0.532828688621521,
+      "learning_rate": 0.0001,
+      "loss": 1.438,
+      "step": 9577
+    },
+    {
+      "epoch": 1.100223996324163,
+      "grad_norm": 0.5136977434158325,
+      "learning_rate": 0.0001,
+      "loss": 1.4017,
+      "step": 9578
+    },
+    {
+      "epoch": 1.10033886623399,
+      "grad_norm": 0.5484635233879089,
+      "learning_rate": 0.0001,
+      "loss": 1.5923,
+      "step": 9579
+    },
+    {
+      "epoch": 1.1004537361438171,
+      "grad_norm": 0.5523190498352051,
+      "learning_rate": 0.0001,
+      "loss": 1.4729,
+      "step": 9580
+    },
+    {
+      "epoch": 1.1005686060536441,
+      "grad_norm": 0.5286394357681274,
+      "learning_rate": 0.0001,
+      "loss": 1.5283,
+      "step": 9581
+    },
+    {
+      "epoch": 1.1006834759634714,
+      "grad_norm": 0.5363751649856567,
+      "learning_rate": 0.0001,
+      "loss": 1.3553,
+      "step": 9582
+    },
+    {
+      "epoch": 1.1007983458732984,
+      "grad_norm": 0.6055129766464233,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 9583
+    },
+    {
+      "epoch": 1.1009132157831256,
+      "grad_norm": 0.5629612803459167,
+      "learning_rate": 0.0001,
+      "loss": 1.1768,
+      "step": 9584
+    },
+    {
+      "epoch": 1.1010280856929526,
+      "grad_norm": 0.5564055442810059,
+      "learning_rate": 0.0001,
+      "loss": 1.4302,
+      "step": 9585
+    },
+    {
+      "epoch": 1.1011429556027799,
+      "grad_norm": 0.6236995458602905,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 9586
+    },
+    {
+      "epoch": 1.1012578255126069,
+      "grad_norm": 0.5221498012542725,
+      "learning_rate": 0.0001,
+      "loss": 1.4249,
+      "step": 9587
+    },
+    {
+      "epoch": 1.101372695422434,
+      "grad_norm": 0.5449710488319397,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 9588
+    },
+    {
+      "epoch": 1.101487565332261,
+      "grad_norm": 0.5365622043609619,
+      "learning_rate": 0.0001,
+      "loss": 1.4603,
+      "step": 9589
+    },
+    {
+      "epoch": 1.1016024352420883,
+      "grad_norm": 0.5306606888771057,
+      "learning_rate": 0.0001,
+      "loss": 1.232,
+      "step": 9590
+    },
+    {
+      "epoch": 1.1017173051519156,
+      "grad_norm": 0.5195592045783997,
+      "learning_rate": 0.0001,
+      "loss": 1.4095,
+      "step": 9591
+    },
+    {
+      "epoch": 1.1018321750617426,
+      "grad_norm": 0.5111302733421326,
+      "learning_rate": 0.0001,
+      "loss": 1.4183,
+      "step": 9592
+    },
+    {
+      "epoch": 1.1019470449715696,
+      "grad_norm": 0.6306554079055786,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 9593
+    },
+    {
+      "epoch": 1.1020619148813968,
+      "grad_norm": 0.5819112062454224,
+      "learning_rate": 0.0001,
+      "loss": 1.5985,
+      "step": 9594
+    },
+    {
+      "epoch": 1.102176784791224,
+      "grad_norm": 0.5284945368766785,
+      "learning_rate": 0.0001,
+      "loss": 1.457,
+      "step": 9595
+    },
+    {
+      "epoch": 1.102291654701051,
+      "grad_norm": 0.5951351523399353,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 9596
+    },
+    {
+      "epoch": 1.102406524610878,
+      "grad_norm": 0.5186183452606201,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 9597
+    },
+    {
+      "epoch": 1.1025213945207053,
+      "grad_norm": 0.5780056118965149,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 9598
+    },
+    {
+      "epoch": 1.1026362644305325,
+      "grad_norm": 0.5407369136810303,
+      "learning_rate": 0.0001,
+      "loss": 1.4405,
+      "step": 9599
+    },
+    {
+      "epoch": 1.1027511343403595,
+      "grad_norm": 0.5503196120262146,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 9600
+    },
+    {
+      "epoch": 1.1028660042501865,
+      "grad_norm": 0.5217320919036865,
+      "learning_rate": 0.0001,
+      "loss": 1.5486,
+      "step": 9601
+    },
+    {
+      "epoch": 1.1029808741600138,
+      "grad_norm": 0.5520971417427063,
+      "learning_rate": 0.0001,
+      "loss": 1.5373,
+      "step": 9602
+    },
+    {
+      "epoch": 1.103095744069841,
+      "grad_norm": 0.531853973865509,
+      "learning_rate": 0.0001,
+      "loss": 1.2797,
+      "step": 9603
+    },
+    {
+      "epoch": 1.103210613979668,
+      "grad_norm": 0.5419740676879883,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 9604
+    },
+    {
+      "epoch": 1.103325483889495,
+      "grad_norm": 0.5819375514984131,
+      "learning_rate": 0.0001,
+      "loss": 1.3623,
+      "step": 9605
+    },
+    {
+      "epoch": 1.1034403537993223,
+      "grad_norm": 0.5647183656692505,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 9606
+    },
+    {
+      "epoch": 1.1035552237091495,
+      "grad_norm": 0.5452742576599121,
+      "learning_rate": 0.0001,
+      "loss": 1.5939,
+      "step": 9607
+    },
+    {
+      "epoch": 1.1036700936189765,
+      "grad_norm": 0.5725043416023254,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 9608
+    },
+    {
+      "epoch": 1.1037849635288037,
+      "grad_norm": 0.5144510269165039,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 9609
+    },
+    {
+      "epoch": 1.1038998334386307,
+      "grad_norm": 0.5603897571563721,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 9610
+    },
+    {
+      "epoch": 1.104014703348458,
+      "grad_norm": 0.5356127023696899,
+      "learning_rate": 0.0001,
+      "loss": 1.2814,
+      "step": 9611
+    },
+    {
+      "epoch": 1.104129573258285,
+      "grad_norm": 0.5167484879493713,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 9612
+    },
+    {
+      "epoch": 1.1042444431681122,
+      "grad_norm": 0.5493837594985962,
+      "learning_rate": 0.0001,
+      "loss": 1.226,
+      "step": 9613
+    },
+    {
+      "epoch": 1.1043593130779392,
+      "grad_norm": 0.5558966398239136,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 9614
+    },
+    {
+      "epoch": 1.1044741829877665,
+      "grad_norm": 0.5911669135093689,
+      "learning_rate": 0.0001,
+      "loss": 1.4025,
+      "step": 9615
+    },
+    {
+      "epoch": 1.1045890528975935,
+      "grad_norm": 0.5736728310585022,
+      "learning_rate": 0.0001,
+      "loss": 1.2921,
+      "step": 9616
+    },
+    {
+      "epoch": 1.1047039228074207,
+      "grad_norm": 0.55420982837677,
+      "learning_rate": 0.0001,
+      "loss": 1.4201,
+      "step": 9617
+    },
+    {
+      "epoch": 1.1048187927172477,
+      "grad_norm": 0.6312325596809387,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 9618
+    },
+    {
+      "epoch": 1.104933662627075,
+      "grad_norm": 0.5303093194961548,
+      "learning_rate": 0.0001,
+      "loss": 1.3404,
+      "step": 9619
+    },
+    {
+      "epoch": 1.105048532536902,
+      "grad_norm": 0.5138710737228394,
+      "learning_rate": 0.0001,
+      "loss": 1.4007,
+      "step": 9620
+    },
+    {
+      "epoch": 1.1051634024467292,
+      "grad_norm": 0.6012336015701294,
+      "learning_rate": 0.0001,
+      "loss": 1.632,
+      "step": 9621
+    },
+    {
+      "epoch": 1.1052782723565562,
+      "grad_norm": 0.5534988641738892,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 9622
+    },
+    {
+      "epoch": 1.1053931422663834,
+      "grad_norm": 0.5863144993782043,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 9623
+    },
+    {
+      "epoch": 1.1055080121762104,
+      "grad_norm": 0.5527574419975281,
+      "learning_rate": 0.0001,
+      "loss": 1.6297,
+      "step": 9624
+    },
+    {
+      "epoch": 1.1056228820860377,
+      "grad_norm": 0.5487685203552246,
+      "learning_rate": 0.0001,
+      "loss": 1.3806,
+      "step": 9625
+    },
+    {
+      "epoch": 1.1057377519958647,
+      "grad_norm": 0.541068971157074,
+      "learning_rate": 0.0001,
+      "loss": 1.5204,
+      "step": 9626
+    },
+    {
+      "epoch": 1.105852621905692,
+      "grad_norm": 0.5597257018089294,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 9627
+    },
+    {
+      "epoch": 1.105967491815519,
+      "grad_norm": 0.5184458494186401,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 9628
+    },
+    {
+      "epoch": 1.1060823617253461,
+      "grad_norm": 0.6078442931175232,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 9629
+    },
+    {
+      "epoch": 1.1061972316351731,
+      "grad_norm": 0.5217808485031128,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 9630
+    },
+    {
+      "epoch": 1.1063121015450004,
+      "grad_norm": 0.5544619560241699,
+      "learning_rate": 0.0001,
+      "loss": 1.5177,
+      "step": 9631
+    },
+    {
+      "epoch": 1.1064269714548274,
+      "grad_norm": 0.5434312224388123,
+      "learning_rate": 0.0001,
+      "loss": 1.497,
+      "step": 9632
+    },
+    {
+      "epoch": 1.1065418413646546,
+      "grad_norm": 0.5772722959518433,
+      "learning_rate": 0.0001,
+      "loss": 1.2362,
+      "step": 9633
+    },
+    {
+      "epoch": 1.1066567112744816,
+      "grad_norm": 0.571907103061676,
+      "learning_rate": 0.0001,
+      "loss": 1.4295,
+      "step": 9634
+    },
+    {
+      "epoch": 1.1067715811843089,
+      "grad_norm": 0.5757927894592285,
+      "learning_rate": 0.0001,
+      "loss": 1.437,
+      "step": 9635
+    },
+    {
+      "epoch": 1.1068864510941359,
+      "grad_norm": 0.5998075008392334,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 9636
+    },
+    {
+      "epoch": 1.107001321003963,
+      "grad_norm": 0.5600723624229431,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 9637
+    },
+    {
+      "epoch": 1.10711619091379,
+      "grad_norm": 0.5362415909767151,
+      "learning_rate": 0.0001,
+      "loss": 1.4906,
+      "step": 9638
+    },
+    {
+      "epoch": 1.1072310608236173,
+      "grad_norm": 0.5417195558547974,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 9639
+    },
+    {
+      "epoch": 1.1073459307334443,
+      "grad_norm": 0.5452782511711121,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 9640
+    },
+    {
+      "epoch": 1.1074608006432716,
+      "grad_norm": 0.565497100353241,
+      "learning_rate": 0.0001,
+      "loss": 1.626,
+      "step": 9641
+    },
+    {
+      "epoch": 1.1075756705530986,
+      "grad_norm": 0.5098233222961426,
+      "learning_rate": 0.0001,
+      "loss": 1.2684,
+      "step": 9642
+    },
+    {
+      "epoch": 1.1076905404629258,
+      "grad_norm": 0.5035114288330078,
+      "learning_rate": 0.0001,
+      "loss": 1.4774,
+      "step": 9643
+    },
+    {
+      "epoch": 1.1078054103727528,
+      "grad_norm": 0.5948421359062195,
+      "learning_rate": 0.0001,
+      "loss": 1.198,
+      "step": 9644
+    },
+    {
+      "epoch": 1.10792028028258,
+      "grad_norm": 0.5490888357162476,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 9645
+    },
+    {
+      "epoch": 1.108035150192407,
+      "grad_norm": 0.5749881267547607,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 9646
+    },
+    {
+      "epoch": 1.1081500201022343,
+      "grad_norm": 0.5411537885665894,
+      "learning_rate": 0.0001,
+      "loss": 1.4558,
+      "step": 9647
+    },
+    {
+      "epoch": 1.1082648900120613,
+      "grad_norm": 0.5704604387283325,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 9648
+    },
+    {
+      "epoch": 1.1083797599218885,
+      "grad_norm": 0.5247997641563416,
+      "learning_rate": 0.0001,
+      "loss": 1.2063,
+      "step": 9649
+    },
+    {
+      "epoch": 1.1084946298317155,
+      "grad_norm": 0.5762208700180054,
+      "learning_rate": 0.0001,
+      "loss": 1.6697,
+      "step": 9650
+    },
+    {
+      "epoch": 1.1086094997415428,
+      "grad_norm": 0.5289835333824158,
+      "learning_rate": 0.0001,
+      "loss": 1.3045,
+      "step": 9651
+    },
+    {
+      "epoch": 1.1087243696513698,
+      "grad_norm": 0.5333086848258972,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 9652
+    },
+    {
+      "epoch": 1.108839239561197,
+      "grad_norm": 0.5131956338882446,
+      "learning_rate": 0.0001,
+      "loss": 1.3351,
+      "step": 9653
+    },
+    {
+      "epoch": 1.108954109471024,
+      "grad_norm": 0.5169805884361267,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 9654
+    },
+    {
+      "epoch": 1.1090689793808512,
+      "grad_norm": 0.5443220734596252,
+      "learning_rate": 0.0001,
+      "loss": 1.5177,
+      "step": 9655
+    },
+    {
+      "epoch": 1.1091838492906783,
+      "grad_norm": 0.5020636916160583,
+      "learning_rate": 0.0001,
+      "loss": 1.2518,
+      "step": 9656
+    },
+    {
+      "epoch": 1.1092987192005055,
+      "grad_norm": 0.49906498193740845,
+      "learning_rate": 0.0001,
+      "loss": 1.3971,
+      "step": 9657
+    },
+    {
+      "epoch": 1.1094135891103325,
+      "grad_norm": 0.5414373874664307,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 9658
+    },
+    {
+      "epoch": 1.1095284590201597,
+      "grad_norm": 0.5560583472251892,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 9659
+    },
+    {
+      "epoch": 1.1096433289299867,
+      "grad_norm": 0.5157898664474487,
+      "learning_rate": 0.0001,
+      "loss": 1.3337,
+      "step": 9660
+    },
+    {
+      "epoch": 1.109758198839814,
+      "grad_norm": 0.5900468826293945,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 9661
+    },
+    {
+      "epoch": 1.109873068749641,
+      "grad_norm": 0.5296939611434937,
+      "learning_rate": 0.0001,
+      "loss": 1.468,
+      "step": 9662
+    },
+    {
+      "epoch": 1.1099879386594682,
+      "grad_norm": 0.5018689632415771,
+      "learning_rate": 0.0001,
+      "loss": 1.4018,
+      "step": 9663
+    },
+    {
+      "epoch": 1.1101028085692952,
+      "grad_norm": 0.4990765452384949,
+      "learning_rate": 0.0001,
+      "loss": 1.3802,
+      "step": 9664
+    },
+    {
+      "epoch": 1.1102176784791224,
+      "grad_norm": 0.4909820556640625,
+      "learning_rate": 0.0001,
+      "loss": 1.3892,
+      "step": 9665
+    },
+    {
+      "epoch": 1.1103325483889495,
+      "grad_norm": 0.5232742428779602,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 9666
+    },
+    {
+      "epoch": 1.1104474182987767,
+      "grad_norm": 0.5467817783355713,
+      "learning_rate": 0.0001,
+      "loss": 1.3905,
+      "step": 9667
+    },
+    {
+      "epoch": 1.1105622882086037,
+      "grad_norm": 0.49867168068885803,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 9668
+    },
+    {
+      "epoch": 1.110677158118431,
+      "grad_norm": 0.5683446526527405,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 9669
+    },
+    {
+      "epoch": 1.110792028028258,
+      "grad_norm": 0.5260730385780334,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 9670
+    },
+    {
+      "epoch": 1.1109068979380852,
+      "grad_norm": 0.5530440807342529,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 9671
+    },
+    {
+      "epoch": 1.1110217678479122,
+      "grad_norm": 0.597644567489624,
+      "learning_rate": 0.0001,
+      "loss": 1.6696,
+      "step": 9672
+    },
+    {
+      "epoch": 1.1111366377577394,
+      "grad_norm": 0.5474144220352173,
+      "learning_rate": 0.0001,
+      "loss": 1.4307,
+      "step": 9673
+    },
+    {
+      "epoch": 1.1112515076675664,
+      "grad_norm": 0.499236136674881,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 9674
+    },
+    {
+      "epoch": 1.1113663775773936,
+      "grad_norm": 0.5453770756721497,
+      "learning_rate": 0.0001,
+      "loss": 1.3632,
+      "step": 9675
+    },
+    {
+      "epoch": 1.1114812474872207,
+      "grad_norm": 0.5170800685882568,
+      "learning_rate": 0.0001,
+      "loss": 1.4903,
+      "step": 9676
+    },
+    {
+      "epoch": 1.111596117397048,
+      "grad_norm": 0.5183429718017578,
+      "learning_rate": 0.0001,
+      "loss": 1.3797,
+      "step": 9677
+    },
+    {
+      "epoch": 1.111710987306875,
+      "grad_norm": 0.5546143054962158,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 9678
+    },
+    {
+      "epoch": 1.1118258572167021,
+      "grad_norm": 0.5528146028518677,
+      "learning_rate": 0.0001,
+      "loss": 1.3667,
+      "step": 9679
+    },
+    {
+      "epoch": 1.1119407271265291,
+      "grad_norm": 0.5532962679862976,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 9680
+    },
+    {
+      "epoch": 1.1120555970363564,
+      "grad_norm": 0.5805869698524475,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 9681
+    },
+    {
+      "epoch": 1.1121704669461834,
+      "grad_norm": 0.5238086581230164,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 9682
+    },
+    {
+      "epoch": 1.1122853368560106,
+      "grad_norm": 0.5135483741760254,
+      "learning_rate": 0.0001,
+      "loss": 1.3031,
+      "step": 9683
+    },
+    {
+      "epoch": 1.1124002067658376,
+      "grad_norm": 0.535915732383728,
+      "learning_rate": 0.0001,
+      "loss": 1.4422,
+      "step": 9684
+    },
+    {
+      "epoch": 1.1125150766756648,
+      "grad_norm": 0.5621902346611023,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 9685
+    },
+    {
+      "epoch": 1.1126299465854919,
+      "grad_norm": 0.5478909611701965,
+      "learning_rate": 0.0001,
+      "loss": 1.3292,
+      "step": 9686
+    },
+    {
+      "epoch": 1.112744816495319,
+      "grad_norm": 0.5914627313613892,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 9687
+    },
+    {
+      "epoch": 1.112859686405146,
+      "grad_norm": 0.555410623550415,
+      "learning_rate": 0.0001,
+      "loss": 1.4208,
+      "step": 9688
+    },
+    {
+      "epoch": 1.1129745563149733,
+      "grad_norm": 0.6287572979927063,
+      "learning_rate": 0.0001,
+      "loss": 1.5024,
+      "step": 9689
+    },
+    {
+      "epoch": 1.1130894262248003,
+      "grad_norm": 0.5535327792167664,
+      "learning_rate": 0.0001,
+      "loss": 1.283,
+      "step": 9690
+    },
+    {
+      "epoch": 1.1132042961346276,
+      "grad_norm": 0.6567327976226807,
+      "learning_rate": 0.0001,
+      "loss": 1.6511,
+      "step": 9691
+    },
+    {
+      "epoch": 1.1133191660444546,
+      "grad_norm": 0.5440387725830078,
+      "learning_rate": 0.0001,
+      "loss": 1.4729,
+      "step": 9692
+    },
+    {
+      "epoch": 1.1134340359542818,
+      "grad_norm": 0.576004683971405,
+      "learning_rate": 0.0001,
+      "loss": 1.5678,
+      "step": 9693
+    },
+    {
+      "epoch": 1.1135489058641088,
+      "grad_norm": 0.5710541605949402,
+      "learning_rate": 0.0001,
+      "loss": 1.5312,
+      "step": 9694
+    },
+    {
+      "epoch": 1.113663775773936,
+      "grad_norm": 0.5407060384750366,
+      "learning_rate": 0.0001,
+      "loss": 1.4169,
+      "step": 9695
+    },
+    {
+      "epoch": 1.113778645683763,
+      "grad_norm": 0.618705689907074,
+      "learning_rate": 0.0001,
+      "loss": 1.7292,
+      "step": 9696
+    },
+    {
+      "epoch": 1.1138935155935903,
+      "grad_norm": 0.5561045408248901,
+      "learning_rate": 0.0001,
+      "loss": 1.4788,
+      "step": 9697
+    },
+    {
+      "epoch": 1.1140083855034173,
+      "grad_norm": 0.590338945388794,
+      "learning_rate": 0.0001,
+      "loss": 1.4074,
+      "step": 9698
+    },
+    {
+      "epoch": 1.1141232554132445,
+      "grad_norm": 0.5989879369735718,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 9699
+    },
+    {
+      "epoch": 1.1142381253230715,
+      "grad_norm": 0.5720906853675842,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 9700
+    },
+    {
+      "epoch": 1.1143529952328988,
+      "grad_norm": 0.53765869140625,
+      "learning_rate": 0.0001,
+      "loss": 1.2634,
+      "step": 9701
+    },
+    {
+      "epoch": 1.1144678651427258,
+      "grad_norm": 0.5853990316390991,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 9702
+    },
+    {
+      "epoch": 1.114582735052553,
+      "grad_norm": 0.5783793926239014,
+      "learning_rate": 0.0001,
+      "loss": 1.5545,
+      "step": 9703
+    },
+    {
+      "epoch": 1.11469760496238,
+      "grad_norm": 0.5769542455673218,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 9704
+    },
+    {
+      "epoch": 1.1148124748722072,
+      "grad_norm": 0.56031334400177,
+      "learning_rate": 0.0001,
+      "loss": 1.4986,
+      "step": 9705
+    },
+    {
+      "epoch": 1.1149273447820343,
+      "grad_norm": 0.5601339936256409,
+      "learning_rate": 0.0001,
+      "loss": 1.2713,
+      "step": 9706
+    },
+    {
+      "epoch": 1.1150422146918615,
+      "grad_norm": 0.552882730960846,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 9707
+    },
+    {
+      "epoch": 1.1151570846016885,
+      "grad_norm": 0.513989269733429,
+      "learning_rate": 0.0001,
+      "loss": 1.4661,
+      "step": 9708
+    },
+    {
+      "epoch": 1.1152719545115157,
+      "grad_norm": 0.5489285588264465,
+      "learning_rate": 0.0001,
+      "loss": 1.456,
+      "step": 9709
+    },
+    {
+      "epoch": 1.1153868244213427,
+      "grad_norm": 0.5569881200790405,
+      "learning_rate": 0.0001,
+      "loss": 1.5918,
+      "step": 9710
+    },
+    {
+      "epoch": 1.11550169433117,
+      "grad_norm": 0.5199733972549438,
+      "learning_rate": 0.0001,
+      "loss": 1.3066,
+      "step": 9711
+    },
+    {
+      "epoch": 1.115616564240997,
+      "grad_norm": 0.5051583051681519,
+      "learning_rate": 0.0001,
+      "loss": 1.4748,
+      "step": 9712
+    },
+    {
+      "epoch": 1.1157314341508242,
+      "grad_norm": 0.5694435238838196,
+      "learning_rate": 0.0001,
+      "loss": 1.5148,
+      "step": 9713
+    },
+    {
+      "epoch": 1.1158463040606512,
+      "grad_norm": 0.5439704656600952,
+      "learning_rate": 0.0001,
+      "loss": 1.4556,
+      "step": 9714
+    },
+    {
+      "epoch": 1.1159611739704784,
+      "grad_norm": 0.5158405900001526,
+      "learning_rate": 0.0001,
+      "loss": 1.3826,
+      "step": 9715
+    },
+    {
+      "epoch": 1.1160760438803055,
+      "grad_norm": 0.5488100647926331,
+      "learning_rate": 0.0001,
+      "loss": 1.5873,
+      "step": 9716
+    },
+    {
+      "epoch": 1.1161909137901327,
+      "grad_norm": 0.5777077078819275,
+      "learning_rate": 0.0001,
+      "loss": 1.3642,
+      "step": 9717
+    },
+    {
+      "epoch": 1.1163057836999597,
+      "grad_norm": 0.5701835751533508,
+      "learning_rate": 0.0001,
+      "loss": 1.2141,
+      "step": 9718
+    },
+    {
+      "epoch": 1.116420653609787,
+      "grad_norm": 0.5419429540634155,
+      "learning_rate": 0.0001,
+      "loss": 1.4514,
+      "step": 9719
+    },
+    {
+      "epoch": 1.116535523519614,
+      "grad_norm": 0.5637021660804749,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 9720
+    },
+    {
+      "epoch": 1.1166503934294412,
+      "grad_norm": 0.6277347803115845,
+      "learning_rate": 0.0001,
+      "loss": 1.6107,
+      "step": 9721
+    },
+    {
+      "epoch": 1.1167652633392682,
+      "grad_norm": 0.6077524423599243,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 9722
+    },
+    {
+      "epoch": 1.1168801332490954,
+      "grad_norm": 0.579602837562561,
+      "learning_rate": 0.0001,
+      "loss": 1.5203,
+      "step": 9723
+    },
+    {
+      "epoch": 1.1169950031589224,
+      "grad_norm": 0.6332294344902039,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 9724
+    },
+    {
+      "epoch": 1.1171098730687496,
+      "grad_norm": 0.5484510064125061,
+      "learning_rate": 0.0001,
+      "loss": 1.4893,
+      "step": 9725
+    },
+    {
+      "epoch": 1.1172247429785767,
+      "grad_norm": 0.5461903214454651,
+      "learning_rate": 0.0001,
+      "loss": 1.424,
+      "step": 9726
+    },
+    {
+      "epoch": 1.1173396128884039,
+      "grad_norm": 0.5597689151763916,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 9727
+    },
+    {
+      "epoch": 1.1174544827982311,
+      "grad_norm": 0.5503045916557312,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 9728
+    },
+    {
+      "epoch": 1.1175693527080581,
+      "grad_norm": 0.5259599685668945,
+      "learning_rate": 0.0001,
+      "loss": 1.4662,
+      "step": 9729
+    },
+    {
+      "epoch": 1.1176842226178851,
+      "grad_norm": 0.5283674001693726,
+      "learning_rate": 0.0001,
+      "loss": 1.4768,
+      "step": 9730
+    },
+    {
+      "epoch": 1.1177990925277124,
+      "grad_norm": 0.5296016335487366,
+      "learning_rate": 0.0001,
+      "loss": 1.2765,
+      "step": 9731
+    },
+    {
+      "epoch": 1.1179139624375396,
+      "grad_norm": 0.528243899345398,
+      "learning_rate": 0.0001,
+      "loss": 1.3637,
+      "step": 9732
+    },
+    {
+      "epoch": 1.1180288323473666,
+      "grad_norm": 0.5380933880805969,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 9733
+    },
+    {
+      "epoch": 1.1181437022571936,
+      "grad_norm": 0.5619294047355652,
+      "learning_rate": 0.0001,
+      "loss": 1.3127,
+      "step": 9734
+    },
+    {
+      "epoch": 1.1182585721670208,
+      "grad_norm": 0.5538292527198792,
+      "learning_rate": 0.0001,
+      "loss": 1.4159,
+      "step": 9735
+    },
+    {
+      "epoch": 1.118373442076848,
+      "grad_norm": 0.5651335120201111,
+      "learning_rate": 0.0001,
+      "loss": 1.3661,
+      "step": 9736
+    },
+    {
+      "epoch": 1.118488311986675,
+      "grad_norm": 0.5316612720489502,
+      "learning_rate": 0.0001,
+      "loss": 1.5018,
+      "step": 9737
+    },
+    {
+      "epoch": 1.118603181896502,
+      "grad_norm": 0.5468516945838928,
+      "learning_rate": 0.0001,
+      "loss": 1.3585,
+      "step": 9738
+    },
+    {
+      "epoch": 1.1187180518063293,
+      "grad_norm": 0.60384202003479,
+      "learning_rate": 0.0001,
+      "loss": 1.5244,
+      "step": 9739
+    },
+    {
+      "epoch": 1.1188329217161566,
+      "grad_norm": 0.5294783711433411,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 9740
+    },
+    {
+      "epoch": 1.1189477916259836,
+      "grad_norm": 0.5466625690460205,
+      "learning_rate": 0.0001,
+      "loss": 1.5823,
+      "step": 9741
+    },
+    {
+      "epoch": 1.1190626615358106,
+      "grad_norm": 0.5397646427154541,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 9742
+    },
+    {
+      "epoch": 1.1191775314456378,
+      "grad_norm": 0.525026798248291,
+      "learning_rate": 0.0001,
+      "loss": 1.4643,
+      "step": 9743
+    },
+    {
+      "epoch": 1.119292401355465,
+      "grad_norm": 0.5274828672409058,
+      "learning_rate": 0.0001,
+      "loss": 1.2221,
+      "step": 9744
+    },
+    {
+      "epoch": 1.119407271265292,
+      "grad_norm": 0.5290265679359436,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 9745
+    },
+    {
+      "epoch": 1.1195221411751193,
+      "grad_norm": 0.5176373720169067,
+      "learning_rate": 0.0001,
+      "loss": 1.3871,
+      "step": 9746
+    },
+    {
+      "epoch": 1.1196370110849463,
+      "grad_norm": 0.5549317598342896,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 9747
+    },
+    {
+      "epoch": 1.1197518809947735,
+      "grad_norm": 0.5477745532989502,
+      "learning_rate": 0.0001,
+      "loss": 1.4009,
+      "step": 9748
+    },
+    {
+      "epoch": 1.1198667509046005,
+      "grad_norm": 0.5613062381744385,
+      "learning_rate": 0.0001,
+      "loss": 1.3842,
+      "step": 9749
+    },
+    {
+      "epoch": 1.1199816208144278,
+      "grad_norm": 0.5442327260971069,
+      "learning_rate": 0.0001,
+      "loss": 1.4955,
+      "step": 9750
+    },
+    {
+      "epoch": 1.1200964907242548,
+      "grad_norm": 0.5651378631591797,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 9751
+    },
+    {
+      "epoch": 1.120211360634082,
+      "grad_norm": 0.5338380336761475,
+      "learning_rate": 0.0001,
+      "loss": 1.2772,
+      "step": 9752
+    },
+    {
+      "epoch": 1.120326230543909,
+      "grad_norm": 0.5371922850608826,
+      "learning_rate": 0.0001,
+      "loss": 1.4111,
+      "step": 9753
+    },
+    {
+      "epoch": 1.1204411004537362,
+      "grad_norm": 0.49218225479125977,
+      "learning_rate": 0.0001,
+      "loss": 1.3432,
+      "step": 9754
+    },
+    {
+      "epoch": 1.1205559703635632,
+      "grad_norm": 0.5054636001586914,
+      "learning_rate": 0.0001,
+      "loss": 1.2946,
+      "step": 9755
+    },
+    {
+      "epoch": 1.1206708402733905,
+      "grad_norm": 0.5905545949935913,
+      "learning_rate": 0.0001,
+      "loss": 1.5725,
+      "step": 9756
+    },
+    {
+      "epoch": 1.1207857101832175,
+      "grad_norm": 0.5989205241203308,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 9757
+    },
+    {
+      "epoch": 1.1209005800930447,
+      "grad_norm": 0.5933013558387756,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 9758
+    },
+    {
+      "epoch": 1.1210154500028717,
+      "grad_norm": 0.5364729166030884,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 9759
+    },
+    {
+      "epoch": 1.121130319912699,
+      "grad_norm": 0.638589084148407,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 9760
+    },
+    {
+      "epoch": 1.121245189822526,
+      "grad_norm": 0.5210257172584534,
+      "learning_rate": 0.0001,
+      "loss": 1.3005,
+      "step": 9761
+    },
+    {
+      "epoch": 1.1213600597323532,
+      "grad_norm": 0.581584095954895,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 9762
+    },
+    {
+      "epoch": 1.1214749296421802,
+      "grad_norm": 0.5900195240974426,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 9763
+    },
+    {
+      "epoch": 1.1215897995520074,
+      "grad_norm": 0.5471107959747314,
+      "learning_rate": 0.0001,
+      "loss": 1.457,
+      "step": 9764
+    },
+    {
+      "epoch": 1.1217046694618344,
+      "grad_norm": 0.5870269536972046,
+      "learning_rate": 0.0001,
+      "loss": 1.3416,
+      "step": 9765
+    },
+    {
+      "epoch": 1.1218195393716617,
+      "grad_norm": 0.6071326732635498,
+      "learning_rate": 0.0001,
+      "loss": 1.362,
+      "step": 9766
+    },
+    {
+      "epoch": 1.1219344092814887,
+      "grad_norm": 0.5291370749473572,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 9767
+    },
+    {
+      "epoch": 1.122049279191316,
+      "grad_norm": 0.5899519920349121,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 9768
+    },
+    {
+      "epoch": 1.122164149101143,
+      "grad_norm": 0.5426907539367676,
+      "learning_rate": 0.0001,
+      "loss": 1.3719,
+      "step": 9769
+    },
+    {
+      "epoch": 1.1222790190109702,
+      "grad_norm": 0.634436309337616,
+      "learning_rate": 0.0001,
+      "loss": 1.6449,
+      "step": 9770
+    },
+    {
+      "epoch": 1.1223938889207972,
+      "grad_norm": 0.5499168038368225,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 9771
+    },
+    {
+      "epoch": 1.1225087588306244,
+      "grad_norm": 0.5340597629547119,
+      "learning_rate": 0.0001,
+      "loss": 1.5949,
+      "step": 9772
+    },
+    {
+      "epoch": 1.1226236287404514,
+      "grad_norm": 0.5629338622093201,
+      "learning_rate": 0.0001,
+      "loss": 1.4928,
+      "step": 9773
+    },
+    {
+      "epoch": 1.1227384986502786,
+      "grad_norm": 0.5041453242301941,
+      "learning_rate": 0.0001,
+      "loss": 1.2472,
+      "step": 9774
+    },
+    {
+      "epoch": 1.1228533685601056,
+      "grad_norm": 0.5762504935264587,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 9775
+    },
+    {
+      "epoch": 1.1229682384699329,
+      "grad_norm": 0.5604518055915833,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 9776
+    },
+    {
+      "epoch": 1.1230831083797599,
+      "grad_norm": 0.5645705461502075,
+      "learning_rate": 0.0001,
+      "loss": 1.2915,
+      "step": 9777
+    },
+    {
+      "epoch": 1.1231979782895871,
+      "grad_norm": 0.47870761156082153,
+      "learning_rate": 0.0001,
+      "loss": 1.2856,
+      "step": 9778
+    },
+    {
+      "epoch": 1.1233128481994141,
+      "grad_norm": 0.5456746220588684,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 9779
+    },
+    {
+      "epoch": 1.1234277181092414,
+      "grad_norm": 0.5596235394477844,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 9780
+    },
+    {
+      "epoch": 1.1235425880190684,
+      "grad_norm": 0.5665818452835083,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 9781
+    },
+    {
+      "epoch": 1.1236574579288956,
+      "grad_norm": 0.605762779712677,
+      "learning_rate": 0.0001,
+      "loss": 1.3177,
+      "step": 9782
+    },
+    {
+      "epoch": 1.1237723278387226,
+      "grad_norm": 0.5420957207679749,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 9783
+    },
+    {
+      "epoch": 1.1238871977485498,
+      "grad_norm": 0.6387947201728821,
+      "learning_rate": 0.0001,
+      "loss": 1.3037,
+      "step": 9784
+    },
+    {
+      "epoch": 1.1240020676583768,
+      "grad_norm": 0.5580546259880066,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 9785
+    },
+    {
+      "epoch": 1.124116937568204,
+      "grad_norm": 0.5426499843597412,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 9786
+    },
+    {
+      "epoch": 1.124231807478031,
+      "grad_norm": 0.5494298934936523,
+      "learning_rate": 0.0001,
+      "loss": 1.565,
+      "step": 9787
+    },
+    {
+      "epoch": 1.1243466773878583,
+      "grad_norm": 0.5614138245582581,
+      "learning_rate": 0.0001,
+      "loss": 1.4699,
+      "step": 9788
+    },
+    {
+      "epoch": 1.1244615472976853,
+      "grad_norm": 0.5716959238052368,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 9789
+    },
+    {
+      "epoch": 1.1245764172075126,
+      "grad_norm": 0.5259473919868469,
+      "learning_rate": 0.0001,
+      "loss": 1.4069,
+      "step": 9790
+    },
+    {
+      "epoch": 1.1246912871173396,
+      "grad_norm": 0.5336620211601257,
+      "learning_rate": 0.0001,
+      "loss": 1.5229,
+      "step": 9791
+    },
+    {
+      "epoch": 1.1248061570271668,
+      "grad_norm": 0.5299500226974487,
+      "learning_rate": 0.0001,
+      "loss": 1.5239,
+      "step": 9792
+    },
+    {
+      "epoch": 1.1249210269369938,
+      "grad_norm": 0.5537680387496948,
+      "learning_rate": 0.0001,
+      "loss": 1.3462,
+      "step": 9793
+    },
+    {
+      "epoch": 1.125035896846821,
+      "grad_norm": 0.5444726943969727,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 9794
+    },
+    {
+      "epoch": 1.125150766756648,
+      "grad_norm": 0.5653538107872009,
+      "learning_rate": 0.0001,
+      "loss": 1.7283,
+      "step": 9795
+    },
+    {
+      "epoch": 1.1252656366664753,
+      "grad_norm": 0.5434619784355164,
+      "learning_rate": 0.0001,
+      "loss": 1.2987,
+      "step": 9796
+    },
+    {
+      "epoch": 1.1253805065763023,
+      "grad_norm": 0.5725109577178955,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 9797
+    },
+    {
+      "epoch": 1.1254953764861295,
+      "grad_norm": 0.6203802227973938,
+      "learning_rate": 0.0001,
+      "loss": 1.5608,
+      "step": 9798
+    },
+    {
+      "epoch": 1.1256102463959565,
+      "grad_norm": 0.5914912223815918,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 9799
+    },
+    {
+      "epoch": 1.1257251163057838,
+      "grad_norm": 0.7401518225669861,
+      "learning_rate": 0.0001,
+      "loss": 1.5393,
+      "step": 9800
+    },
+    {
+      "epoch": 1.1258399862156108,
+      "grad_norm": 0.54278165102005,
+      "learning_rate": 0.0001,
+      "loss": 1.3736,
+      "step": 9801
+    },
+    {
+      "epoch": 1.125954856125438,
+      "grad_norm": 0.5903399586677551,
+      "learning_rate": 0.0001,
+      "loss": 1.6015,
+      "step": 9802
+    },
+    {
+      "epoch": 1.126069726035265,
+      "grad_norm": 0.5835331678390503,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 9803
+    },
+    {
+      "epoch": 1.1261845959450922,
+      "grad_norm": 0.5370138883590698,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 9804
+    },
+    {
+      "epoch": 1.1262994658549192,
+      "grad_norm": 0.6110140681266785,
+      "learning_rate": 0.0001,
+      "loss": 1.7323,
+      "step": 9805
+    },
+    {
+      "epoch": 1.1264143357647465,
+      "grad_norm": 0.5909807085990906,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 9806
+    },
+    {
+      "epoch": 1.1265292056745735,
+      "grad_norm": 0.5648689270019531,
+      "learning_rate": 0.0001,
+      "loss": 1.4891,
+      "step": 9807
+    },
+    {
+      "epoch": 1.1266440755844007,
+      "grad_norm": 0.6406177282333374,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 9808
+    },
+    {
+      "epoch": 1.1267589454942277,
+      "grad_norm": 0.5629665851593018,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 9809
+    },
+    {
+      "epoch": 1.126873815404055,
+      "grad_norm": 0.5698572993278503,
+      "learning_rate": 0.0001,
+      "loss": 1.456,
+      "step": 9810
+    },
+    {
+      "epoch": 1.126988685313882,
+      "grad_norm": 0.5605424046516418,
+      "learning_rate": 0.0001,
+      "loss": 1.3712,
+      "step": 9811
+    },
+    {
+      "epoch": 1.1271035552237092,
+      "grad_norm": 0.5104329586029053,
+      "learning_rate": 0.0001,
+      "loss": 1.372,
+      "step": 9812
+    },
+    {
+      "epoch": 1.1272184251335362,
+      "grad_norm": 0.5657603144645691,
+      "learning_rate": 0.0001,
+      "loss": 1.3791,
+      "step": 9813
+    },
+    {
+      "epoch": 1.1273332950433634,
+      "grad_norm": 0.5272459983825684,
+      "learning_rate": 0.0001,
+      "loss": 1.3004,
+      "step": 9814
+    },
+    {
+      "epoch": 1.1274481649531904,
+      "grad_norm": 0.5541790127754211,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 9815
+    },
+    {
+      "epoch": 1.1275630348630177,
+      "grad_norm": 0.5871829986572266,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 9816
+    },
+    {
+      "epoch": 1.1276779047728447,
+      "grad_norm": 0.5780866742134094,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 9817
+    },
+    {
+      "epoch": 1.127792774682672,
+      "grad_norm": 0.606268048286438,
+      "learning_rate": 0.0001,
+      "loss": 1.733,
+      "step": 9818
+    },
+    {
+      "epoch": 1.127907644592499,
+      "grad_norm": 0.559883713722229,
+      "learning_rate": 0.0001,
+      "loss": 1.4383,
+      "step": 9819
+    },
+    {
+      "epoch": 1.1280225145023262,
+      "grad_norm": 0.5222516655921936,
+      "learning_rate": 0.0001,
+      "loss": 1.3852,
+      "step": 9820
+    },
+    {
+      "epoch": 1.1281373844121532,
+      "grad_norm": 0.5751352310180664,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 9821
+    },
+    {
+      "epoch": 1.1282522543219804,
+      "grad_norm": 0.5874543786048889,
+      "learning_rate": 0.0001,
+      "loss": 1.6203,
+      "step": 9822
+    },
+    {
+      "epoch": 1.1283671242318074,
+      "grad_norm": 0.5984853506088257,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 9823
+    },
+    {
+      "epoch": 1.1284819941416346,
+      "grad_norm": 0.5243591070175171,
+      "learning_rate": 0.0001,
+      "loss": 1.2917,
+      "step": 9824
+    },
+    {
+      "epoch": 1.1285968640514616,
+      "grad_norm": 0.5501663088798523,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 9825
+    },
+    {
+      "epoch": 1.1287117339612889,
+      "grad_norm": 0.5242180824279785,
+      "learning_rate": 0.0001,
+      "loss": 1.3522,
+      "step": 9826
+    },
+    {
+      "epoch": 1.1288266038711159,
+      "grad_norm": 0.5722183585166931,
+      "learning_rate": 0.0001,
+      "loss": 1.4354,
+      "step": 9827
+    },
+    {
+      "epoch": 1.1289414737809431,
+      "grad_norm": 0.4917822778224945,
+      "learning_rate": 0.0001,
+      "loss": 1.3709,
+      "step": 9828
+    },
+    {
+      "epoch": 1.1290563436907701,
+      "grad_norm": 0.5814813375473022,
+      "learning_rate": 0.0001,
+      "loss": 1.7127,
+      "step": 9829
+    },
+    {
+      "epoch": 1.1291712136005974,
+      "grad_norm": 0.5022294521331787,
+      "learning_rate": 0.0001,
+      "loss": 1.4454,
+      "step": 9830
+    },
+    {
+      "epoch": 1.1292860835104244,
+      "grad_norm": 0.5753988027572632,
+      "learning_rate": 0.0001,
+      "loss": 1.5266,
+      "step": 9831
+    },
+    {
+      "epoch": 1.1294009534202516,
+      "grad_norm": 0.5708458423614502,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 9832
+    },
+    {
+      "epoch": 1.1295158233300786,
+      "grad_norm": 0.5471024513244629,
+      "learning_rate": 0.0001,
+      "loss": 1.2932,
+      "step": 9833
+    },
+    {
+      "epoch": 1.1296306932399058,
+      "grad_norm": 0.5528780817985535,
+      "learning_rate": 0.0001,
+      "loss": 1.3153,
+      "step": 9834
+    },
+    {
+      "epoch": 1.1297455631497328,
+      "grad_norm": 0.5226410627365112,
+      "learning_rate": 0.0001,
+      "loss": 1.4946,
+      "step": 9835
+    },
+    {
+      "epoch": 1.12986043305956,
+      "grad_norm": 0.5486363768577576,
+      "learning_rate": 0.0001,
+      "loss": 1.4376,
+      "step": 9836
+    },
+    {
+      "epoch": 1.129975302969387,
+      "grad_norm": 0.526906430721283,
+      "learning_rate": 0.0001,
+      "loss": 1.4453,
+      "step": 9837
+    },
+    {
+      "epoch": 1.1300901728792143,
+      "grad_norm": 0.5345343351364136,
+      "learning_rate": 0.0001,
+      "loss": 1.4311,
+      "step": 9838
+    },
+    {
+      "epoch": 1.1302050427890413,
+      "grad_norm": 0.6694420576095581,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 9839
+    },
+    {
+      "epoch": 1.1303199126988686,
+      "grad_norm": 0.5467296242713928,
+      "learning_rate": 0.0001,
+      "loss": 1.4171,
+      "step": 9840
+    },
+    {
+      "epoch": 1.1304347826086956,
+      "grad_norm": 0.6066096425056458,
+      "learning_rate": 0.0001,
+      "loss": 1.5134,
+      "step": 9841
+    },
+    {
+      "epoch": 1.1305496525185228,
+      "grad_norm": 0.5474829077720642,
+      "learning_rate": 0.0001,
+      "loss": 1.4532,
+      "step": 9842
+    },
+    {
+      "epoch": 1.1306645224283498,
+      "grad_norm": 0.5534715056419373,
+      "learning_rate": 0.0001,
+      "loss": 1.5475,
+      "step": 9843
+    },
+    {
+      "epoch": 1.130779392338177,
+      "grad_norm": 0.5239380598068237,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 9844
+    },
+    {
+      "epoch": 1.130894262248004,
+      "grad_norm": 0.6033841371536255,
+      "learning_rate": 0.0001,
+      "loss": 1.7539,
+      "step": 9845
+    },
+    {
+      "epoch": 1.1310091321578313,
+      "grad_norm": 0.5651752352714539,
+      "learning_rate": 0.0001,
+      "loss": 1.3981,
+      "step": 9846
+    },
+    {
+      "epoch": 1.1311240020676583,
+      "grad_norm": 0.5665850043296814,
+      "learning_rate": 0.0001,
+      "loss": 1.3451,
+      "step": 9847
+    },
+    {
+      "epoch": 1.1312388719774855,
+      "grad_norm": 0.5850709676742554,
+      "learning_rate": 0.0001,
+      "loss": 1.4922,
+      "step": 9848
+    },
+    {
+      "epoch": 1.1313537418873125,
+      "grad_norm": 0.5818759799003601,
+      "learning_rate": 0.0001,
+      "loss": 1.4969,
+      "step": 9849
+    },
+    {
+      "epoch": 1.1314686117971398,
+      "grad_norm": 0.59281986951828,
+      "learning_rate": 0.0001,
+      "loss": 1.4706,
+      "step": 9850
+    },
+    {
+      "epoch": 1.1315834817069668,
+      "grad_norm": 0.5399245023727417,
+      "learning_rate": 0.0001,
+      "loss": 1.3788,
+      "step": 9851
+    },
+    {
+      "epoch": 1.131698351616794,
+      "grad_norm": 0.5174872875213623,
+      "learning_rate": 0.0001,
+      "loss": 1.3944,
+      "step": 9852
+    },
+    {
+      "epoch": 1.131813221526621,
+      "grad_norm": 0.5072240829467773,
+      "learning_rate": 0.0001,
+      "loss": 1.3622,
+      "step": 9853
+    },
+    {
+      "epoch": 1.1319280914364482,
+      "grad_norm": 0.5090083479881287,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 9854
+    },
+    {
+      "epoch": 1.1320429613462752,
+      "grad_norm": 0.5711216926574707,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 9855
+    },
+    {
+      "epoch": 1.1321578312561025,
+      "grad_norm": 0.5663666129112244,
+      "learning_rate": 0.0001,
+      "loss": 1.4702,
+      "step": 9856
+    },
+    {
+      "epoch": 1.1322727011659297,
+      "grad_norm": 0.5527740120887756,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 9857
+    },
+    {
+      "epoch": 1.1323875710757567,
+      "grad_norm": 0.5693678855895996,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 9858
+    },
+    {
+      "epoch": 1.1325024409855837,
+      "grad_norm": 0.5752173066139221,
+      "learning_rate": 0.0001,
+      "loss": 1.4533,
+      "step": 9859
+    },
+    {
+      "epoch": 1.132617310895411,
+      "grad_norm": 0.5374181866645813,
+      "learning_rate": 0.0001,
+      "loss": 1.2613,
+      "step": 9860
+    },
+    {
+      "epoch": 1.1327321808052382,
+      "grad_norm": 0.5573873519897461,
+      "learning_rate": 0.0001,
+      "loss": 1.4246,
+      "step": 9861
+    },
+    {
+      "epoch": 1.1328470507150652,
+      "grad_norm": 0.5458858609199524,
+      "learning_rate": 0.0001,
+      "loss": 1.355,
+      "step": 9862
+    },
+    {
+      "epoch": 1.1329619206248922,
+      "grad_norm": 0.5189163684844971,
+      "learning_rate": 0.0001,
+      "loss": 1.4186,
+      "step": 9863
+    },
+    {
+      "epoch": 1.1330767905347194,
+      "grad_norm": 0.5939661860466003,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 9864
+    },
+    {
+      "epoch": 1.1331916604445467,
+      "grad_norm": 0.5526455044746399,
+      "learning_rate": 0.0001,
+      "loss": 1.6,
+      "step": 9865
+    },
+    {
+      "epoch": 1.1333065303543737,
+      "grad_norm": 0.5535423755645752,
+      "learning_rate": 0.0001,
+      "loss": 1.5184,
+      "step": 9866
+    },
+    {
+      "epoch": 1.1334214002642007,
+      "grad_norm": 0.5559777617454529,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 9867
+    },
+    {
+      "epoch": 1.133536270174028,
+      "grad_norm": 0.5639746785163879,
+      "learning_rate": 0.0001,
+      "loss": 1.3472,
+      "step": 9868
+    },
+    {
+      "epoch": 1.1336511400838551,
+      "grad_norm": 0.5288705229759216,
+      "learning_rate": 0.0001,
+      "loss": 1.3665,
+      "step": 9869
+    },
+    {
+      "epoch": 1.1337660099936822,
+      "grad_norm": 0.6040565371513367,
+      "learning_rate": 0.0001,
+      "loss": 1.7257,
+      "step": 9870
+    },
+    {
+      "epoch": 1.1338808799035092,
+      "grad_norm": 0.5919288992881775,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 9871
+    },
+    {
+      "epoch": 1.1339957498133364,
+      "grad_norm": 0.5653237104415894,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 9872
+    },
+    {
+      "epoch": 1.1341106197231636,
+      "grad_norm": 0.5587108731269836,
+      "learning_rate": 0.0001,
+      "loss": 1.5307,
+      "step": 9873
+    },
+    {
+      "epoch": 1.1342254896329906,
+      "grad_norm": 0.5890237092971802,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 9874
+    },
+    {
+      "epoch": 1.1343403595428176,
+      "grad_norm": 0.5263911485671997,
+      "learning_rate": 0.0001,
+      "loss": 1.4927,
+      "step": 9875
+    },
+    {
+      "epoch": 1.1344552294526449,
+      "grad_norm": 0.494157075881958,
+      "learning_rate": 0.0001,
+      "loss": 1.3288,
+      "step": 9876
+    },
+    {
+      "epoch": 1.134570099362472,
+      "grad_norm": 0.5565462112426758,
+      "learning_rate": 0.0001,
+      "loss": 1.1901,
+      "step": 9877
+    },
+    {
+      "epoch": 1.1346849692722991,
+      "grad_norm": 0.5866313576698303,
+      "learning_rate": 0.0001,
+      "loss": 1.3815,
+      "step": 9878
+    },
+    {
+      "epoch": 1.1347998391821261,
+      "grad_norm": 0.5734443068504333,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 9879
+    },
+    {
+      "epoch": 1.1349147090919534,
+      "grad_norm": 0.533345103263855,
+      "learning_rate": 0.0001,
+      "loss": 1.4551,
+      "step": 9880
+    },
+    {
+      "epoch": 1.1350295790017806,
+      "grad_norm": 0.5391797423362732,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 9881
+    },
+    {
+      "epoch": 1.1351444489116076,
+      "grad_norm": 0.5512882471084595,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 9882
+    },
+    {
+      "epoch": 1.1352593188214346,
+      "grad_norm": 0.565726637840271,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 9883
+    },
+    {
+      "epoch": 1.1353741887312618,
+      "grad_norm": 0.5509957671165466,
+      "learning_rate": 0.0001,
+      "loss": 1.4008,
+      "step": 9884
+    },
+    {
+      "epoch": 1.135489058641089,
+      "grad_norm": 0.6037254333496094,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 9885
+    },
+    {
+      "epoch": 1.135603928550916,
+      "grad_norm": 0.5137581825256348,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 9886
+    },
+    {
+      "epoch": 1.135718798460743,
+      "grad_norm": 0.5548390746116638,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 9887
+    },
+    {
+      "epoch": 1.1358336683705703,
+      "grad_norm": 0.538205087184906,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 9888
+    },
+    {
+      "epoch": 1.1359485382803975,
+      "grad_norm": 0.5580369830131531,
+      "learning_rate": 0.0001,
+      "loss": 1.4531,
+      "step": 9889
+    },
+    {
+      "epoch": 1.1360634081902246,
+      "grad_norm": 0.5717020034790039,
+      "learning_rate": 0.0001,
+      "loss": 1.3808,
+      "step": 9890
+    },
+    {
+      "epoch": 1.1361782781000518,
+      "grad_norm": 0.5775260329246521,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 9891
+    },
+    {
+      "epoch": 1.1362931480098788,
+      "grad_norm": 0.6119628548622131,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 9892
+    },
+    {
+      "epoch": 1.136408017919706,
+      "grad_norm": 0.533612072467804,
+      "learning_rate": 0.0001,
+      "loss": 1.2585,
+      "step": 9893
+    },
+    {
+      "epoch": 1.136522887829533,
+      "grad_norm": 0.5262476205825806,
+      "learning_rate": 0.0001,
+      "loss": 1.3074,
+      "step": 9894
+    },
+    {
+      "epoch": 1.1366377577393603,
+      "grad_norm": 0.5193493366241455,
+      "learning_rate": 0.0001,
+      "loss": 1.43,
+      "step": 9895
+    },
+    {
+      "epoch": 1.1367526276491873,
+      "grad_norm": 0.5410826206207275,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 9896
+    },
+    {
+      "epoch": 1.1368674975590145,
+      "grad_norm": 0.5434339642524719,
+      "learning_rate": 0.0001,
+      "loss": 1.4059,
+      "step": 9897
+    },
+    {
+      "epoch": 1.1369823674688415,
+      "grad_norm": 0.6266562342643738,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 9898
+    },
+    {
+      "epoch": 1.1370972373786687,
+      "grad_norm": 0.5530896782875061,
+      "learning_rate": 0.0001,
+      "loss": 1.4496,
+      "step": 9899
+    },
+    {
+      "epoch": 1.1372121072884958,
+      "grad_norm": 0.5319846868515015,
+      "learning_rate": 0.0001,
+      "loss": 1.4595,
+      "step": 9900
+    },
+    {
+      "epoch": 1.137326977198323,
+      "grad_norm": 0.5379259586334229,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 9901
+    },
+    {
+      "epoch": 1.13744184710815,
+      "grad_norm": 0.5139333009719849,
+      "learning_rate": 0.0001,
+      "loss": 1.3436,
+      "step": 9902
+    },
+    {
+      "epoch": 1.1375567170179772,
+      "grad_norm": 0.5401899218559265,
+      "learning_rate": 0.0001,
+      "loss": 1.7349,
+      "step": 9903
+    },
+    {
+      "epoch": 1.1376715869278042,
+      "grad_norm": 0.551706075668335,
+      "learning_rate": 0.0001,
+      "loss": 1.4543,
+      "step": 9904
+    },
+    {
+      "epoch": 1.1377864568376315,
+      "grad_norm": 0.5094159245491028,
+      "learning_rate": 0.0001,
+      "loss": 1.4717,
+      "step": 9905
+    },
+    {
+      "epoch": 1.1379013267474585,
+      "grad_norm": 0.514320433139801,
+      "learning_rate": 0.0001,
+      "loss": 1.275,
+      "step": 9906
+    },
+    {
+      "epoch": 1.1380161966572857,
+      "grad_norm": 0.5413798689842224,
+      "learning_rate": 0.0001,
+      "loss": 1.4261,
+      "step": 9907
+    },
+    {
+      "epoch": 1.1381310665671127,
+      "grad_norm": 0.6346654295921326,
+      "learning_rate": 0.0001,
+      "loss": 1.4869,
+      "step": 9908
+    },
+    {
+      "epoch": 1.13824593647694,
+      "grad_norm": 0.5805730223655701,
+      "learning_rate": 0.0001,
+      "loss": 1.4811,
+      "step": 9909
+    },
+    {
+      "epoch": 1.138360806386767,
+      "grad_norm": 0.5956717133522034,
+      "learning_rate": 0.0001,
+      "loss": 1.6149,
+      "step": 9910
+    },
+    {
+      "epoch": 1.1384756762965942,
+      "grad_norm": 0.5871737599372864,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 9911
+    },
+    {
+      "epoch": 1.1385905462064212,
+      "grad_norm": 0.6298748254776001,
+      "learning_rate": 0.0001,
+      "loss": 1.5124,
+      "step": 9912
+    },
+    {
+      "epoch": 1.1387054161162484,
+      "grad_norm": 0.5252795219421387,
+      "learning_rate": 0.0001,
+      "loss": 1.4858,
+      "step": 9913
+    },
+    {
+      "epoch": 1.1388202860260754,
+      "grad_norm": 0.5625512003898621,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 9914
+    },
+    {
+      "epoch": 1.1389351559359027,
+      "grad_norm": 0.47873494029045105,
+      "learning_rate": 0.0001,
+      "loss": 1.2768,
+      "step": 9915
+    },
+    {
+      "epoch": 1.1390500258457297,
+      "grad_norm": 0.5676445960998535,
+      "learning_rate": 0.0001,
+      "loss": 1.6487,
+      "step": 9916
+    },
+    {
+      "epoch": 1.139164895755557,
+      "grad_norm": 0.5738489031791687,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 9917
+    },
+    {
+      "epoch": 1.139279765665384,
+      "grad_norm": 0.5460545420646667,
+      "learning_rate": 0.0001,
+      "loss": 1.375,
+      "step": 9918
+    },
+    {
+      "epoch": 1.1393946355752111,
+      "grad_norm": 0.5179741382598877,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 9919
+    },
+    {
+      "epoch": 1.1395095054850382,
+      "grad_norm": 0.6625398993492126,
+      "learning_rate": 0.0001,
+      "loss": 1.7408,
+      "step": 9920
+    },
+    {
+      "epoch": 1.1396243753948654,
+      "grad_norm": 0.55277019739151,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 9921
+    },
+    {
+      "epoch": 1.1397392453046924,
+      "grad_norm": 0.5727016925811768,
+      "learning_rate": 0.0001,
+      "loss": 1.4051,
+      "step": 9922
+    },
+    {
+      "epoch": 1.1398541152145196,
+      "grad_norm": 0.5403141975402832,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 9923
+    },
+    {
+      "epoch": 1.1399689851243466,
+      "grad_norm": 0.5404056906700134,
+      "learning_rate": 0.0001,
+      "loss": 1.4899,
+      "step": 9924
+    },
+    {
+      "epoch": 1.1400838550341739,
+      "grad_norm": 0.5808719992637634,
+      "learning_rate": 0.0001,
+      "loss": 1.4641,
+      "step": 9925
+    },
+    {
+      "epoch": 1.1401987249440009,
+      "grad_norm": 0.5395628809928894,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 9926
+    },
+    {
+      "epoch": 1.140313594853828,
+      "grad_norm": 0.5930881500244141,
+      "learning_rate": 0.0001,
+      "loss": 1.52,
+      "step": 9927
+    },
+    {
+      "epoch": 1.1404284647636551,
+      "grad_norm": 0.5815712809562683,
+      "learning_rate": 0.0001,
+      "loss": 1.2868,
+      "step": 9928
+    },
+    {
+      "epoch": 1.1405433346734823,
+      "grad_norm": 0.5318073630332947,
+      "learning_rate": 0.0001,
+      "loss": 1.4842,
+      "step": 9929
+    },
+    {
+      "epoch": 1.1406582045833094,
+      "grad_norm": 0.5083526968955994,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 9930
+    },
+    {
+      "epoch": 1.1407730744931366,
+      "grad_norm": 0.6382246017456055,
+      "learning_rate": 0.0001,
+      "loss": 1.6084,
+      "step": 9931
+    },
+    {
+      "epoch": 1.1408879444029636,
+      "grad_norm": 0.5410823225975037,
+      "learning_rate": 0.0001,
+      "loss": 1.3782,
+      "step": 9932
+    },
+    {
+      "epoch": 1.1410028143127908,
+      "grad_norm": 0.5354570746421814,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 9933
+    },
+    {
+      "epoch": 1.1411176842226178,
+      "grad_norm": 0.5121961832046509,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 9934
+    },
+    {
+      "epoch": 1.141232554132445,
+      "grad_norm": 0.5164211392402649,
+      "learning_rate": 0.0001,
+      "loss": 1.3373,
+      "step": 9935
+    },
+    {
+      "epoch": 1.141347424042272,
+      "grad_norm": 0.5433701276779175,
+      "learning_rate": 0.0001,
+      "loss": 1.4922,
+      "step": 9936
+    },
+    {
+      "epoch": 1.1414622939520993,
+      "grad_norm": 0.5362388491630554,
+      "learning_rate": 0.0001,
+      "loss": 1.2922,
+      "step": 9937
+    },
+    {
+      "epoch": 1.1415771638619263,
+      "grad_norm": 0.5343228578567505,
+      "learning_rate": 0.0001,
+      "loss": 1.3815,
+      "step": 9938
+    },
+    {
+      "epoch": 1.1416920337717535,
+      "grad_norm": 0.5795764327049255,
+      "learning_rate": 0.0001,
+      "loss": 1.5437,
+      "step": 9939
+    },
+    {
+      "epoch": 1.1418069036815806,
+      "grad_norm": 0.5852019190788269,
+      "learning_rate": 0.0001,
+      "loss": 1.463,
+      "step": 9940
+    },
+    {
+      "epoch": 1.1419217735914078,
+      "grad_norm": 0.5105196237564087,
+      "learning_rate": 0.0001,
+      "loss": 1.3497,
+      "step": 9941
+    },
+    {
+      "epoch": 1.1420366435012348,
+      "grad_norm": 0.5408787131309509,
+      "learning_rate": 0.0001,
+      "loss": 1.3446,
+      "step": 9942
+    },
+    {
+      "epoch": 1.142151513411062,
+      "grad_norm": 0.552517294883728,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 9943
+    },
+    {
+      "epoch": 1.142266383320889,
+      "grad_norm": 0.5696591138839722,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 9944
+    },
+    {
+      "epoch": 1.1423812532307163,
+      "grad_norm": 0.560500979423523,
+      "learning_rate": 0.0001,
+      "loss": 1.5635,
+      "step": 9945
+    },
+    {
+      "epoch": 1.1424961231405433,
+      "grad_norm": 0.6407963037490845,
+      "learning_rate": 0.0001,
+      "loss": 1.5365,
+      "step": 9946
+    },
+    {
+      "epoch": 1.1426109930503705,
+      "grad_norm": 0.5425936579704285,
+      "learning_rate": 0.0001,
+      "loss": 1.3761,
+      "step": 9947
+    },
+    {
+      "epoch": 1.1427258629601975,
+      "grad_norm": 0.5866653323173523,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 9948
+    },
+    {
+      "epoch": 1.1428407328700247,
+      "grad_norm": 0.5960055589675903,
+      "learning_rate": 0.0001,
+      "loss": 1.3274,
+      "step": 9949
+    },
+    {
+      "epoch": 1.1429556027798518,
+      "grad_norm": 0.5556841492652893,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 9950
+    },
+    {
+      "epoch": 1.143070472689679,
+      "grad_norm": 0.5279529094696045,
+      "learning_rate": 0.0001,
+      "loss": 1.2829,
+      "step": 9951
+    },
+    {
+      "epoch": 1.143185342599506,
+      "grad_norm": 0.5272794961929321,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 9952
+    },
+    {
+      "epoch": 1.1433002125093332,
+      "grad_norm": 0.5824765563011169,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 9953
+    },
+    {
+      "epoch": 1.1434150824191602,
+      "grad_norm": 0.5903116464614868,
+      "learning_rate": 0.0001,
+      "loss": 1.6437,
+      "step": 9954
+    },
+    {
+      "epoch": 1.1435299523289875,
+      "grad_norm": 0.525478184223175,
+      "learning_rate": 0.0001,
+      "loss": 1.4775,
+      "step": 9955
+    },
+    {
+      "epoch": 1.1436448222388145,
+      "grad_norm": 0.4939769208431244,
+      "learning_rate": 0.0001,
+      "loss": 1.2772,
+      "step": 9956
+    },
+    {
+      "epoch": 1.1437596921486417,
+      "grad_norm": 0.5516688823699951,
+      "learning_rate": 0.0001,
+      "loss": 1.3937,
+      "step": 9957
+    },
+    {
+      "epoch": 1.1438745620584687,
+      "grad_norm": 0.5349141955375671,
+      "learning_rate": 0.0001,
+      "loss": 1.4102,
+      "step": 9958
+    },
+    {
+      "epoch": 1.143989431968296,
+      "grad_norm": 0.5700792670249939,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 9959
+    },
+    {
+      "epoch": 1.144104301878123,
+      "grad_norm": 0.5753479599952698,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 9960
+    },
+    {
+      "epoch": 1.1442191717879502,
+      "grad_norm": 0.49369415640830994,
+      "learning_rate": 0.0001,
+      "loss": 1.3815,
+      "step": 9961
+    },
+    {
+      "epoch": 1.1443340416977772,
+      "grad_norm": 0.5094558000564575,
+      "learning_rate": 0.0001,
+      "loss": 1.3144,
+      "step": 9962
+    },
+    {
+      "epoch": 1.1444489116076044,
+      "grad_norm": 0.5119041800498962,
+      "learning_rate": 0.0001,
+      "loss": 1.3715,
+      "step": 9963
+    },
+    {
+      "epoch": 1.1445637815174314,
+      "grad_norm": 0.7021076083183289,
+      "learning_rate": 0.0001,
+      "loss": 1.818,
+      "step": 9964
+    },
+    {
+      "epoch": 1.1446786514272587,
+      "grad_norm": 0.5357758402824402,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 9965
+    },
+    {
+      "epoch": 1.1447935213370857,
+      "grad_norm": 0.5504557490348816,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 9966
+    },
+    {
+      "epoch": 1.144908391246913,
+      "grad_norm": 0.5839464068412781,
+      "learning_rate": 0.0001,
+      "loss": 1.5019,
+      "step": 9967
+    },
+    {
+      "epoch": 1.14502326115674,
+      "grad_norm": 0.5492793917655945,
+      "learning_rate": 0.0001,
+      "loss": 1.6016,
+      "step": 9968
+    },
+    {
+      "epoch": 1.1451381310665671,
+      "grad_norm": 0.5798627138137817,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 9969
+    },
+    {
+      "epoch": 1.1452530009763942,
+      "grad_norm": 0.5143793225288391,
+      "learning_rate": 0.0001,
+      "loss": 1.429,
+      "step": 9970
+    },
+    {
+      "epoch": 1.1453678708862214,
+      "grad_norm": 0.5312758684158325,
+      "learning_rate": 0.0001,
+      "loss": 1.4962,
+      "step": 9971
+    },
+    {
+      "epoch": 1.1454827407960484,
+      "grad_norm": 0.5357524156570435,
+      "learning_rate": 0.0001,
+      "loss": 1.4713,
+      "step": 9972
+    },
+    {
+      "epoch": 1.1455976107058756,
+      "grad_norm": 0.5727779269218445,
+      "learning_rate": 0.0001,
+      "loss": 1.3171,
+      "step": 9973
+    },
+    {
+      "epoch": 1.1457124806157026,
+      "grad_norm": 0.5089569091796875,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 9974
+    },
+    {
+      "epoch": 1.1458273505255299,
+      "grad_norm": 0.587568998336792,
+      "learning_rate": 0.0001,
+      "loss": 1.4017,
+      "step": 9975
+    },
+    {
+      "epoch": 1.1459422204353569,
+      "grad_norm": 0.5800287127494812,
+      "learning_rate": 0.0001,
+      "loss": 1.5692,
+      "step": 9976
+    },
+    {
+      "epoch": 1.146057090345184,
+      "grad_norm": 0.5658348798751831,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 9977
+    },
+    {
+      "epoch": 1.1461719602550111,
+      "grad_norm": 0.5689564347267151,
+      "learning_rate": 0.0001,
+      "loss": 1.4476,
+      "step": 9978
+    },
+    {
+      "epoch": 1.1462868301648383,
+      "grad_norm": 0.5329977869987488,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 9979
+    },
+    {
+      "epoch": 1.1464017000746654,
+      "grad_norm": 0.5485842227935791,
+      "learning_rate": 0.0001,
+      "loss": 1.494,
+      "step": 9980
+    },
+    {
+      "epoch": 1.1465165699844926,
+      "grad_norm": 0.5831310153007507,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 9981
+    },
+    {
+      "epoch": 1.1466314398943196,
+      "grad_norm": 0.5669857263565063,
+      "learning_rate": 0.0001,
+      "loss": 1.436,
+      "step": 9982
+    },
+    {
+      "epoch": 1.1467463098041468,
+      "grad_norm": 0.569514274597168,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 9983
+    },
+    {
+      "epoch": 1.1468611797139738,
+      "grad_norm": 0.5632662773132324,
+      "learning_rate": 0.0001,
+      "loss": 1.345,
+      "step": 9984
+    },
+    {
+      "epoch": 1.146976049623801,
+      "grad_norm": 0.5908129215240479,
+      "learning_rate": 0.0001,
+      "loss": 1.4914,
+      "step": 9985
+    },
+    {
+      "epoch": 1.147090919533628,
+      "grad_norm": 0.6112525463104248,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 9986
+    },
+    {
+      "epoch": 1.1472057894434553,
+      "grad_norm": 0.5455618500709534,
+      "learning_rate": 0.0001,
+      "loss": 1.4891,
+      "step": 9987
+    },
+    {
+      "epoch": 1.1473206593532823,
+      "grad_norm": 0.5896759033203125,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 9988
+    },
+    {
+      "epoch": 1.1474355292631095,
+      "grad_norm": 0.571235179901123,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 9989
+    },
+    {
+      "epoch": 1.1475503991729366,
+      "grad_norm": 0.554031252861023,
+      "learning_rate": 0.0001,
+      "loss": 1.4231,
+      "step": 9990
+    },
+    {
+      "epoch": 1.1476652690827638,
+      "grad_norm": 0.5455310344696045,
+      "learning_rate": 0.0001,
+      "loss": 1.4302,
+      "step": 9991
+    },
+    {
+      "epoch": 1.1477801389925908,
+      "grad_norm": 0.5141569972038269,
+      "learning_rate": 0.0001,
+      "loss": 1.377,
+      "step": 9992
+    },
+    {
+      "epoch": 1.147895008902418,
+      "grad_norm": 0.5764080286026001,
+      "learning_rate": 0.0001,
+      "loss": 1.4583,
+      "step": 9993
+    },
+    {
+      "epoch": 1.1480098788122453,
+      "grad_norm": 0.5481144189834595,
+      "learning_rate": 0.0001,
+      "loss": 1.3084,
+      "step": 9994
+    },
+    {
+      "epoch": 1.1481247487220723,
+      "grad_norm": 0.5942389965057373,
+      "learning_rate": 0.0001,
+      "loss": 1.3149,
+      "step": 9995
+    },
+    {
+      "epoch": 1.1482396186318993,
+      "grad_norm": 0.5685340166091919,
+      "learning_rate": 0.0001,
+      "loss": 1.6043,
+      "step": 9996
+    },
+    {
+      "epoch": 1.1483544885417265,
+      "grad_norm": 0.5472145080566406,
+      "learning_rate": 0.0001,
+      "loss": 1.3976,
+      "step": 9997
+    },
+    {
+      "epoch": 1.1484693584515537,
+      "grad_norm": 0.6134904623031616,
+      "learning_rate": 0.0001,
+      "loss": 1.3928,
+      "step": 9998
+    },
+    {
+      "epoch": 1.1485842283613807,
+      "grad_norm": 0.5569812655448914,
+      "learning_rate": 0.0001,
+      "loss": 1.3422,
+      "step": 9999
+    },
+    {
+      "epoch": 1.1486990982712078,
+      "grad_norm": 0.5773634314537048,
+      "learning_rate": 0.0001,
+      "loss": 1.2924,
+      "step": 10000
+    },
+    {
+      "epoch": 1.148813968181035,
+      "grad_norm": 0.5295365452766418,
+      "learning_rate": 0.0001,
+      "loss": 1.4651,
+      "step": 10001
+    },
+    {
+      "epoch": 1.1489288380908622,
+      "grad_norm": 0.5416709780693054,
+      "learning_rate": 0.0001,
+      "loss": 1.3973,
+      "step": 10002
+    },
+    {
+      "epoch": 1.1490437080006892,
+      "grad_norm": 0.5793564319610596,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 10003
+    },
+    {
+      "epoch": 1.1491585779105162,
+      "grad_norm": 0.6285029649734497,
+      "learning_rate": 0.0001,
+      "loss": 1.4183,
+      "step": 10004
+    },
+    {
+      "epoch": 1.1492734478203435,
+      "grad_norm": 0.5785372853279114,
+      "learning_rate": 0.0001,
+      "loss": 1.2509,
+      "step": 10005
+    },
+    {
+      "epoch": 1.1493883177301707,
+      "grad_norm": 0.6486859321594238,
+      "learning_rate": 0.0001,
+      "loss": 1.6246,
+      "step": 10006
+    },
+    {
+      "epoch": 1.1495031876399977,
+      "grad_norm": 0.5618028044700623,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 10007
+    },
+    {
+      "epoch": 1.1496180575498247,
+      "grad_norm": 0.5516600012779236,
+      "learning_rate": 0.0001,
+      "loss": 1.3189,
+      "step": 10008
+    },
+    {
+      "epoch": 1.149732927459652,
+      "grad_norm": 0.5638642907142639,
+      "learning_rate": 0.0001,
+      "loss": 1.4698,
+      "step": 10009
+    },
+    {
+      "epoch": 1.1498477973694792,
+      "grad_norm": 0.5339157581329346,
+      "learning_rate": 0.0001,
+      "loss": 1.4751,
+      "step": 10010
+    },
+    {
+      "epoch": 1.1499626672793062,
+      "grad_norm": 0.5690274834632874,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 10011
+    },
+    {
+      "epoch": 1.1500775371891332,
+      "grad_norm": 0.5620351433753967,
+      "learning_rate": 0.0001,
+      "loss": 1.5383,
+      "step": 10012
+    },
+    {
+      "epoch": 1.1501924070989604,
+      "grad_norm": 0.5315772294998169,
+      "learning_rate": 0.0001,
+      "loss": 1.5289,
+      "step": 10013
+    },
+    {
+      "epoch": 1.1503072770087877,
+      "grad_norm": 0.5182132124900818,
+      "learning_rate": 0.0001,
+      "loss": 1.3737,
+      "step": 10014
+    },
+    {
+      "epoch": 1.1504221469186147,
+      "grad_norm": 0.5324050188064575,
+      "learning_rate": 0.0001,
+      "loss": 1.3595,
+      "step": 10015
+    },
+    {
+      "epoch": 1.1505370168284417,
+      "grad_norm": 0.5690122842788696,
+      "learning_rate": 0.0001,
+      "loss": 1.6227,
+      "step": 10016
+    },
+    {
+      "epoch": 1.150651886738269,
+      "grad_norm": 0.5757220983505249,
+      "learning_rate": 0.0001,
+      "loss": 1.4787,
+      "step": 10017
+    },
+    {
+      "epoch": 1.1507667566480961,
+      "grad_norm": 0.5967593789100647,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 10018
+    },
+    {
+      "epoch": 1.1508816265579231,
+      "grad_norm": 0.5282356142997742,
+      "learning_rate": 0.0001,
+      "loss": 1.4053,
+      "step": 10019
+    },
+    {
+      "epoch": 1.1509964964677502,
+      "grad_norm": 0.5435159206390381,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 10020
+    },
+    {
+      "epoch": 1.1511113663775774,
+      "grad_norm": 0.5466963052749634,
+      "learning_rate": 0.0001,
+      "loss": 1.3257,
+      "step": 10021
+    },
+    {
+      "epoch": 1.1512262362874046,
+      "grad_norm": 0.5418528914451599,
+      "learning_rate": 0.0001,
+      "loss": 1.4969,
+      "step": 10022
+    },
+    {
+      "epoch": 1.1513411061972316,
+      "grad_norm": 0.6105648875236511,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 10023
+    },
+    {
+      "epoch": 1.1514559761070586,
+      "grad_norm": 0.5551570057868958,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 10024
+    },
+    {
+      "epoch": 1.1515708460168859,
+      "grad_norm": 0.5221659541130066,
+      "learning_rate": 0.0001,
+      "loss": 1.3441,
+      "step": 10025
+    },
+    {
+      "epoch": 1.151685715926713,
+      "grad_norm": 0.6351895332336426,
+      "learning_rate": 0.0001,
+      "loss": 1.5949,
+      "step": 10026
+    },
+    {
+      "epoch": 1.15180058583654,
+      "grad_norm": 0.5757087469100952,
+      "learning_rate": 0.0001,
+      "loss": 1.4843,
+      "step": 10027
+    },
+    {
+      "epoch": 1.1519154557463673,
+      "grad_norm": 0.5461332201957703,
+      "learning_rate": 0.0001,
+      "loss": 1.4231,
+      "step": 10028
+    },
+    {
+      "epoch": 1.1520303256561943,
+      "grad_norm": 0.5618088841438293,
+      "learning_rate": 0.0001,
+      "loss": 1.2567,
+      "step": 10029
+    },
+    {
+      "epoch": 1.1521451955660216,
+      "grad_norm": 0.668555736541748,
+      "learning_rate": 0.0001,
+      "loss": 1.4302,
+      "step": 10030
+    },
+    {
+      "epoch": 1.1522600654758486,
+      "grad_norm": 0.5682727694511414,
+      "learning_rate": 0.0001,
+      "loss": 1.3702,
+      "step": 10031
+    },
+    {
+      "epoch": 1.1523749353856758,
+      "grad_norm": 0.531745195388794,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 10032
+    },
+    {
+      "epoch": 1.1524898052955028,
+      "grad_norm": 0.5723196864128113,
+      "learning_rate": 0.0001,
+      "loss": 1.4564,
+      "step": 10033
+    },
+    {
+      "epoch": 1.15260467520533,
+      "grad_norm": 0.5728488564491272,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 10034
+    },
+    {
+      "epoch": 1.152719545115157,
+      "grad_norm": 0.528705894947052,
+      "learning_rate": 0.0001,
+      "loss": 1.4234,
+      "step": 10035
+    },
+    {
+      "epoch": 1.1528344150249843,
+      "grad_norm": 0.5178189873695374,
+      "learning_rate": 0.0001,
+      "loss": 1.3147,
+      "step": 10036
+    },
+    {
+      "epoch": 1.1529492849348113,
+      "grad_norm": 0.6135823130607605,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 10037
+    },
+    {
+      "epoch": 1.1530641548446385,
+      "grad_norm": 0.5911353230476379,
+      "learning_rate": 0.0001,
+      "loss": 1.1557,
+      "step": 10038
+    },
+    {
+      "epoch": 1.1531790247544655,
+      "grad_norm": 0.5582436919212341,
+      "learning_rate": 0.0001,
+      "loss": 1.2749,
+      "step": 10039
+    },
+    {
+      "epoch": 1.1532938946642928,
+      "grad_norm": 0.5512915849685669,
+      "learning_rate": 0.0001,
+      "loss": 1.4394,
+      "step": 10040
+    },
+    {
+      "epoch": 1.1534087645741198,
+      "grad_norm": 0.5502411127090454,
+      "learning_rate": 0.0001,
+      "loss": 1.2865,
+      "step": 10041
+    },
+    {
+      "epoch": 1.153523634483947,
+      "grad_norm": 0.5328064560890198,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 10042
+    },
+    {
+      "epoch": 1.153638504393774,
+      "grad_norm": 0.4983726441860199,
+      "learning_rate": 0.0001,
+      "loss": 1.2559,
+      "step": 10043
+    },
+    {
+      "epoch": 1.1537533743036013,
+      "grad_norm": 0.6484810709953308,
+      "learning_rate": 0.0001,
+      "loss": 1.77,
+      "step": 10044
+    },
+    {
+      "epoch": 1.1538682442134283,
+      "grad_norm": 0.5316923260688782,
+      "learning_rate": 0.0001,
+      "loss": 1.3421,
+      "step": 10045
+    },
+    {
+      "epoch": 1.1539831141232555,
+      "grad_norm": 0.5837815403938293,
+      "learning_rate": 0.0001,
+      "loss": 1.5662,
+      "step": 10046
+    },
+    {
+      "epoch": 1.1540979840330825,
+      "grad_norm": 0.557544469833374,
+      "learning_rate": 0.0001,
+      "loss": 1.6774,
+      "step": 10047
+    },
+    {
+      "epoch": 1.1542128539429097,
+      "grad_norm": 0.5785412192344666,
+      "learning_rate": 0.0001,
+      "loss": 1.4416,
+      "step": 10048
+    },
+    {
+      "epoch": 1.1543277238527367,
+      "grad_norm": 0.5959882140159607,
+      "learning_rate": 0.0001,
+      "loss": 1.5142,
+      "step": 10049
+    },
+    {
+      "epoch": 1.154442593762564,
+      "grad_norm": 0.5870964527130127,
+      "learning_rate": 0.0001,
+      "loss": 1.4598,
+      "step": 10050
+    },
+    {
+      "epoch": 1.154557463672391,
+      "grad_norm": 0.6367591619491577,
+      "learning_rate": 0.0001,
+      "loss": 1.3788,
+      "step": 10051
+    },
+    {
+      "epoch": 1.1546723335822182,
+      "grad_norm": 0.5441646575927734,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 10052
+    },
+    {
+      "epoch": 1.1547872034920452,
+      "grad_norm": 0.5900537967681885,
+      "learning_rate": 0.0001,
+      "loss": 1.2413,
+      "step": 10053
+    },
+    {
+      "epoch": 1.1549020734018725,
+      "grad_norm": 0.6054137945175171,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 10054
+    },
+    {
+      "epoch": 1.1550169433116995,
+      "grad_norm": 0.5724448561668396,
+      "learning_rate": 0.0001,
+      "loss": 1.3456,
+      "step": 10055
+    },
+    {
+      "epoch": 1.1551318132215267,
+      "grad_norm": 0.6357987523078918,
+      "learning_rate": 0.0001,
+      "loss": 1.5517,
+      "step": 10056
+    },
+    {
+      "epoch": 1.1552466831313537,
+      "grad_norm": 0.5552554130554199,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 10057
+    },
+    {
+      "epoch": 1.155361553041181,
+      "grad_norm": 0.5502124428749084,
+      "learning_rate": 0.0001,
+      "loss": 1.3949,
+      "step": 10058
+    },
+    {
+      "epoch": 1.155476422951008,
+      "grad_norm": 0.5750370025634766,
+      "learning_rate": 0.0001,
+      "loss": 1.4175,
+      "step": 10059
+    },
+    {
+      "epoch": 1.1555912928608352,
+      "grad_norm": 0.5867468118667603,
+      "learning_rate": 0.0001,
+      "loss": 1.5483,
+      "step": 10060
+    },
+    {
+      "epoch": 1.1557061627706622,
+      "grad_norm": 0.5584085583686829,
+      "learning_rate": 0.0001,
+      "loss": 1.3498,
+      "step": 10061
+    },
+    {
+      "epoch": 1.1558210326804894,
+      "grad_norm": 0.5806280970573425,
+      "learning_rate": 0.0001,
+      "loss": 1.4329,
+      "step": 10062
+    },
+    {
+      "epoch": 1.1559359025903164,
+      "grad_norm": 0.6064799427986145,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 10063
+    },
+    {
+      "epoch": 1.1560507725001437,
+      "grad_norm": 0.5346385836601257,
+      "learning_rate": 0.0001,
+      "loss": 1.4026,
+      "step": 10064
+    },
+    {
+      "epoch": 1.1561656424099707,
+      "grad_norm": 0.5625292062759399,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 10065
+    },
+    {
+      "epoch": 1.156280512319798,
+      "grad_norm": 0.5931135416030884,
+      "learning_rate": 0.0001,
+      "loss": 1.4473,
+      "step": 10066
+    },
+    {
+      "epoch": 1.156395382229625,
+      "grad_norm": 0.5838515162467957,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 10067
+    },
+    {
+      "epoch": 1.1565102521394521,
+      "grad_norm": 0.5748473405838013,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 10068
+    },
+    {
+      "epoch": 1.1566251220492791,
+      "grad_norm": 0.5717789530754089,
+      "learning_rate": 0.0001,
+      "loss": 1.359,
+      "step": 10069
+    },
+    {
+      "epoch": 1.1567399919591064,
+      "grad_norm": 0.5722107887268066,
+      "learning_rate": 0.0001,
+      "loss": 1.6154,
+      "step": 10070
+    },
+    {
+      "epoch": 1.1568548618689334,
+      "grad_norm": 0.5895611643791199,
+      "learning_rate": 0.0001,
+      "loss": 1.4712,
+      "step": 10071
+    },
+    {
+      "epoch": 1.1569697317787606,
+      "grad_norm": 0.5189610719680786,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 10072
+    },
+    {
+      "epoch": 1.1570846016885876,
+      "grad_norm": 0.5800937414169312,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 10073
+    },
+    {
+      "epoch": 1.1571994715984149,
+      "grad_norm": 0.5774216055870056,
+      "learning_rate": 0.0001,
+      "loss": 1.3827,
+      "step": 10074
+    },
+    {
+      "epoch": 1.1573143415082419,
+      "grad_norm": 0.6525434255599976,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 10075
+    },
+    {
+      "epoch": 1.157429211418069,
+      "grad_norm": 0.5682554244995117,
+      "learning_rate": 0.0001,
+      "loss": 1.5512,
+      "step": 10076
+    },
+    {
+      "epoch": 1.157544081327896,
+      "grad_norm": 0.606096088886261,
+      "learning_rate": 0.0001,
+      "loss": 1.4762,
+      "step": 10077
+    },
+    {
+      "epoch": 1.1576589512377233,
+      "grad_norm": 0.5647655129432678,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 10078
+    },
+    {
+      "epoch": 1.1577738211475503,
+      "grad_norm": 0.580194354057312,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 10079
+    },
+    {
+      "epoch": 1.1578886910573776,
+      "grad_norm": 0.5711573958396912,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 10080
+    },
+    {
+      "epoch": 1.1580035609672046,
+      "grad_norm": 0.5933459401130676,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 10081
+    },
+    {
+      "epoch": 1.1581184308770318,
+      "grad_norm": 0.5333853960037231,
+      "learning_rate": 0.0001,
+      "loss": 1.3838,
+      "step": 10082
+    },
+    {
+      "epoch": 1.1582333007868588,
+      "grad_norm": 0.5227363705635071,
+      "learning_rate": 0.0001,
+      "loss": 1.5504,
+      "step": 10083
+    },
+    {
+      "epoch": 1.158348170696686,
+      "grad_norm": 0.5608334541320801,
+      "learning_rate": 0.0001,
+      "loss": 1.5402,
+      "step": 10084
+    },
+    {
+      "epoch": 1.158463040606513,
+      "grad_norm": 0.5283211469650269,
+      "learning_rate": 0.0001,
+      "loss": 1.2809,
+      "step": 10085
+    },
+    {
+      "epoch": 1.1585779105163403,
+      "grad_norm": 0.5522334575653076,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 10086
+    },
+    {
+      "epoch": 1.1586927804261673,
+      "grad_norm": 0.5605513453483582,
+      "learning_rate": 0.0001,
+      "loss": 1.3811,
+      "step": 10087
+    },
+    {
+      "epoch": 1.1588076503359945,
+      "grad_norm": 0.584088146686554,
+      "learning_rate": 0.0001,
+      "loss": 1.3513,
+      "step": 10088
+    },
+    {
+      "epoch": 1.1589225202458215,
+      "grad_norm": 0.5514279007911682,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 10089
+    },
+    {
+      "epoch": 1.1590373901556488,
+      "grad_norm": 0.5283199548721313,
+      "learning_rate": 0.0001,
+      "loss": 1.4285,
+      "step": 10090
+    },
+    {
+      "epoch": 1.1591522600654758,
+      "grad_norm": 0.5820270776748657,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 10091
+    },
+    {
+      "epoch": 1.159267129975303,
+      "grad_norm": 0.5973749160766602,
+      "learning_rate": 0.0001,
+      "loss": 1.3123,
+      "step": 10092
+    },
+    {
+      "epoch": 1.15938199988513,
+      "grad_norm": 0.5737087726593018,
+      "learning_rate": 0.0001,
+      "loss": 1.5244,
+      "step": 10093
+    },
+    {
+      "epoch": 1.1594968697949573,
+      "grad_norm": 0.5605661273002625,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 10094
+    },
+    {
+      "epoch": 1.1596117397047843,
+      "grad_norm": 0.5315190553665161,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 10095
+    },
+    {
+      "epoch": 1.1597266096146115,
+      "grad_norm": 0.5676159262657166,
+      "learning_rate": 0.0001,
+      "loss": 1.48,
+      "step": 10096
+    },
+    {
+      "epoch": 1.1598414795244385,
+      "grad_norm": 0.5562459826469421,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 10097
+    },
+    {
+      "epoch": 1.1599563494342657,
+      "grad_norm": 0.5750492215156555,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 10098
+    },
+    {
+      "epoch": 1.1600712193440927,
+      "grad_norm": 0.565864622592926,
+      "learning_rate": 0.0001,
+      "loss": 1.3542,
+      "step": 10099
+    },
+    {
+      "epoch": 1.16018608925392,
+      "grad_norm": 0.5741795897483826,
+      "learning_rate": 0.0001,
+      "loss": 1.5441,
+      "step": 10100
+    },
+    {
+      "epoch": 1.160300959163747,
+      "grad_norm": 0.5474345088005066,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 10101
+    },
+    {
+      "epoch": 1.1604158290735742,
+      "grad_norm": 0.5959522724151611,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 10102
+    },
+    {
+      "epoch": 1.1605306989834012,
+      "grad_norm": 0.5602507591247559,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 10103
+    },
+    {
+      "epoch": 1.1606455688932285,
+      "grad_norm": 0.5457338690757751,
+      "learning_rate": 0.0001,
+      "loss": 1.4527,
+      "step": 10104
+    },
+    {
+      "epoch": 1.1607604388030555,
+      "grad_norm": 0.5657529830932617,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 10105
+    },
+    {
+      "epoch": 1.1608753087128827,
+      "grad_norm": 0.5475678443908691,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 10106
+    },
+    {
+      "epoch": 1.1609901786227097,
+      "grad_norm": 0.5341169238090515,
+      "learning_rate": 0.0001,
+      "loss": 1.5948,
+      "step": 10107
+    },
+    {
+      "epoch": 1.161105048532537,
+      "grad_norm": 0.5745497345924377,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 10108
+    },
+    {
+      "epoch": 1.161219918442364,
+      "grad_norm": 0.56389319896698,
+      "learning_rate": 0.0001,
+      "loss": 1.4402,
+      "step": 10109
+    },
+    {
+      "epoch": 1.1613347883521912,
+      "grad_norm": 0.5795552134513855,
+      "learning_rate": 0.0001,
+      "loss": 1.4112,
+      "step": 10110
+    },
+    {
+      "epoch": 1.1614496582620182,
+      "grad_norm": 0.5938699841499329,
+      "learning_rate": 0.0001,
+      "loss": 1.453,
+      "step": 10111
+    },
+    {
+      "epoch": 1.1615645281718454,
+      "grad_norm": 0.5595149993896484,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 10112
+    },
+    {
+      "epoch": 1.1616793980816724,
+      "grad_norm": 0.6180020570755005,
+      "learning_rate": 0.0001,
+      "loss": 1.6828,
+      "step": 10113
+    },
+    {
+      "epoch": 1.1617942679914997,
+      "grad_norm": 0.5853464603424072,
+      "learning_rate": 0.0001,
+      "loss": 1.3735,
+      "step": 10114
+    },
+    {
+      "epoch": 1.1619091379013267,
+      "grad_norm": 0.5351958870887756,
+      "learning_rate": 0.0001,
+      "loss": 1.3615,
+      "step": 10115
+    },
+    {
+      "epoch": 1.162024007811154,
+      "grad_norm": 0.5981417894363403,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 10116
+    },
+    {
+      "epoch": 1.162138877720981,
+      "grad_norm": 0.5747444033622742,
+      "learning_rate": 0.0001,
+      "loss": 1.5427,
+      "step": 10117
+    },
+    {
+      "epoch": 1.1622537476308081,
+      "grad_norm": 0.5646340250968933,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 10118
+    },
+    {
+      "epoch": 1.1623686175406351,
+      "grad_norm": 0.5687159299850464,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 10119
+    },
+    {
+      "epoch": 1.1624834874504624,
+      "grad_norm": 0.5587418079376221,
+      "learning_rate": 0.0001,
+      "loss": 1.4571,
+      "step": 10120
+    },
+    {
+      "epoch": 1.1625983573602894,
+      "grad_norm": 0.575074315071106,
+      "learning_rate": 0.0001,
+      "loss": 1.3044,
+      "step": 10121
+    },
+    {
+      "epoch": 1.1627132272701166,
+      "grad_norm": 0.6066795587539673,
+      "learning_rate": 0.0001,
+      "loss": 1.7487,
+      "step": 10122
+    },
+    {
+      "epoch": 1.1628280971799436,
+      "grad_norm": 0.6262649893760681,
+      "learning_rate": 0.0001,
+      "loss": 1.5702,
+      "step": 10123
+    },
+    {
+      "epoch": 1.1629429670897709,
+      "grad_norm": 0.5629507303237915,
+      "learning_rate": 0.0001,
+      "loss": 1.4479,
+      "step": 10124
+    },
+    {
+      "epoch": 1.1630578369995979,
+      "grad_norm": 0.6446200013160706,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 10125
+    },
+    {
+      "epoch": 1.163172706909425,
+      "grad_norm": 0.567599356174469,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 10126
+    },
+    {
+      "epoch": 1.163287576819252,
+      "grad_norm": 0.5245000720024109,
+      "learning_rate": 0.0001,
+      "loss": 1.3432,
+      "step": 10127
+    },
+    {
+      "epoch": 1.1634024467290793,
+      "grad_norm": 0.6057835817337036,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 10128
+    },
+    {
+      "epoch": 1.1635173166389063,
+      "grad_norm": 0.5868533253669739,
+      "learning_rate": 0.0001,
+      "loss": 1.6142,
+      "step": 10129
+    },
+    {
+      "epoch": 1.1636321865487336,
+      "grad_norm": 0.533150315284729,
+      "learning_rate": 0.0001,
+      "loss": 1.4869,
+      "step": 10130
+    },
+    {
+      "epoch": 1.1637470564585608,
+      "grad_norm": 0.6112939715385437,
+      "learning_rate": 0.0001,
+      "loss": 1.3912,
+      "step": 10131
+    },
+    {
+      "epoch": 1.1638619263683878,
+      "grad_norm": 0.5236315131187439,
+      "learning_rate": 0.0001,
+      "loss": 1.4242,
+      "step": 10132
+    },
+    {
+      "epoch": 1.1639767962782148,
+      "grad_norm": 0.5508266091346741,
+      "learning_rate": 0.0001,
+      "loss": 1.582,
+      "step": 10133
+    },
+    {
+      "epoch": 1.164091666188042,
+      "grad_norm": 0.5487030744552612,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 10134
+    },
+    {
+      "epoch": 1.1642065360978693,
+      "grad_norm": 0.6089741587638855,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 10135
+    },
+    {
+      "epoch": 1.1643214060076963,
+      "grad_norm": 0.6037748456001282,
+      "learning_rate": 0.0001,
+      "loss": 1.3875,
+      "step": 10136
+    },
+    {
+      "epoch": 1.1644362759175233,
+      "grad_norm": 0.5815845131874084,
+      "learning_rate": 0.0001,
+      "loss": 1.5777,
+      "step": 10137
+    },
+    {
+      "epoch": 1.1645511458273505,
+      "grad_norm": 0.5638243556022644,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 10138
+    },
+    {
+      "epoch": 1.1646660157371778,
+      "grad_norm": 0.546511173248291,
+      "learning_rate": 0.0001,
+      "loss": 1.4114,
+      "step": 10139
+    },
+    {
+      "epoch": 1.1647808856470048,
+      "grad_norm": 0.5502268671989441,
+      "learning_rate": 0.0001,
+      "loss": 1.4463,
+      "step": 10140
+    },
+    {
+      "epoch": 1.1648957555568318,
+      "grad_norm": 0.5502709150314331,
+      "learning_rate": 0.0001,
+      "loss": 1.4657,
+      "step": 10141
+    },
+    {
+      "epoch": 1.165010625466659,
+      "grad_norm": 0.5224112868309021,
+      "learning_rate": 0.0001,
+      "loss": 1.2935,
+      "step": 10142
+    },
+    {
+      "epoch": 1.1651254953764862,
+      "grad_norm": 0.5977753400802612,
+      "learning_rate": 0.0001,
+      "loss": 1.4873,
+      "step": 10143
+    },
+    {
+      "epoch": 1.1652403652863133,
+      "grad_norm": 0.5752722024917603,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 10144
+    },
+    {
+      "epoch": 1.1653552351961403,
+      "grad_norm": 0.5518738031387329,
+      "learning_rate": 0.0001,
+      "loss": 1.3145,
+      "step": 10145
+    },
+    {
+      "epoch": 1.1654701051059675,
+      "grad_norm": 0.5660110712051392,
+      "learning_rate": 0.0001,
+      "loss": 1.4595,
+      "step": 10146
+    },
+    {
+      "epoch": 1.1655849750157947,
+      "grad_norm": 0.6433327198028564,
+      "learning_rate": 0.0001,
+      "loss": 1.7201,
+      "step": 10147
+    },
+    {
+      "epoch": 1.1656998449256217,
+      "grad_norm": 0.5778366327285767,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 10148
+    },
+    {
+      "epoch": 1.1658147148354487,
+      "grad_norm": 0.5929848551750183,
+      "learning_rate": 0.0001,
+      "loss": 1.3827,
+      "step": 10149
+    },
+    {
+      "epoch": 1.165929584745276,
+      "grad_norm": 0.5661731958389282,
+      "learning_rate": 0.0001,
+      "loss": 1.3457,
+      "step": 10150
+    },
+    {
+      "epoch": 1.1660444546551032,
+      "grad_norm": 0.5540391802787781,
+      "learning_rate": 0.0001,
+      "loss": 1.513,
+      "step": 10151
+    },
+    {
+      "epoch": 1.1661593245649302,
+      "grad_norm": 0.5562129616737366,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 10152
+    },
+    {
+      "epoch": 1.1662741944747572,
+      "grad_norm": 0.5246435403823853,
+      "learning_rate": 0.0001,
+      "loss": 1.4335,
+      "step": 10153
+    },
+    {
+      "epoch": 1.1663890643845845,
+      "grad_norm": 0.650452733039856,
+      "learning_rate": 0.0001,
+      "loss": 1.7906,
+      "step": 10154
+    },
+    {
+      "epoch": 1.1665039342944117,
+      "grad_norm": 0.5996583104133606,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 10155
+    },
+    {
+      "epoch": 1.1666188042042387,
+      "grad_norm": 0.52683424949646,
+      "learning_rate": 0.0001,
+      "loss": 1.2967,
+      "step": 10156
+    },
+    {
+      "epoch": 1.1667336741140657,
+      "grad_norm": 0.5419752597808838,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 10157
+    },
+    {
+      "epoch": 1.166848544023893,
+      "grad_norm": 0.5371859073638916,
+      "learning_rate": 0.0001,
+      "loss": 1.2567,
+      "step": 10158
+    },
+    {
+      "epoch": 1.1669634139337202,
+      "grad_norm": 0.6510452032089233,
+      "learning_rate": 0.0001,
+      "loss": 1.7457,
+      "step": 10159
+    },
+    {
+      "epoch": 1.1670782838435472,
+      "grad_norm": 0.5702193379402161,
+      "learning_rate": 0.0001,
+      "loss": 1.5441,
+      "step": 10160
+    },
+    {
+      "epoch": 1.1671931537533742,
+      "grad_norm": 0.5725999474525452,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 10161
+    },
+    {
+      "epoch": 1.1673080236632014,
+      "grad_norm": 0.5554792284965515,
+      "learning_rate": 0.0001,
+      "loss": 1.4167,
+      "step": 10162
+    },
+    {
+      "epoch": 1.1674228935730286,
+      "grad_norm": 0.5548241138458252,
+      "learning_rate": 0.0001,
+      "loss": 1.3767,
+      "step": 10163
+    },
+    {
+      "epoch": 1.1675377634828557,
+      "grad_norm": 0.6183516979217529,
+      "learning_rate": 0.0001,
+      "loss": 1.626,
+      "step": 10164
+    },
+    {
+      "epoch": 1.1676526333926829,
+      "grad_norm": 0.5450295209884644,
+      "learning_rate": 0.0001,
+      "loss": 1.5589,
+      "step": 10165
+    },
+    {
+      "epoch": 1.16776750330251,
+      "grad_norm": 0.5990351438522339,
+      "learning_rate": 0.0001,
+      "loss": 1.544,
+      "step": 10166
+    },
+    {
+      "epoch": 1.1678823732123371,
+      "grad_norm": 0.5059851408004761,
+      "learning_rate": 0.0001,
+      "loss": 1.4095,
+      "step": 10167
+    },
+    {
+      "epoch": 1.1679972431221641,
+      "grad_norm": 0.534441351890564,
+      "learning_rate": 0.0001,
+      "loss": 1.5801,
+      "step": 10168
+    },
+    {
+      "epoch": 1.1681121130319914,
+      "grad_norm": 0.5403814911842346,
+      "learning_rate": 0.0001,
+      "loss": 1.4503,
+      "step": 10169
+    },
+    {
+      "epoch": 1.1682269829418184,
+      "grad_norm": 0.5525118112564087,
+      "learning_rate": 0.0001,
+      "loss": 1.2902,
+      "step": 10170
+    },
+    {
+      "epoch": 1.1683418528516456,
+      "grad_norm": 0.5069050788879395,
+      "learning_rate": 0.0001,
+      "loss": 1.3002,
+      "step": 10171
+    },
+    {
+      "epoch": 1.1684567227614726,
+      "grad_norm": 0.5735329389572144,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 10172
+    },
+    {
+      "epoch": 1.1685715926712998,
+      "grad_norm": 0.5312854647636414,
+      "learning_rate": 0.0001,
+      "loss": 1.6037,
+      "step": 10173
+    },
+    {
+      "epoch": 1.1686864625811269,
+      "grad_norm": 0.5686579346656799,
+      "learning_rate": 0.0001,
+      "loss": 1.5586,
+      "step": 10174
+    },
+    {
+      "epoch": 1.168801332490954,
+      "grad_norm": 0.5913828015327454,
+      "learning_rate": 0.0001,
+      "loss": 1.4098,
+      "step": 10175
+    },
+    {
+      "epoch": 1.168916202400781,
+      "grad_norm": 0.5945882797241211,
+      "learning_rate": 0.0001,
+      "loss": 1.1721,
+      "step": 10176
+    },
+    {
+      "epoch": 1.1690310723106083,
+      "grad_norm": 0.565756618976593,
+      "learning_rate": 0.0001,
+      "loss": 1.3638,
+      "step": 10177
+    },
+    {
+      "epoch": 1.1691459422204353,
+      "grad_norm": 0.557446300983429,
+      "learning_rate": 0.0001,
+      "loss": 1.3487,
+      "step": 10178
+    },
+    {
+      "epoch": 1.1692608121302626,
+      "grad_norm": 0.5504045486450195,
+      "learning_rate": 0.0001,
+      "loss": 1.4271,
+      "step": 10179
+    },
+    {
+      "epoch": 1.1693756820400896,
+      "grad_norm": 0.576124906539917,
+      "learning_rate": 0.0001,
+      "loss": 1.5313,
+      "step": 10180
+    },
+    {
+      "epoch": 1.1694905519499168,
+      "grad_norm": 0.5811876058578491,
+      "learning_rate": 0.0001,
+      "loss": 1.3186,
+      "step": 10181
+    },
+    {
+      "epoch": 1.1696054218597438,
+      "grad_norm": 0.5304403305053711,
+      "learning_rate": 0.0001,
+      "loss": 1.4356,
+      "step": 10182
+    },
+    {
+      "epoch": 1.169720291769571,
+      "grad_norm": 0.5441917777061462,
+      "learning_rate": 0.0001,
+      "loss": 1.4705,
+      "step": 10183
+    },
+    {
+      "epoch": 1.169835161679398,
+      "grad_norm": 0.5729200839996338,
+      "learning_rate": 0.0001,
+      "loss": 1.5841,
+      "step": 10184
+    },
+    {
+      "epoch": 1.1699500315892253,
+      "grad_norm": 0.5269560813903809,
+      "learning_rate": 0.0001,
+      "loss": 1.4486,
+      "step": 10185
+    },
+    {
+      "epoch": 1.1700649014990523,
+      "grad_norm": 0.6391506791114807,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 10186
+    },
+    {
+      "epoch": 1.1701797714088795,
+      "grad_norm": 0.5685237646102905,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 10187
+    },
+    {
+      "epoch": 1.1702946413187065,
+      "grad_norm": 0.6011033654212952,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 10188
+    },
+    {
+      "epoch": 1.1704095112285338,
+      "grad_norm": 0.5902635455131531,
+      "learning_rate": 0.0001,
+      "loss": 1.3985,
+      "step": 10189
+    },
+    {
+      "epoch": 1.1705243811383608,
+      "grad_norm": 0.5687843561172485,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 10190
+    },
+    {
+      "epoch": 1.170639251048188,
+      "grad_norm": 0.5689520239830017,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 10191
+    },
+    {
+      "epoch": 1.170754120958015,
+      "grad_norm": 0.5470170378684998,
+      "learning_rate": 0.0001,
+      "loss": 1.2694,
+      "step": 10192
+    },
+    {
+      "epoch": 1.1708689908678422,
+      "grad_norm": 0.5402204394340515,
+      "learning_rate": 0.0001,
+      "loss": 1.3891,
+      "step": 10193
+    },
+    {
+      "epoch": 1.1709838607776692,
+      "grad_norm": 0.5700439214706421,
+      "learning_rate": 0.0001,
+      "loss": 1.2554,
+      "step": 10194
+    },
+    {
+      "epoch": 1.1710987306874965,
+      "grad_norm": 0.5934173464775085,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 10195
+    },
+    {
+      "epoch": 1.1712136005973235,
+      "grad_norm": 0.5891930460929871,
+      "learning_rate": 0.0001,
+      "loss": 1.5339,
+      "step": 10196
+    },
+    {
+      "epoch": 1.1713284705071507,
+      "grad_norm": 0.5680025219917297,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 10197
+    },
+    {
+      "epoch": 1.1714433404169777,
+      "grad_norm": 0.557145893573761,
+      "learning_rate": 0.0001,
+      "loss": 1.363,
+      "step": 10198
+    },
+    {
+      "epoch": 1.171558210326805,
+      "grad_norm": 0.5671913623809814,
+      "learning_rate": 0.0001,
+      "loss": 1.1882,
+      "step": 10199
+    },
+    {
+      "epoch": 1.171673080236632,
+      "grad_norm": 0.5525150895118713,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 10200
+    },
+    {
+      "epoch": 1.1717879501464592,
+      "grad_norm": 0.5642703771591187,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 10201
+    },
+    {
+      "epoch": 1.1719028200562862,
+      "grad_norm": 0.5610083341598511,
+      "learning_rate": 0.0001,
+      "loss": 1.5124,
+      "step": 10202
+    },
+    {
+      "epoch": 1.1720176899661134,
+      "grad_norm": 0.5838199257850647,
+      "learning_rate": 0.0001,
+      "loss": 1.4971,
+      "step": 10203
+    },
+    {
+      "epoch": 1.1721325598759404,
+      "grad_norm": 0.6114639639854431,
+      "learning_rate": 0.0001,
+      "loss": 1.45,
+      "step": 10204
+    },
+    {
+      "epoch": 1.1722474297857677,
+      "grad_norm": 0.6125414371490479,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 10205
+    },
+    {
+      "epoch": 1.1723622996955947,
+      "grad_norm": 0.6367550492286682,
+      "learning_rate": 0.0001,
+      "loss": 1.7869,
+      "step": 10206
+    },
+    {
+      "epoch": 1.172477169605422,
+      "grad_norm": 0.5995863080024719,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 10207
+    },
+    {
+      "epoch": 1.172592039515249,
+      "grad_norm": 0.5278496742248535,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 10208
+    },
+    {
+      "epoch": 1.1727069094250762,
+      "grad_norm": 0.5485237240791321,
+      "learning_rate": 0.0001,
+      "loss": 1.5175,
+      "step": 10209
+    },
+    {
+      "epoch": 1.1728217793349032,
+      "grad_norm": 0.6074742674827576,
+      "learning_rate": 0.0001,
+      "loss": 1.3414,
+      "step": 10210
+    },
+    {
+      "epoch": 1.1729366492447304,
+      "grad_norm": 0.5964166522026062,
+      "learning_rate": 0.0001,
+      "loss": 1.3031,
+      "step": 10211
+    },
+    {
+      "epoch": 1.1730515191545574,
+      "grad_norm": 0.5800673365592957,
+      "learning_rate": 0.0001,
+      "loss": 1.4228,
+      "step": 10212
+    },
+    {
+      "epoch": 1.1731663890643846,
+      "grad_norm": 0.5409739017486572,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 10213
+    },
+    {
+      "epoch": 1.1732812589742116,
+      "grad_norm": 0.5506752729415894,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 10214
+    },
+    {
+      "epoch": 1.1733961288840389,
+      "grad_norm": 0.551644504070282,
+      "learning_rate": 0.0001,
+      "loss": 1.4612,
+      "step": 10215
+    },
+    {
+      "epoch": 1.173510998793866,
+      "grad_norm": 0.5593010783195496,
+      "learning_rate": 0.0001,
+      "loss": 1.4069,
+      "step": 10216
+    },
+    {
+      "epoch": 1.1736258687036931,
+      "grad_norm": 0.5570374131202698,
+      "learning_rate": 0.0001,
+      "loss": 1.1389,
+      "step": 10217
+    },
+    {
+      "epoch": 1.1737407386135201,
+      "grad_norm": 0.6743999719619751,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 10218
+    },
+    {
+      "epoch": 1.1738556085233474,
+      "grad_norm": 0.5389916300773621,
+      "learning_rate": 0.0001,
+      "loss": 1.3792,
+      "step": 10219
+    },
+    {
+      "epoch": 1.1739704784331744,
+      "grad_norm": 0.5596214532852173,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 10220
+    },
+    {
+      "epoch": 1.1740853483430016,
+      "grad_norm": 0.6074917912483215,
+      "learning_rate": 0.0001,
+      "loss": 1.2292,
+      "step": 10221
+    },
+    {
+      "epoch": 1.1742002182528286,
+      "grad_norm": 0.5609689950942993,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 10222
+    },
+    {
+      "epoch": 1.1743150881626558,
+      "grad_norm": 0.554728627204895,
+      "learning_rate": 0.0001,
+      "loss": 1.4622,
+      "step": 10223
+    },
+    {
+      "epoch": 1.1744299580724828,
+      "grad_norm": 0.548478901386261,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 10224
+    },
+    {
+      "epoch": 1.17454482798231,
+      "grad_norm": 0.5518718957901001,
+      "learning_rate": 0.0001,
+      "loss": 1.4001,
+      "step": 10225
+    },
+    {
+      "epoch": 1.174659697892137,
+      "grad_norm": 0.6477909088134766,
+      "learning_rate": 0.0001,
+      "loss": 1.5887,
+      "step": 10226
+    },
+    {
+      "epoch": 1.1747745678019643,
+      "grad_norm": 0.5640280842781067,
+      "learning_rate": 0.0001,
+      "loss": 1.4278,
+      "step": 10227
+    },
+    {
+      "epoch": 1.1748894377117913,
+      "grad_norm": 0.5902442932128906,
+      "learning_rate": 0.0001,
+      "loss": 1.3603,
+      "step": 10228
+    },
+    {
+      "epoch": 1.1750043076216186,
+      "grad_norm": 0.575197160243988,
+      "learning_rate": 0.0001,
+      "loss": 1.4094,
+      "step": 10229
+    },
+    {
+      "epoch": 1.1751191775314456,
+      "grad_norm": 0.5721563696861267,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 10230
+    },
+    {
+      "epoch": 1.1752340474412728,
+      "grad_norm": 0.7208542823791504,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 10231
+    },
+    {
+      "epoch": 1.1753489173510998,
+      "grad_norm": 0.5777501463890076,
+      "learning_rate": 0.0001,
+      "loss": 1.3529,
+      "step": 10232
+    },
+    {
+      "epoch": 1.175463787260927,
+      "grad_norm": 0.5787646174430847,
+      "learning_rate": 0.0001,
+      "loss": 1.4706,
+      "step": 10233
+    },
+    {
+      "epoch": 1.175578657170754,
+      "grad_norm": 0.5515586733818054,
+      "learning_rate": 0.0001,
+      "loss": 1.5613,
+      "step": 10234
+    },
+    {
+      "epoch": 1.1756935270805813,
+      "grad_norm": 0.5776956677436829,
+      "learning_rate": 0.0001,
+      "loss": 1.4891,
+      "step": 10235
+    },
+    {
+      "epoch": 1.1758083969904083,
+      "grad_norm": 0.6034742593765259,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 10236
+    },
+    {
+      "epoch": 1.1759232669002355,
+      "grad_norm": 0.5228462219238281,
+      "learning_rate": 0.0001,
+      "loss": 1.5063,
+      "step": 10237
+    },
+    {
+      "epoch": 1.1760381368100625,
+      "grad_norm": 0.5508900880813599,
+      "learning_rate": 0.0001,
+      "loss": 1.477,
+      "step": 10238
+    },
+    {
+      "epoch": 1.1761530067198898,
+      "grad_norm": 0.5309219360351562,
+      "learning_rate": 0.0001,
+      "loss": 1.446,
+      "step": 10239
+    },
+    {
+      "epoch": 1.1762678766297168,
+      "grad_norm": 0.5277041792869568,
+      "learning_rate": 0.0001,
+      "loss": 1.405,
+      "step": 10240
+    },
+    {
+      "epoch": 1.176382746539544,
+      "grad_norm": 0.5799806118011475,
+      "learning_rate": 0.0001,
+      "loss": 1.342,
+      "step": 10241
+    },
+    {
+      "epoch": 1.176497616449371,
+      "grad_norm": 0.5781192183494568,
+      "learning_rate": 0.0001,
+      "loss": 1.356,
+      "step": 10242
+    },
+    {
+      "epoch": 1.1766124863591982,
+      "grad_norm": 0.5948365330696106,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 10243
+    },
+    {
+      "epoch": 1.1767273562690252,
+      "grad_norm": 0.5791916251182556,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 10244
+    },
+    {
+      "epoch": 1.1768422261788525,
+      "grad_norm": 0.60085129737854,
+      "learning_rate": 0.0001,
+      "loss": 1.4501,
+      "step": 10245
+    },
+    {
+      "epoch": 1.1769570960886795,
+      "grad_norm": 0.5680752396583557,
+      "learning_rate": 0.0001,
+      "loss": 1.5569,
+      "step": 10246
+    },
+    {
+      "epoch": 1.1770719659985067,
+      "grad_norm": 0.5210364460945129,
+      "learning_rate": 0.0001,
+      "loss": 1.4383,
+      "step": 10247
+    },
+    {
+      "epoch": 1.1771868359083337,
+      "grad_norm": 0.5816622972488403,
+      "learning_rate": 0.0001,
+      "loss": 1.543,
+      "step": 10248
+    },
+    {
+      "epoch": 1.177301705818161,
+      "grad_norm": 0.5394155383110046,
+      "learning_rate": 0.0001,
+      "loss": 1.4946,
+      "step": 10249
+    },
+    {
+      "epoch": 1.177416575727988,
+      "grad_norm": 0.5853523015975952,
+      "learning_rate": 0.0001,
+      "loss": 1.6292,
+      "step": 10250
+    },
+    {
+      "epoch": 1.1775314456378152,
+      "grad_norm": 0.5694888234138489,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 10251
+    },
+    {
+      "epoch": 1.1776463155476422,
+      "grad_norm": 0.5495486855506897,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 10252
+    },
+    {
+      "epoch": 1.1777611854574694,
+      "grad_norm": 0.5514094233512878,
+      "learning_rate": 0.0001,
+      "loss": 1.1921,
+      "step": 10253
+    },
+    {
+      "epoch": 1.1778760553672964,
+      "grad_norm": 0.5321869850158691,
+      "learning_rate": 0.0001,
+      "loss": 1.4492,
+      "step": 10254
+    },
+    {
+      "epoch": 1.1779909252771237,
+      "grad_norm": 0.5383161902427673,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 10255
+    },
+    {
+      "epoch": 1.1781057951869507,
+      "grad_norm": 0.5531543493270874,
+      "learning_rate": 0.0001,
+      "loss": 1.494,
+      "step": 10256
+    },
+    {
+      "epoch": 1.178220665096778,
+      "grad_norm": 0.5813853144645691,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 10257
+    },
+    {
+      "epoch": 1.178335535006605,
+      "grad_norm": 0.6172130703926086,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 10258
+    },
+    {
+      "epoch": 1.1784504049164322,
+      "grad_norm": 0.6372018456459045,
+      "learning_rate": 0.0001,
+      "loss": 1.4326,
+      "step": 10259
+    },
+    {
+      "epoch": 1.1785652748262592,
+      "grad_norm": 0.5197803974151611,
+      "learning_rate": 0.0001,
+      "loss": 1.2676,
+      "step": 10260
+    },
+    {
+      "epoch": 1.1786801447360864,
+      "grad_norm": 0.549474835395813,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 10261
+    },
+    {
+      "epoch": 1.1787950146459134,
+      "grad_norm": 0.5710943937301636,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 10262
+    },
+    {
+      "epoch": 1.1789098845557406,
+      "grad_norm": 0.5572322010993958,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 10263
+    },
+    {
+      "epoch": 1.1790247544655676,
+      "grad_norm": 0.5510053038597107,
+      "learning_rate": 0.0001,
+      "loss": 1.2321,
+      "step": 10264
+    },
+    {
+      "epoch": 1.1791396243753949,
+      "grad_norm": 0.5195276141166687,
+      "learning_rate": 0.0001,
+      "loss": 1.2855,
+      "step": 10265
+    },
+    {
+      "epoch": 1.1792544942852219,
+      "grad_norm": 0.5562936663627625,
+      "learning_rate": 0.0001,
+      "loss": 1.3052,
+      "step": 10266
+    },
+    {
+      "epoch": 1.1793693641950491,
+      "grad_norm": 0.5841695666313171,
+      "learning_rate": 0.0001,
+      "loss": 1.6325,
+      "step": 10267
+    },
+    {
+      "epoch": 1.1794842341048764,
+      "grad_norm": 0.5963363647460938,
+      "learning_rate": 0.0001,
+      "loss": 1.4882,
+      "step": 10268
+    },
+    {
+      "epoch": 1.1795991040147034,
+      "grad_norm": 0.5893837809562683,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 10269
+    },
+    {
+      "epoch": 1.1797139739245304,
+      "grad_norm": 0.5855959057807922,
+      "learning_rate": 0.0001,
+      "loss": 1.4191,
+      "step": 10270
+    },
+    {
+      "epoch": 1.1798288438343576,
+      "grad_norm": 0.5364851355552673,
+      "learning_rate": 0.0001,
+      "loss": 1.3232,
+      "step": 10271
+    },
+    {
+      "epoch": 1.1799437137441848,
+      "grad_norm": 0.574670135974884,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 10272
+    },
+    {
+      "epoch": 1.1800585836540118,
+      "grad_norm": 0.5624828934669495,
+      "learning_rate": 0.0001,
+      "loss": 1.4357,
+      "step": 10273
+    },
+    {
+      "epoch": 1.1801734535638388,
+      "grad_norm": 0.585806131362915,
+      "learning_rate": 0.0001,
+      "loss": 1.4264,
+      "step": 10274
+    },
+    {
+      "epoch": 1.180288323473666,
+      "grad_norm": 0.6353849172592163,
+      "learning_rate": 0.0001,
+      "loss": 1.76,
+      "step": 10275
+    },
+    {
+      "epoch": 1.1804031933834933,
+      "grad_norm": 0.5217344760894775,
+      "learning_rate": 0.0001,
+      "loss": 1.4722,
+      "step": 10276
+    },
+    {
+      "epoch": 1.1805180632933203,
+      "grad_norm": 0.5620312690734863,
+      "learning_rate": 0.0001,
+      "loss": 1.3626,
+      "step": 10277
+    },
+    {
+      "epoch": 1.1806329332031473,
+      "grad_norm": 0.6162904500961304,
+      "learning_rate": 0.0001,
+      "loss": 1.2,
+      "step": 10278
+    },
+    {
+      "epoch": 1.1807478031129746,
+      "grad_norm": 0.5573971271514893,
+      "learning_rate": 0.0001,
+      "loss": 1.2377,
+      "step": 10279
+    },
+    {
+      "epoch": 1.1808626730228018,
+      "grad_norm": 0.5593479871749878,
+      "learning_rate": 0.0001,
+      "loss": 1.2591,
+      "step": 10280
+    },
+    {
+      "epoch": 1.1809775429326288,
+      "grad_norm": 0.5311997532844543,
+      "learning_rate": 0.0001,
+      "loss": 1.4939,
+      "step": 10281
+    },
+    {
+      "epoch": 1.1810924128424558,
+      "grad_norm": 0.5329908728599548,
+      "learning_rate": 0.0001,
+      "loss": 1.4148,
+      "step": 10282
+    },
+    {
+      "epoch": 1.181207282752283,
+      "grad_norm": 0.6186027526855469,
+      "learning_rate": 0.0001,
+      "loss": 1.7251,
+      "step": 10283
+    },
+    {
+      "epoch": 1.1813221526621103,
+      "grad_norm": 0.5550754070281982,
+      "learning_rate": 0.0001,
+      "loss": 1.3355,
+      "step": 10284
+    },
+    {
+      "epoch": 1.1814370225719373,
+      "grad_norm": 0.5469382405281067,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 10285
+    },
+    {
+      "epoch": 1.1815518924817643,
+      "grad_norm": 0.5690637826919556,
+      "learning_rate": 0.0001,
+      "loss": 1.4224,
+      "step": 10286
+    },
+    {
+      "epoch": 1.1816667623915915,
+      "grad_norm": 0.5610548853874207,
+      "learning_rate": 0.0001,
+      "loss": 1.4282,
+      "step": 10287
+    },
+    {
+      "epoch": 1.1817816323014187,
+      "grad_norm": 0.5904365181922913,
+      "learning_rate": 0.0001,
+      "loss": 1.5588,
+      "step": 10288
+    },
+    {
+      "epoch": 1.1818965022112458,
+      "grad_norm": 0.538159966468811,
+      "learning_rate": 0.0001,
+      "loss": 1.3842,
+      "step": 10289
+    },
+    {
+      "epoch": 1.1820113721210728,
+      "grad_norm": 0.5710332989692688,
+      "learning_rate": 0.0001,
+      "loss": 1.441,
+      "step": 10290
+    },
+    {
+      "epoch": 1.1821262420309,
+      "grad_norm": 0.5427202582359314,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 10291
+    },
+    {
+      "epoch": 1.1822411119407272,
+      "grad_norm": 0.5821972489356995,
+      "learning_rate": 0.0001,
+      "loss": 1.3847,
+      "step": 10292
+    },
+    {
+      "epoch": 1.1823559818505542,
+      "grad_norm": 0.5653535723686218,
+      "learning_rate": 0.0001,
+      "loss": 1.5676,
+      "step": 10293
+    },
+    {
+      "epoch": 1.1824708517603812,
+      "grad_norm": 0.5518774390220642,
+      "learning_rate": 0.0001,
+      "loss": 1.4604,
+      "step": 10294
+    },
+    {
+      "epoch": 1.1825857216702085,
+      "grad_norm": 0.506719172000885,
+      "learning_rate": 0.0001,
+      "loss": 1.4198,
+      "step": 10295
+    },
+    {
+      "epoch": 1.1827005915800357,
+      "grad_norm": 0.5615983009338379,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 10296
+    },
+    {
+      "epoch": 1.1828154614898627,
+      "grad_norm": 0.5364108681678772,
+      "learning_rate": 0.0001,
+      "loss": 1.3754,
+      "step": 10297
+    },
+    {
+      "epoch": 1.1829303313996897,
+      "grad_norm": 0.552790105342865,
+      "learning_rate": 0.0001,
+      "loss": 1.3189,
+      "step": 10298
+    },
+    {
+      "epoch": 1.183045201309517,
+      "grad_norm": 0.5753331780433655,
+      "learning_rate": 0.0001,
+      "loss": 1.6094,
+      "step": 10299
+    },
+    {
+      "epoch": 1.1831600712193442,
+      "grad_norm": 0.5673770308494568,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 10300
+    },
+    {
+      "epoch": 1.1832749411291712,
+      "grad_norm": 0.5506584644317627,
+      "learning_rate": 0.0001,
+      "loss": 1.4499,
+      "step": 10301
+    },
+    {
+      "epoch": 1.1833898110389984,
+      "grad_norm": 0.5558772087097168,
+      "learning_rate": 0.0001,
+      "loss": 1.5703,
+      "step": 10302
+    },
+    {
+      "epoch": 1.1835046809488254,
+      "grad_norm": 0.5837011337280273,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 10303
+    },
+    {
+      "epoch": 1.1836195508586527,
+      "grad_norm": 0.5474368333816528,
+      "learning_rate": 0.0001,
+      "loss": 1.4428,
+      "step": 10304
+    },
+    {
+      "epoch": 1.1837344207684797,
+      "grad_norm": 0.5366905331611633,
+      "learning_rate": 0.0001,
+      "loss": 1.4069,
+      "step": 10305
+    },
+    {
+      "epoch": 1.183849290678307,
+      "grad_norm": 0.5617532134056091,
+      "learning_rate": 0.0001,
+      "loss": 1.4165,
+      "step": 10306
+    },
+    {
+      "epoch": 1.183964160588134,
+      "grad_norm": 0.5564953684806824,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 10307
+    },
+    {
+      "epoch": 1.1840790304979611,
+      "grad_norm": 0.5196337699890137,
+      "learning_rate": 0.0001,
+      "loss": 1.3772,
+      "step": 10308
+    },
+    {
+      "epoch": 1.1841939004077882,
+      "grad_norm": 0.5515973567962646,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 10309
+    },
+    {
+      "epoch": 1.1843087703176154,
+      "grad_norm": 0.5052083730697632,
+      "learning_rate": 0.0001,
+      "loss": 1.4214,
+      "step": 10310
+    },
+    {
+      "epoch": 1.1844236402274424,
+      "grad_norm": 0.6033722162246704,
+      "learning_rate": 0.0001,
+      "loss": 1.7061,
+      "step": 10311
+    },
+    {
+      "epoch": 1.1845385101372696,
+      "grad_norm": 0.535667359828949,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 10312
+    },
+    {
+      "epoch": 1.1846533800470966,
+      "grad_norm": 0.5910223722457886,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 10313
+    },
+    {
+      "epoch": 1.1847682499569239,
+      "grad_norm": 0.542930543422699,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 10314
+    },
+    {
+      "epoch": 1.1848831198667509,
+      "grad_norm": 0.5212183594703674,
+      "learning_rate": 0.0001,
+      "loss": 1.5306,
+      "step": 10315
+    },
+    {
+      "epoch": 1.184997989776578,
+      "grad_norm": 0.5621635317802429,
+      "learning_rate": 0.0001,
+      "loss": 1.5478,
+      "step": 10316
+    },
+    {
+      "epoch": 1.1851128596864051,
+      "grad_norm": 0.6084299087524414,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 10317
+    },
+    {
+      "epoch": 1.1852277295962323,
+      "grad_norm": 0.5304902195930481,
+      "learning_rate": 0.0001,
+      "loss": 1.3968,
+      "step": 10318
+    },
+    {
+      "epoch": 1.1853425995060594,
+      "grad_norm": 0.5503674745559692,
+      "learning_rate": 0.0001,
+      "loss": 1.5019,
+      "step": 10319
+    },
+    {
+      "epoch": 1.1854574694158866,
+      "grad_norm": 0.5336641073226929,
+      "learning_rate": 0.0001,
+      "loss": 1.3079,
+      "step": 10320
+    },
+    {
+      "epoch": 1.1855723393257136,
+      "grad_norm": 0.6056388020515442,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 10321
+    },
+    {
+      "epoch": 1.1856872092355408,
+      "grad_norm": 0.5721645355224609,
+      "learning_rate": 0.0001,
+      "loss": 1.4073,
+      "step": 10322
+    },
+    {
+      "epoch": 1.1858020791453678,
+      "grad_norm": 0.6094290018081665,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 10323
+    },
+    {
+      "epoch": 1.185916949055195,
+      "grad_norm": 0.558312177658081,
+      "learning_rate": 0.0001,
+      "loss": 1.4004,
+      "step": 10324
+    },
+    {
+      "epoch": 1.186031818965022,
+      "grad_norm": 0.5515640377998352,
+      "learning_rate": 0.0001,
+      "loss": 1.4245,
+      "step": 10325
+    },
+    {
+      "epoch": 1.1861466888748493,
+      "grad_norm": 0.6152453422546387,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 10326
+    },
+    {
+      "epoch": 1.1862615587846763,
+      "grad_norm": 0.5758329629898071,
+      "learning_rate": 0.0001,
+      "loss": 1.3415,
+      "step": 10327
+    },
+    {
+      "epoch": 1.1863764286945035,
+      "grad_norm": 0.5209949612617493,
+      "learning_rate": 0.0001,
+      "loss": 1.3905,
+      "step": 10328
+    },
+    {
+      "epoch": 1.1864912986043306,
+      "grad_norm": 0.5001648664474487,
+      "learning_rate": 0.0001,
+      "loss": 1.31,
+      "step": 10329
+    },
+    {
+      "epoch": 1.1866061685141578,
+      "grad_norm": 0.5593850612640381,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 10330
+    },
+    {
+      "epoch": 1.1867210384239848,
+      "grad_norm": 0.5500064492225647,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 10331
+    },
+    {
+      "epoch": 1.186835908333812,
+      "grad_norm": 0.5763779282569885,
+      "learning_rate": 0.0001,
+      "loss": 1.3441,
+      "step": 10332
+    },
+    {
+      "epoch": 1.186950778243639,
+      "grad_norm": 0.5549381375312805,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 10333
+    },
+    {
+      "epoch": 1.1870656481534663,
+      "grad_norm": 0.5587765574455261,
+      "learning_rate": 0.0001,
+      "loss": 1.3678,
+      "step": 10334
+    },
+    {
+      "epoch": 1.1871805180632933,
+      "grad_norm": 0.5837385654449463,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 10335
+    },
+    {
+      "epoch": 1.1872953879731205,
+      "grad_norm": 0.5909486413002014,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 10336
+    },
+    {
+      "epoch": 1.1874102578829475,
+      "grad_norm": 0.5553216338157654,
+      "learning_rate": 0.0001,
+      "loss": 1.4709,
+      "step": 10337
+    },
+    {
+      "epoch": 1.1875251277927747,
+      "grad_norm": 0.5260200500488281,
+      "learning_rate": 0.0001,
+      "loss": 1.3144,
+      "step": 10338
+    },
+    {
+      "epoch": 1.1876399977026018,
+      "grad_norm": 0.5829170346260071,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 10339
+    },
+    {
+      "epoch": 1.187754867612429,
+      "grad_norm": 0.5121257305145264,
+      "learning_rate": 0.0001,
+      "loss": 1.3636,
+      "step": 10340
+    },
+    {
+      "epoch": 1.187869737522256,
+      "grad_norm": 0.5356987714767456,
+      "learning_rate": 0.0001,
+      "loss": 1.2993,
+      "step": 10341
+    },
+    {
+      "epoch": 1.1879846074320832,
+      "grad_norm": 0.5900508761405945,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 10342
+    },
+    {
+      "epoch": 1.1880994773419102,
+      "grad_norm": 0.5535077452659607,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 10343
+    },
+    {
+      "epoch": 1.1882143472517375,
+      "grad_norm": 0.5894918441772461,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 10344
+    },
+    {
+      "epoch": 1.1883292171615645,
+      "grad_norm": 0.5290895104408264,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 10345
+    },
+    {
+      "epoch": 1.1884440870713917,
+      "grad_norm": 0.5344952344894409,
+      "learning_rate": 0.0001,
+      "loss": 1.3112,
+      "step": 10346
+    },
+    {
+      "epoch": 1.1885589569812187,
+      "grad_norm": 0.5518721342086792,
+      "learning_rate": 0.0001,
+      "loss": 1.386,
+      "step": 10347
+    },
+    {
+      "epoch": 1.188673826891046,
+      "grad_norm": 0.5459201335906982,
+      "learning_rate": 0.0001,
+      "loss": 1.3449,
+      "step": 10348
+    },
+    {
+      "epoch": 1.188788696800873,
+      "grad_norm": 0.5421302318572998,
+      "learning_rate": 0.0001,
+      "loss": 1.4268,
+      "step": 10349
+    },
+    {
+      "epoch": 1.1889035667107002,
+      "grad_norm": 0.5335027575492859,
+      "learning_rate": 0.0001,
+      "loss": 1.4623,
+      "step": 10350
+    },
+    {
+      "epoch": 1.1890184366205272,
+      "grad_norm": 0.5534794926643372,
+      "learning_rate": 0.0001,
+      "loss": 1.4581,
+      "step": 10351
+    },
+    {
+      "epoch": 1.1891333065303544,
+      "grad_norm": 0.5756953954696655,
+      "learning_rate": 0.0001,
+      "loss": 1.4053,
+      "step": 10352
+    },
+    {
+      "epoch": 1.1892481764401814,
+      "grad_norm": 0.6161737442016602,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 10353
+    },
+    {
+      "epoch": 1.1893630463500087,
+      "grad_norm": 0.6521564722061157,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 10354
+    },
+    {
+      "epoch": 1.1894779162598357,
+      "grad_norm": 0.5857300162315369,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 10355
+    },
+    {
+      "epoch": 1.189592786169663,
+      "grad_norm": 0.6228790283203125,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 10356
+    },
+    {
+      "epoch": 1.18970765607949,
+      "grad_norm": 0.5708868503570557,
+      "learning_rate": 0.0001,
+      "loss": 1.13,
+      "step": 10357
+    },
+    {
+      "epoch": 1.1898225259893171,
+      "grad_norm": 0.6147711873054504,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 10358
+    },
+    {
+      "epoch": 1.1899373958991442,
+      "grad_norm": 0.5267451405525208,
+      "learning_rate": 0.0001,
+      "loss": 1.4593,
+      "step": 10359
+    },
+    {
+      "epoch": 1.1900522658089714,
+      "grad_norm": 0.5641406178474426,
+      "learning_rate": 0.0001,
+      "loss": 1.6509,
+      "step": 10360
+    },
+    {
+      "epoch": 1.1901671357187984,
+      "grad_norm": 0.6026042699813843,
+      "learning_rate": 0.0001,
+      "loss": 1.5547,
+      "step": 10361
+    },
+    {
+      "epoch": 1.1902820056286256,
+      "grad_norm": 0.5601838231086731,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 10362
+    },
+    {
+      "epoch": 1.1903968755384526,
+      "grad_norm": 0.505669355392456,
+      "learning_rate": 0.0001,
+      "loss": 1.3774,
+      "step": 10363
+    },
+    {
+      "epoch": 1.1905117454482799,
+      "grad_norm": 0.5833644270896912,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 10364
+    },
+    {
+      "epoch": 1.1906266153581069,
+      "grad_norm": 0.5416715145111084,
+      "learning_rate": 0.0001,
+      "loss": 1.3589,
+      "step": 10365
+    },
+    {
+      "epoch": 1.190741485267934,
+      "grad_norm": 0.5769158601760864,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 10366
+    },
+    {
+      "epoch": 1.1908563551777611,
+      "grad_norm": 0.5252044200897217,
+      "learning_rate": 0.0001,
+      "loss": 1.2886,
+      "step": 10367
+    },
+    {
+      "epoch": 1.1909712250875883,
+      "grad_norm": 0.5603752136230469,
+      "learning_rate": 0.0001,
+      "loss": 1.4126,
+      "step": 10368
+    },
+    {
+      "epoch": 1.1910860949974154,
+      "grad_norm": 0.5847504138946533,
+      "learning_rate": 0.0001,
+      "loss": 1.4077,
+      "step": 10369
+    },
+    {
+      "epoch": 1.1912009649072426,
+      "grad_norm": 0.5426949262619019,
+      "learning_rate": 0.0001,
+      "loss": 1.3664,
+      "step": 10370
+    },
+    {
+      "epoch": 1.1913158348170696,
+      "grad_norm": 0.5475144386291504,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 10371
+    },
+    {
+      "epoch": 1.1914307047268968,
+      "grad_norm": 0.6072627305984497,
+      "learning_rate": 0.0001,
+      "loss": 1.6914,
+      "step": 10372
+    },
+    {
+      "epoch": 1.1915455746367238,
+      "grad_norm": 0.5702595710754395,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 10373
+    },
+    {
+      "epoch": 1.191660444546551,
+      "grad_norm": 0.5092419385910034,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 10374
+    },
+    {
+      "epoch": 1.191775314456378,
+      "grad_norm": 0.5371357202529907,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 10375
+    },
+    {
+      "epoch": 1.1918901843662053,
+      "grad_norm": 0.6508344411849976,
+      "learning_rate": 0.0001,
+      "loss": 1.4152,
+      "step": 10376
+    },
+    {
+      "epoch": 1.1920050542760323,
+      "grad_norm": 0.5292761921882629,
+      "learning_rate": 0.0001,
+      "loss": 1.2684,
+      "step": 10377
+    },
+    {
+      "epoch": 1.1921199241858595,
+      "grad_norm": 0.5798367857933044,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 10378
+    },
+    {
+      "epoch": 1.1922347940956866,
+      "grad_norm": 0.5566953420639038,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 10379
+    },
+    {
+      "epoch": 1.1923496640055138,
+      "grad_norm": 0.6179848909378052,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 10380
+    },
+    {
+      "epoch": 1.1924645339153408,
+      "grad_norm": 0.5988585948944092,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 10381
+    },
+    {
+      "epoch": 1.192579403825168,
+      "grad_norm": 0.6032318472862244,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 10382
+    },
+    {
+      "epoch": 1.192694273734995,
+      "grad_norm": 0.5561224818229675,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 10383
+    },
+    {
+      "epoch": 1.1928091436448223,
+      "grad_norm": 0.5613009333610535,
+      "learning_rate": 0.0001,
+      "loss": 1.5307,
+      "step": 10384
+    },
+    {
+      "epoch": 1.1929240135546493,
+      "grad_norm": 0.5561355352401733,
+      "learning_rate": 0.0001,
+      "loss": 1.5057,
+      "step": 10385
+    },
+    {
+      "epoch": 1.1930388834644765,
+      "grad_norm": 0.5625793933868408,
+      "learning_rate": 0.0001,
+      "loss": 1.397,
+      "step": 10386
+    },
+    {
+      "epoch": 1.1931537533743035,
+      "grad_norm": 0.5584040284156799,
+      "learning_rate": 0.0001,
+      "loss": 1.529,
+      "step": 10387
+    },
+    {
+      "epoch": 1.1932686232841307,
+      "grad_norm": 0.5510308146476746,
+      "learning_rate": 0.0001,
+      "loss": 1.3537,
+      "step": 10388
+    },
+    {
+      "epoch": 1.1933834931939578,
+      "grad_norm": 0.5637242794036865,
+      "learning_rate": 0.0001,
+      "loss": 1.4676,
+      "step": 10389
+    },
+    {
+      "epoch": 1.193498363103785,
+      "grad_norm": 0.604081392288208,
+      "learning_rate": 0.0001,
+      "loss": 1.4873,
+      "step": 10390
+    },
+    {
+      "epoch": 1.193613233013612,
+      "grad_norm": 0.5362832546234131,
+      "learning_rate": 0.0001,
+      "loss": 1.2121,
+      "step": 10391
+    },
+    {
+      "epoch": 1.1937281029234392,
+      "grad_norm": 0.6020187735557556,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 10392
+    },
+    {
+      "epoch": 1.1938429728332662,
+      "grad_norm": 0.5550323724746704,
+      "learning_rate": 0.0001,
+      "loss": 1.3623,
+      "step": 10393
+    },
+    {
+      "epoch": 1.1939578427430935,
+      "grad_norm": 0.6165339350700378,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 10394
+    },
+    {
+      "epoch": 1.1940727126529205,
+      "grad_norm": 0.5294722318649292,
+      "learning_rate": 0.0001,
+      "loss": 1.3107,
+      "step": 10395
+    },
+    {
+      "epoch": 1.1941875825627477,
+      "grad_norm": 0.5582475662231445,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 10396
+    },
+    {
+      "epoch": 1.1943024524725747,
+      "grad_norm": 0.6636613607406616,
+      "learning_rate": 0.0001,
+      "loss": 1.2665,
+      "step": 10397
+    },
+    {
+      "epoch": 1.194417322382402,
+      "grad_norm": 0.5724130868911743,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 10398
+    },
+    {
+      "epoch": 1.194532192292229,
+      "grad_norm": 0.5657904148101807,
+      "learning_rate": 0.0001,
+      "loss": 1.5283,
+      "step": 10399
+    },
+    {
+      "epoch": 1.1946470622020562,
+      "grad_norm": 0.5498049259185791,
+      "learning_rate": 0.0001,
+      "loss": 1.4346,
+      "step": 10400
+    },
+    {
+      "epoch": 1.1947619321118832,
+      "grad_norm": 0.5673407912254333,
+      "learning_rate": 0.0001,
+      "loss": 1.3824,
+      "step": 10401
+    },
+    {
+      "epoch": 1.1948768020217104,
+      "grad_norm": 0.609406054019928,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 10402
+    },
+    {
+      "epoch": 1.1949916719315374,
+      "grad_norm": 0.5271400809288025,
+      "learning_rate": 0.0001,
+      "loss": 1.3876,
+      "step": 10403
+    },
+    {
+      "epoch": 1.1951065418413647,
+      "grad_norm": 0.5551247596740723,
+      "learning_rate": 0.0001,
+      "loss": 1.4809,
+      "step": 10404
+    },
+    {
+      "epoch": 1.195221411751192,
+      "grad_norm": 0.6490895748138428,
+      "learning_rate": 0.0001,
+      "loss": 1.5965,
+      "step": 10405
+    },
+    {
+      "epoch": 1.195336281661019,
+      "grad_norm": 0.5571427941322327,
+      "learning_rate": 0.0001,
+      "loss": 1.4025,
+      "step": 10406
+    },
+    {
+      "epoch": 1.195451151570846,
+      "grad_norm": 0.5582363605499268,
+      "learning_rate": 0.0001,
+      "loss": 1.4128,
+      "step": 10407
+    },
+    {
+      "epoch": 1.1955660214806731,
+      "grad_norm": 0.5777066946029663,
+      "learning_rate": 0.0001,
+      "loss": 1.4456,
+      "step": 10408
+    },
+    {
+      "epoch": 1.1956808913905004,
+      "grad_norm": 0.5664290189743042,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 10409
+    },
+    {
+      "epoch": 1.1957957613003274,
+      "grad_norm": 0.6081710457801819,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 10410
+    },
+    {
+      "epoch": 1.1959106312101544,
+      "grad_norm": 0.5311821103096008,
+      "learning_rate": 0.0001,
+      "loss": 1.1295,
+      "step": 10411
+    },
+    {
+      "epoch": 1.1960255011199816,
+      "grad_norm": 0.5118587017059326,
+      "learning_rate": 0.0001,
+      "loss": 1.2627,
+      "step": 10412
+    },
+    {
+      "epoch": 1.1961403710298089,
+      "grad_norm": 0.5950908660888672,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 10413
+    },
+    {
+      "epoch": 1.1962552409396359,
+      "grad_norm": 0.5565327405929565,
+      "learning_rate": 0.0001,
+      "loss": 1.4616,
+      "step": 10414
+    },
+    {
+      "epoch": 1.1963701108494629,
+      "grad_norm": 0.5890604853630066,
+      "learning_rate": 0.0001,
+      "loss": 1.4465,
+      "step": 10415
+    },
+    {
+      "epoch": 1.19648498075929,
+      "grad_norm": 0.5750487446784973,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 10416
+    },
+    {
+      "epoch": 1.1965998506691173,
+      "grad_norm": 0.5400042533874512,
+      "learning_rate": 0.0001,
+      "loss": 1.2722,
+      "step": 10417
+    },
+    {
+      "epoch": 1.1967147205789443,
+      "grad_norm": 0.5932859778404236,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 10418
+    },
+    {
+      "epoch": 1.1968295904887714,
+      "grad_norm": 0.5861518383026123,
+      "learning_rate": 0.0001,
+      "loss": 1.4621,
+      "step": 10419
+    },
+    {
+      "epoch": 1.1969444603985986,
+      "grad_norm": 0.5601950883865356,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 10420
+    },
+    {
+      "epoch": 1.1970593303084258,
+      "grad_norm": 0.5883669257164001,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 10421
+    },
+    {
+      "epoch": 1.1971742002182528,
+      "grad_norm": 0.5568323135375977,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 10422
+    },
+    {
+      "epoch": 1.1972890701280798,
+      "grad_norm": 0.5244744420051575,
+      "learning_rate": 0.0001,
+      "loss": 1.3016,
+      "step": 10423
+    },
+    {
+      "epoch": 1.197403940037907,
+      "grad_norm": 0.5605520606040955,
+      "learning_rate": 0.0001,
+      "loss": 1.4803,
+      "step": 10424
+    },
+    {
+      "epoch": 1.1975188099477343,
+      "grad_norm": 0.5980091691017151,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 10425
+    },
+    {
+      "epoch": 1.1976336798575613,
+      "grad_norm": 0.607279896736145,
+      "learning_rate": 0.0001,
+      "loss": 1.5404,
+      "step": 10426
+    },
+    {
+      "epoch": 1.1977485497673883,
+      "grad_norm": 0.5929856896400452,
+      "learning_rate": 0.0001,
+      "loss": 1.5402,
+      "step": 10427
+    },
+    {
+      "epoch": 1.1978634196772155,
+      "grad_norm": 0.5797104835510254,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 10428
+    },
+    {
+      "epoch": 1.1979782895870428,
+      "grad_norm": 0.5740180611610413,
+      "learning_rate": 0.0001,
+      "loss": 1.5289,
+      "step": 10429
+    },
+    {
+      "epoch": 1.1980931594968698,
+      "grad_norm": 0.5844727158546448,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 10430
+    },
+    {
+      "epoch": 1.1982080294066968,
+      "grad_norm": 0.542769193649292,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 10431
+    },
+    {
+      "epoch": 1.198322899316524,
+      "grad_norm": 0.5846665501594543,
+      "learning_rate": 0.0001,
+      "loss": 1.4399,
+      "step": 10432
+    },
+    {
+      "epoch": 1.1984377692263513,
+      "grad_norm": 0.5460920333862305,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 10433
+    },
+    {
+      "epoch": 1.1985526391361783,
+      "grad_norm": 0.582313597202301,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 10434
+    },
+    {
+      "epoch": 1.1986675090460053,
+      "grad_norm": 0.6199240684509277,
+      "learning_rate": 0.0001,
+      "loss": 1.4142,
+      "step": 10435
+    },
+    {
+      "epoch": 1.1987823789558325,
+      "grad_norm": 0.6095178127288818,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 10436
+    },
+    {
+      "epoch": 1.1988972488656597,
+      "grad_norm": 0.5635156631469727,
+      "learning_rate": 0.0001,
+      "loss": 1.2746,
+      "step": 10437
+    },
+    {
+      "epoch": 1.1990121187754867,
+      "grad_norm": 0.5342568755149841,
+      "learning_rate": 0.0001,
+      "loss": 1.3511,
+      "step": 10438
+    },
+    {
+      "epoch": 1.199126988685314,
+      "grad_norm": 0.5932930111885071,
+      "learning_rate": 0.0001,
+      "loss": 1.3836,
+      "step": 10439
+    },
+    {
+      "epoch": 1.199241858595141,
+      "grad_norm": 0.5218791365623474,
+      "learning_rate": 0.0001,
+      "loss": 1.169,
+      "step": 10440
+    },
+    {
+      "epoch": 1.1993567285049682,
+      "grad_norm": 0.5568528771400452,
+      "learning_rate": 0.0001,
+      "loss": 1.5939,
+      "step": 10441
+    },
+    {
+      "epoch": 1.1994715984147952,
+      "grad_norm": 0.512910008430481,
+      "learning_rate": 0.0001,
+      "loss": 1.34,
+      "step": 10442
+    },
+    {
+      "epoch": 1.1995864683246225,
+      "grad_norm": 0.5277653336524963,
+      "learning_rate": 0.0001,
+      "loss": 1.4526,
+      "step": 10443
+    },
+    {
+      "epoch": 1.1997013382344495,
+      "grad_norm": 0.5597155094146729,
+      "learning_rate": 0.0001,
+      "loss": 1.3142,
+      "step": 10444
+    },
+    {
+      "epoch": 1.1998162081442767,
+      "grad_norm": 0.5273895859718323,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 10445
+    },
+    {
+      "epoch": 1.1999310780541037,
+      "grad_norm": 0.5639094114303589,
+      "learning_rate": 0.0001,
+      "loss": 1.2597,
+      "step": 10446
+    },
+    {
+      "epoch": 1.200045947963931,
+      "grad_norm": 0.6367859244346619,
+      "learning_rate": 0.0001,
+      "loss": 1.6992,
+      "step": 10447
+    },
+    {
+      "epoch": 1.200160817873758,
+      "grad_norm": 0.6159301400184631,
+      "learning_rate": 0.0001,
+      "loss": 1.475,
+      "step": 10448
+    },
+    {
+      "epoch": 1.2002756877835852,
+      "grad_norm": 0.5858900547027588,
+      "learning_rate": 0.0001,
+      "loss": 1.4077,
+      "step": 10449
+    },
+    {
+      "epoch": 1.2003905576934122,
+      "grad_norm": 0.5637074112892151,
+      "learning_rate": 0.0001,
+      "loss": 1.4567,
+      "step": 10450
+    },
+    {
+      "epoch": 1.2005054276032394,
+      "grad_norm": 0.5393563508987427,
+      "learning_rate": 0.0001,
+      "loss": 1.3936,
+      "step": 10451
+    },
+    {
+      "epoch": 1.2006202975130664,
+      "grad_norm": 0.530194103717804,
+      "learning_rate": 0.0001,
+      "loss": 1.4639,
+      "step": 10452
+    },
+    {
+      "epoch": 1.2007351674228937,
+      "grad_norm": 0.5299816727638245,
+      "learning_rate": 0.0001,
+      "loss": 1.4348,
+      "step": 10453
+    },
+    {
+      "epoch": 1.2008500373327207,
+      "grad_norm": 0.5166096687316895,
+      "learning_rate": 0.0001,
+      "loss": 1.331,
+      "step": 10454
+    },
+    {
+      "epoch": 1.200964907242548,
+      "grad_norm": 0.5923041105270386,
+      "learning_rate": 0.0001,
+      "loss": 1.4583,
+      "step": 10455
+    },
+    {
+      "epoch": 1.201079777152375,
+      "grad_norm": 0.5478973388671875,
+      "learning_rate": 0.0001,
+      "loss": 1.3418,
+      "step": 10456
+    },
+    {
+      "epoch": 1.2011946470622021,
+      "grad_norm": 0.5634967684745789,
+      "learning_rate": 0.0001,
+      "loss": 1.3768,
+      "step": 10457
+    },
+    {
+      "epoch": 1.2013095169720291,
+      "grad_norm": 0.5654754042625427,
+      "learning_rate": 0.0001,
+      "loss": 1.6154,
+      "step": 10458
+    },
+    {
+      "epoch": 1.2014243868818564,
+      "grad_norm": 0.6268435716629028,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 10459
+    },
+    {
+      "epoch": 1.2015392567916834,
+      "grad_norm": 0.5597686171531677,
+      "learning_rate": 0.0001,
+      "loss": 1.2409,
+      "step": 10460
+    },
+    {
+      "epoch": 1.2016541267015106,
+      "grad_norm": 0.55586177110672,
+      "learning_rate": 0.0001,
+      "loss": 1.4184,
+      "step": 10461
+    },
+    {
+      "epoch": 1.2017689966113376,
+      "grad_norm": 0.6155359745025635,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 10462
+    },
+    {
+      "epoch": 1.2018838665211649,
+      "grad_norm": 0.6004906296730042,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 10463
+    },
+    {
+      "epoch": 1.2019987364309919,
+      "grad_norm": 0.5957759022712708,
+      "learning_rate": 0.0001,
+      "loss": 1.6768,
+      "step": 10464
+    },
+    {
+      "epoch": 1.202113606340819,
+      "grad_norm": 0.5964832901954651,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 10465
+    },
+    {
+      "epoch": 1.202228476250646,
+      "grad_norm": 0.6004605889320374,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 10466
+    },
+    {
+      "epoch": 1.2023433461604733,
+      "grad_norm": 0.5793516635894775,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 10467
+    },
+    {
+      "epoch": 1.2024582160703003,
+      "grad_norm": 0.565511167049408,
+      "learning_rate": 0.0001,
+      "loss": 1.3453,
+      "step": 10468
+    },
+    {
+      "epoch": 1.2025730859801276,
+      "grad_norm": 0.6401596069335938,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 10469
+    },
+    {
+      "epoch": 1.2026879558899546,
+      "grad_norm": 0.6024700999259949,
+      "learning_rate": 0.0001,
+      "loss": 1.6374,
+      "step": 10470
+    },
+    {
+      "epoch": 1.2028028257997818,
+      "grad_norm": 0.5722721219062805,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 10471
+    },
+    {
+      "epoch": 1.2029176957096088,
+      "grad_norm": 0.5830640196800232,
+      "learning_rate": 0.0001,
+      "loss": 1.4463,
+      "step": 10472
+    },
+    {
+      "epoch": 1.203032565619436,
+      "grad_norm": 0.5298201441764832,
+      "learning_rate": 0.0001,
+      "loss": 1.4246,
+      "step": 10473
+    },
+    {
+      "epoch": 1.203147435529263,
+      "grad_norm": 0.5793371796607971,
+      "learning_rate": 0.0001,
+      "loss": 1.4691,
+      "step": 10474
+    },
+    {
+      "epoch": 1.2032623054390903,
+      "grad_norm": 0.5723811984062195,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 10475
+    },
+    {
+      "epoch": 1.2033771753489173,
+      "grad_norm": 0.5795249342918396,
+      "learning_rate": 0.0001,
+      "loss": 1.398,
+      "step": 10476
+    },
+    {
+      "epoch": 1.2034920452587445,
+      "grad_norm": 0.5610243678092957,
+      "learning_rate": 0.0001,
+      "loss": 1.6885,
+      "step": 10477
+    },
+    {
+      "epoch": 1.2036069151685715,
+      "grad_norm": 0.6054256558418274,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 10478
+    },
+    {
+      "epoch": 1.2037217850783988,
+      "grad_norm": 0.5674859881401062,
+      "learning_rate": 0.0001,
+      "loss": 1.4247,
+      "step": 10479
+    },
+    {
+      "epoch": 1.2038366549882258,
+      "grad_norm": 0.5590078830718994,
+      "learning_rate": 0.0001,
+      "loss": 1.4417,
+      "step": 10480
+    },
+    {
+      "epoch": 1.203951524898053,
+      "grad_norm": 0.5892341732978821,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 10481
+    },
+    {
+      "epoch": 1.20406639480788,
+      "grad_norm": 0.5569952130317688,
+      "learning_rate": 0.0001,
+      "loss": 1.324,
+      "step": 10482
+    },
+    {
+      "epoch": 1.2041812647177073,
+      "grad_norm": 0.5327668190002441,
+      "learning_rate": 0.0001,
+      "loss": 1.369,
+      "step": 10483
+    },
+    {
+      "epoch": 1.2042961346275343,
+      "grad_norm": 0.5070328116416931,
+      "learning_rate": 0.0001,
+      "loss": 1.2728,
+      "step": 10484
+    },
+    {
+      "epoch": 1.2044110045373615,
+      "grad_norm": 0.5609564781188965,
+      "learning_rate": 0.0001,
+      "loss": 1.4026,
+      "step": 10485
+    },
+    {
+      "epoch": 1.2045258744471885,
+      "grad_norm": 0.5414237380027771,
+      "learning_rate": 0.0001,
+      "loss": 1.3149,
+      "step": 10486
+    },
+    {
+      "epoch": 1.2046407443570157,
+      "grad_norm": 0.5648483633995056,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 10487
+    },
+    {
+      "epoch": 1.2047556142668427,
+      "grad_norm": 0.5687791705131531,
+      "learning_rate": 0.0001,
+      "loss": 1.5176,
+      "step": 10488
+    },
+    {
+      "epoch": 1.20487048417667,
+      "grad_norm": 0.5945368409156799,
+      "learning_rate": 0.0001,
+      "loss": 1.4052,
+      "step": 10489
+    },
+    {
+      "epoch": 1.204985354086497,
+      "grad_norm": 0.5893955826759338,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 10490
+    },
+    {
+      "epoch": 1.2051002239963242,
+      "grad_norm": 0.5878959894180298,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 10491
+    },
+    {
+      "epoch": 1.2052150939061512,
+      "grad_norm": 0.5943409204483032,
+      "learning_rate": 0.0001,
+      "loss": 1.3367,
+      "step": 10492
+    },
+    {
+      "epoch": 1.2053299638159785,
+      "grad_norm": 0.5430295467376709,
+      "learning_rate": 0.0001,
+      "loss": 1.46,
+      "step": 10493
+    },
+    {
+      "epoch": 1.2054448337258055,
+      "grad_norm": 0.5535471439361572,
+      "learning_rate": 0.0001,
+      "loss": 1.3129,
+      "step": 10494
+    },
+    {
+      "epoch": 1.2055597036356327,
+      "grad_norm": 0.540138840675354,
+      "learning_rate": 0.0001,
+      "loss": 1.3614,
+      "step": 10495
+    },
+    {
+      "epoch": 1.2056745735454597,
+      "grad_norm": 0.6055250763893127,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 10496
+    },
+    {
+      "epoch": 1.205789443455287,
+      "grad_norm": 0.6143799424171448,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 10497
+    },
+    {
+      "epoch": 1.205904313365114,
+      "grad_norm": 0.5408201813697815,
+      "learning_rate": 0.0001,
+      "loss": 1.4474,
+      "step": 10498
+    },
+    {
+      "epoch": 1.2060191832749412,
+      "grad_norm": 0.5724786520004272,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 10499
+    },
+    {
+      "epoch": 1.2061340531847682,
+      "grad_norm": 0.523482620716095,
+      "learning_rate": 0.0001,
+      "loss": 1.155,
+      "step": 10500
+    },
+    {
+      "epoch": 1.2062489230945954,
+      "grad_norm": 0.5779107809066772,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 10501
+    },
+    {
+      "epoch": 1.2063637930044224,
+      "grad_norm": 0.5509933233261108,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 10502
+    },
+    {
+      "epoch": 1.2064786629142497,
+      "grad_norm": 0.55925053358078,
+      "learning_rate": 0.0001,
+      "loss": 1.2454,
+      "step": 10503
+    },
+    {
+      "epoch": 1.2065935328240767,
+      "grad_norm": 0.5909457802772522,
+      "learning_rate": 0.0001,
+      "loss": 1.4099,
+      "step": 10504
+    },
+    {
+      "epoch": 1.206708402733904,
+      "grad_norm": 0.6092385649681091,
+      "learning_rate": 0.0001,
+      "loss": 1.5369,
+      "step": 10505
+    },
+    {
+      "epoch": 1.206823272643731,
+      "grad_norm": 0.5715480446815491,
+      "learning_rate": 0.0001,
+      "loss": 1.4217,
+      "step": 10506
+    },
+    {
+      "epoch": 1.2069381425535581,
+      "grad_norm": 0.5696176290512085,
+      "learning_rate": 0.0001,
+      "loss": 1.4611,
+      "step": 10507
+    },
+    {
+      "epoch": 1.2070530124633851,
+      "grad_norm": 0.5600457787513733,
+      "learning_rate": 0.0001,
+      "loss": 1.3629,
+      "step": 10508
+    },
+    {
+      "epoch": 1.2071678823732124,
+      "grad_norm": 0.5562861561775208,
+      "learning_rate": 0.0001,
+      "loss": 1.6141,
+      "step": 10509
+    },
+    {
+      "epoch": 1.2072827522830394,
+      "grad_norm": 0.5927658081054688,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 10510
+    },
+    {
+      "epoch": 1.2073976221928666,
+      "grad_norm": 0.5737072825431824,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 10511
+    },
+    {
+      "epoch": 1.2075124921026936,
+      "grad_norm": 0.5820686221122742,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 10512
+    },
+    {
+      "epoch": 1.2076273620125209,
+      "grad_norm": 0.5760904550552368,
+      "learning_rate": 0.0001,
+      "loss": 1.3553,
+      "step": 10513
+    },
+    {
+      "epoch": 1.2077422319223479,
+      "grad_norm": 0.5971365571022034,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 10514
+    },
+    {
+      "epoch": 1.207857101832175,
+      "grad_norm": 0.5970271825790405,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 10515
+    },
+    {
+      "epoch": 1.207971971742002,
+      "grad_norm": 0.5633013248443604,
+      "learning_rate": 0.0001,
+      "loss": 1.3834,
+      "step": 10516
+    },
+    {
+      "epoch": 1.2080868416518293,
+      "grad_norm": 0.5619766116142273,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 10517
+    },
+    {
+      "epoch": 1.2082017115616563,
+      "grad_norm": 0.5610624551773071,
+      "learning_rate": 0.0001,
+      "loss": 1.456,
+      "step": 10518
+    },
+    {
+      "epoch": 1.2083165814714836,
+      "grad_norm": 0.5785056948661804,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 10519
+    },
+    {
+      "epoch": 1.2084314513813106,
+      "grad_norm": 0.6023985147476196,
+      "learning_rate": 0.0001,
+      "loss": 1.4268,
+      "step": 10520
+    },
+    {
+      "epoch": 1.2085463212911378,
+      "grad_norm": 0.6198663115501404,
+      "learning_rate": 0.0001,
+      "loss": 1.4412,
+      "step": 10521
+    },
+    {
+      "epoch": 1.2086611912009648,
+      "grad_norm": 0.5379393100738525,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 10522
+    },
+    {
+      "epoch": 1.208776061110792,
+      "grad_norm": 0.5652551651000977,
+      "learning_rate": 0.0001,
+      "loss": 1.3385,
+      "step": 10523
+    },
+    {
+      "epoch": 1.208890931020619,
+      "grad_norm": 0.5869433879852295,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 10524
+    },
+    {
+      "epoch": 1.2090058009304463,
+      "grad_norm": 0.5975571870803833,
+      "learning_rate": 0.0001,
+      "loss": 1.4304,
+      "step": 10525
+    },
+    {
+      "epoch": 1.2091206708402733,
+      "grad_norm": 0.5863459706306458,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 10526
+    },
+    {
+      "epoch": 1.2092355407501005,
+      "grad_norm": 0.5686241984367371,
+      "learning_rate": 0.0001,
+      "loss": 1.4045,
+      "step": 10527
+    },
+    {
+      "epoch": 1.2093504106599275,
+      "grad_norm": 0.6071587204933167,
+      "learning_rate": 0.0001,
+      "loss": 1.438,
+      "step": 10528
+    },
+    {
+      "epoch": 1.2094652805697548,
+      "grad_norm": 0.5743688344955444,
+      "learning_rate": 0.0001,
+      "loss": 1.388,
+      "step": 10529
+    },
+    {
+      "epoch": 1.2095801504795818,
+      "grad_norm": 0.5639120936393738,
+      "learning_rate": 0.0001,
+      "loss": 1.538,
+      "step": 10530
+    },
+    {
+      "epoch": 1.209695020389409,
+      "grad_norm": 0.5790785551071167,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 10531
+    },
+    {
+      "epoch": 1.209809890299236,
+      "grad_norm": 0.5463698506355286,
+      "learning_rate": 0.0001,
+      "loss": 1.3599,
+      "step": 10532
+    },
+    {
+      "epoch": 1.2099247602090633,
+      "grad_norm": 0.5384880304336548,
+      "learning_rate": 0.0001,
+      "loss": 1.5287,
+      "step": 10533
+    },
+    {
+      "epoch": 1.2100396301188903,
+      "grad_norm": 0.5078497529029846,
+      "learning_rate": 0.0001,
+      "loss": 1.4054,
+      "step": 10534
+    },
+    {
+      "epoch": 1.2101545000287175,
+      "grad_norm": 0.5871212482452393,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 10535
+    },
+    {
+      "epoch": 1.2102693699385445,
+      "grad_norm": 0.5731205344200134,
+      "learning_rate": 0.0001,
+      "loss": 1.6316,
+      "step": 10536
+    },
+    {
+      "epoch": 1.2103842398483717,
+      "grad_norm": 0.5511032938957214,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 10537
+    },
+    {
+      "epoch": 1.2104991097581987,
+      "grad_norm": 0.6102611422538757,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 10538
+    },
+    {
+      "epoch": 1.210613979668026,
+      "grad_norm": 0.5358269214630127,
+      "learning_rate": 0.0001,
+      "loss": 1.2765,
+      "step": 10539
+    },
+    {
+      "epoch": 1.210728849577853,
+      "grad_norm": 0.5827724933624268,
+      "learning_rate": 0.0001,
+      "loss": 1.4465,
+      "step": 10540
+    },
+    {
+      "epoch": 1.2108437194876802,
+      "grad_norm": 0.6176273822784424,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 10541
+    },
+    {
+      "epoch": 1.2109585893975074,
+      "grad_norm": 0.6200592517852783,
+      "learning_rate": 0.0001,
+      "loss": 1.7385,
+      "step": 10542
+    },
+    {
+      "epoch": 1.2110734593073345,
+      "grad_norm": 0.5881513357162476,
+      "learning_rate": 0.0001,
+      "loss": 1.4508,
+      "step": 10543
+    },
+    {
+      "epoch": 1.2111883292171615,
+      "grad_norm": 0.5368597507476807,
+      "learning_rate": 0.0001,
+      "loss": 1.2807,
+      "step": 10544
+    },
+    {
+      "epoch": 1.2113031991269887,
+      "grad_norm": 0.5737367868423462,
+      "learning_rate": 0.0001,
+      "loss": 1.3851,
+      "step": 10545
+    },
+    {
+      "epoch": 1.211418069036816,
+      "grad_norm": 0.5376740097999573,
+      "learning_rate": 0.0001,
+      "loss": 1.5095,
+      "step": 10546
+    },
+    {
+      "epoch": 1.211532938946643,
+      "grad_norm": 0.5448538064956665,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 10547
+    },
+    {
+      "epoch": 1.21164780885647,
+      "grad_norm": 0.5198777914047241,
+      "learning_rate": 0.0001,
+      "loss": 1.4856,
+      "step": 10548
+    },
+    {
+      "epoch": 1.2117626787662972,
+      "grad_norm": 0.5620352625846863,
+      "learning_rate": 0.0001,
+      "loss": 1.4172,
+      "step": 10549
+    },
+    {
+      "epoch": 1.2118775486761244,
+      "grad_norm": 0.6086106896400452,
+      "learning_rate": 0.0001,
+      "loss": 1.3698,
+      "step": 10550
+    },
+    {
+      "epoch": 1.2119924185859514,
+      "grad_norm": 0.5318285822868347,
+      "learning_rate": 0.0001,
+      "loss": 1.1826,
+      "step": 10551
+    },
+    {
+      "epoch": 1.2121072884957784,
+      "grad_norm": 0.6011495590209961,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 10552
+    },
+    {
+      "epoch": 1.2122221584056057,
+      "grad_norm": 0.536434531211853,
+      "learning_rate": 0.0001,
+      "loss": 1.3322,
+      "step": 10553
+    },
+    {
+      "epoch": 1.2123370283154329,
+      "grad_norm": 0.57552170753479,
+      "learning_rate": 0.0001,
+      "loss": 1.7569,
+      "step": 10554
+    },
+    {
+      "epoch": 1.21245189822526,
+      "grad_norm": 0.5894425511360168,
+      "learning_rate": 0.0001,
+      "loss": 1.4857,
+      "step": 10555
+    },
+    {
+      "epoch": 1.212566768135087,
+      "grad_norm": 0.6130958795547485,
+      "learning_rate": 0.0001,
+      "loss": 1.7492,
+      "step": 10556
+    },
+    {
+      "epoch": 1.2126816380449141,
+      "grad_norm": 0.6135485172271729,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 10557
+    },
+    {
+      "epoch": 1.2127965079547414,
+      "grad_norm": 0.5616319179534912,
+      "learning_rate": 0.0001,
+      "loss": 1.4481,
+      "step": 10558
+    },
+    {
+      "epoch": 1.2129113778645684,
+      "grad_norm": 0.567959725856781,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 10559
+    },
+    {
+      "epoch": 1.2130262477743954,
+      "grad_norm": 0.5984885692596436,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 10560
+    },
+    {
+      "epoch": 1.2131411176842226,
+      "grad_norm": 0.5240800380706787,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 10561
+    },
+    {
+      "epoch": 1.2132559875940498,
+      "grad_norm": 0.5749354362487793,
+      "learning_rate": 0.0001,
+      "loss": 1.4552,
+      "step": 10562
+    },
+    {
+      "epoch": 1.2133708575038769,
+      "grad_norm": 0.5432642698287964,
+      "learning_rate": 0.0001,
+      "loss": 1.5985,
+      "step": 10563
+    },
+    {
+      "epoch": 1.2134857274137039,
+      "grad_norm": 0.5396465063095093,
+      "learning_rate": 0.0001,
+      "loss": 1.4983,
+      "step": 10564
+    },
+    {
+      "epoch": 1.213600597323531,
+      "grad_norm": 0.5398960113525391,
+      "learning_rate": 0.0001,
+      "loss": 1.3812,
+      "step": 10565
+    },
+    {
+      "epoch": 1.2137154672333583,
+      "grad_norm": 0.5659125447273254,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 10566
+    },
+    {
+      "epoch": 1.2138303371431853,
+      "grad_norm": 0.5628279447555542,
+      "learning_rate": 0.0001,
+      "loss": 1.2602,
+      "step": 10567
+    },
+    {
+      "epoch": 1.2139452070530123,
+      "grad_norm": 0.5381669998168945,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 10568
+    },
+    {
+      "epoch": 1.2140600769628396,
+      "grad_norm": 0.6219281554222107,
+      "learning_rate": 0.0001,
+      "loss": 1.5175,
+      "step": 10569
+    },
+    {
+      "epoch": 1.2141749468726668,
+      "grad_norm": 0.5203022956848145,
+      "learning_rate": 0.0001,
+      "loss": 1.3109,
+      "step": 10570
+    },
+    {
+      "epoch": 1.2142898167824938,
+      "grad_norm": 0.6202723383903503,
+      "learning_rate": 0.0001,
+      "loss": 1.5858,
+      "step": 10571
+    },
+    {
+      "epoch": 1.2144046866923208,
+      "grad_norm": 0.6326547861099243,
+      "learning_rate": 0.0001,
+      "loss": 1.4975,
+      "step": 10572
+    },
+    {
+      "epoch": 1.214519556602148,
+      "grad_norm": 0.5709686875343323,
+      "learning_rate": 0.0001,
+      "loss": 1.3928,
+      "step": 10573
+    },
+    {
+      "epoch": 1.2146344265119753,
+      "grad_norm": 0.5582396984100342,
+      "learning_rate": 0.0001,
+      "loss": 1.4887,
+      "step": 10574
+    },
+    {
+      "epoch": 1.2147492964218023,
+      "grad_norm": 0.678507387638092,
+      "learning_rate": 0.0001,
+      "loss": 1.7535,
+      "step": 10575
+    },
+    {
+      "epoch": 1.2148641663316295,
+      "grad_norm": 0.5705639123916626,
+      "learning_rate": 0.0001,
+      "loss": 1.6121,
+      "step": 10576
+    },
+    {
+      "epoch": 1.2149790362414565,
+      "grad_norm": 0.5790746212005615,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 10577
+    },
+    {
+      "epoch": 1.2150939061512838,
+      "grad_norm": 0.5600179433822632,
+      "learning_rate": 0.0001,
+      "loss": 1.4223,
+      "step": 10578
+    },
+    {
+      "epoch": 1.2152087760611108,
+      "grad_norm": 0.5656793713569641,
+      "learning_rate": 0.0001,
+      "loss": 1.4859,
+      "step": 10579
+    },
+    {
+      "epoch": 1.215323645970938,
+      "grad_norm": 0.537078857421875,
+      "learning_rate": 0.0001,
+      "loss": 1.4168,
+      "step": 10580
+    },
+    {
+      "epoch": 1.215438515880765,
+      "grad_norm": 0.5688410401344299,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 10581
+    },
+    {
+      "epoch": 1.2155533857905922,
+      "grad_norm": 0.6211932897567749,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 10582
+    },
+    {
+      "epoch": 1.2156682557004193,
+      "grad_norm": 0.5459011793136597,
+      "learning_rate": 0.0001,
+      "loss": 1.3783,
+      "step": 10583
+    },
+    {
+      "epoch": 1.2157831256102465,
+      "grad_norm": 0.540130615234375,
+      "learning_rate": 0.0001,
+      "loss": 1.35,
+      "step": 10584
+    },
+    {
+      "epoch": 1.2158979955200735,
+      "grad_norm": 0.5064961910247803,
+      "learning_rate": 0.0001,
+      "loss": 1.3146,
+      "step": 10585
+    },
+    {
+      "epoch": 1.2160128654299007,
+      "grad_norm": 0.5856577157974243,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 10586
+    },
+    {
+      "epoch": 1.2161277353397277,
+      "grad_norm": 0.5436236262321472,
+      "learning_rate": 0.0001,
+      "loss": 1.2324,
+      "step": 10587
+    },
+    {
+      "epoch": 1.216242605249555,
+      "grad_norm": 0.5454403162002563,
+      "learning_rate": 0.0001,
+      "loss": 1.5701,
+      "step": 10588
+    },
+    {
+      "epoch": 1.216357475159382,
+      "grad_norm": 0.5347267985343933,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 10589
+    },
+    {
+      "epoch": 1.2164723450692092,
+      "grad_norm": 0.5841572284698486,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 10590
+    },
+    {
+      "epoch": 1.2165872149790362,
+      "grad_norm": 0.5742143988609314,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 10591
+    },
+    {
+      "epoch": 1.2167020848888634,
+      "grad_norm": 0.5593157410621643,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 10592
+    },
+    {
+      "epoch": 1.2168169547986905,
+      "grad_norm": 0.58852219581604,
+      "learning_rate": 0.0001,
+      "loss": 1.4253,
+      "step": 10593
+    },
+    {
+      "epoch": 1.2169318247085177,
+      "grad_norm": 0.5854587554931641,
+      "learning_rate": 0.0001,
+      "loss": 1.4844,
+      "step": 10594
+    },
+    {
+      "epoch": 1.2170466946183447,
+      "grad_norm": 0.6186012625694275,
+      "learning_rate": 0.0001,
+      "loss": 1.5365,
+      "step": 10595
+    },
+    {
+      "epoch": 1.217161564528172,
+      "grad_norm": 0.5782433748245239,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 10596
+    },
+    {
+      "epoch": 1.217276434437999,
+      "grad_norm": 0.5456607341766357,
+      "learning_rate": 0.0001,
+      "loss": 1.383,
+      "step": 10597
+    },
+    {
+      "epoch": 1.2173913043478262,
+      "grad_norm": 0.5986420512199402,
+      "learning_rate": 0.0001,
+      "loss": 1.746,
+      "step": 10598
+    },
+    {
+      "epoch": 1.2175061742576532,
+      "grad_norm": 0.5691726207733154,
+      "learning_rate": 0.0001,
+      "loss": 1.4222,
+      "step": 10599
+    },
+    {
+      "epoch": 1.2176210441674804,
+      "grad_norm": 0.5510737299919128,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 10600
+    },
+    {
+      "epoch": 1.2177359140773074,
+      "grad_norm": 0.528681755065918,
+      "learning_rate": 0.0001,
+      "loss": 1.2332,
+      "step": 10601
+    },
+    {
+      "epoch": 1.2178507839871346,
+      "grad_norm": 0.5856773853302002,
+      "learning_rate": 0.0001,
+      "loss": 1.3116,
+      "step": 10602
+    },
+    {
+      "epoch": 1.2179656538969617,
+      "grad_norm": 0.6299785375595093,
+      "learning_rate": 0.0001,
+      "loss": 1.3567,
+      "step": 10603
+    },
+    {
+      "epoch": 1.2180805238067889,
+      "grad_norm": 0.5841706395149231,
+      "learning_rate": 0.0001,
+      "loss": 1.2655,
+      "step": 10604
+    },
+    {
+      "epoch": 1.218195393716616,
+      "grad_norm": 0.5885977745056152,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 10605
+    },
+    {
+      "epoch": 1.2183102636264431,
+      "grad_norm": 0.5559094548225403,
+      "learning_rate": 0.0001,
+      "loss": 1.2172,
+      "step": 10606
+    },
+    {
+      "epoch": 1.2184251335362701,
+      "grad_norm": 0.5585634112358093,
+      "learning_rate": 0.0001,
+      "loss": 1.504,
+      "step": 10607
+    },
+    {
+      "epoch": 1.2185400034460974,
+      "grad_norm": 0.5708191394805908,
+      "learning_rate": 0.0001,
+      "loss": 1.549,
+      "step": 10608
+    },
+    {
+      "epoch": 1.2186548733559244,
+      "grad_norm": 0.5848360657691956,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 10609
+    },
+    {
+      "epoch": 1.2187697432657516,
+      "grad_norm": 0.702562153339386,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 10610
+    },
+    {
+      "epoch": 1.2188846131755786,
+      "grad_norm": 0.5700719952583313,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 10611
+    },
+    {
+      "epoch": 1.2189994830854058,
+      "grad_norm": 0.5429587960243225,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 10612
+    },
+    {
+      "epoch": 1.2191143529952329,
+      "grad_norm": 0.5469475388526917,
+      "learning_rate": 0.0001,
+      "loss": 1.5313,
+      "step": 10613
+    },
+    {
+      "epoch": 1.21922922290506,
+      "grad_norm": 0.5817015171051025,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 10614
+    },
+    {
+      "epoch": 1.219344092814887,
+      "grad_norm": 0.6177971363067627,
+      "learning_rate": 0.0001,
+      "loss": 1.3462,
+      "step": 10615
+    },
+    {
+      "epoch": 1.2194589627247143,
+      "grad_norm": 0.5310166478157043,
+      "learning_rate": 0.0001,
+      "loss": 1.3836,
+      "step": 10616
+    },
+    {
+      "epoch": 1.2195738326345413,
+      "grad_norm": 0.5617627501487732,
+      "learning_rate": 0.0001,
+      "loss": 1.4244,
+      "step": 10617
+    },
+    {
+      "epoch": 1.2196887025443686,
+      "grad_norm": 0.6105504035949707,
+      "learning_rate": 0.0001,
+      "loss": 1.3383,
+      "step": 10618
+    },
+    {
+      "epoch": 1.2198035724541956,
+      "grad_norm": 0.5250139832496643,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 10619
+    },
+    {
+      "epoch": 1.2199184423640228,
+      "grad_norm": 0.5927942991256714,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 10620
+    },
+    {
+      "epoch": 1.2200333122738498,
+      "grad_norm": 0.6836938858032227,
+      "learning_rate": 0.0001,
+      "loss": 1.6504,
+      "step": 10621
+    },
+    {
+      "epoch": 1.220148182183677,
+      "grad_norm": 0.6172232627868652,
+      "learning_rate": 0.0001,
+      "loss": 1.5157,
+      "step": 10622
+    },
+    {
+      "epoch": 1.220263052093504,
+      "grad_norm": 0.5897935032844543,
+      "learning_rate": 0.0001,
+      "loss": 1.3833,
+      "step": 10623
+    },
+    {
+      "epoch": 1.2203779220033313,
+      "grad_norm": 0.5973081588745117,
+      "learning_rate": 0.0001,
+      "loss": 1.3658,
+      "step": 10624
+    },
+    {
+      "epoch": 1.2204927919131583,
+      "grad_norm": 0.6020708680152893,
+      "learning_rate": 0.0001,
+      "loss": 1.4735,
+      "step": 10625
+    },
+    {
+      "epoch": 1.2206076618229855,
+      "grad_norm": 0.5734338164329529,
+      "learning_rate": 0.0001,
+      "loss": 1.5679,
+      "step": 10626
+    },
+    {
+      "epoch": 1.2207225317328125,
+      "grad_norm": 0.6005342602729797,
+      "learning_rate": 0.0001,
+      "loss": 1.6655,
+      "step": 10627
+    },
+    {
+      "epoch": 1.2208374016426398,
+      "grad_norm": 0.6020346283912659,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 10628
+    },
+    {
+      "epoch": 1.2209522715524668,
+      "grad_norm": 0.5923011302947998,
+      "learning_rate": 0.0001,
+      "loss": 1.471,
+      "step": 10629
+    },
+    {
+      "epoch": 1.221067141462294,
+      "grad_norm": 0.5160320997238159,
+      "learning_rate": 0.0001,
+      "loss": 1.4516,
+      "step": 10630
+    },
+    {
+      "epoch": 1.221182011372121,
+      "grad_norm": 0.6067989468574524,
+      "learning_rate": 0.0001,
+      "loss": 1.3357,
+      "step": 10631
+    },
+    {
+      "epoch": 1.2212968812819482,
+      "grad_norm": 0.5493841767311096,
+      "learning_rate": 0.0001,
+      "loss": 1.0806,
+      "step": 10632
+    },
+    {
+      "epoch": 1.2214117511917753,
+      "grad_norm": 0.6001129150390625,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 10633
+    },
+    {
+      "epoch": 1.2215266211016025,
+      "grad_norm": 0.5696105360984802,
+      "learning_rate": 0.0001,
+      "loss": 1.4366,
+      "step": 10634
+    },
+    {
+      "epoch": 1.2216414910114295,
+      "grad_norm": 0.5546556711196899,
+      "learning_rate": 0.0001,
+      "loss": 1.224,
+      "step": 10635
+    },
+    {
+      "epoch": 1.2217563609212567,
+      "grad_norm": 0.5779848098754883,
+      "learning_rate": 0.0001,
+      "loss": 1.4388,
+      "step": 10636
+    },
+    {
+      "epoch": 1.2218712308310837,
+      "grad_norm": 0.6062044501304626,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 10637
+    },
+    {
+      "epoch": 1.221986100740911,
+      "grad_norm": 0.530096173286438,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 10638
+    },
+    {
+      "epoch": 1.222100970650738,
+      "grad_norm": 0.5306719541549683,
+      "learning_rate": 0.0001,
+      "loss": 1.4092,
+      "step": 10639
+    },
+    {
+      "epoch": 1.2222158405605652,
+      "grad_norm": 0.5447390675544739,
+      "learning_rate": 0.0001,
+      "loss": 1.3903,
+      "step": 10640
+    },
+    {
+      "epoch": 1.2223307104703922,
+      "grad_norm": 0.5845821499824524,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 10641
+    },
+    {
+      "epoch": 1.2224455803802194,
+      "grad_norm": 0.5478098392486572,
+      "learning_rate": 0.0001,
+      "loss": 1.3178,
+      "step": 10642
+    },
+    {
+      "epoch": 1.2225604502900465,
+      "grad_norm": 0.5850222706794739,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 10643
+    },
+    {
+      "epoch": 1.2226753201998737,
+      "grad_norm": 0.533097505569458,
+      "learning_rate": 0.0001,
+      "loss": 1.3787,
+      "step": 10644
+    },
+    {
+      "epoch": 1.2227901901097007,
+      "grad_norm": 0.6224300861358643,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 10645
+    },
+    {
+      "epoch": 1.222905060019528,
+      "grad_norm": 0.5365055203437805,
+      "learning_rate": 0.0001,
+      "loss": 1.3455,
+      "step": 10646
+    },
+    {
+      "epoch": 1.223019929929355,
+      "grad_norm": 0.5840891599655151,
+      "learning_rate": 0.0001,
+      "loss": 1.5427,
+      "step": 10647
+    },
+    {
+      "epoch": 1.2231347998391822,
+      "grad_norm": 0.5483365654945374,
+      "learning_rate": 0.0001,
+      "loss": 1.4997,
+      "step": 10648
+    },
+    {
+      "epoch": 1.2232496697490092,
+      "grad_norm": 0.5992576479911804,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 10649
+    },
+    {
+      "epoch": 1.2233645396588364,
+      "grad_norm": 0.5553631782531738,
+      "learning_rate": 0.0001,
+      "loss": 1.4211,
+      "step": 10650
+    },
+    {
+      "epoch": 1.2234794095686634,
+      "grad_norm": 0.5532668232917786,
+      "learning_rate": 0.0001,
+      "loss": 1.526,
+      "step": 10651
+    },
+    {
+      "epoch": 1.2235942794784906,
+      "grad_norm": 0.6031765341758728,
+      "learning_rate": 0.0001,
+      "loss": 1.4622,
+      "step": 10652
+    },
+    {
+      "epoch": 1.2237091493883177,
+      "grad_norm": 0.5665226578712463,
+      "learning_rate": 0.0001,
+      "loss": 1.4103,
+      "step": 10653
+    },
+    {
+      "epoch": 1.2238240192981449,
+      "grad_norm": 0.5733132362365723,
+      "learning_rate": 0.0001,
+      "loss": 1.451,
+      "step": 10654
+    },
+    {
+      "epoch": 1.223938889207972,
+      "grad_norm": 0.5637727379798889,
+      "learning_rate": 0.0001,
+      "loss": 1.4781,
+      "step": 10655
+    },
+    {
+      "epoch": 1.2240537591177991,
+      "grad_norm": 0.5689738392829895,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 10656
+    },
+    {
+      "epoch": 1.2241686290276261,
+      "grad_norm": 0.5613055229187012,
+      "learning_rate": 0.0001,
+      "loss": 1.5456,
+      "step": 10657
+    },
+    {
+      "epoch": 1.2242834989374534,
+      "grad_norm": 0.6242927312850952,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 10658
+    },
+    {
+      "epoch": 1.2243983688472804,
+      "grad_norm": 0.558158278465271,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 10659
+    },
+    {
+      "epoch": 1.2245132387571076,
+      "grad_norm": 0.5290167331695557,
+      "learning_rate": 0.0001,
+      "loss": 1.3642,
+      "step": 10660
+    },
+    {
+      "epoch": 1.2246281086669346,
+      "grad_norm": 0.530127763748169,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 10661
+    },
+    {
+      "epoch": 1.2247429785767618,
+      "grad_norm": 0.6060307621955872,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 10662
+    },
+    {
+      "epoch": 1.2248578484865889,
+      "grad_norm": 0.5839899182319641,
+      "learning_rate": 0.0001,
+      "loss": 1.5879,
+      "step": 10663
+    },
+    {
+      "epoch": 1.224972718396416,
+      "grad_norm": 0.5494903326034546,
+      "learning_rate": 0.0001,
+      "loss": 1.5572,
+      "step": 10664
+    },
+    {
+      "epoch": 1.225087588306243,
+      "grad_norm": 0.5795137882232666,
+      "learning_rate": 0.0001,
+      "loss": 1.4283,
+      "step": 10665
+    },
+    {
+      "epoch": 1.2252024582160703,
+      "grad_norm": 0.5595057606697083,
+      "learning_rate": 0.0001,
+      "loss": 1.3211,
+      "step": 10666
+    },
+    {
+      "epoch": 1.2253173281258973,
+      "grad_norm": 0.535707414150238,
+      "learning_rate": 0.0001,
+      "loss": 1.3843,
+      "step": 10667
+    },
+    {
+      "epoch": 1.2254321980357246,
+      "grad_norm": 0.555937647819519,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 10668
+    },
+    {
+      "epoch": 1.2255470679455516,
+      "grad_norm": 0.568341076374054,
+      "learning_rate": 0.0001,
+      "loss": 1.5307,
+      "step": 10669
+    },
+    {
+      "epoch": 1.2256619378553788,
+      "grad_norm": 0.5604590177536011,
+      "learning_rate": 0.0001,
+      "loss": 1.3813,
+      "step": 10670
+    },
+    {
+      "epoch": 1.2257768077652058,
+      "grad_norm": 0.5631393790245056,
+      "learning_rate": 0.0001,
+      "loss": 1.3685,
+      "step": 10671
+    },
+    {
+      "epoch": 1.225891677675033,
+      "grad_norm": 0.579447329044342,
+      "learning_rate": 0.0001,
+      "loss": 1.4172,
+      "step": 10672
+    },
+    {
+      "epoch": 1.22600654758486,
+      "grad_norm": 0.635651171207428,
+      "learning_rate": 0.0001,
+      "loss": 1.4747,
+      "step": 10673
+    },
+    {
+      "epoch": 1.2261214174946873,
+      "grad_norm": 0.5449775457382202,
+      "learning_rate": 0.0001,
+      "loss": 1.3641,
+      "step": 10674
+    },
+    {
+      "epoch": 1.2262362874045143,
+      "grad_norm": 0.5698413252830505,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 10675
+    },
+    {
+      "epoch": 1.2263511573143415,
+      "grad_norm": 0.5639238953590393,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 10676
+    },
+    {
+      "epoch": 1.2264660272241685,
+      "grad_norm": 0.6942810416221619,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 10677
+    },
+    {
+      "epoch": 1.2265808971339958,
+      "grad_norm": 0.5638768076896667,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 10678
+    },
+    {
+      "epoch": 1.226695767043823,
+      "grad_norm": 0.5303918123245239,
+      "learning_rate": 0.0001,
+      "loss": 1.1953,
+      "step": 10679
+    },
+    {
+      "epoch": 1.22681063695365,
+      "grad_norm": 0.5451003313064575,
+      "learning_rate": 0.0001,
+      "loss": 1.3523,
+      "step": 10680
+    },
+    {
+      "epoch": 1.226925506863477,
+      "grad_norm": 0.5491355657577515,
+      "learning_rate": 0.0001,
+      "loss": 1.4684,
+      "step": 10681
+    },
+    {
+      "epoch": 1.2270403767733042,
+      "grad_norm": 0.5453058481216431,
+      "learning_rate": 0.0001,
+      "loss": 1.4057,
+      "step": 10682
+    },
+    {
+      "epoch": 1.2271552466831315,
+      "grad_norm": 0.5452755093574524,
+      "learning_rate": 0.0001,
+      "loss": 1.2485,
+      "step": 10683
+    },
+    {
+      "epoch": 1.2272701165929585,
+      "grad_norm": 0.5503637194633484,
+      "learning_rate": 0.0001,
+      "loss": 1.4669,
+      "step": 10684
+    },
+    {
+      "epoch": 1.2273849865027855,
+      "grad_norm": 0.5912861227989197,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 10685
+    },
+    {
+      "epoch": 1.2274998564126127,
+      "grad_norm": 0.5763379335403442,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 10686
+    },
+    {
+      "epoch": 1.22761472632244,
+      "grad_norm": 0.590811550617218,
+      "learning_rate": 0.0001,
+      "loss": 1.4122,
+      "step": 10687
+    },
+    {
+      "epoch": 1.227729596232267,
+      "grad_norm": 0.6356752514839172,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 10688
+    },
+    {
+      "epoch": 1.227844466142094,
+      "grad_norm": 0.5507737398147583,
+      "learning_rate": 0.0001,
+      "loss": 1.4867,
+      "step": 10689
+    },
+    {
+      "epoch": 1.2279593360519212,
+      "grad_norm": 0.5941469073295593,
+      "learning_rate": 0.0001,
+      "loss": 1.5367,
+      "step": 10690
+    },
+    {
+      "epoch": 1.2280742059617484,
+      "grad_norm": 0.5983887314796448,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 10691
+    },
+    {
+      "epoch": 1.2281890758715754,
+      "grad_norm": 0.5874168276786804,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 10692
+    },
+    {
+      "epoch": 1.2283039457814025,
+      "grad_norm": 0.6247429847717285,
+      "learning_rate": 0.0001,
+      "loss": 1.712,
+      "step": 10693
+    },
+    {
+      "epoch": 1.2284188156912297,
+      "grad_norm": 0.5645884275436401,
+      "learning_rate": 0.0001,
+      "loss": 1.5444,
+      "step": 10694
+    },
+    {
+      "epoch": 1.228533685601057,
+      "grad_norm": 0.6020731925964355,
+      "learning_rate": 0.0001,
+      "loss": 1.4052,
+      "step": 10695
+    },
+    {
+      "epoch": 1.228648555510884,
+      "grad_norm": 0.6159355044364929,
+      "learning_rate": 0.0001,
+      "loss": 1.5297,
+      "step": 10696
+    },
+    {
+      "epoch": 1.228763425420711,
+      "grad_norm": 0.5802619457244873,
+      "learning_rate": 0.0001,
+      "loss": 1.4455,
+      "step": 10697
+    },
+    {
+      "epoch": 1.2288782953305382,
+      "grad_norm": 0.5384595990180969,
+      "learning_rate": 0.0001,
+      "loss": 1.3547,
+      "step": 10698
+    },
+    {
+      "epoch": 1.2289931652403654,
+      "grad_norm": 0.5957247018814087,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 10699
+    },
+    {
+      "epoch": 1.2291080351501924,
+      "grad_norm": 0.5814695954322815,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 10700
+    },
+    {
+      "epoch": 1.2292229050600194,
+      "grad_norm": 0.5483696460723877,
+      "learning_rate": 0.0001,
+      "loss": 1.3991,
+      "step": 10701
+    },
+    {
+      "epoch": 1.2293377749698466,
+      "grad_norm": 0.5747151374816895,
+      "learning_rate": 0.0001,
+      "loss": 1.4609,
+      "step": 10702
+    },
+    {
+      "epoch": 1.2294526448796739,
+      "grad_norm": 0.5760491490364075,
+      "learning_rate": 0.0001,
+      "loss": 1.5331,
+      "step": 10703
+    },
+    {
+      "epoch": 1.2295675147895009,
+      "grad_norm": 0.5944879055023193,
+      "learning_rate": 0.0001,
+      "loss": 1.2962,
+      "step": 10704
+    },
+    {
+      "epoch": 1.229682384699328,
+      "grad_norm": 0.5719342231750488,
+      "learning_rate": 0.0001,
+      "loss": 1.3065,
+      "step": 10705
+    },
+    {
+      "epoch": 1.2297972546091551,
+      "grad_norm": 0.6149895787239075,
+      "learning_rate": 0.0001,
+      "loss": 1.4154,
+      "step": 10706
+    },
+    {
+      "epoch": 1.2299121245189824,
+      "grad_norm": 0.6030192971229553,
+      "learning_rate": 0.0001,
+      "loss": 1.5497,
+      "step": 10707
+    },
+    {
+      "epoch": 1.2300269944288094,
+      "grad_norm": 0.6043549180030823,
+      "learning_rate": 0.0001,
+      "loss": 1.5306,
+      "step": 10708
+    },
+    {
+      "epoch": 1.2301418643386364,
+      "grad_norm": 0.5450412034988403,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 10709
+    },
+    {
+      "epoch": 1.2302567342484636,
+      "grad_norm": 0.5756334066390991,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 10710
+    },
+    {
+      "epoch": 1.2303716041582908,
+      "grad_norm": 0.5773504376411438,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 10711
+    },
+    {
+      "epoch": 1.2304864740681178,
+      "grad_norm": 0.5983827710151672,
+      "learning_rate": 0.0001,
+      "loss": 1.4885,
+      "step": 10712
+    },
+    {
+      "epoch": 1.230601343977945,
+      "grad_norm": 0.5947871804237366,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 10713
+    },
+    {
+      "epoch": 1.230716213887772,
+      "grad_norm": 0.5693453550338745,
+      "learning_rate": 0.0001,
+      "loss": 1.4284,
+      "step": 10714
+    },
+    {
+      "epoch": 1.2308310837975993,
+      "grad_norm": 0.616321325302124,
+      "learning_rate": 0.0001,
+      "loss": 1.59,
+      "step": 10715
+    },
+    {
+      "epoch": 1.2309459537074263,
+      "grad_norm": 0.581452488899231,
+      "learning_rate": 0.0001,
+      "loss": 1.4594,
+      "step": 10716
+    },
+    {
+      "epoch": 1.2310608236172536,
+      "grad_norm": 0.5225065350532532,
+      "learning_rate": 0.0001,
+      "loss": 1.4939,
+      "step": 10717
+    },
+    {
+      "epoch": 1.2311756935270806,
+      "grad_norm": 0.5418626666069031,
+      "learning_rate": 0.0001,
+      "loss": 1.3131,
+      "step": 10718
+    },
+    {
+      "epoch": 1.2312905634369078,
+      "grad_norm": 0.5273642539978027,
+      "learning_rate": 0.0001,
+      "loss": 1.4067,
+      "step": 10719
+    },
+    {
+      "epoch": 1.2314054333467348,
+      "grad_norm": 0.5336480140686035,
+      "learning_rate": 0.0001,
+      "loss": 1.3815,
+      "step": 10720
+    },
+    {
+      "epoch": 1.231520303256562,
+      "grad_norm": 0.6629604697227478,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 10721
+    },
+    {
+      "epoch": 1.231635173166389,
+      "grad_norm": 0.5382062792778015,
+      "learning_rate": 0.0001,
+      "loss": 1.1965,
+      "step": 10722
+    },
+    {
+      "epoch": 1.2317500430762163,
+      "grad_norm": 0.5966381430625916,
+      "learning_rate": 0.0001,
+      "loss": 1.3991,
+      "step": 10723
+    },
+    {
+      "epoch": 1.2318649129860433,
+      "grad_norm": 0.5476511716842651,
+      "learning_rate": 0.0001,
+      "loss": 1.4263,
+      "step": 10724
+    },
+    {
+      "epoch": 1.2319797828958705,
+      "grad_norm": 0.5930783152580261,
+      "learning_rate": 0.0001,
+      "loss": 1.3903,
+      "step": 10725
+    },
+    {
+      "epoch": 1.2320946528056975,
+      "grad_norm": 0.5383990406990051,
+      "learning_rate": 0.0001,
+      "loss": 1.3619,
+      "step": 10726
+    },
+    {
+      "epoch": 1.2322095227155248,
+      "grad_norm": 0.5706599950790405,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 10727
+    },
+    {
+      "epoch": 1.2323243926253518,
+      "grad_norm": 0.5487735867500305,
+      "learning_rate": 0.0001,
+      "loss": 1.4059,
+      "step": 10728
+    },
+    {
+      "epoch": 1.232439262535179,
+      "grad_norm": 0.5330200791358948,
+      "learning_rate": 0.0001,
+      "loss": 1.4389,
+      "step": 10729
+    },
+    {
+      "epoch": 1.232554132445006,
+      "grad_norm": 0.5702281594276428,
+      "learning_rate": 0.0001,
+      "loss": 1.632,
+      "step": 10730
+    },
+    {
+      "epoch": 1.2326690023548332,
+      "grad_norm": 0.5802901983261108,
+      "learning_rate": 0.0001,
+      "loss": 1.4692,
+      "step": 10731
+    },
+    {
+      "epoch": 1.2327838722646602,
+      "grad_norm": 0.6011072993278503,
+      "learning_rate": 0.0001,
+      "loss": 1.3248,
+      "step": 10732
+    },
+    {
+      "epoch": 1.2328987421744875,
+      "grad_norm": 0.6271477937698364,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 10733
+    },
+    {
+      "epoch": 1.2330136120843145,
+      "grad_norm": 0.5693656802177429,
+      "learning_rate": 0.0001,
+      "loss": 1.4817,
+      "step": 10734
+    },
+    {
+      "epoch": 1.2331284819941417,
+      "grad_norm": 0.625735342502594,
+      "learning_rate": 0.0001,
+      "loss": 1.5831,
+      "step": 10735
+    },
+    {
+      "epoch": 1.2332433519039687,
+      "grad_norm": 0.5748286247253418,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 10736
+    },
+    {
+      "epoch": 1.233358221813796,
+      "grad_norm": 0.636698842048645,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 10737
+    },
+    {
+      "epoch": 1.233473091723623,
+      "grad_norm": 0.5615163445472717,
+      "learning_rate": 0.0001,
+      "loss": 1.2444,
+      "step": 10738
+    },
+    {
+      "epoch": 1.2335879616334502,
+      "grad_norm": 0.572391927242279,
+      "learning_rate": 0.0001,
+      "loss": 1.3733,
+      "step": 10739
+    },
+    {
+      "epoch": 1.2337028315432772,
+      "grad_norm": 0.5233102440834045,
+      "learning_rate": 0.0001,
+      "loss": 1.2587,
+      "step": 10740
+    },
+    {
+      "epoch": 1.2338177014531044,
+      "grad_norm": 0.5385754108428955,
+      "learning_rate": 0.0001,
+      "loss": 1.3457,
+      "step": 10741
+    },
+    {
+      "epoch": 1.2339325713629314,
+      "grad_norm": 0.5545738339424133,
+      "learning_rate": 0.0001,
+      "loss": 1.4354,
+      "step": 10742
+    },
+    {
+      "epoch": 1.2340474412727587,
+      "grad_norm": 0.5675808787345886,
+      "learning_rate": 0.0001,
+      "loss": 1.4972,
+      "step": 10743
+    },
+    {
+      "epoch": 1.2341623111825857,
+      "grad_norm": 0.5583345293998718,
+      "learning_rate": 0.0001,
+      "loss": 1.4697,
+      "step": 10744
+    },
+    {
+      "epoch": 1.234277181092413,
+      "grad_norm": 0.5754501819610596,
+      "learning_rate": 0.0001,
+      "loss": 1.4148,
+      "step": 10745
+    },
+    {
+      "epoch": 1.23439205100224,
+      "grad_norm": 0.5526403784751892,
+      "learning_rate": 0.0001,
+      "loss": 1.4903,
+      "step": 10746
+    },
+    {
+      "epoch": 1.2345069209120672,
+      "grad_norm": 0.6143956184387207,
+      "learning_rate": 0.0001,
+      "loss": 1.4575,
+      "step": 10747
+    },
+    {
+      "epoch": 1.2346217908218942,
+      "grad_norm": 0.556839644908905,
+      "learning_rate": 0.0001,
+      "loss": 1.4682,
+      "step": 10748
+    },
+    {
+      "epoch": 1.2347366607317214,
+      "grad_norm": 0.5772037506103516,
+      "learning_rate": 0.0001,
+      "loss": 1.2345,
+      "step": 10749
+    },
+    {
+      "epoch": 1.2348515306415484,
+      "grad_norm": 0.5617235898971558,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 10750
+    },
+    {
+      "epoch": 1.2349664005513756,
+      "grad_norm": 0.5470720529556274,
+      "learning_rate": 0.0001,
+      "loss": 1.3394,
+      "step": 10751
+    },
+    {
+      "epoch": 1.2350812704612026,
+      "grad_norm": 0.6235065460205078,
+      "learning_rate": 0.0001,
+      "loss": 1.3382,
+      "step": 10752
+    },
+    {
+      "epoch": 1.2351961403710299,
+      "grad_norm": 0.5779804587364197,
+      "learning_rate": 0.0001,
+      "loss": 1.1888,
+      "step": 10753
+    },
+    {
+      "epoch": 1.2353110102808569,
+      "grad_norm": 0.6196669340133667,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 10754
+    },
+    {
+      "epoch": 1.2354258801906841,
+      "grad_norm": 0.6144828200340271,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 10755
+    },
+    {
+      "epoch": 1.2355407501005111,
+      "grad_norm": 0.5465558171272278,
+      "learning_rate": 0.0001,
+      "loss": 1.3457,
+      "step": 10756
+    },
+    {
+      "epoch": 1.2356556200103384,
+      "grad_norm": 0.6325592398643494,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 10757
+    },
+    {
+      "epoch": 1.2357704899201654,
+      "grad_norm": 0.5570918917655945,
+      "learning_rate": 0.0001,
+      "loss": 1.453,
+      "step": 10758
+    },
+    {
+      "epoch": 1.2358853598299926,
+      "grad_norm": 0.5771710872650146,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 10759
+    },
+    {
+      "epoch": 1.2360002297398196,
+      "grad_norm": 0.5577232837677002,
+      "learning_rate": 0.0001,
+      "loss": 1.4324,
+      "step": 10760
+    },
+    {
+      "epoch": 1.2361150996496468,
+      "grad_norm": 0.5303571224212646,
+      "learning_rate": 0.0001,
+      "loss": 1.2595,
+      "step": 10761
+    },
+    {
+      "epoch": 1.2362299695594738,
+      "grad_norm": 0.6245179176330566,
+      "learning_rate": 0.0001,
+      "loss": 1.4688,
+      "step": 10762
+    },
+    {
+      "epoch": 1.236344839469301,
+      "grad_norm": 0.5816663503646851,
+      "learning_rate": 0.0001,
+      "loss": 1.5545,
+      "step": 10763
+    },
+    {
+      "epoch": 1.236459709379128,
+      "grad_norm": 0.5651383996009827,
+      "learning_rate": 0.0001,
+      "loss": 1.293,
+      "step": 10764
+    },
+    {
+      "epoch": 1.2365745792889553,
+      "grad_norm": 0.563153862953186,
+      "learning_rate": 0.0001,
+      "loss": 1.3759,
+      "step": 10765
+    },
+    {
+      "epoch": 1.2366894491987823,
+      "grad_norm": 0.5659855604171753,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 10766
+    },
+    {
+      "epoch": 1.2368043191086096,
+      "grad_norm": 0.564038872718811,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 10767
+    },
+    {
+      "epoch": 1.2369191890184366,
+      "grad_norm": 0.6177818775177002,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 10768
+    },
+    {
+      "epoch": 1.2370340589282638,
+      "grad_norm": 0.5920203328132629,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 10769
+    },
+    {
+      "epoch": 1.2371489288380908,
+      "grad_norm": 0.6030647158622742,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 10770
+    },
+    {
+      "epoch": 1.237263798747918,
+      "grad_norm": 0.5532615780830383,
+      "learning_rate": 0.0001,
+      "loss": 1.3566,
+      "step": 10771
+    },
+    {
+      "epoch": 1.237378668657745,
+      "grad_norm": 0.6269826292991638,
+      "learning_rate": 0.0001,
+      "loss": 1.3191,
+      "step": 10772
+    },
+    {
+      "epoch": 1.2374935385675723,
+      "grad_norm": 0.6108141541481018,
+      "learning_rate": 0.0001,
+      "loss": 1.3749,
+      "step": 10773
+    },
+    {
+      "epoch": 1.2376084084773993,
+      "grad_norm": 0.557407796382904,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 10774
+    },
+    {
+      "epoch": 1.2377232783872265,
+      "grad_norm": 0.6010729074478149,
+      "learning_rate": 0.0001,
+      "loss": 1.4623,
+      "step": 10775
+    },
+    {
+      "epoch": 1.2378381482970535,
+      "grad_norm": 0.6081424355506897,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 10776
+    },
+    {
+      "epoch": 1.2379530182068808,
+      "grad_norm": 0.6339059472084045,
+      "learning_rate": 0.0001,
+      "loss": 1.5383,
+      "step": 10777
+    },
+    {
+      "epoch": 1.2380678881167078,
+      "grad_norm": 0.5645229816436768,
+      "learning_rate": 0.0001,
+      "loss": 1.479,
+      "step": 10778
+    },
+    {
+      "epoch": 1.238182758026535,
+      "grad_norm": 0.5893407464027405,
+      "learning_rate": 0.0001,
+      "loss": 1.5504,
+      "step": 10779
+    },
+    {
+      "epoch": 1.238297627936362,
+      "grad_norm": 0.6675935387611389,
+      "learning_rate": 0.0001,
+      "loss": 1.7331,
+      "step": 10780
+    },
+    {
+      "epoch": 1.2384124978461892,
+      "grad_norm": 0.5694357752799988,
+      "learning_rate": 0.0001,
+      "loss": 1.3615,
+      "step": 10781
+    },
+    {
+      "epoch": 1.2385273677560162,
+      "grad_norm": 0.5906005501747131,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 10782
+    },
+    {
+      "epoch": 1.2386422376658435,
+      "grad_norm": 0.6008540391921997,
+      "learning_rate": 0.0001,
+      "loss": 1.6614,
+      "step": 10783
+    },
+    {
+      "epoch": 1.2387571075756705,
+      "grad_norm": 0.5501822233200073,
+      "learning_rate": 0.0001,
+      "loss": 1.5858,
+      "step": 10784
+    },
+    {
+      "epoch": 1.2388719774854977,
+      "grad_norm": 0.5909137725830078,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 10785
+    },
+    {
+      "epoch": 1.2389868473953247,
+      "grad_norm": 0.5546433329582214,
+      "learning_rate": 0.0001,
+      "loss": 1.4521,
+      "step": 10786
+    },
+    {
+      "epoch": 1.239101717305152,
+      "grad_norm": 0.5489922165870667,
+      "learning_rate": 0.0001,
+      "loss": 1.2556,
+      "step": 10787
+    },
+    {
+      "epoch": 1.239216587214979,
+      "grad_norm": 0.5586884617805481,
+      "learning_rate": 0.0001,
+      "loss": 1.3457,
+      "step": 10788
+    },
+    {
+      "epoch": 1.2393314571248062,
+      "grad_norm": 0.5664530992507935,
+      "learning_rate": 0.0001,
+      "loss": 1.4248,
+      "step": 10789
+    },
+    {
+      "epoch": 1.2394463270346332,
+      "grad_norm": 0.5709450244903564,
+      "learning_rate": 0.0001,
+      "loss": 1.4569,
+      "step": 10790
+    },
+    {
+      "epoch": 1.2395611969444604,
+      "grad_norm": 0.514665424823761,
+      "learning_rate": 0.0001,
+      "loss": 1.394,
+      "step": 10791
+    },
+    {
+      "epoch": 1.2396760668542874,
+      "grad_norm": 0.5852550268173218,
+      "learning_rate": 0.0001,
+      "loss": 1.6148,
+      "step": 10792
+    },
+    {
+      "epoch": 1.2397909367641147,
+      "grad_norm": 0.5568869113922119,
+      "learning_rate": 0.0001,
+      "loss": 1.3949,
+      "step": 10793
+    },
+    {
+      "epoch": 1.2399058066739417,
+      "grad_norm": 0.5907896161079407,
+      "learning_rate": 0.0001,
+      "loss": 1.383,
+      "step": 10794
+    },
+    {
+      "epoch": 1.240020676583769,
+      "grad_norm": 0.6120040416717529,
+      "learning_rate": 0.0001,
+      "loss": 1.4843,
+      "step": 10795
+    },
+    {
+      "epoch": 1.240135546493596,
+      "grad_norm": 0.5711389183998108,
+      "learning_rate": 0.0001,
+      "loss": 1.4388,
+      "step": 10796
+    },
+    {
+      "epoch": 1.2402504164034232,
+      "grad_norm": 0.5505643486976624,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 10797
+    },
+    {
+      "epoch": 1.2403652863132502,
+      "grad_norm": 0.5786750316619873,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 10798
+    },
+    {
+      "epoch": 1.2404801562230774,
+      "grad_norm": 0.5783752799034119,
+      "learning_rate": 0.0001,
+      "loss": 1.399,
+      "step": 10799
+    },
+    {
+      "epoch": 1.2405950261329044,
+      "grad_norm": 0.6035143733024597,
+      "learning_rate": 0.0001,
+      "loss": 1.4854,
+      "step": 10800
+    },
+    {
+      "epoch": 1.2407098960427316,
+      "grad_norm": 0.5874049663543701,
+      "learning_rate": 0.0001,
+      "loss": 1.5683,
+      "step": 10801
+    },
+    {
+      "epoch": 1.2408247659525586,
+      "grad_norm": 0.5582104921340942,
+      "learning_rate": 0.0001,
+      "loss": 1.3253,
+      "step": 10802
+    },
+    {
+      "epoch": 1.2409396358623859,
+      "grad_norm": 0.5821365118026733,
+      "learning_rate": 0.0001,
+      "loss": 1.396,
+      "step": 10803
+    },
+    {
+      "epoch": 1.2410545057722129,
+      "grad_norm": 0.5929169654846191,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 10804
+    },
+    {
+      "epoch": 1.24116937568204,
+      "grad_norm": 0.5773475766181946,
+      "learning_rate": 0.0001,
+      "loss": 1.6268,
+      "step": 10805
+    },
+    {
+      "epoch": 1.2412842455918671,
+      "grad_norm": 0.5636146068572998,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 10806
+    },
+    {
+      "epoch": 1.2413991155016944,
+      "grad_norm": 0.5757436156272888,
+      "learning_rate": 0.0001,
+      "loss": 1.6073,
+      "step": 10807
+    },
+    {
+      "epoch": 1.2415139854115214,
+      "grad_norm": 0.5821229815483093,
+      "learning_rate": 0.0001,
+      "loss": 1.4745,
+      "step": 10808
+    },
+    {
+      "epoch": 1.2416288553213486,
+      "grad_norm": 0.5732831954956055,
+      "learning_rate": 0.0001,
+      "loss": 1.4471,
+      "step": 10809
+    },
+    {
+      "epoch": 1.2417437252311756,
+      "grad_norm": 0.5851945877075195,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 10810
+    },
+    {
+      "epoch": 1.2418585951410028,
+      "grad_norm": 0.5864282846450806,
+      "learning_rate": 0.0001,
+      "loss": 1.6848,
+      "step": 10811
+    },
+    {
+      "epoch": 1.2419734650508298,
+      "grad_norm": 0.5669234395027161,
+      "learning_rate": 0.0001,
+      "loss": 1.2759,
+      "step": 10812
+    },
+    {
+      "epoch": 1.242088334960657,
+      "grad_norm": 0.5712340474128723,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 10813
+    },
+    {
+      "epoch": 1.242203204870484,
+      "grad_norm": 0.5951799750328064,
+      "learning_rate": 0.0001,
+      "loss": 1.4003,
+      "step": 10814
+    },
+    {
+      "epoch": 1.2423180747803113,
+      "grad_norm": 0.5672787427902222,
+      "learning_rate": 0.0001,
+      "loss": 1.5504,
+      "step": 10815
+    },
+    {
+      "epoch": 1.2424329446901385,
+      "grad_norm": 0.6768470406532288,
+      "learning_rate": 0.0001,
+      "loss": 1.7115,
+      "step": 10816
+    },
+    {
+      "epoch": 1.2425478145999655,
+      "grad_norm": 0.5624891519546509,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 10817
+    },
+    {
+      "epoch": 1.2426626845097926,
+      "grad_norm": 0.6003620624542236,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 10818
+    },
+    {
+      "epoch": 1.2427775544196198,
+      "grad_norm": 0.61441570520401,
+      "learning_rate": 0.0001,
+      "loss": 1.668,
+      "step": 10819
+    },
+    {
+      "epoch": 1.242892424329447,
+      "grad_norm": 0.5454705953598022,
+      "learning_rate": 0.0001,
+      "loss": 1.695,
+      "step": 10820
+    },
+    {
+      "epoch": 1.243007294239274,
+      "grad_norm": 0.6024851202964783,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 10821
+    },
+    {
+      "epoch": 1.243122164149101,
+      "grad_norm": 0.560492992401123,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 10822
+    },
+    {
+      "epoch": 1.2432370340589283,
+      "grad_norm": 0.5419345498085022,
+      "learning_rate": 0.0001,
+      "loss": 1.4609,
+      "step": 10823
+    },
+    {
+      "epoch": 1.2433519039687555,
+      "grad_norm": 0.6079069972038269,
+      "learning_rate": 0.0001,
+      "loss": 1.4584,
+      "step": 10824
+    },
+    {
+      "epoch": 1.2434667738785825,
+      "grad_norm": 0.519904613494873,
+      "learning_rate": 0.0001,
+      "loss": 1.4182,
+      "step": 10825
+    },
+    {
+      "epoch": 1.2435816437884095,
+      "grad_norm": 0.5554570555686951,
+      "learning_rate": 0.0001,
+      "loss": 1.5727,
+      "step": 10826
+    },
+    {
+      "epoch": 1.2436965136982367,
+      "grad_norm": 0.5826308131217957,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 10827
+    },
+    {
+      "epoch": 1.243811383608064,
+      "grad_norm": 0.5417545437812805,
+      "learning_rate": 0.0001,
+      "loss": 1.3831,
+      "step": 10828
+    },
+    {
+      "epoch": 1.243926253517891,
+      "grad_norm": 0.5289231538772583,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 10829
+    },
+    {
+      "epoch": 1.244041123427718,
+      "grad_norm": 0.5962126851081848,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 10830
+    },
+    {
+      "epoch": 1.2441559933375452,
+      "grad_norm": 0.5800266861915588,
+      "learning_rate": 0.0001,
+      "loss": 1.5021,
+      "step": 10831
+    },
+    {
+      "epoch": 1.2442708632473725,
+      "grad_norm": 0.567566990852356,
+      "learning_rate": 0.0001,
+      "loss": 1.4349,
+      "step": 10832
+    },
+    {
+      "epoch": 1.2443857331571995,
+      "grad_norm": 0.6255682110786438,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 10833
+    },
+    {
+      "epoch": 1.2445006030670265,
+      "grad_norm": 0.5554635524749756,
+      "learning_rate": 0.0001,
+      "loss": 1.307,
+      "step": 10834
+    },
+    {
+      "epoch": 1.2446154729768537,
+      "grad_norm": 0.5715488195419312,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 10835
+    },
+    {
+      "epoch": 1.244730342886681,
+      "grad_norm": 0.5674731731414795,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 10836
+    },
+    {
+      "epoch": 1.244845212796508,
+      "grad_norm": 0.5632114410400391,
+      "learning_rate": 0.0001,
+      "loss": 1.4986,
+      "step": 10837
+    },
+    {
+      "epoch": 1.244960082706335,
+      "grad_norm": 0.5295623540878296,
+      "learning_rate": 0.0001,
+      "loss": 1.4137,
+      "step": 10838
+    },
+    {
+      "epoch": 1.2450749526161622,
+      "grad_norm": 0.5888751149177551,
+      "learning_rate": 0.0001,
+      "loss": 1.562,
+      "step": 10839
+    },
+    {
+      "epoch": 1.2451898225259894,
+      "grad_norm": 0.6099748015403748,
+      "learning_rate": 0.0001,
+      "loss": 1.4734,
+      "step": 10840
+    },
+    {
+      "epoch": 1.2453046924358164,
+      "grad_norm": 0.5546545386314392,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 10841
+    },
+    {
+      "epoch": 1.2454195623456434,
+      "grad_norm": 0.5606866478919983,
+      "learning_rate": 0.0001,
+      "loss": 1.5588,
+      "step": 10842
+    },
+    {
+      "epoch": 1.2455344322554707,
+      "grad_norm": 0.6139029860496521,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 10843
+    },
+    {
+      "epoch": 1.245649302165298,
+      "grad_norm": 0.6030558347702026,
+      "learning_rate": 0.0001,
+      "loss": 1.362,
+      "step": 10844
+    },
+    {
+      "epoch": 1.245764172075125,
+      "grad_norm": 0.5658564567565918,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 10845
+    },
+    {
+      "epoch": 1.245879041984952,
+      "grad_norm": 0.5419941544532776,
+      "learning_rate": 0.0001,
+      "loss": 1.3872,
+      "step": 10846
+    },
+    {
+      "epoch": 1.2459939118947791,
+      "grad_norm": 0.6134659647941589,
+      "learning_rate": 0.0001,
+      "loss": 1.6605,
+      "step": 10847
+    },
+    {
+      "epoch": 1.2461087818046064,
+      "grad_norm": 0.5531185269355774,
+      "learning_rate": 0.0001,
+      "loss": 1.4083,
+      "step": 10848
+    },
+    {
+      "epoch": 1.2462236517144334,
+      "grad_norm": 0.5249099731445312,
+      "learning_rate": 0.0001,
+      "loss": 1.4228,
+      "step": 10849
+    },
+    {
+      "epoch": 1.2463385216242606,
+      "grad_norm": 0.5550907254219055,
+      "learning_rate": 0.0001,
+      "loss": 1.314,
+      "step": 10850
+    },
+    {
+      "epoch": 1.2464533915340876,
+      "grad_norm": 0.6986422538757324,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 10851
+    },
+    {
+      "epoch": 1.2465682614439149,
+      "grad_norm": 0.5527095794677734,
+      "learning_rate": 0.0001,
+      "loss": 1.4435,
+      "step": 10852
+    },
+    {
+      "epoch": 1.2466831313537419,
+      "grad_norm": 0.6040549278259277,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 10853
+    },
+    {
+      "epoch": 1.246798001263569,
+      "grad_norm": 0.5868404507637024,
+      "learning_rate": 0.0001,
+      "loss": 1.3417,
+      "step": 10854
+    },
+    {
+      "epoch": 1.246912871173396,
+      "grad_norm": 0.5885109305381775,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 10855
+    },
+    {
+      "epoch": 1.2470277410832233,
+      "grad_norm": 0.6273307800292969,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 10856
+    },
+    {
+      "epoch": 1.2471426109930503,
+      "grad_norm": 0.6067867279052734,
+      "learning_rate": 0.0001,
+      "loss": 1.4738,
+      "step": 10857
+    },
+    {
+      "epoch": 1.2472574809028776,
+      "grad_norm": 0.6141306757926941,
+      "learning_rate": 0.0001,
+      "loss": 1.3835,
+      "step": 10858
+    },
+    {
+      "epoch": 1.2473723508127046,
+      "grad_norm": 0.5698774456977844,
+      "learning_rate": 0.0001,
+      "loss": 1.3689,
+      "step": 10859
+    },
+    {
+      "epoch": 1.2474872207225318,
+      "grad_norm": 0.6184090375900269,
+      "learning_rate": 0.0001,
+      "loss": 1.549,
+      "step": 10860
+    },
+    {
+      "epoch": 1.2476020906323588,
+      "grad_norm": 0.5597769021987915,
+      "learning_rate": 0.0001,
+      "loss": 1.3235,
+      "step": 10861
+    },
+    {
+      "epoch": 1.247716960542186,
+      "grad_norm": 0.5379227995872498,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 10862
+    },
+    {
+      "epoch": 1.247831830452013,
+      "grad_norm": 0.5401409864425659,
+      "learning_rate": 0.0001,
+      "loss": 1.3701,
+      "step": 10863
+    },
+    {
+      "epoch": 1.2479467003618403,
+      "grad_norm": 0.5571491718292236,
+      "learning_rate": 0.0001,
+      "loss": 1.4373,
+      "step": 10864
+    },
+    {
+      "epoch": 1.2480615702716673,
+      "grad_norm": 0.550502359867096,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 10865
+    },
+    {
+      "epoch": 1.2481764401814945,
+      "grad_norm": 0.5631131529808044,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 10866
+    },
+    {
+      "epoch": 1.2482913100913215,
+      "grad_norm": 0.5850018858909607,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 10867
+    },
+    {
+      "epoch": 1.2484061800011488,
+      "grad_norm": 0.6137135028839111,
+      "learning_rate": 0.0001,
+      "loss": 1.5496,
+      "step": 10868
+    },
+    {
+      "epoch": 1.2485210499109758,
+      "grad_norm": 0.5973943471908569,
+      "learning_rate": 0.0001,
+      "loss": 1.4159,
+      "step": 10869
+    },
+    {
+      "epoch": 1.248635919820803,
+      "grad_norm": 0.5690673589706421,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 10870
+    },
+    {
+      "epoch": 1.24875078973063,
+      "grad_norm": 0.5665297508239746,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 10871
+    },
+    {
+      "epoch": 1.2488656596404573,
+      "grad_norm": 0.5932878851890564,
+      "learning_rate": 0.0001,
+      "loss": 1.5818,
+      "step": 10872
+    },
+    {
+      "epoch": 1.2489805295502843,
+      "grad_norm": 0.634212076663971,
+      "learning_rate": 0.0001,
+      "loss": 1.4781,
+      "step": 10873
+    },
+    {
+      "epoch": 1.2490953994601115,
+      "grad_norm": 0.6295340061187744,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 10874
+    },
+    {
+      "epoch": 1.2492102693699385,
+      "grad_norm": 0.5564239025115967,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 10875
+    },
+    {
+      "epoch": 1.2493251392797657,
+      "grad_norm": 0.591235339641571,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 10876
+    },
+    {
+      "epoch": 1.2494400091895927,
+      "grad_norm": 0.540838360786438,
+      "learning_rate": 0.0001,
+      "loss": 1.4963,
+      "step": 10877
+    },
+    {
+      "epoch": 1.24955487909942,
+      "grad_norm": 0.5274412631988525,
+      "learning_rate": 0.0001,
+      "loss": 1.3721,
+      "step": 10878
+    },
+    {
+      "epoch": 1.249669749009247,
+      "grad_norm": 0.532894492149353,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 10879
+    },
+    {
+      "epoch": 1.2497846189190742,
+      "grad_norm": 0.5777339339256287,
+      "learning_rate": 0.0001,
+      "loss": 1.4037,
+      "step": 10880
+    },
+    {
+      "epoch": 1.2498994888289012,
+      "grad_norm": 0.6504830718040466,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 10881
+    },
+    {
+      "epoch": 1.2500143587387285,
+      "grad_norm": 0.5571553111076355,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 10882
+    },
+    {
+      "epoch": 1.2501292286485555,
+      "grad_norm": 0.6024576425552368,
+      "learning_rate": 0.0001,
+      "loss": 1.5162,
+      "step": 10883
+    },
+    {
+      "epoch": 1.2502440985583827,
+      "grad_norm": 0.5811855792999268,
+      "learning_rate": 0.0001,
+      "loss": 1.3607,
+      "step": 10884
+    },
+    {
+      "epoch": 1.2503589684682097,
+      "grad_norm": 0.571258544921875,
+      "learning_rate": 0.0001,
+      "loss": 1.6321,
+      "step": 10885
+    },
+    {
+      "epoch": 1.250473838378037,
+      "grad_norm": 0.5692873001098633,
+      "learning_rate": 0.0001,
+      "loss": 1.4705,
+      "step": 10886
+    },
+    {
+      "epoch": 1.250588708287864,
+      "grad_norm": 0.6267750263214111,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 10887
+    },
+    {
+      "epoch": 1.2507035781976912,
+      "grad_norm": 0.6186516284942627,
+      "learning_rate": 0.0001,
+      "loss": 1.4483,
+      "step": 10888
+    },
+    {
+      "epoch": 1.2508184481075182,
+      "grad_norm": 0.5934524536132812,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 10889
+    },
+    {
+      "epoch": 1.2509333180173454,
+      "grad_norm": 0.6174366474151611,
+      "learning_rate": 0.0001,
+      "loss": 1.6715,
+      "step": 10890
+    },
+    {
+      "epoch": 1.2510481879271724,
+      "grad_norm": 0.5978173613548279,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 10891
+    },
+    {
+      "epoch": 1.2511630578369997,
+      "grad_norm": 0.563639760017395,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 10892
+    },
+    {
+      "epoch": 1.2512779277468267,
+      "grad_norm": 0.6402945518493652,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 10893
+    },
+    {
+      "epoch": 1.251392797656654,
+      "grad_norm": 0.589114248752594,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 10894
+    },
+    {
+      "epoch": 1.251507667566481,
+      "grad_norm": 0.5770435929298401,
+      "learning_rate": 0.0001,
+      "loss": 1.4116,
+      "step": 10895
+    },
+    {
+      "epoch": 1.2516225374763081,
+      "grad_norm": 0.6756162643432617,
+      "learning_rate": 0.0001,
+      "loss": 1.7188,
+      "step": 10896
+    },
+    {
+      "epoch": 1.2517374073861351,
+      "grad_norm": 0.5813696384429932,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 10897
+    },
+    {
+      "epoch": 1.2518522772959624,
+      "grad_norm": 0.5571926832199097,
+      "learning_rate": 0.0001,
+      "loss": 1.392,
+      "step": 10898
+    },
+    {
+      "epoch": 1.2519671472057894,
+      "grad_norm": 0.5992504954338074,
+      "learning_rate": 0.0001,
+      "loss": 1.6786,
+      "step": 10899
+    },
+    {
+      "epoch": 1.2520820171156166,
+      "grad_norm": 0.5757802724838257,
+      "learning_rate": 0.0001,
+      "loss": 1.4874,
+      "step": 10900
+    },
+    {
+      "epoch": 1.2521968870254436,
+      "grad_norm": 0.5644787549972534,
+      "learning_rate": 0.0001,
+      "loss": 1.3862,
+      "step": 10901
+    },
+    {
+      "epoch": 1.2523117569352709,
+      "grad_norm": 0.5546348690986633,
+      "learning_rate": 0.0001,
+      "loss": 1.4577,
+      "step": 10902
+    },
+    {
+      "epoch": 1.2524266268450979,
+      "grad_norm": 0.5690358281135559,
+      "learning_rate": 0.0001,
+      "loss": 1.3687,
+      "step": 10903
+    },
+    {
+      "epoch": 1.252541496754925,
+      "grad_norm": 0.5834867358207703,
+      "learning_rate": 0.0001,
+      "loss": 1.4238,
+      "step": 10904
+    },
+    {
+      "epoch": 1.252656366664752,
+      "grad_norm": 0.558188796043396,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 10905
+    },
+    {
+      "epoch": 1.2527712365745793,
+      "grad_norm": 0.5730266571044922,
+      "learning_rate": 0.0001,
+      "loss": 1.1854,
+      "step": 10906
+    },
+    {
+      "epoch": 1.2528861064844063,
+      "grad_norm": 0.6047393083572388,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 10907
+    },
+    {
+      "epoch": 1.2530009763942336,
+      "grad_norm": 0.5652978420257568,
+      "learning_rate": 0.0001,
+      "loss": 1.3756,
+      "step": 10908
+    },
+    {
+      "epoch": 1.2531158463040606,
+      "grad_norm": 0.5942120552062988,
+      "learning_rate": 0.0001,
+      "loss": 1.3795,
+      "step": 10909
+    },
+    {
+      "epoch": 1.2532307162138878,
+      "grad_norm": 0.5671000480651855,
+      "learning_rate": 0.0001,
+      "loss": 1.4261,
+      "step": 10910
+    },
+    {
+      "epoch": 1.2533455861237148,
+      "grad_norm": 0.5476558208465576,
+      "learning_rate": 0.0001,
+      "loss": 1.3019,
+      "step": 10911
+    },
+    {
+      "epoch": 1.253460456033542,
+      "grad_norm": 0.6260891556739807,
+      "learning_rate": 0.0001,
+      "loss": 1.7735,
+      "step": 10912
+    },
+    {
+      "epoch": 1.253575325943369,
+      "grad_norm": 0.6707521677017212,
+      "learning_rate": 0.0001,
+      "loss": 1.3852,
+      "step": 10913
+    },
+    {
+      "epoch": 1.2536901958531963,
+      "grad_norm": 0.5497795939445496,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 10914
+    },
+    {
+      "epoch": 1.2538050657630233,
+      "grad_norm": 0.6137018203735352,
+      "learning_rate": 0.0001,
+      "loss": 1.449,
+      "step": 10915
+    },
+    {
+      "epoch": 1.2539199356728505,
+      "grad_norm": 0.575927734375,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 10916
+    },
+    {
+      "epoch": 1.2540348055826775,
+      "grad_norm": 0.5461201667785645,
+      "learning_rate": 0.0001,
+      "loss": 1.571,
+      "step": 10917
+    },
+    {
+      "epoch": 1.2541496754925048,
+      "grad_norm": 0.5775430202484131,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 10918
+    },
+    {
+      "epoch": 1.2542645454023318,
+      "grad_norm": 0.596335768699646,
+      "learning_rate": 0.0001,
+      "loss": 1.2492,
+      "step": 10919
+    },
+    {
+      "epoch": 1.254379415312159,
+      "grad_norm": 0.552144467830658,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 10920
+    },
+    {
+      "epoch": 1.254494285221986,
+      "grad_norm": 0.6232661008834839,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 10921
+    },
+    {
+      "epoch": 1.2546091551318133,
+      "grad_norm": 0.579709529876709,
+      "learning_rate": 0.0001,
+      "loss": 1.3082,
+      "step": 10922
+    },
+    {
+      "epoch": 1.2547240250416403,
+      "grad_norm": 0.5315948724746704,
+      "learning_rate": 0.0001,
+      "loss": 1.3547,
+      "step": 10923
+    },
+    {
+      "epoch": 1.2548388949514675,
+      "grad_norm": 0.5723145008087158,
+      "learning_rate": 0.0001,
+      "loss": 1.4858,
+      "step": 10924
+    },
+    {
+      "epoch": 1.2549537648612945,
+      "grad_norm": 0.5732467770576477,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 10925
+    },
+    {
+      "epoch": 1.2550686347711217,
+      "grad_norm": 0.5439171195030212,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 10926
+    },
+    {
+      "epoch": 1.2551835046809487,
+      "grad_norm": 0.5616052150726318,
+      "learning_rate": 0.0001,
+      "loss": 1.5706,
+      "step": 10927
+    },
+    {
+      "epoch": 1.255298374590776,
+      "grad_norm": 0.5952640175819397,
+      "learning_rate": 0.0001,
+      "loss": 1.3995,
+      "step": 10928
+    },
+    {
+      "epoch": 1.255413244500603,
+      "grad_norm": 0.5831541419029236,
+      "learning_rate": 0.0001,
+      "loss": 1.6304,
+      "step": 10929
+    },
+    {
+      "epoch": 1.2555281144104302,
+      "grad_norm": 0.5609625577926636,
+      "learning_rate": 0.0001,
+      "loss": 1.4818,
+      "step": 10930
+    },
+    {
+      "epoch": 1.2556429843202572,
+      "grad_norm": 0.5512393116950989,
+      "learning_rate": 0.0001,
+      "loss": 1.2979,
+      "step": 10931
+    },
+    {
+      "epoch": 1.2557578542300845,
+      "grad_norm": 0.5595287084579468,
+      "learning_rate": 0.0001,
+      "loss": 1.4539,
+      "step": 10932
+    },
+    {
+      "epoch": 1.2558727241399115,
+      "grad_norm": 0.5650918483734131,
+      "learning_rate": 0.0001,
+      "loss": 1.4097,
+      "step": 10933
+    },
+    {
+      "epoch": 1.2559875940497387,
+      "grad_norm": 0.5434353351593018,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 10934
+    },
+    {
+      "epoch": 1.2561024639595657,
+      "grad_norm": 0.565133810043335,
+      "learning_rate": 0.0001,
+      "loss": 1.5029,
+      "step": 10935
+    },
+    {
+      "epoch": 1.256217333869393,
+      "grad_norm": 0.6342281699180603,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 10936
+    },
+    {
+      "epoch": 1.2563322037792202,
+      "grad_norm": 0.5778996348381042,
+      "learning_rate": 0.0001,
+      "loss": 1.4856,
+      "step": 10937
+    },
+    {
+      "epoch": 1.2564470736890472,
+      "grad_norm": 0.5874305367469788,
+      "learning_rate": 0.0001,
+      "loss": 1.6482,
+      "step": 10938
+    },
+    {
+      "epoch": 1.2565619435988742,
+      "grad_norm": 0.5764653086662292,
+      "learning_rate": 0.0001,
+      "loss": 1.4978,
+      "step": 10939
+    },
+    {
+      "epoch": 1.2566768135087014,
+      "grad_norm": 0.5709093809127808,
+      "learning_rate": 0.0001,
+      "loss": 1.3895,
+      "step": 10940
+    },
+    {
+      "epoch": 1.2567916834185286,
+      "grad_norm": 0.610541045665741,
+      "learning_rate": 0.0001,
+      "loss": 1.5036,
+      "step": 10941
+    },
+    {
+      "epoch": 1.2569065533283557,
+      "grad_norm": 0.6028667688369751,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 10942
+    },
+    {
+      "epoch": 1.2570214232381827,
+      "grad_norm": 0.5337639451026917,
+      "learning_rate": 0.0001,
+      "loss": 1.4331,
+      "step": 10943
+    },
+    {
+      "epoch": 1.25713629314801,
+      "grad_norm": 0.6048827767372131,
+      "learning_rate": 0.0001,
+      "loss": 1.4235,
+      "step": 10944
+    },
+    {
+      "epoch": 1.2572511630578371,
+      "grad_norm": 0.5719192028045654,
+      "learning_rate": 0.0001,
+      "loss": 1.4289,
+      "step": 10945
+    },
+    {
+      "epoch": 1.2573660329676641,
+      "grad_norm": 0.5911000967025757,
+      "learning_rate": 0.0001,
+      "loss": 1.5606,
+      "step": 10946
+    },
+    {
+      "epoch": 1.2574809028774911,
+      "grad_norm": 0.5696734189987183,
+      "learning_rate": 0.0001,
+      "loss": 1.2368,
+      "step": 10947
+    },
+    {
+      "epoch": 1.2575957727873184,
+      "grad_norm": 0.5841726064682007,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 10948
+    },
+    {
+      "epoch": 1.2577106426971456,
+      "grad_norm": 0.6017117500305176,
+      "learning_rate": 0.0001,
+      "loss": 1.3813,
+      "step": 10949
+    },
+    {
+      "epoch": 1.2578255126069726,
+      "grad_norm": 0.5943053364753723,
+      "learning_rate": 0.0001,
+      "loss": 1.3885,
+      "step": 10950
+    },
+    {
+      "epoch": 1.2579403825167996,
+      "grad_norm": 0.5891920924186707,
+      "learning_rate": 0.0001,
+      "loss": 1.4698,
+      "step": 10951
+    },
+    {
+      "epoch": 1.2580552524266269,
+      "grad_norm": 0.5655304193496704,
+      "learning_rate": 0.0001,
+      "loss": 1.4617,
+      "step": 10952
+    },
+    {
+      "epoch": 1.258170122336454,
+      "grad_norm": 0.6333246231079102,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 10953
+    },
+    {
+      "epoch": 1.258284992246281,
+      "grad_norm": 0.6080989837646484,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 10954
+    },
+    {
+      "epoch": 1.258399862156108,
+      "grad_norm": 0.5369704961776733,
+      "learning_rate": 0.0001,
+      "loss": 1.418,
+      "step": 10955
+    },
+    {
+      "epoch": 1.2585147320659353,
+      "grad_norm": 0.5709648132324219,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 10956
+    },
+    {
+      "epoch": 1.2586296019757626,
+      "grad_norm": 0.5845744609832764,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 10957
+    },
+    {
+      "epoch": 1.2587444718855896,
+      "grad_norm": 0.588251531124115,
+      "learning_rate": 0.0001,
+      "loss": 1.5864,
+      "step": 10958
+    },
+    {
+      "epoch": 1.2588593417954166,
+      "grad_norm": 0.5609616041183472,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 10959
+    },
+    {
+      "epoch": 1.2589742117052438,
+      "grad_norm": 0.5940161347389221,
+      "learning_rate": 0.0001,
+      "loss": 1.631,
+      "step": 10960
+    },
+    {
+      "epoch": 1.259089081615071,
+      "grad_norm": 0.6487417817115784,
+      "learning_rate": 0.0001,
+      "loss": 1.6715,
+      "step": 10961
+    },
+    {
+      "epoch": 1.259203951524898,
+      "grad_norm": 0.5806244015693665,
+      "learning_rate": 0.0001,
+      "loss": 1.3649,
+      "step": 10962
+    },
+    {
+      "epoch": 1.259318821434725,
+      "grad_norm": 0.5590049624443054,
+      "learning_rate": 0.0001,
+      "loss": 1.3849,
+      "step": 10963
+    },
+    {
+      "epoch": 1.2594336913445523,
+      "grad_norm": 0.57107013463974,
+      "learning_rate": 0.0001,
+      "loss": 1.3517,
+      "step": 10964
+    },
+    {
+      "epoch": 1.2595485612543795,
+      "grad_norm": 0.5718011260032654,
+      "learning_rate": 0.0001,
+      "loss": 1.3671,
+      "step": 10965
+    },
+    {
+      "epoch": 1.2596634311642065,
+      "grad_norm": 0.530774712562561,
+      "learning_rate": 0.0001,
+      "loss": 1.3329,
+      "step": 10966
+    },
+    {
+      "epoch": 1.2597783010740335,
+      "grad_norm": 0.5350990891456604,
+      "learning_rate": 0.0001,
+      "loss": 1.3828,
+      "step": 10967
+    },
+    {
+      "epoch": 1.2598931709838608,
+      "grad_norm": 0.5452684760093689,
+      "learning_rate": 0.0001,
+      "loss": 1.46,
+      "step": 10968
+    },
+    {
+      "epoch": 1.260008040893688,
+      "grad_norm": 0.5853266716003418,
+      "learning_rate": 0.0001,
+      "loss": 1.6132,
+      "step": 10969
+    },
+    {
+      "epoch": 1.260122910803515,
+      "grad_norm": 0.6207892298698425,
+      "learning_rate": 0.0001,
+      "loss": 1.5122,
+      "step": 10970
+    },
+    {
+      "epoch": 1.260237780713342,
+      "grad_norm": 0.6362786293029785,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 10971
+    },
+    {
+      "epoch": 1.2603526506231693,
+      "grad_norm": 0.5912286043167114,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 10972
+    },
+    {
+      "epoch": 1.2604675205329965,
+      "grad_norm": 0.5660232305526733,
+      "learning_rate": 0.0001,
+      "loss": 1.4575,
+      "step": 10973
+    },
+    {
+      "epoch": 1.2605823904428235,
+      "grad_norm": 0.5991741418838501,
+      "learning_rate": 0.0001,
+      "loss": 1.4294,
+      "step": 10974
+    },
+    {
+      "epoch": 1.2606972603526505,
+      "grad_norm": 0.6811667680740356,
+      "learning_rate": 0.0001,
+      "loss": 1.2061,
+      "step": 10975
+    },
+    {
+      "epoch": 1.2608121302624777,
+      "grad_norm": 0.5778743028640747,
+      "learning_rate": 0.0001,
+      "loss": 1.4363,
+      "step": 10976
+    },
+    {
+      "epoch": 1.260927000172305,
+      "grad_norm": 0.593360960483551,
+      "learning_rate": 0.0001,
+      "loss": 1.5062,
+      "step": 10977
+    },
+    {
+      "epoch": 1.261041870082132,
+      "grad_norm": 0.5722994804382324,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 10978
+    },
+    {
+      "epoch": 1.261156739991959,
+      "grad_norm": 0.5691931843757629,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 10979
+    },
+    {
+      "epoch": 1.2612716099017862,
+      "grad_norm": 0.5399587750434875,
+      "learning_rate": 0.0001,
+      "loss": 1.3153,
+      "step": 10980
+    },
+    {
+      "epoch": 1.2613864798116134,
+      "grad_norm": 0.5468644499778748,
+      "learning_rate": 0.0001,
+      "loss": 1.4229,
+      "step": 10981
+    },
+    {
+      "epoch": 1.2615013497214405,
+      "grad_norm": 0.5714578032493591,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 10982
+    },
+    {
+      "epoch": 1.2616162196312675,
+      "grad_norm": 0.5381194353103638,
+      "learning_rate": 0.0001,
+      "loss": 1.4761,
+      "step": 10983
+    },
+    {
+      "epoch": 1.2617310895410947,
+      "grad_norm": 0.5348224639892578,
+      "learning_rate": 0.0001,
+      "loss": 1.328,
+      "step": 10984
+    },
+    {
+      "epoch": 1.261845959450922,
+      "grad_norm": 0.570966899394989,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 10985
+    },
+    {
+      "epoch": 1.261960829360749,
+      "grad_norm": 0.5382642149925232,
+      "learning_rate": 0.0001,
+      "loss": 1.4547,
+      "step": 10986
+    },
+    {
+      "epoch": 1.262075699270576,
+      "grad_norm": 0.5181246995925903,
+      "learning_rate": 0.0001,
+      "loss": 1.3591,
+      "step": 10987
+    },
+    {
+      "epoch": 1.2621905691804032,
+      "grad_norm": 0.5006334781646729,
+      "learning_rate": 0.0001,
+      "loss": 1.2591,
+      "step": 10988
+    },
+    {
+      "epoch": 1.2623054390902304,
+      "grad_norm": 0.5715638995170593,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 10989
+    },
+    {
+      "epoch": 1.2624203090000574,
+      "grad_norm": 0.5791033506393433,
+      "learning_rate": 0.0001,
+      "loss": 1.4059,
+      "step": 10990
+    },
+    {
+      "epoch": 1.2625351789098844,
+      "grad_norm": 0.5769622325897217,
+      "learning_rate": 0.0001,
+      "loss": 1.474,
+      "step": 10991
+    },
+    {
+      "epoch": 1.2626500488197117,
+      "grad_norm": 0.5328226685523987,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 10992
+    },
+    {
+      "epoch": 1.2627649187295389,
+      "grad_norm": 0.5586057901382446,
+      "learning_rate": 0.0001,
+      "loss": 1.3952,
+      "step": 10993
+    },
+    {
+      "epoch": 1.262879788639366,
+      "grad_norm": 0.5298300385475159,
+      "learning_rate": 0.0001,
+      "loss": 1.2045,
+      "step": 10994
+    },
+    {
+      "epoch": 1.262994658549193,
+      "grad_norm": 0.5825002193450928,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 10995
+    },
+    {
+      "epoch": 1.2631095284590201,
+      "grad_norm": 0.5873317122459412,
+      "learning_rate": 0.0001,
+      "loss": 1.5954,
+      "step": 10996
+    },
+    {
+      "epoch": 1.2632243983688474,
+      "grad_norm": 0.5621634125709534,
+      "learning_rate": 0.0001,
+      "loss": 1.4467,
+      "step": 10997
+    },
+    {
+      "epoch": 1.2633392682786744,
+      "grad_norm": 0.6157379150390625,
+      "learning_rate": 0.0001,
+      "loss": 1.4748,
+      "step": 10998
+    },
+    {
+      "epoch": 1.2634541381885014,
+      "grad_norm": 0.5493670105934143,
+      "learning_rate": 0.0001,
+      "loss": 1.3428,
+      "step": 10999
+    },
+    {
+      "epoch": 1.2635690080983286,
+      "grad_norm": 0.5555403828620911,
+      "learning_rate": 0.0001,
+      "loss": 1.2846,
+      "step": 11000
+    },
+    {
+      "epoch": 1.2636838780081558,
+      "grad_norm": 0.5692359209060669,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 11001
+    },
+    {
+      "epoch": 1.2637987479179829,
+      "grad_norm": 0.5852733850479126,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 11002
+    },
+    {
+      "epoch": 1.26391361782781,
+      "grad_norm": 0.5792815685272217,
+      "learning_rate": 0.0001,
+      "loss": 1.41,
+      "step": 11003
+    },
+    {
+      "epoch": 1.264028487737637,
+      "grad_norm": 0.6380968689918518,
+      "learning_rate": 0.0001,
+      "loss": 1.3436,
+      "step": 11004
+    },
+    {
+      "epoch": 1.2641433576474643,
+      "grad_norm": 0.6811634302139282,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 11005
+    },
+    {
+      "epoch": 1.2642582275572913,
+      "grad_norm": 0.6941441297531128,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 11006
+    },
+    {
+      "epoch": 1.2643730974671186,
+      "grad_norm": 0.5787838697433472,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 11007
+    },
+    {
+      "epoch": 1.2644879673769456,
+      "grad_norm": 0.5790321230888367,
+      "learning_rate": 0.0001,
+      "loss": 1.477,
+      "step": 11008
+    },
+    {
+      "epoch": 1.2646028372867728,
+      "grad_norm": 0.55388343334198,
+      "learning_rate": 0.0001,
+      "loss": 1.4449,
+      "step": 11009
+    },
+    {
+      "epoch": 1.2647177071965998,
+      "grad_norm": 0.5964010953903198,
+      "learning_rate": 0.0001,
+      "loss": 1.3159,
+      "step": 11010
+    },
+    {
+      "epoch": 1.264832577106427,
+      "grad_norm": 0.5677118897438049,
+      "learning_rate": 0.0001,
+      "loss": 1.3851,
+      "step": 11011
+    },
+    {
+      "epoch": 1.264947447016254,
+      "grad_norm": 0.5453557372093201,
+      "learning_rate": 0.0001,
+      "loss": 1.3942,
+      "step": 11012
+    },
+    {
+      "epoch": 1.2650623169260813,
+      "grad_norm": 0.594939649105072,
+      "learning_rate": 0.0001,
+      "loss": 1.6116,
+      "step": 11013
+    },
+    {
+      "epoch": 1.2651771868359083,
+      "grad_norm": 0.5862032175064087,
+      "learning_rate": 0.0001,
+      "loss": 1.438,
+      "step": 11014
+    },
+    {
+      "epoch": 1.2652920567457355,
+      "grad_norm": 0.6072475910186768,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 11015
+    },
+    {
+      "epoch": 1.2654069266555625,
+      "grad_norm": 0.5791646242141724,
+      "learning_rate": 0.0001,
+      "loss": 1.4486,
+      "step": 11016
+    },
+    {
+      "epoch": 1.2655217965653898,
+      "grad_norm": 0.5972772240638733,
+      "learning_rate": 0.0001,
+      "loss": 1.6154,
+      "step": 11017
+    },
+    {
+      "epoch": 1.2656366664752168,
+      "grad_norm": 0.5665194988250732,
+      "learning_rate": 0.0001,
+      "loss": 1.2982,
+      "step": 11018
+    },
+    {
+      "epoch": 1.265751536385044,
+      "grad_norm": 0.5202224254608154,
+      "learning_rate": 0.0001,
+      "loss": 1.081,
+      "step": 11019
+    },
+    {
+      "epoch": 1.265866406294871,
+      "grad_norm": 0.5774416327476501,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 11020
+    },
+    {
+      "epoch": 1.2659812762046982,
+      "grad_norm": 0.6238007545471191,
+      "learning_rate": 0.0001,
+      "loss": 1.5547,
+      "step": 11021
+    },
+    {
+      "epoch": 1.2660961461145253,
+      "grad_norm": 0.6121378540992737,
+      "learning_rate": 0.0001,
+      "loss": 1.3435,
+      "step": 11022
+    },
+    {
+      "epoch": 1.2662110160243525,
+      "grad_norm": 0.6211857795715332,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 11023
+    },
+    {
+      "epoch": 1.2663258859341795,
+      "grad_norm": 0.6037236452102661,
+      "learning_rate": 0.0001,
+      "loss": 1.5751,
+      "step": 11024
+    },
+    {
+      "epoch": 1.2664407558440067,
+      "grad_norm": 0.5778083801269531,
+      "learning_rate": 0.0001,
+      "loss": 1.4495,
+      "step": 11025
+    },
+    {
+      "epoch": 1.2665556257538337,
+      "grad_norm": 0.5363320112228394,
+      "learning_rate": 0.0001,
+      "loss": 1.3852,
+      "step": 11026
+    },
+    {
+      "epoch": 1.266670495663661,
+      "grad_norm": 0.5823166966438293,
+      "learning_rate": 0.0001,
+      "loss": 1.5257,
+      "step": 11027
+    },
+    {
+      "epoch": 1.266785365573488,
+      "grad_norm": 0.5804757475852966,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 11028
+    },
+    {
+      "epoch": 1.2669002354833152,
+      "grad_norm": 0.5479066967964172,
+      "learning_rate": 0.0001,
+      "loss": 1.3971,
+      "step": 11029
+    },
+    {
+      "epoch": 1.2670151053931422,
+      "grad_norm": 0.5440239310264587,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 11030
+    },
+    {
+      "epoch": 1.2671299753029694,
+      "grad_norm": 0.5420060157775879,
+      "learning_rate": 0.0001,
+      "loss": 1.4851,
+      "step": 11031
+    },
+    {
+      "epoch": 1.2672448452127965,
+      "grad_norm": 0.551177442073822,
+      "learning_rate": 0.0001,
+      "loss": 1.3622,
+      "step": 11032
+    },
+    {
+      "epoch": 1.2673597151226237,
+      "grad_norm": 0.6188027262687683,
+      "learning_rate": 0.0001,
+      "loss": 1.4396,
+      "step": 11033
+    },
+    {
+      "epoch": 1.2674745850324507,
+      "grad_norm": 0.5277479887008667,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 11034
+    },
+    {
+      "epoch": 1.267589454942278,
+      "grad_norm": 0.5837298631668091,
+      "learning_rate": 0.0001,
+      "loss": 1.5163,
+      "step": 11035
+    },
+    {
+      "epoch": 1.267704324852105,
+      "grad_norm": 0.5999343991279602,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 11036
+    },
+    {
+      "epoch": 1.2678191947619322,
+      "grad_norm": 0.5406692624092102,
+      "learning_rate": 0.0001,
+      "loss": 1.3183,
+      "step": 11037
+    },
+    {
+      "epoch": 1.2679340646717592,
+      "grad_norm": 0.5225005745887756,
+      "learning_rate": 0.0001,
+      "loss": 1.3585,
+      "step": 11038
+    },
+    {
+      "epoch": 1.2680489345815864,
+      "grad_norm": 0.5517762899398804,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 11039
+    },
+    {
+      "epoch": 1.2681638044914134,
+      "grad_norm": 0.5250986218452454,
+      "learning_rate": 0.0001,
+      "loss": 1.4083,
+      "step": 11040
+    },
+    {
+      "epoch": 1.2682786744012406,
+      "grad_norm": 0.5599402189254761,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 11041
+    },
+    {
+      "epoch": 1.2683935443110677,
+      "grad_norm": 0.6003100872039795,
+      "learning_rate": 0.0001,
+      "loss": 1.3209,
+      "step": 11042
+    },
+    {
+      "epoch": 1.2685084142208949,
+      "grad_norm": 0.6202194690704346,
+      "learning_rate": 0.0001,
+      "loss": 1.3633,
+      "step": 11043
+    },
+    {
+      "epoch": 1.268623284130722,
+      "grad_norm": 0.5758548974990845,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 11044
+    },
+    {
+      "epoch": 1.2687381540405491,
+      "grad_norm": 0.6529328227043152,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 11045
+    },
+    {
+      "epoch": 1.2688530239503761,
+      "grad_norm": 0.617214560508728,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 11046
+    },
+    {
+      "epoch": 1.2689678938602034,
+      "grad_norm": 0.5643380880355835,
+      "learning_rate": 0.0001,
+      "loss": 1.5033,
+      "step": 11047
+    },
+    {
+      "epoch": 1.2690827637700304,
+      "grad_norm": 0.5501247048377991,
+      "learning_rate": 0.0001,
+      "loss": 1.5196,
+      "step": 11048
+    },
+    {
+      "epoch": 1.2691976336798576,
+      "grad_norm": 0.5517723560333252,
+      "learning_rate": 0.0001,
+      "loss": 1.2386,
+      "step": 11049
+    },
+    {
+      "epoch": 1.2693125035896846,
+      "grad_norm": 0.6187138557434082,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 11050
+    },
+    {
+      "epoch": 1.2694273734995118,
+      "grad_norm": 0.5243611931800842,
+      "learning_rate": 0.0001,
+      "loss": 1.2427,
+      "step": 11051
+    },
+    {
+      "epoch": 1.2695422434093389,
+      "grad_norm": 0.5593788623809814,
+      "learning_rate": 0.0001,
+      "loss": 1.3313,
+      "step": 11052
+    },
+    {
+      "epoch": 1.269657113319166,
+      "grad_norm": 0.6037552356719971,
+      "learning_rate": 0.0001,
+      "loss": 1.2074,
+      "step": 11053
+    },
+    {
+      "epoch": 1.269771983228993,
+      "grad_norm": 0.5670240521430969,
+      "learning_rate": 0.0001,
+      "loss": 1.327,
+      "step": 11054
+    },
+    {
+      "epoch": 1.2698868531388203,
+      "grad_norm": 0.6029371023178101,
+      "learning_rate": 0.0001,
+      "loss": 1.6885,
+      "step": 11055
+    },
+    {
+      "epoch": 1.2700017230486473,
+      "grad_norm": 0.6077609062194824,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 11056
+    },
+    {
+      "epoch": 1.2701165929584746,
+      "grad_norm": 0.5685033202171326,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 11057
+    },
+    {
+      "epoch": 1.2702314628683016,
+      "grad_norm": 0.6237457990646362,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 11058
+    },
+    {
+      "epoch": 1.2703463327781288,
+      "grad_norm": 0.5337412357330322,
+      "learning_rate": 0.0001,
+      "loss": 1.2787,
+      "step": 11059
+    },
+    {
+      "epoch": 1.2704612026879558,
+      "grad_norm": 0.5789129137992859,
+      "learning_rate": 0.0001,
+      "loss": 1.5369,
+      "step": 11060
+    },
+    {
+      "epoch": 1.270576072597783,
+      "grad_norm": 0.6278301477432251,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 11061
+    },
+    {
+      "epoch": 1.27069094250761,
+      "grad_norm": 0.5445199608802795,
+      "learning_rate": 0.0001,
+      "loss": 1.4081,
+      "step": 11062
+    },
+    {
+      "epoch": 1.2708058124174373,
+      "grad_norm": 0.5890607833862305,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 11063
+    },
+    {
+      "epoch": 1.2709206823272643,
+      "grad_norm": 0.5365835428237915,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 11064
+    },
+    {
+      "epoch": 1.2710355522370915,
+      "grad_norm": 0.5734820365905762,
+      "learning_rate": 0.0001,
+      "loss": 1.6119,
+      "step": 11065
+    },
+    {
+      "epoch": 1.2711504221469185,
+      "grad_norm": 0.5551184415817261,
+      "learning_rate": 0.0001,
+      "loss": 1.4606,
+      "step": 11066
+    },
+    {
+      "epoch": 1.2712652920567458,
+      "grad_norm": 0.5175613760948181,
+      "learning_rate": 0.0001,
+      "loss": 1.4638,
+      "step": 11067
+    },
+    {
+      "epoch": 1.2713801619665728,
+      "grad_norm": 0.5665346384048462,
+      "learning_rate": 0.0001,
+      "loss": 1.3524,
+      "step": 11068
+    },
+    {
+      "epoch": 1.2714950318764,
+      "grad_norm": 0.6107744574546814,
+      "learning_rate": 0.0001,
+      "loss": 1.528,
+      "step": 11069
+    },
+    {
+      "epoch": 1.271609901786227,
+      "grad_norm": 0.5816113948822021,
+      "learning_rate": 0.0001,
+      "loss": 1.4268,
+      "step": 11070
+    },
+    {
+      "epoch": 1.2717247716960542,
+      "grad_norm": 0.5861494541168213,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 11071
+    },
+    {
+      "epoch": 1.2718396416058813,
+      "grad_norm": 0.5912983417510986,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 11072
+    },
+    {
+      "epoch": 1.2719545115157085,
+      "grad_norm": 0.5545825362205505,
+      "learning_rate": 0.0001,
+      "loss": 1.2985,
+      "step": 11073
+    },
+    {
+      "epoch": 1.2720693814255357,
+      "grad_norm": 0.5643607974052429,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 11074
+    },
+    {
+      "epoch": 1.2721842513353627,
+      "grad_norm": 0.5997006297111511,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 11075
+    },
+    {
+      "epoch": 1.2722991212451897,
+      "grad_norm": 0.5586094260215759,
+      "learning_rate": 0.0001,
+      "loss": 1.453,
+      "step": 11076
+    },
+    {
+      "epoch": 1.272413991155017,
+      "grad_norm": 0.5602713227272034,
+      "learning_rate": 0.0001,
+      "loss": 1.4295,
+      "step": 11077
+    },
+    {
+      "epoch": 1.2725288610648442,
+      "grad_norm": 0.543787956237793,
+      "learning_rate": 0.0001,
+      "loss": 1.4595,
+      "step": 11078
+    },
+    {
+      "epoch": 1.2726437309746712,
+      "grad_norm": 0.5545603036880493,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 11079
+    },
+    {
+      "epoch": 1.2727586008844982,
+      "grad_norm": 0.6172767281532288,
+      "learning_rate": 0.0001,
+      "loss": 1.6377,
+      "step": 11080
+    },
+    {
+      "epoch": 1.2728734707943254,
+      "grad_norm": 0.5919816493988037,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 11081
+    },
+    {
+      "epoch": 1.2729883407041527,
+      "grad_norm": 0.5594301819801331,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 11082
+    },
+    {
+      "epoch": 1.2731032106139797,
+      "grad_norm": 0.6125577092170715,
+      "learning_rate": 0.0001,
+      "loss": 1.468,
+      "step": 11083
+    },
+    {
+      "epoch": 1.2732180805238067,
+      "grad_norm": 0.5834252834320068,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 11084
+    },
+    {
+      "epoch": 1.273332950433634,
+      "grad_norm": 0.5693926215171814,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 11085
+    },
+    {
+      "epoch": 1.2734478203434612,
+      "grad_norm": 0.5403253436088562,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 11086
+    },
+    {
+      "epoch": 1.2735626902532882,
+      "grad_norm": 0.5751697421073914,
+      "learning_rate": 0.0001,
+      "loss": 1.5373,
+      "step": 11087
+    },
+    {
+      "epoch": 1.2736775601631152,
+      "grad_norm": 0.5498262643814087,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 11088
+    },
+    {
+      "epoch": 1.2737924300729424,
+      "grad_norm": 0.5573551654815674,
+      "learning_rate": 0.0001,
+      "loss": 1.5071,
+      "step": 11089
+    },
+    {
+      "epoch": 1.2739072999827696,
+      "grad_norm": 0.5851514339447021,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 11090
+    },
+    {
+      "epoch": 1.2740221698925966,
+      "grad_norm": 0.5665884017944336,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 11091
+    },
+    {
+      "epoch": 1.2741370398024237,
+      "grad_norm": 0.5661911368370056,
+      "learning_rate": 0.0001,
+      "loss": 1.4927,
+      "step": 11092
+    },
+    {
+      "epoch": 1.2742519097122509,
+      "grad_norm": 0.5676097869873047,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 11093
+    },
+    {
+      "epoch": 1.2743667796220781,
+      "grad_norm": 0.62990403175354,
+      "learning_rate": 0.0001,
+      "loss": 1.6448,
+      "step": 11094
+    },
+    {
+      "epoch": 1.2744816495319051,
+      "grad_norm": 0.5617504715919495,
+      "learning_rate": 0.0001,
+      "loss": 1.3303,
+      "step": 11095
+    },
+    {
+      "epoch": 1.2745965194417321,
+      "grad_norm": 0.603344738483429,
+      "learning_rate": 0.0001,
+      "loss": 1.3515,
+      "step": 11096
+    },
+    {
+      "epoch": 1.2747113893515594,
+      "grad_norm": 0.5757971405982971,
+      "learning_rate": 0.0001,
+      "loss": 1.53,
+      "step": 11097
+    },
+    {
+      "epoch": 1.2748262592613866,
+      "grad_norm": 0.5930181741714478,
+      "learning_rate": 0.0001,
+      "loss": 1.4135,
+      "step": 11098
+    },
+    {
+      "epoch": 1.2749411291712136,
+      "grad_norm": 0.5848099589347839,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 11099
+    },
+    {
+      "epoch": 1.2750559990810406,
+      "grad_norm": 0.585987389087677,
+      "learning_rate": 0.0001,
+      "loss": 1.5055,
+      "step": 11100
+    },
+    {
+      "epoch": 1.2751708689908678,
+      "grad_norm": 0.6561366319656372,
+      "learning_rate": 0.0001,
+      "loss": 1.7621,
+      "step": 11101
+    },
+    {
+      "epoch": 1.275285738900695,
+      "grad_norm": 0.6223732233047485,
+      "learning_rate": 0.0001,
+      "loss": 1.5776,
+      "step": 11102
+    },
+    {
+      "epoch": 1.275400608810522,
+      "grad_norm": 0.5525573492050171,
+      "learning_rate": 0.0001,
+      "loss": 1.4327,
+      "step": 11103
+    },
+    {
+      "epoch": 1.275515478720349,
+      "grad_norm": 0.5294277667999268,
+      "learning_rate": 0.0001,
+      "loss": 1.3882,
+      "step": 11104
+    },
+    {
+      "epoch": 1.2756303486301763,
+      "grad_norm": 0.5459890961647034,
+      "learning_rate": 0.0001,
+      "loss": 1.3725,
+      "step": 11105
+    },
+    {
+      "epoch": 1.2757452185400036,
+      "grad_norm": 0.5299192667007446,
+      "learning_rate": 0.0001,
+      "loss": 1.4483,
+      "step": 11106
+    },
+    {
+      "epoch": 1.2758600884498306,
+      "grad_norm": 0.5696981549263,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 11107
+    },
+    {
+      "epoch": 1.2759749583596576,
+      "grad_norm": 0.5575034022331238,
+      "learning_rate": 0.0001,
+      "loss": 1.2812,
+      "step": 11108
+    },
+    {
+      "epoch": 1.2760898282694848,
+      "grad_norm": 0.570679783821106,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 11109
+    },
+    {
+      "epoch": 1.276204698179312,
+      "grad_norm": 0.5604519248008728,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 11110
+    },
+    {
+      "epoch": 1.276319568089139,
+      "grad_norm": 0.5516342520713806,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 11111
+    },
+    {
+      "epoch": 1.276434437998966,
+      "grad_norm": 0.5541993379592896,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 11112
+    },
+    {
+      "epoch": 1.2765493079087933,
+      "grad_norm": 0.6293795108795166,
+      "learning_rate": 0.0001,
+      "loss": 1.6179,
+      "step": 11113
+    },
+    {
+      "epoch": 1.2766641778186205,
+      "grad_norm": 0.5466127395629883,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 11114
+    },
+    {
+      "epoch": 1.2767790477284475,
+      "grad_norm": 0.5696339011192322,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 11115
+    },
+    {
+      "epoch": 1.2768939176382745,
+      "grad_norm": 0.5600918531417847,
+      "learning_rate": 0.0001,
+      "loss": 1.3857,
+      "step": 11116
+    },
+    {
+      "epoch": 1.2770087875481018,
+      "grad_norm": 0.5669732093811035,
+      "learning_rate": 0.0001,
+      "loss": 1.3401,
+      "step": 11117
+    },
+    {
+      "epoch": 1.277123657457929,
+      "grad_norm": 0.5944907665252686,
+      "learning_rate": 0.0001,
+      "loss": 1.4226,
+      "step": 11118
+    },
+    {
+      "epoch": 1.277238527367756,
+      "grad_norm": 0.5800361633300781,
+      "learning_rate": 0.0001,
+      "loss": 1.4727,
+      "step": 11119
+    },
+    {
+      "epoch": 1.277353397277583,
+      "grad_norm": 0.5681383609771729,
+      "learning_rate": 0.0001,
+      "loss": 1.4153,
+      "step": 11120
+    },
+    {
+      "epoch": 1.2774682671874102,
+      "grad_norm": 0.5693331956863403,
+      "learning_rate": 0.0001,
+      "loss": 1.7115,
+      "step": 11121
+    },
+    {
+      "epoch": 1.2775831370972375,
+      "grad_norm": 0.5463958978652954,
+      "learning_rate": 0.0001,
+      "loss": 1.5031,
+      "step": 11122
+    },
+    {
+      "epoch": 1.2776980070070645,
+      "grad_norm": 0.5928342342376709,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 11123
+    },
+    {
+      "epoch": 1.2778128769168915,
+      "grad_norm": 0.5508242845535278,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 11124
+    },
+    {
+      "epoch": 1.2779277468267187,
+      "grad_norm": 0.5686891078948975,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 11125
+    },
+    {
+      "epoch": 1.278042616736546,
+      "grad_norm": 0.5789403915405273,
+      "learning_rate": 0.0001,
+      "loss": 1.4,
+      "step": 11126
+    },
+    {
+      "epoch": 1.278157486646373,
+      "grad_norm": 0.6012910604476929,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 11127
+    },
+    {
+      "epoch": 1.2782723565562,
+      "grad_norm": 0.5584916472434998,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 11128
+    },
+    {
+      "epoch": 1.2783872264660272,
+      "grad_norm": 0.6591640710830688,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 11129
+    },
+    {
+      "epoch": 1.2785020963758544,
+      "grad_norm": 0.611392080783844,
+      "learning_rate": 0.0001,
+      "loss": 1.3733,
+      "step": 11130
+    },
+    {
+      "epoch": 1.2786169662856814,
+      "grad_norm": 0.5938810706138611,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 11131
+    },
+    {
+      "epoch": 1.2787318361955085,
+      "grad_norm": 0.5487352609634399,
+      "learning_rate": 0.0001,
+      "loss": 1.2881,
+      "step": 11132
+    },
+    {
+      "epoch": 1.2788467061053357,
+      "grad_norm": 0.5437043905258179,
+      "learning_rate": 0.0001,
+      "loss": 1.394,
+      "step": 11133
+    },
+    {
+      "epoch": 1.278961576015163,
+      "grad_norm": 0.6171312928199768,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 11134
+    },
+    {
+      "epoch": 1.27907644592499,
+      "grad_norm": 0.5747122168540955,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 11135
+    },
+    {
+      "epoch": 1.279191315834817,
+      "grad_norm": 0.5911594033241272,
+      "learning_rate": 0.0001,
+      "loss": 1.4412,
+      "step": 11136
+    },
+    {
+      "epoch": 1.2793061857446442,
+      "grad_norm": 0.5997012257575989,
+      "learning_rate": 0.0001,
+      "loss": 1.3998,
+      "step": 11137
+    },
+    {
+      "epoch": 1.2794210556544714,
+      "grad_norm": 0.5709753632545471,
+      "learning_rate": 0.0001,
+      "loss": 1.37,
+      "step": 11138
+    },
+    {
+      "epoch": 1.2795359255642984,
+      "grad_norm": 0.5752970576286316,
+      "learning_rate": 0.0001,
+      "loss": 1.2787,
+      "step": 11139
+    },
+    {
+      "epoch": 1.2796507954741256,
+      "grad_norm": 0.7025997042655945,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 11140
+    },
+    {
+      "epoch": 1.2797656653839526,
+      "grad_norm": 0.5793277025222778,
+      "learning_rate": 0.0001,
+      "loss": 1.4309,
+      "step": 11141
+    },
+    {
+      "epoch": 1.2798805352937799,
+      "grad_norm": 0.550739049911499,
+      "learning_rate": 0.0001,
+      "loss": 1.471,
+      "step": 11142
+    },
+    {
+      "epoch": 1.2799954052036069,
+      "grad_norm": 0.5863038897514343,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 11143
+    },
+    {
+      "epoch": 1.2801102751134341,
+      "grad_norm": 0.5753504633903503,
+      "learning_rate": 0.0001,
+      "loss": 1.3984,
+      "step": 11144
+    },
+    {
+      "epoch": 1.2802251450232611,
+      "grad_norm": 0.566723644733429,
+      "learning_rate": 0.0001,
+      "loss": 1.3528,
+      "step": 11145
+    },
+    {
+      "epoch": 1.2803400149330884,
+      "grad_norm": 0.600090503692627,
+      "learning_rate": 0.0001,
+      "loss": 1.5385,
+      "step": 11146
+    },
+    {
+      "epoch": 1.2804548848429154,
+      "grad_norm": 0.6001926064491272,
+      "learning_rate": 0.0001,
+      "loss": 1.3591,
+      "step": 11147
+    },
+    {
+      "epoch": 1.2805697547527426,
+      "grad_norm": 0.5610960721969604,
+      "learning_rate": 0.0001,
+      "loss": 1.4732,
+      "step": 11148
+    },
+    {
+      "epoch": 1.2806846246625696,
+      "grad_norm": 0.5722346305847168,
+      "learning_rate": 0.0001,
+      "loss": 1.5572,
+      "step": 11149
+    },
+    {
+      "epoch": 1.2807994945723968,
+      "grad_norm": 0.6043455600738525,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 11150
+    },
+    {
+      "epoch": 1.2809143644822238,
+      "grad_norm": 0.6156902313232422,
+      "learning_rate": 0.0001,
+      "loss": 1.6608,
+      "step": 11151
+    },
+    {
+      "epoch": 1.281029234392051,
+      "grad_norm": 0.531518280506134,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 11152
+    },
+    {
+      "epoch": 1.281144104301878,
+      "grad_norm": 0.5361231565475464,
+      "learning_rate": 0.0001,
+      "loss": 1.4514,
+      "step": 11153
+    },
+    {
+      "epoch": 1.2812589742117053,
+      "grad_norm": 0.6165813207626343,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 11154
+    },
+    {
+      "epoch": 1.2813738441215323,
+      "grad_norm": 0.5662755966186523,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 11155
+    },
+    {
+      "epoch": 1.2814887140313596,
+      "grad_norm": 0.5308471322059631,
+      "learning_rate": 0.0001,
+      "loss": 1.3735,
+      "step": 11156
+    },
+    {
+      "epoch": 1.2816035839411866,
+      "grad_norm": 0.7272038459777832,
+      "learning_rate": 0.0001,
+      "loss": 1.5976,
+      "step": 11157
+    },
+    {
+      "epoch": 1.2817184538510138,
+      "grad_norm": 0.606468141078949,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 11158
+    },
+    {
+      "epoch": 1.2818333237608408,
+      "grad_norm": 0.5892067551612854,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 11159
+    },
+    {
+      "epoch": 1.281948193670668,
+      "grad_norm": 0.5960896015167236,
+      "learning_rate": 0.0001,
+      "loss": 1.4519,
+      "step": 11160
+    },
+    {
+      "epoch": 1.282063063580495,
+      "grad_norm": 0.5746586322784424,
+      "learning_rate": 0.0001,
+      "loss": 1.5134,
+      "step": 11161
+    },
+    {
+      "epoch": 1.2821779334903223,
+      "grad_norm": 0.5626453757286072,
+      "learning_rate": 0.0001,
+      "loss": 1.3905,
+      "step": 11162
+    },
+    {
+      "epoch": 1.2822928034001493,
+      "grad_norm": 0.615481972694397,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 11163
+    },
+    {
+      "epoch": 1.2824076733099765,
+      "grad_norm": 0.56350177526474,
+      "learning_rate": 0.0001,
+      "loss": 1.5846,
+      "step": 11164
+    },
+    {
+      "epoch": 1.2825225432198035,
+      "grad_norm": 0.5434756278991699,
+      "learning_rate": 0.0001,
+      "loss": 1.2041,
+      "step": 11165
+    },
+    {
+      "epoch": 1.2826374131296308,
+      "grad_norm": 0.5499529838562012,
+      "learning_rate": 0.0001,
+      "loss": 1.3228,
+      "step": 11166
+    },
+    {
+      "epoch": 1.2827522830394578,
+      "grad_norm": 0.6841591000556946,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 11167
+    },
+    {
+      "epoch": 1.282867152949285,
+      "grad_norm": 0.5592834949493408,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 11168
+    },
+    {
+      "epoch": 1.282982022859112,
+      "grad_norm": 0.5657499432563782,
+      "learning_rate": 0.0001,
+      "loss": 1.3523,
+      "step": 11169
+    },
+    {
+      "epoch": 1.2830968927689392,
+      "grad_norm": 0.7231560945510864,
+      "learning_rate": 0.0001,
+      "loss": 1.4121,
+      "step": 11170
+    },
+    {
+      "epoch": 1.2832117626787662,
+      "grad_norm": 0.6643901467323303,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 11171
+    },
+    {
+      "epoch": 1.2833266325885935,
+      "grad_norm": 0.5481140613555908,
+      "learning_rate": 0.0001,
+      "loss": 1.4084,
+      "step": 11172
+    },
+    {
+      "epoch": 1.2834415024984205,
+      "grad_norm": 0.5619171857833862,
+      "learning_rate": 0.0001,
+      "loss": 1.5306,
+      "step": 11173
+    },
+    {
+      "epoch": 1.2835563724082477,
+      "grad_norm": 0.6013187170028687,
+      "learning_rate": 0.0001,
+      "loss": 1.3116,
+      "step": 11174
+    },
+    {
+      "epoch": 1.2836712423180747,
+      "grad_norm": 0.5309972167015076,
+      "learning_rate": 0.0001,
+      "loss": 1.3814,
+      "step": 11175
+    },
+    {
+      "epoch": 1.283786112227902,
+      "grad_norm": 0.5112704038619995,
+      "learning_rate": 0.0001,
+      "loss": 1.4198,
+      "step": 11176
+    },
+    {
+      "epoch": 1.283900982137729,
+      "grad_norm": 0.5378767251968384,
+      "learning_rate": 0.0001,
+      "loss": 1.4499,
+      "step": 11177
+    },
+    {
+      "epoch": 1.2840158520475562,
+      "grad_norm": 0.5731784105300903,
+      "learning_rate": 0.0001,
+      "loss": 1.4214,
+      "step": 11178
+    },
+    {
+      "epoch": 1.2841307219573832,
+      "grad_norm": 0.520551860332489,
+      "learning_rate": 0.0001,
+      "loss": 1.3355,
+      "step": 11179
+    },
+    {
+      "epoch": 1.2842455918672104,
+      "grad_norm": 0.6834689378738403,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 11180
+    },
+    {
+      "epoch": 1.2843604617770374,
+      "grad_norm": 0.6506545543670654,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 11181
+    },
+    {
+      "epoch": 1.2844753316868647,
+      "grad_norm": 0.5278921723365784,
+      "learning_rate": 0.0001,
+      "loss": 1.2855,
+      "step": 11182
+    },
+    {
+      "epoch": 1.2845902015966917,
+      "grad_norm": 0.6012974977493286,
+      "learning_rate": 0.0001,
+      "loss": 1.3572,
+      "step": 11183
+    },
+    {
+      "epoch": 1.284705071506519,
+      "grad_norm": 0.59954434633255,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 11184
+    },
+    {
+      "epoch": 1.284819941416346,
+      "grad_norm": 0.5746206045150757,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 11185
+    },
+    {
+      "epoch": 1.2849348113261732,
+      "grad_norm": 0.5760904550552368,
+      "learning_rate": 0.0001,
+      "loss": 1.4865,
+      "step": 11186
+    },
+    {
+      "epoch": 1.2850496812360002,
+      "grad_norm": 0.6236897110939026,
+      "learning_rate": 0.0001,
+      "loss": 1.446,
+      "step": 11187
+    },
+    {
+      "epoch": 1.2851645511458274,
+      "grad_norm": 0.5215889811515808,
+      "learning_rate": 0.0001,
+      "loss": 1.4459,
+      "step": 11188
+    },
+    {
+      "epoch": 1.2852794210556544,
+      "grad_norm": 0.5659036040306091,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 11189
+    },
+    {
+      "epoch": 1.2853942909654816,
+      "grad_norm": 0.510826587677002,
+      "learning_rate": 0.0001,
+      "loss": 1.3339,
+      "step": 11190
+    },
+    {
+      "epoch": 1.2855091608753086,
+      "grad_norm": 0.5990018844604492,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 11191
+    },
+    {
+      "epoch": 1.2856240307851359,
+      "grad_norm": 0.6071823835372925,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 11192
+    },
+    {
+      "epoch": 1.2857389006949629,
+      "grad_norm": 0.6266348958015442,
+      "learning_rate": 0.0001,
+      "loss": 1.7044,
+      "step": 11193
+    },
+    {
+      "epoch": 1.2858537706047901,
+      "grad_norm": 0.6025568246841431,
+      "learning_rate": 0.0001,
+      "loss": 1.2896,
+      "step": 11194
+    },
+    {
+      "epoch": 1.2859686405146171,
+      "grad_norm": 0.6044636368751526,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 11195
+    },
+    {
+      "epoch": 1.2860835104244444,
+      "grad_norm": 0.7049621939659119,
+      "learning_rate": 0.0001,
+      "loss": 1.8122,
+      "step": 11196
+    },
+    {
+      "epoch": 1.2861983803342714,
+      "grad_norm": 0.5462698936462402,
+      "learning_rate": 0.0001,
+      "loss": 1.484,
+      "step": 11197
+    },
+    {
+      "epoch": 1.2863132502440986,
+      "grad_norm": 0.5640134811401367,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 11198
+    },
+    {
+      "epoch": 1.2864281201539256,
+      "grad_norm": 0.5560702681541443,
+      "learning_rate": 0.0001,
+      "loss": 1.4912,
+      "step": 11199
+    },
+    {
+      "epoch": 1.2865429900637528,
+      "grad_norm": 0.5284111499786377,
+      "learning_rate": 0.0001,
+      "loss": 1.2599,
+      "step": 11200
+    },
+    {
+      "epoch": 1.2866578599735798,
+      "grad_norm": 0.5436888933181763,
+      "learning_rate": 0.0001,
+      "loss": 1.3779,
+      "step": 11201
+    },
+    {
+      "epoch": 1.286772729883407,
+      "grad_norm": 0.5959984660148621,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 11202
+    },
+    {
+      "epoch": 1.286887599793234,
+      "grad_norm": 0.5684484839439392,
+      "learning_rate": 0.0001,
+      "loss": 1.4591,
+      "step": 11203
+    },
+    {
+      "epoch": 1.2870024697030613,
+      "grad_norm": 0.5544134974479675,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 11204
+    },
+    {
+      "epoch": 1.2871173396128883,
+      "grad_norm": 0.5449683666229248,
+      "learning_rate": 0.0001,
+      "loss": 1.4292,
+      "step": 11205
+    },
+    {
+      "epoch": 1.2872322095227156,
+      "grad_norm": 0.5800986289978027,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 11206
+    },
+    {
+      "epoch": 1.2873470794325426,
+      "grad_norm": 0.5813232064247131,
+      "learning_rate": 0.0001,
+      "loss": 1.5989,
+      "step": 11207
+    },
+    {
+      "epoch": 1.2874619493423698,
+      "grad_norm": 0.5666254758834839,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 11208
+    },
+    {
+      "epoch": 1.2875768192521968,
+      "grad_norm": 0.6252794861793518,
+      "learning_rate": 0.0001,
+      "loss": 1.6814,
+      "step": 11209
+    },
+    {
+      "epoch": 1.287691689162024,
+      "grad_norm": 0.5321157574653625,
+      "learning_rate": 0.0001,
+      "loss": 1.3024,
+      "step": 11210
+    },
+    {
+      "epoch": 1.2878065590718513,
+      "grad_norm": 0.5458829998970032,
+      "learning_rate": 0.0001,
+      "loss": 1.3577,
+      "step": 11211
+    },
+    {
+      "epoch": 1.2879214289816783,
+      "grad_norm": 0.5970067977905273,
+      "learning_rate": 0.0001,
+      "loss": 1.4105,
+      "step": 11212
+    },
+    {
+      "epoch": 1.2880362988915053,
+      "grad_norm": 0.5711007118225098,
+      "learning_rate": 0.0001,
+      "loss": 1.4712,
+      "step": 11213
+    },
+    {
+      "epoch": 1.2881511688013325,
+      "grad_norm": 0.5796260833740234,
+      "learning_rate": 0.0001,
+      "loss": 1.5902,
+      "step": 11214
+    },
+    {
+      "epoch": 1.2882660387111597,
+      "grad_norm": 0.6137135028839111,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 11215
+    },
+    {
+      "epoch": 1.2883809086209868,
+      "grad_norm": 0.6448894739151001,
+      "learning_rate": 0.0001,
+      "loss": 1.4776,
+      "step": 11216
+    },
+    {
+      "epoch": 1.2884957785308138,
+      "grad_norm": 0.5258020758628845,
+      "learning_rate": 0.0001,
+      "loss": 1.396,
+      "step": 11217
+    },
+    {
+      "epoch": 1.288610648440641,
+      "grad_norm": 0.5507315397262573,
+      "learning_rate": 0.0001,
+      "loss": 1.3656,
+      "step": 11218
+    },
+    {
+      "epoch": 1.2887255183504682,
+      "grad_norm": 0.5842088460922241,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 11219
+    },
+    {
+      "epoch": 1.2888403882602952,
+      "grad_norm": 0.6149991154670715,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 11220
+    },
+    {
+      "epoch": 1.2889552581701222,
+      "grad_norm": 0.5865837335586548,
+      "learning_rate": 0.0001,
+      "loss": 1.3607,
+      "step": 11221
+    },
+    {
+      "epoch": 1.2890701280799495,
+      "grad_norm": 0.5908994078636169,
+      "learning_rate": 0.0001,
+      "loss": 1.403,
+      "step": 11222
+    },
+    {
+      "epoch": 1.2891849979897767,
+      "grad_norm": 0.5459616184234619,
+      "learning_rate": 0.0001,
+      "loss": 1.3524,
+      "step": 11223
+    },
+    {
+      "epoch": 1.2892998678996037,
+      "grad_norm": 0.6645961403846741,
+      "learning_rate": 0.0001,
+      "loss": 1.2631,
+      "step": 11224
+    },
+    {
+      "epoch": 1.2894147378094307,
+      "grad_norm": 0.6288965344429016,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 11225
+    },
+    {
+      "epoch": 1.289529607719258,
+      "grad_norm": 0.6147708892822266,
+      "learning_rate": 0.0001,
+      "loss": 1.4303,
+      "step": 11226
+    },
+    {
+      "epoch": 1.2896444776290852,
+      "grad_norm": 0.6296586394309998,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 11227
+    },
+    {
+      "epoch": 1.2897593475389122,
+      "grad_norm": 0.5921497344970703,
+      "learning_rate": 0.0001,
+      "loss": 1.4274,
+      "step": 11228
+    },
+    {
+      "epoch": 1.2898742174487392,
+      "grad_norm": 0.6173571348190308,
+      "learning_rate": 0.0001,
+      "loss": 1.6724,
+      "step": 11229
+    },
+    {
+      "epoch": 1.2899890873585664,
+      "grad_norm": 0.6189887523651123,
+      "learning_rate": 0.0001,
+      "loss": 1.3725,
+      "step": 11230
+    },
+    {
+      "epoch": 1.2901039572683937,
+      "grad_norm": 0.569510817527771,
+      "learning_rate": 0.0001,
+      "loss": 1.4343,
+      "step": 11231
+    },
+    {
+      "epoch": 1.2902188271782207,
+      "grad_norm": 0.622203528881073,
+      "learning_rate": 0.0001,
+      "loss": 1.4382,
+      "step": 11232
+    },
+    {
+      "epoch": 1.2903336970880477,
+      "grad_norm": 0.6754775643348694,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 11233
+    },
+    {
+      "epoch": 1.290448566997875,
+      "grad_norm": 0.5515174865722656,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 11234
+    },
+    {
+      "epoch": 1.2905634369077021,
+      "grad_norm": 0.6426066160202026,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 11235
+    },
+    {
+      "epoch": 1.2906783068175292,
+      "grad_norm": 0.6196022033691406,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 11236
+    },
+    {
+      "epoch": 1.2907931767273562,
+      "grad_norm": 0.5864408612251282,
+      "learning_rate": 0.0001,
+      "loss": 1.4135,
+      "step": 11237
+    },
+    {
+      "epoch": 1.2909080466371834,
+      "grad_norm": 0.607479989528656,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 11238
+    },
+    {
+      "epoch": 1.2910229165470106,
+      "grad_norm": 0.5427273511886597,
+      "learning_rate": 0.0001,
+      "loss": 1.374,
+      "step": 11239
+    },
+    {
+      "epoch": 1.2911377864568376,
+      "grad_norm": 0.5896614789962769,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 11240
+    },
+    {
+      "epoch": 1.2912526563666646,
+      "grad_norm": 0.5615909695625305,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 11241
+    },
+    {
+      "epoch": 1.2913675262764919,
+      "grad_norm": 0.6087133884429932,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 11242
+    },
+    {
+      "epoch": 1.291482396186319,
+      "grad_norm": 0.5094661712646484,
+      "learning_rate": 0.0001,
+      "loss": 1.2658,
+      "step": 11243
+    },
+    {
+      "epoch": 1.2915972660961461,
+      "grad_norm": 0.6043559908866882,
+      "learning_rate": 0.0001,
+      "loss": 1.4343,
+      "step": 11244
+    },
+    {
+      "epoch": 1.2917121360059731,
+      "grad_norm": 0.586310625076294,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 11245
+    },
+    {
+      "epoch": 1.2918270059158004,
+      "grad_norm": 0.5929206013679504,
+      "learning_rate": 0.0001,
+      "loss": 1.2632,
+      "step": 11246
+    },
+    {
+      "epoch": 1.2919418758256276,
+      "grad_norm": 0.5981222987174988,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 11247
+    },
+    {
+      "epoch": 1.2920567457354546,
+      "grad_norm": 0.5489922761917114,
+      "learning_rate": 0.0001,
+      "loss": 1.3284,
+      "step": 11248
+    },
+    {
+      "epoch": 1.2921716156452816,
+      "grad_norm": 0.5560479164123535,
+      "learning_rate": 0.0001,
+      "loss": 1.2838,
+      "step": 11249
+    },
+    {
+      "epoch": 1.2922864855551088,
+      "grad_norm": 0.5626586675643921,
+      "learning_rate": 0.0001,
+      "loss": 1.5256,
+      "step": 11250
+    },
+    {
+      "epoch": 1.292401355464936,
+      "grad_norm": 0.5712599158287048,
+      "learning_rate": 0.0001,
+      "loss": 1.598,
+      "step": 11251
+    },
+    {
+      "epoch": 1.292516225374763,
+      "grad_norm": 0.5926998257637024,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 11252
+    },
+    {
+      "epoch": 1.29263109528459,
+      "grad_norm": 0.5249360799789429,
+      "learning_rate": 0.0001,
+      "loss": 1.2447,
+      "step": 11253
+    },
+    {
+      "epoch": 1.2927459651944173,
+      "grad_norm": 0.5579542517662048,
+      "learning_rate": 0.0001,
+      "loss": 1.3727,
+      "step": 11254
+    },
+    {
+      "epoch": 1.2928608351042445,
+      "grad_norm": 0.5543383955955505,
+      "learning_rate": 0.0001,
+      "loss": 1.3233,
+      "step": 11255
+    },
+    {
+      "epoch": 1.2929757050140716,
+      "grad_norm": 0.600277304649353,
+      "learning_rate": 0.0001,
+      "loss": 1.5099,
+      "step": 11256
+    },
+    {
+      "epoch": 1.2930905749238986,
+      "grad_norm": 0.5769601464271545,
+      "learning_rate": 0.0001,
+      "loss": 1.1063,
+      "step": 11257
+    },
+    {
+      "epoch": 1.2932054448337258,
+      "grad_norm": 0.5800927877426147,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 11258
+    },
+    {
+      "epoch": 1.293320314743553,
+      "grad_norm": 0.5866220593452454,
+      "learning_rate": 0.0001,
+      "loss": 1.4052,
+      "step": 11259
+    },
+    {
+      "epoch": 1.29343518465338,
+      "grad_norm": 0.5848959684371948,
+      "learning_rate": 0.0001,
+      "loss": 1.3863,
+      "step": 11260
+    },
+    {
+      "epoch": 1.293550054563207,
+      "grad_norm": 0.5541538000106812,
+      "learning_rate": 0.0001,
+      "loss": 1.325,
+      "step": 11261
+    },
+    {
+      "epoch": 1.2936649244730343,
+      "grad_norm": 0.6487392783164978,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 11262
+    },
+    {
+      "epoch": 1.2937797943828615,
+      "grad_norm": 0.6783632040023804,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 11263
+    },
+    {
+      "epoch": 1.2938946642926885,
+      "grad_norm": 0.5799053907394409,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 11264
+    },
+    {
+      "epoch": 1.2940095342025155,
+      "grad_norm": 0.5588035583496094,
+      "learning_rate": 0.0001,
+      "loss": 1.3215,
+      "step": 11265
+    },
+    {
+      "epoch": 1.2941244041123428,
+      "grad_norm": 0.5953567028045654,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 11266
+    },
+    {
+      "epoch": 1.29423927402217,
+      "grad_norm": 0.5587005615234375,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 11267
+    },
+    {
+      "epoch": 1.294354143931997,
+      "grad_norm": 0.6000686287879944,
+      "learning_rate": 0.0001,
+      "loss": 1.3503,
+      "step": 11268
+    },
+    {
+      "epoch": 1.294469013841824,
+      "grad_norm": 0.5800769329071045,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 11269
+    },
+    {
+      "epoch": 1.2945838837516512,
+      "grad_norm": 0.5538628101348877,
+      "learning_rate": 0.0001,
+      "loss": 1.2169,
+      "step": 11270
+    },
+    {
+      "epoch": 1.2946987536614785,
+      "grad_norm": 0.5448545217514038,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 11271
+    },
+    {
+      "epoch": 1.2948136235713055,
+      "grad_norm": 0.5209142565727234,
+      "learning_rate": 0.0001,
+      "loss": 1.3355,
+      "step": 11272
+    },
+    {
+      "epoch": 1.2949284934811325,
+      "grad_norm": 0.5593119859695435,
+      "learning_rate": 0.0001,
+      "loss": 1.4323,
+      "step": 11273
+    },
+    {
+      "epoch": 1.2950433633909597,
+      "grad_norm": 0.5502519607543945,
+      "learning_rate": 0.0001,
+      "loss": 1.4902,
+      "step": 11274
+    },
+    {
+      "epoch": 1.295158233300787,
+      "grad_norm": 0.5910590887069702,
+      "learning_rate": 0.0001,
+      "loss": 1.4953,
+      "step": 11275
+    },
+    {
+      "epoch": 1.295273103210614,
+      "grad_norm": 0.5811395645141602,
+      "learning_rate": 0.0001,
+      "loss": 1.6886,
+      "step": 11276
+    },
+    {
+      "epoch": 1.2953879731204412,
+      "grad_norm": 0.582584798336029,
+      "learning_rate": 0.0001,
+      "loss": 1.4012,
+      "step": 11277
+    },
+    {
+      "epoch": 1.2955028430302682,
+      "grad_norm": 0.5751772522926331,
+      "learning_rate": 0.0001,
+      "loss": 1.3962,
+      "step": 11278
+    },
+    {
+      "epoch": 1.2956177129400954,
+      "grad_norm": 0.5952954888343811,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 11279
+    },
+    {
+      "epoch": 1.2957325828499224,
+      "grad_norm": 0.5689733028411865,
+      "learning_rate": 0.0001,
+      "loss": 1.43,
+      "step": 11280
+    },
+    {
+      "epoch": 1.2958474527597497,
+      "grad_norm": 0.5460832715034485,
+      "learning_rate": 0.0001,
+      "loss": 1.4803,
+      "step": 11281
+    },
+    {
+      "epoch": 1.2959623226695767,
+      "grad_norm": 0.604256808757782,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 11282
+    },
+    {
+      "epoch": 1.296077192579404,
+      "grad_norm": 0.5800360441207886,
+      "learning_rate": 0.0001,
+      "loss": 1.4641,
+      "step": 11283
+    },
+    {
+      "epoch": 1.296192062489231,
+      "grad_norm": 0.529690146446228,
+      "learning_rate": 0.0001,
+      "loss": 1.3187,
+      "step": 11284
+    },
+    {
+      "epoch": 1.2963069323990581,
+      "grad_norm": 0.5543114542961121,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 11285
+    },
+    {
+      "epoch": 1.2964218023088852,
+      "grad_norm": 0.5886980295181274,
+      "learning_rate": 0.0001,
+      "loss": 1.5753,
+      "step": 11286
+    },
+    {
+      "epoch": 1.2965366722187124,
+      "grad_norm": 0.5955923795700073,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 11287
+    },
+    {
+      "epoch": 1.2966515421285394,
+      "grad_norm": 0.6192274689674377,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 11288
+    },
+    {
+      "epoch": 1.2967664120383666,
+      "grad_norm": 0.5242689251899719,
+      "learning_rate": 0.0001,
+      "loss": 1.3975,
+      "step": 11289
+    },
+    {
+      "epoch": 1.2968812819481936,
+      "grad_norm": 0.6103929877281189,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 11290
+    },
+    {
+      "epoch": 1.2969961518580209,
+      "grad_norm": 0.5915534496307373,
+      "learning_rate": 0.0001,
+      "loss": 1.4537,
+      "step": 11291
+    },
+    {
+      "epoch": 1.2971110217678479,
+      "grad_norm": 0.5651288032531738,
+      "learning_rate": 0.0001,
+      "loss": 1.2593,
+      "step": 11292
+    },
+    {
+      "epoch": 1.297225891677675,
+      "grad_norm": 0.5925031304359436,
+      "learning_rate": 0.0001,
+      "loss": 1.4754,
+      "step": 11293
+    },
+    {
+      "epoch": 1.2973407615875021,
+      "grad_norm": 0.5318403840065002,
+      "learning_rate": 0.0001,
+      "loss": 1.3499,
+      "step": 11294
+    },
+    {
+      "epoch": 1.2974556314973293,
+      "grad_norm": 0.5483918786048889,
+      "learning_rate": 0.0001,
+      "loss": 1.45,
+      "step": 11295
+    },
+    {
+      "epoch": 1.2975705014071564,
+      "grad_norm": 0.561732828617096,
+      "learning_rate": 0.0001,
+      "loss": 1.4525,
+      "step": 11296
+    },
+    {
+      "epoch": 1.2976853713169836,
+      "grad_norm": 0.5497708916664124,
+      "learning_rate": 0.0001,
+      "loss": 1.2997,
+      "step": 11297
+    },
+    {
+      "epoch": 1.2978002412268106,
+      "grad_norm": 0.5718294382095337,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 11298
+    },
+    {
+      "epoch": 1.2979151111366378,
+      "grad_norm": 0.5569791793823242,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 11299
+    },
+    {
+      "epoch": 1.2980299810464648,
+      "grad_norm": 0.6240097284317017,
+      "learning_rate": 0.0001,
+      "loss": 1.7817,
+      "step": 11300
+    },
+    {
+      "epoch": 1.298144850956292,
+      "grad_norm": 0.5455806851387024,
+      "learning_rate": 0.0001,
+      "loss": 1.517,
+      "step": 11301
+    },
+    {
+      "epoch": 1.298259720866119,
+      "grad_norm": 0.6025298833847046,
+      "learning_rate": 0.0001,
+      "loss": 1.4755,
+      "step": 11302
+    },
+    {
+      "epoch": 1.2983745907759463,
+      "grad_norm": 0.5894249677658081,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 11303
+    },
+    {
+      "epoch": 1.2984894606857733,
+      "grad_norm": 0.5962689518928528,
+      "learning_rate": 0.0001,
+      "loss": 1.6333,
+      "step": 11304
+    },
+    {
+      "epoch": 1.2986043305956005,
+      "grad_norm": 0.6713701486587524,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 11305
+    },
+    {
+      "epoch": 1.2987192005054276,
+      "grad_norm": 0.5443199276924133,
+      "learning_rate": 0.0001,
+      "loss": 1.4608,
+      "step": 11306
+    },
+    {
+      "epoch": 1.2988340704152548,
+      "grad_norm": 0.5849419236183167,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 11307
+    },
+    {
+      "epoch": 1.2989489403250818,
+      "grad_norm": 0.5818788409233093,
+      "learning_rate": 0.0001,
+      "loss": 1.5971,
+      "step": 11308
+    },
+    {
+      "epoch": 1.299063810234909,
+      "grad_norm": 0.5864542126655579,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 11309
+    },
+    {
+      "epoch": 1.299178680144736,
+      "grad_norm": 0.5585055947303772,
+      "learning_rate": 0.0001,
+      "loss": 1.3616,
+      "step": 11310
+    },
+    {
+      "epoch": 1.2992935500545633,
+      "grad_norm": 0.725004255771637,
+      "learning_rate": 0.0001,
+      "loss": 1.7103,
+      "step": 11311
+    },
+    {
+      "epoch": 1.2994084199643903,
+      "grad_norm": 0.6212374567985535,
+      "learning_rate": 0.0001,
+      "loss": 1.6987,
+      "step": 11312
+    },
+    {
+      "epoch": 1.2995232898742175,
+      "grad_norm": 0.5447000861167908,
+      "learning_rate": 0.0001,
+      "loss": 1.474,
+      "step": 11313
+    },
+    {
+      "epoch": 1.2996381597840445,
+      "grad_norm": 0.6027454137802124,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 11314
+    },
+    {
+      "epoch": 1.2997530296938717,
+      "grad_norm": 0.568374514579773,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 11315
+    },
+    {
+      "epoch": 1.2998678996036988,
+      "grad_norm": 0.6000039577484131,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 11316
+    },
+    {
+      "epoch": 1.299982769513526,
+      "grad_norm": 0.5799726247787476,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 11317
+    },
+    {
+      "epoch": 1.300097639423353,
+      "grad_norm": 0.5564817190170288,
+      "learning_rate": 0.0001,
+      "loss": 1.3677,
+      "step": 11318
+    },
+    {
+      "epoch": 1.3002125093331802,
+      "grad_norm": 0.537472665309906,
+      "learning_rate": 0.0001,
+      "loss": 1.3931,
+      "step": 11319
+    },
+    {
+      "epoch": 1.3003273792430072,
+      "grad_norm": 0.577150821685791,
+      "learning_rate": 0.0001,
+      "loss": 1.42,
+      "step": 11320
+    },
+    {
+      "epoch": 1.3004422491528345,
+      "grad_norm": 0.572995126247406,
+      "learning_rate": 0.0001,
+      "loss": 1.3415,
+      "step": 11321
+    },
+    {
+      "epoch": 1.3005571190626615,
+      "grad_norm": 0.5590890645980835,
+      "learning_rate": 0.0001,
+      "loss": 1.321,
+      "step": 11322
+    },
+    {
+      "epoch": 1.3006719889724887,
+      "grad_norm": 0.5591031908988953,
+      "learning_rate": 0.0001,
+      "loss": 1.4025,
+      "step": 11323
+    },
+    {
+      "epoch": 1.3007868588823157,
+      "grad_norm": 0.5289561748504639,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 11324
+    },
+    {
+      "epoch": 1.300901728792143,
+      "grad_norm": 0.6011853814125061,
+      "learning_rate": 0.0001,
+      "loss": 1.7713,
+      "step": 11325
+    },
+    {
+      "epoch": 1.30101659870197,
+      "grad_norm": 0.5395965576171875,
+      "learning_rate": 0.0001,
+      "loss": 1.4244,
+      "step": 11326
+    },
+    {
+      "epoch": 1.3011314686117972,
+      "grad_norm": 0.5598213076591492,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 11327
+    },
+    {
+      "epoch": 1.3012463385216242,
+      "grad_norm": 0.5253582000732422,
+      "learning_rate": 0.0001,
+      "loss": 1.3091,
+      "step": 11328
+    },
+    {
+      "epoch": 1.3013612084314514,
+      "grad_norm": 0.57694011926651,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 11329
+    },
+    {
+      "epoch": 1.3014760783412784,
+      "grad_norm": 0.6319670677185059,
+      "learning_rate": 0.0001,
+      "loss": 1.6043,
+      "step": 11330
+    },
+    {
+      "epoch": 1.3015909482511057,
+      "grad_norm": 0.6250157356262207,
+      "learning_rate": 0.0001,
+      "loss": 1.5442,
+      "step": 11331
+    },
+    {
+      "epoch": 1.3017058181609327,
+      "grad_norm": 0.5917470455169678,
+      "learning_rate": 0.0001,
+      "loss": 1.3269,
+      "step": 11332
+    },
+    {
+      "epoch": 1.30182068807076,
+      "grad_norm": 0.546095609664917,
+      "learning_rate": 0.0001,
+      "loss": 1.186,
+      "step": 11333
+    },
+    {
+      "epoch": 1.301935557980587,
+      "grad_norm": 0.6022743582725525,
+      "learning_rate": 0.0001,
+      "loss": 1.4801,
+      "step": 11334
+    },
+    {
+      "epoch": 1.3020504278904141,
+      "grad_norm": 0.6155171394348145,
+      "learning_rate": 0.0001,
+      "loss": 1.7134,
+      "step": 11335
+    },
+    {
+      "epoch": 1.3021652978002412,
+      "grad_norm": 0.6018640995025635,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 11336
+    },
+    {
+      "epoch": 1.3022801677100684,
+      "grad_norm": 0.5561271905899048,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 11337
+    },
+    {
+      "epoch": 1.3023950376198954,
+      "grad_norm": 0.5770154595375061,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 11338
+    },
+    {
+      "epoch": 1.3025099075297226,
+      "grad_norm": 0.6154848337173462,
+      "learning_rate": 0.0001,
+      "loss": 1.3091,
+      "step": 11339
+    },
+    {
+      "epoch": 1.3026247774395496,
+      "grad_norm": 0.5868239402770996,
+      "learning_rate": 0.0001,
+      "loss": 1.5571,
+      "step": 11340
+    },
+    {
+      "epoch": 1.3027396473493769,
+      "grad_norm": 0.6514554619789124,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 11341
+    },
+    {
+      "epoch": 1.3028545172592039,
+      "grad_norm": 0.60183185338974,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 11342
+    },
+    {
+      "epoch": 1.302969387169031,
+      "grad_norm": 0.5899741649627686,
+      "learning_rate": 0.0001,
+      "loss": 1.4028,
+      "step": 11343
+    },
+    {
+      "epoch": 1.303084257078858,
+      "grad_norm": 0.5583611726760864,
+      "learning_rate": 0.0001,
+      "loss": 1.3219,
+      "step": 11344
+    },
+    {
+      "epoch": 1.3031991269886853,
+      "grad_norm": 0.5439897179603577,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 11345
+    },
+    {
+      "epoch": 1.3033139968985124,
+      "grad_norm": 0.5689780712127686,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 11346
+    },
+    {
+      "epoch": 1.3034288668083396,
+      "grad_norm": 0.5775777697563171,
+      "learning_rate": 0.0001,
+      "loss": 1.562,
+      "step": 11347
+    },
+    {
+      "epoch": 1.3035437367181668,
+      "grad_norm": 0.6249666810035706,
+      "learning_rate": 0.0001,
+      "loss": 1.2675,
+      "step": 11348
+    },
+    {
+      "epoch": 1.3036586066279938,
+      "grad_norm": 0.6017664074897766,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 11349
+    },
+    {
+      "epoch": 1.3037734765378208,
+      "grad_norm": 0.5548145174980164,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 11350
+    },
+    {
+      "epoch": 1.303888346447648,
+      "grad_norm": 0.6580796241760254,
+      "learning_rate": 0.0001,
+      "loss": 1.7596,
+      "step": 11351
+    },
+    {
+      "epoch": 1.3040032163574753,
+      "grad_norm": 0.5205984711647034,
+      "learning_rate": 0.0001,
+      "loss": 1.3289,
+      "step": 11352
+    },
+    {
+      "epoch": 1.3041180862673023,
+      "grad_norm": 0.5620521903038025,
+      "learning_rate": 0.0001,
+      "loss": 1.4247,
+      "step": 11353
+    },
+    {
+      "epoch": 1.3042329561771293,
+      "grad_norm": 0.6419723033905029,
+      "learning_rate": 0.0001,
+      "loss": 1.6835,
+      "step": 11354
+    },
+    {
+      "epoch": 1.3043478260869565,
+      "grad_norm": 0.5205932855606079,
+      "learning_rate": 0.0001,
+      "loss": 1.3938,
+      "step": 11355
+    },
+    {
+      "epoch": 1.3044626959967838,
+      "grad_norm": 0.5548444390296936,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 11356
+    },
+    {
+      "epoch": 1.3045775659066108,
+      "grad_norm": 0.6600873470306396,
+      "learning_rate": 0.0001,
+      "loss": 1.6834,
+      "step": 11357
+    },
+    {
+      "epoch": 1.3046924358164378,
+      "grad_norm": 0.5854970812797546,
+      "learning_rate": 0.0001,
+      "loss": 1.4424,
+      "step": 11358
+    },
+    {
+      "epoch": 1.304807305726265,
+      "grad_norm": 0.5743281841278076,
+      "learning_rate": 0.0001,
+      "loss": 1.5258,
+      "step": 11359
+    },
+    {
+      "epoch": 1.3049221756360923,
+      "grad_norm": 0.5808170437812805,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 11360
+    },
+    {
+      "epoch": 1.3050370455459193,
+      "grad_norm": 0.5460790395736694,
+      "learning_rate": 0.0001,
+      "loss": 1.2765,
+      "step": 11361
+    },
+    {
+      "epoch": 1.3051519154557463,
+      "grad_norm": 0.5941900610923767,
+      "learning_rate": 0.0001,
+      "loss": 1.6998,
+      "step": 11362
+    },
+    {
+      "epoch": 1.3052667853655735,
+      "grad_norm": 0.5910804271697998,
+      "learning_rate": 0.0001,
+      "loss": 1.6016,
+      "step": 11363
+    },
+    {
+      "epoch": 1.3053816552754007,
+      "grad_norm": 0.5705919861793518,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 11364
+    },
+    {
+      "epoch": 1.3054965251852277,
+      "grad_norm": 0.5501276850700378,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 11365
+    },
+    {
+      "epoch": 1.3056113950950547,
+      "grad_norm": 0.5633885860443115,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 11366
+    },
+    {
+      "epoch": 1.305726265004882,
+      "grad_norm": 0.6157667636871338,
+      "learning_rate": 0.0001,
+      "loss": 1.4968,
+      "step": 11367
+    },
+    {
+      "epoch": 1.3058411349147092,
+      "grad_norm": 0.5917145609855652,
+      "learning_rate": 0.0001,
+      "loss": 1.3983,
+      "step": 11368
+    },
+    {
+      "epoch": 1.3059560048245362,
+      "grad_norm": 0.6143923997879028,
+      "learning_rate": 0.0001,
+      "loss": 1.172,
+      "step": 11369
+    },
+    {
+      "epoch": 1.3060708747343632,
+      "grad_norm": 0.5878526568412781,
+      "learning_rate": 0.0001,
+      "loss": 1.3639,
+      "step": 11370
+    },
+    {
+      "epoch": 1.3061857446441905,
+      "grad_norm": 0.5434131026268005,
+      "learning_rate": 0.0001,
+      "loss": 1.4643,
+      "step": 11371
+    },
+    {
+      "epoch": 1.3063006145540177,
+      "grad_norm": 0.5768989324569702,
+      "learning_rate": 0.0001,
+      "loss": 1.4185,
+      "step": 11372
+    },
+    {
+      "epoch": 1.3064154844638447,
+      "grad_norm": 0.5958837270736694,
+      "learning_rate": 0.0001,
+      "loss": 1.4573,
+      "step": 11373
+    },
+    {
+      "epoch": 1.3065303543736717,
+      "grad_norm": 0.5841070413589478,
+      "learning_rate": 0.0001,
+      "loss": 1.5135,
+      "step": 11374
+    },
+    {
+      "epoch": 1.306645224283499,
+      "grad_norm": 0.5326519012451172,
+      "learning_rate": 0.0001,
+      "loss": 1.1777,
+      "step": 11375
+    },
+    {
+      "epoch": 1.3067600941933262,
+      "grad_norm": 0.640945553779602,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 11376
+    },
+    {
+      "epoch": 1.3068749641031532,
+      "grad_norm": 0.546720564365387,
+      "learning_rate": 0.0001,
+      "loss": 1.2742,
+      "step": 11377
+    },
+    {
+      "epoch": 1.3069898340129802,
+      "grad_norm": 0.546595573425293,
+      "learning_rate": 0.0001,
+      "loss": 1.1787,
+      "step": 11378
+    },
+    {
+      "epoch": 1.3071047039228074,
+      "grad_norm": 0.6050646901130676,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 11379
+    },
+    {
+      "epoch": 1.3072195738326347,
+      "grad_norm": 0.5861079096794128,
+      "learning_rate": 0.0001,
+      "loss": 1.6063,
+      "step": 11380
+    },
+    {
+      "epoch": 1.3073344437424617,
+      "grad_norm": 0.6234222650527954,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 11381
+    },
+    {
+      "epoch": 1.3074493136522887,
+      "grad_norm": 0.5523126721382141,
+      "learning_rate": 0.0001,
+      "loss": 1.3926,
+      "step": 11382
+    },
+    {
+      "epoch": 1.307564183562116,
+      "grad_norm": 0.6148357391357422,
+      "learning_rate": 0.0001,
+      "loss": 1.7076,
+      "step": 11383
+    },
+    {
+      "epoch": 1.3076790534719431,
+      "grad_norm": 0.5839380025863647,
+      "learning_rate": 0.0001,
+      "loss": 1.4059,
+      "step": 11384
+    },
+    {
+      "epoch": 1.3077939233817701,
+      "grad_norm": 0.559718132019043,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 11385
+    },
+    {
+      "epoch": 1.3079087932915971,
+      "grad_norm": 0.5904785990715027,
+      "learning_rate": 0.0001,
+      "loss": 1.4762,
+      "step": 11386
+    },
+    {
+      "epoch": 1.3080236632014244,
+      "grad_norm": 0.5947718620300293,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 11387
+    },
+    {
+      "epoch": 1.3081385331112516,
+      "grad_norm": 0.5415857434272766,
+      "learning_rate": 0.0001,
+      "loss": 1.1498,
+      "step": 11388
+    },
+    {
+      "epoch": 1.3082534030210786,
+      "grad_norm": 0.5549929738044739,
+      "learning_rate": 0.0001,
+      "loss": 1.256,
+      "step": 11389
+    },
+    {
+      "epoch": 1.3083682729309056,
+      "grad_norm": 0.5811049938201904,
+      "learning_rate": 0.0001,
+      "loss": 1.4635,
+      "step": 11390
+    },
+    {
+      "epoch": 1.3084831428407329,
+      "grad_norm": 0.597584068775177,
+      "learning_rate": 0.0001,
+      "loss": 1.3548,
+      "step": 11391
+    },
+    {
+      "epoch": 1.30859801275056,
+      "grad_norm": 0.6055376529693604,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 11392
+    },
+    {
+      "epoch": 1.308712882660387,
+      "grad_norm": 0.5904808044433594,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 11393
+    },
+    {
+      "epoch": 1.308827752570214,
+      "grad_norm": 0.6280949711799622,
+      "learning_rate": 0.0001,
+      "loss": 1.5001,
+      "step": 11394
+    },
+    {
+      "epoch": 1.3089426224800413,
+      "grad_norm": 0.6043780446052551,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 11395
+    },
+    {
+      "epoch": 1.3090574923898686,
+      "grad_norm": 0.5859163403511047,
+      "learning_rate": 0.0001,
+      "loss": 1.4466,
+      "step": 11396
+    },
+    {
+      "epoch": 1.3091723622996956,
+      "grad_norm": 0.5889775156974792,
+      "learning_rate": 0.0001,
+      "loss": 1.4596,
+      "step": 11397
+    },
+    {
+      "epoch": 1.3092872322095226,
+      "grad_norm": 0.5822187066078186,
+      "learning_rate": 0.0001,
+      "loss": 1.4784,
+      "step": 11398
+    },
+    {
+      "epoch": 1.3094021021193498,
+      "grad_norm": 0.5596886873245239,
+      "learning_rate": 0.0001,
+      "loss": 1.4668,
+      "step": 11399
+    },
+    {
+      "epoch": 1.309516972029177,
+      "grad_norm": 0.6127359867095947,
+      "learning_rate": 0.0001,
+      "loss": 1.4651,
+      "step": 11400
+    },
+    {
+      "epoch": 1.309631841939004,
+      "grad_norm": 0.5832725167274475,
+      "learning_rate": 0.0001,
+      "loss": 1.4728,
+      "step": 11401
+    },
+    {
+      "epoch": 1.309746711848831,
+      "grad_norm": 0.5246071219444275,
+      "learning_rate": 0.0001,
+      "loss": 1.3994,
+      "step": 11402
+    },
+    {
+      "epoch": 1.3098615817586583,
+      "grad_norm": 0.5589280724525452,
+      "learning_rate": 0.0001,
+      "loss": 1.3416,
+      "step": 11403
+    },
+    {
+      "epoch": 1.3099764516684855,
+      "grad_norm": 0.5817440748214722,
+      "learning_rate": 0.0001,
+      "loss": 1.357,
+      "step": 11404
+    },
+    {
+      "epoch": 1.3100913215783125,
+      "grad_norm": 0.5823844075202942,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 11405
+    },
+    {
+      "epoch": 1.3102061914881395,
+      "grad_norm": 0.590492308139801,
+      "learning_rate": 0.0001,
+      "loss": 1.3188,
+      "step": 11406
+    },
+    {
+      "epoch": 1.3103210613979668,
+      "grad_norm": 0.5652885437011719,
+      "learning_rate": 0.0001,
+      "loss": 1.3556,
+      "step": 11407
+    },
+    {
+      "epoch": 1.310435931307794,
+      "grad_norm": 0.5821486711502075,
+      "learning_rate": 0.0001,
+      "loss": 1.5111,
+      "step": 11408
+    },
+    {
+      "epoch": 1.310550801217621,
+      "grad_norm": 0.585404098033905,
+      "learning_rate": 0.0001,
+      "loss": 1.506,
+      "step": 11409
+    },
+    {
+      "epoch": 1.310665671127448,
+      "grad_norm": 0.5636382699012756,
+      "learning_rate": 0.0001,
+      "loss": 1.3113,
+      "step": 11410
+    },
+    {
+      "epoch": 1.3107805410372753,
+      "grad_norm": 0.5605599880218506,
+      "learning_rate": 0.0001,
+      "loss": 1.3142,
+      "step": 11411
+    },
+    {
+      "epoch": 1.3108954109471025,
+      "grad_norm": 0.6051444411277771,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 11412
+    },
+    {
+      "epoch": 1.3110102808569295,
+      "grad_norm": 0.6009870171546936,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 11413
+    },
+    {
+      "epoch": 1.3111251507667567,
+      "grad_norm": 0.5619489550590515,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 11414
+    },
+    {
+      "epoch": 1.3112400206765837,
+      "grad_norm": 0.5233682990074158,
+      "learning_rate": 0.0001,
+      "loss": 1.3234,
+      "step": 11415
+    },
+    {
+      "epoch": 1.311354890586411,
+      "grad_norm": 0.643240749835968,
+      "learning_rate": 0.0001,
+      "loss": 1.6655,
+      "step": 11416
+    },
+    {
+      "epoch": 1.311469760496238,
+      "grad_norm": 0.633753776550293,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 11417
+    },
+    {
+      "epoch": 1.3115846304060652,
+      "grad_norm": 0.6189075708389282,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 11418
+    },
+    {
+      "epoch": 1.3116995003158922,
+      "grad_norm": 0.5767892003059387,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 11419
+    },
+    {
+      "epoch": 1.3118143702257195,
+      "grad_norm": 0.5835198163986206,
+      "learning_rate": 0.0001,
+      "loss": 1.5653,
+      "step": 11420
+    },
+    {
+      "epoch": 1.3119292401355465,
+      "grad_norm": 0.5322223901748657,
+      "learning_rate": 0.0001,
+      "loss": 1.4651,
+      "step": 11421
+    },
+    {
+      "epoch": 1.3120441100453737,
+      "grad_norm": 0.5825353264808655,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 11422
+    },
+    {
+      "epoch": 1.3121589799552007,
+      "grad_norm": 0.5791360139846802,
+      "learning_rate": 0.0001,
+      "loss": 1.3845,
+      "step": 11423
+    },
+    {
+      "epoch": 1.312273849865028,
+      "grad_norm": 0.637932538986206,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 11424
+    },
+    {
+      "epoch": 1.312388719774855,
+      "grad_norm": 0.6340711116790771,
+      "learning_rate": 0.0001,
+      "loss": 1.6321,
+      "step": 11425
+    },
+    {
+      "epoch": 1.3125035896846822,
+      "grad_norm": 0.5915501117706299,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 11426
+    },
+    {
+      "epoch": 1.3126184595945092,
+      "grad_norm": 0.6027012467384338,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 11427
+    },
+    {
+      "epoch": 1.3127333295043364,
+      "grad_norm": 0.5210409760475159,
+      "learning_rate": 0.0001,
+      "loss": 1.2863,
+      "step": 11428
+    },
+    {
+      "epoch": 1.3128481994141634,
+      "grad_norm": 0.5848100185394287,
+      "learning_rate": 0.0001,
+      "loss": 1.4554,
+      "step": 11429
+    },
+    {
+      "epoch": 1.3129630693239907,
+      "grad_norm": 0.552432119846344,
+      "learning_rate": 0.0001,
+      "loss": 1.4561,
+      "step": 11430
+    },
+    {
+      "epoch": 1.3130779392338177,
+      "grad_norm": 0.5485931038856506,
+      "learning_rate": 0.0001,
+      "loss": 1.2885,
+      "step": 11431
+    },
+    {
+      "epoch": 1.313192809143645,
+      "grad_norm": 0.5956612825393677,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 11432
+    },
+    {
+      "epoch": 1.313307679053472,
+      "grad_norm": 0.5833698511123657,
+      "learning_rate": 0.0001,
+      "loss": 1.338,
+      "step": 11433
+    },
+    {
+      "epoch": 1.3134225489632991,
+      "grad_norm": 0.6002306342124939,
+      "learning_rate": 0.0001,
+      "loss": 1.4549,
+      "step": 11434
+    },
+    {
+      "epoch": 1.3135374188731261,
+      "grad_norm": 0.6598058938980103,
+      "learning_rate": 0.0001,
+      "loss": 1.5414,
+      "step": 11435
+    },
+    {
+      "epoch": 1.3136522887829534,
+      "grad_norm": 0.6079245805740356,
+      "learning_rate": 0.0001,
+      "loss": 1.4909,
+      "step": 11436
+    },
+    {
+      "epoch": 1.3137671586927804,
+      "grad_norm": 0.5811184048652649,
+      "learning_rate": 0.0001,
+      "loss": 1.4411,
+      "step": 11437
+    },
+    {
+      "epoch": 1.3138820286026076,
+      "grad_norm": 0.5552648305892944,
+      "learning_rate": 0.0001,
+      "loss": 1.4865,
+      "step": 11438
+    },
+    {
+      "epoch": 1.3139968985124346,
+      "grad_norm": 0.5896119475364685,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 11439
+    },
+    {
+      "epoch": 1.3141117684222619,
+      "grad_norm": 0.5828573703765869,
+      "learning_rate": 0.0001,
+      "loss": 1.5706,
+      "step": 11440
+    },
+    {
+      "epoch": 1.3142266383320889,
+      "grad_norm": 0.5936744809150696,
+      "learning_rate": 0.0001,
+      "loss": 1.5319,
+      "step": 11441
+    },
+    {
+      "epoch": 1.314341508241916,
+      "grad_norm": 0.672229528427124,
+      "learning_rate": 0.0001,
+      "loss": 1.7404,
+      "step": 11442
+    },
+    {
+      "epoch": 1.314456378151743,
+      "grad_norm": 0.5854552388191223,
+      "learning_rate": 0.0001,
+      "loss": 1.4428,
+      "step": 11443
+    },
+    {
+      "epoch": 1.3145712480615703,
+      "grad_norm": 0.5884828567504883,
+      "learning_rate": 0.0001,
+      "loss": 1.34,
+      "step": 11444
+    },
+    {
+      "epoch": 1.3146861179713973,
+      "grad_norm": 0.6672284007072449,
+      "learning_rate": 0.0001,
+      "loss": 1.3559,
+      "step": 11445
+    },
+    {
+      "epoch": 1.3148009878812246,
+      "grad_norm": 0.5890071988105774,
+      "learning_rate": 0.0001,
+      "loss": 1.4041,
+      "step": 11446
+    },
+    {
+      "epoch": 1.3149158577910516,
+      "grad_norm": 0.573523998260498,
+      "learning_rate": 0.0001,
+      "loss": 1.2297,
+      "step": 11447
+    },
+    {
+      "epoch": 1.3150307277008788,
+      "grad_norm": 0.6235065460205078,
+      "learning_rate": 0.0001,
+      "loss": 1.4743,
+      "step": 11448
+    },
+    {
+      "epoch": 1.3151455976107058,
+      "grad_norm": 0.5744754672050476,
+      "learning_rate": 0.0001,
+      "loss": 1.3518,
+      "step": 11449
+    },
+    {
+      "epoch": 1.315260467520533,
+      "grad_norm": 0.5936892032623291,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 11450
+    },
+    {
+      "epoch": 1.31537533743036,
+      "grad_norm": 0.5716831684112549,
+      "learning_rate": 0.0001,
+      "loss": 1.4176,
+      "step": 11451
+    },
+    {
+      "epoch": 1.3154902073401873,
+      "grad_norm": 0.5791821479797363,
+      "learning_rate": 0.0001,
+      "loss": 1.5311,
+      "step": 11452
+    },
+    {
+      "epoch": 1.3156050772500143,
+      "grad_norm": 0.591475248336792,
+      "learning_rate": 0.0001,
+      "loss": 1.6355,
+      "step": 11453
+    },
+    {
+      "epoch": 1.3157199471598415,
+      "grad_norm": 0.5741580128669739,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 11454
+    },
+    {
+      "epoch": 1.3158348170696685,
+      "grad_norm": 0.5884466767311096,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 11455
+    },
+    {
+      "epoch": 1.3159496869794958,
+      "grad_norm": 0.5626681447029114,
+      "learning_rate": 0.0001,
+      "loss": 1.3815,
+      "step": 11456
+    },
+    {
+      "epoch": 1.3160645568893228,
+      "grad_norm": 0.6762535572052002,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 11457
+    },
+    {
+      "epoch": 1.31617942679915,
+      "grad_norm": 0.5966523885726929,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 11458
+    },
+    {
+      "epoch": 1.316294296708977,
+      "grad_norm": 0.5997388958930969,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 11459
+    },
+    {
+      "epoch": 1.3164091666188042,
+      "grad_norm": 0.6016896367073059,
+      "learning_rate": 0.0001,
+      "loss": 1.1723,
+      "step": 11460
+    },
+    {
+      "epoch": 1.3165240365286313,
+      "grad_norm": 0.5532729625701904,
+      "learning_rate": 0.0001,
+      "loss": 1.2546,
+      "step": 11461
+    },
+    {
+      "epoch": 1.3166389064384585,
+      "grad_norm": 0.5814374685287476,
+      "learning_rate": 0.0001,
+      "loss": 1.4586,
+      "step": 11462
+    },
+    {
+      "epoch": 1.3167537763482855,
+      "grad_norm": 0.5885840058326721,
+      "learning_rate": 0.0001,
+      "loss": 1.4208,
+      "step": 11463
+    },
+    {
+      "epoch": 1.3168686462581127,
+      "grad_norm": 0.5459346175193787,
+      "learning_rate": 0.0001,
+      "loss": 1.2433,
+      "step": 11464
+    },
+    {
+      "epoch": 1.3169835161679397,
+      "grad_norm": 0.633172333240509,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 11465
+    },
+    {
+      "epoch": 1.317098386077767,
+      "grad_norm": 0.5619357228279114,
+      "learning_rate": 0.0001,
+      "loss": 1.3453,
+      "step": 11466
+    },
+    {
+      "epoch": 1.317213255987594,
+      "grad_norm": 0.6191542744636536,
+      "learning_rate": 0.0001,
+      "loss": 1.6639,
+      "step": 11467
+    },
+    {
+      "epoch": 1.3173281258974212,
+      "grad_norm": 0.6198133230209351,
+      "learning_rate": 0.0001,
+      "loss": 1.3868,
+      "step": 11468
+    },
+    {
+      "epoch": 1.3174429958072482,
+      "grad_norm": 0.6131882667541504,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 11469
+    },
+    {
+      "epoch": 1.3175578657170754,
+      "grad_norm": 0.6336104869842529,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 11470
+    },
+    {
+      "epoch": 1.3176727356269025,
+      "grad_norm": 0.5683992505073547,
+      "learning_rate": 0.0001,
+      "loss": 1.6024,
+      "step": 11471
+    },
+    {
+      "epoch": 1.3177876055367297,
+      "grad_norm": 0.5928875207901001,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 11472
+    },
+    {
+      "epoch": 1.3179024754465567,
+      "grad_norm": 0.5696129202842712,
+      "learning_rate": 0.0001,
+      "loss": 1.427,
+      "step": 11473
+    },
+    {
+      "epoch": 1.318017345356384,
+      "grad_norm": 0.5560370683670044,
+      "learning_rate": 0.0001,
+      "loss": 1.3691,
+      "step": 11474
+    },
+    {
+      "epoch": 1.318132215266211,
+      "grad_norm": 0.6248264312744141,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 11475
+    },
+    {
+      "epoch": 1.3182470851760382,
+      "grad_norm": 0.6087796688079834,
+      "learning_rate": 0.0001,
+      "loss": 1.4265,
+      "step": 11476
+    },
+    {
+      "epoch": 1.3183619550858652,
+      "grad_norm": 0.5595715641975403,
+      "learning_rate": 0.0001,
+      "loss": 1.4056,
+      "step": 11477
+    },
+    {
+      "epoch": 1.3184768249956924,
+      "grad_norm": 0.5776646733283997,
+      "learning_rate": 0.0001,
+      "loss": 1.5342,
+      "step": 11478
+    },
+    {
+      "epoch": 1.3185916949055194,
+      "grad_norm": 0.6008871793746948,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 11479
+    },
+    {
+      "epoch": 1.3187065648153466,
+      "grad_norm": 0.6645510792732239,
+      "learning_rate": 0.0001,
+      "loss": 1.7724,
+      "step": 11480
+    },
+    {
+      "epoch": 1.3188214347251737,
+      "grad_norm": 0.5654270052909851,
+      "learning_rate": 0.0001,
+      "loss": 1.3849,
+      "step": 11481
+    },
+    {
+      "epoch": 1.318936304635001,
+      "grad_norm": 0.6282428503036499,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 11482
+    },
+    {
+      "epoch": 1.319051174544828,
+      "grad_norm": 0.5395941138267517,
+      "learning_rate": 0.0001,
+      "loss": 1.4415,
+      "step": 11483
+    },
+    {
+      "epoch": 1.3191660444546551,
+      "grad_norm": 0.6053428649902344,
+      "learning_rate": 0.0001,
+      "loss": 1.4817,
+      "step": 11484
+    },
+    {
+      "epoch": 1.3192809143644824,
+      "grad_norm": 0.6166926622390747,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 11485
+    },
+    {
+      "epoch": 1.3193957842743094,
+      "grad_norm": 0.5764713883399963,
+      "learning_rate": 0.0001,
+      "loss": 1.2562,
+      "step": 11486
+    },
+    {
+      "epoch": 1.3195106541841364,
+      "grad_norm": 0.5882161855697632,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 11487
+    },
+    {
+      "epoch": 1.3196255240939636,
+      "grad_norm": 0.5338685512542725,
+      "learning_rate": 0.0001,
+      "loss": 1.3349,
+      "step": 11488
+    },
+    {
+      "epoch": 1.3197403940037908,
+      "grad_norm": 0.6274675130844116,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 11489
+    },
+    {
+      "epoch": 1.3198552639136178,
+      "grad_norm": 0.6399273872375488,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 11490
+    },
+    {
+      "epoch": 1.3199701338234449,
+      "grad_norm": 0.5669889450073242,
+      "learning_rate": 0.0001,
+      "loss": 1.4638,
+      "step": 11491
+    },
+    {
+      "epoch": 1.320085003733272,
+      "grad_norm": 0.6101594567298889,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 11492
+    },
+    {
+      "epoch": 1.3201998736430993,
+      "grad_norm": 0.5899582505226135,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 11493
+    },
+    {
+      "epoch": 1.3203147435529263,
+      "grad_norm": 0.6240655779838562,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 11494
+    },
+    {
+      "epoch": 1.3204296134627533,
+      "grad_norm": 0.5994274616241455,
+      "learning_rate": 0.0001,
+      "loss": 1.6436,
+      "step": 11495
+    },
+    {
+      "epoch": 1.3205444833725806,
+      "grad_norm": 0.5786739587783813,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 11496
+    },
+    {
+      "epoch": 1.3206593532824078,
+      "grad_norm": 0.5860279202461243,
+      "learning_rate": 0.0001,
+      "loss": 1.3348,
+      "step": 11497
+    },
+    {
+      "epoch": 1.3207742231922348,
+      "grad_norm": 0.5858715772628784,
+      "learning_rate": 0.0001,
+      "loss": 1.3238,
+      "step": 11498
+    },
+    {
+      "epoch": 1.3208890931020618,
+      "grad_norm": 0.6237417459487915,
+      "learning_rate": 0.0001,
+      "loss": 1.7704,
+      "step": 11499
+    },
+    {
+      "epoch": 1.321003963011889,
+      "grad_norm": 0.5686200857162476,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 11500
+    },
+    {
+      "epoch": 1.3211188329217163,
+      "grad_norm": 0.6311827301979065,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 11501
+    },
+    {
+      "epoch": 1.3212337028315433,
+      "grad_norm": 0.5898725986480713,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 11502
+    },
+    {
+      "epoch": 1.3213485727413703,
+      "grad_norm": 0.543509304523468,
+      "learning_rate": 0.0001,
+      "loss": 1.4023,
+      "step": 11503
+    },
+    {
+      "epoch": 1.3214634426511975,
+      "grad_norm": 0.6843796372413635,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 11504
+    },
+    {
+      "epoch": 1.3215783125610248,
+      "grad_norm": 0.5604381561279297,
+      "learning_rate": 0.0001,
+      "loss": 1.4211,
+      "step": 11505
+    },
+    {
+      "epoch": 1.3216931824708518,
+      "grad_norm": 0.6264784336090088,
+      "learning_rate": 0.0001,
+      "loss": 1.6592,
+      "step": 11506
+    },
+    {
+      "epoch": 1.3218080523806788,
+      "grad_norm": 0.5538937449455261,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 11507
+    },
+    {
+      "epoch": 1.321922922290506,
+      "grad_norm": 0.5639132857322693,
+      "learning_rate": 0.0001,
+      "loss": 1.3514,
+      "step": 11508
+    },
+    {
+      "epoch": 1.3220377922003332,
+      "grad_norm": 0.5447933077812195,
+      "learning_rate": 0.0001,
+      "loss": 1.3798,
+      "step": 11509
+    },
+    {
+      "epoch": 1.3221526621101602,
+      "grad_norm": 0.5847629308700562,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 11510
+    },
+    {
+      "epoch": 1.3222675320199873,
+      "grad_norm": 0.5697048306465149,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 11511
+    },
+    {
+      "epoch": 1.3223824019298145,
+      "grad_norm": 0.6301547884941101,
+      "learning_rate": 0.0001,
+      "loss": 1.4724,
+      "step": 11512
+    },
+    {
+      "epoch": 1.3224972718396417,
+      "grad_norm": 0.5512596368789673,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 11513
+    },
+    {
+      "epoch": 1.3226121417494687,
+      "grad_norm": 0.5693661570549011,
+      "learning_rate": 0.0001,
+      "loss": 1.3942,
+      "step": 11514
+    },
+    {
+      "epoch": 1.3227270116592957,
+      "grad_norm": 0.5459357500076294,
+      "learning_rate": 0.0001,
+      "loss": 1.4969,
+      "step": 11515
+    },
+    {
+      "epoch": 1.322841881569123,
+      "grad_norm": 0.5842916965484619,
+      "learning_rate": 0.0001,
+      "loss": 1.4668,
+      "step": 11516
+    },
+    {
+      "epoch": 1.3229567514789502,
+      "grad_norm": 0.5588902235031128,
+      "learning_rate": 0.0001,
+      "loss": 1.397,
+      "step": 11517
+    },
+    {
+      "epoch": 1.3230716213887772,
+      "grad_norm": 0.5641937255859375,
+      "learning_rate": 0.0001,
+      "loss": 1.2773,
+      "step": 11518
+    },
+    {
+      "epoch": 1.3231864912986042,
+      "grad_norm": 0.5515016317367554,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 11519
+    },
+    {
+      "epoch": 1.3233013612084314,
+      "grad_norm": 0.6019710898399353,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 11520
+    },
+    {
+      "epoch": 1.3234162311182587,
+      "grad_norm": 0.6003029942512512,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 11521
+    },
+    {
+      "epoch": 1.3235311010280857,
+      "grad_norm": 0.5781972408294678,
+      "learning_rate": 0.0001,
+      "loss": 1.4488,
+      "step": 11522
+    },
+    {
+      "epoch": 1.3236459709379127,
+      "grad_norm": 0.597213089466095,
+      "learning_rate": 0.0001,
+      "loss": 1.4833,
+      "step": 11523
+    },
+    {
+      "epoch": 1.32376084084774,
+      "grad_norm": 0.6153642535209656,
+      "learning_rate": 0.0001,
+      "loss": 1.5441,
+      "step": 11524
+    },
+    {
+      "epoch": 1.3238757107575672,
+      "grad_norm": 0.5358846783638,
+      "learning_rate": 0.0001,
+      "loss": 1.2119,
+      "step": 11525
+    },
+    {
+      "epoch": 1.3239905806673942,
+      "grad_norm": 0.6239076852798462,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 11526
+    },
+    {
+      "epoch": 1.3241054505772212,
+      "grad_norm": 0.5684323906898499,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 11527
+    },
+    {
+      "epoch": 1.3242203204870484,
+      "grad_norm": 0.5422741174697876,
+      "learning_rate": 0.0001,
+      "loss": 1.3541,
+      "step": 11528
+    },
+    {
+      "epoch": 1.3243351903968756,
+      "grad_norm": 0.5979777574539185,
+      "learning_rate": 0.0001,
+      "loss": 1.5381,
+      "step": 11529
+    },
+    {
+      "epoch": 1.3244500603067026,
+      "grad_norm": 0.5492232441902161,
+      "learning_rate": 0.0001,
+      "loss": 1.2141,
+      "step": 11530
+    },
+    {
+      "epoch": 1.3245649302165297,
+      "grad_norm": 0.5755845904350281,
+      "learning_rate": 0.0001,
+      "loss": 1.4904,
+      "step": 11531
+    },
+    {
+      "epoch": 1.3246798001263569,
+      "grad_norm": 0.5299710035324097,
+      "learning_rate": 0.0001,
+      "loss": 1.3358,
+      "step": 11532
+    },
+    {
+      "epoch": 1.3247946700361841,
+      "grad_norm": 0.5345584154129028,
+      "learning_rate": 0.0001,
+      "loss": 1.4269,
+      "step": 11533
+    },
+    {
+      "epoch": 1.3249095399460111,
+      "grad_norm": 0.5578746199607849,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 11534
+    },
+    {
+      "epoch": 1.3250244098558381,
+      "grad_norm": 0.6171008944511414,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 11535
+    },
+    {
+      "epoch": 1.3251392797656654,
+      "grad_norm": 0.5888642072677612,
+      "learning_rate": 0.0001,
+      "loss": 1.4443,
+      "step": 11536
+    },
+    {
+      "epoch": 1.3252541496754926,
+      "grad_norm": 0.5797832012176514,
+      "learning_rate": 0.0001,
+      "loss": 1.5776,
+      "step": 11537
+    },
+    {
+      "epoch": 1.3253690195853196,
+      "grad_norm": 0.5411096215248108,
+      "learning_rate": 0.0001,
+      "loss": 1.2369,
+      "step": 11538
+    },
+    {
+      "epoch": 1.3254838894951466,
+      "grad_norm": 0.5486440062522888,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 11539
+    },
+    {
+      "epoch": 1.3255987594049738,
+      "grad_norm": 0.6069177985191345,
+      "learning_rate": 0.0001,
+      "loss": 1.3719,
+      "step": 11540
+    },
+    {
+      "epoch": 1.325713629314801,
+      "grad_norm": 0.5846525430679321,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 11541
+    },
+    {
+      "epoch": 1.325828499224628,
+      "grad_norm": 0.5994956493377686,
+      "learning_rate": 0.0001,
+      "loss": 1.6159,
+      "step": 11542
+    },
+    {
+      "epoch": 1.325943369134455,
+      "grad_norm": 0.6577488780021667,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 11543
+    },
+    {
+      "epoch": 1.3260582390442823,
+      "grad_norm": 0.5551082491874695,
+      "learning_rate": 0.0001,
+      "loss": 1.4653,
+      "step": 11544
+    },
+    {
+      "epoch": 1.3261731089541096,
+      "grad_norm": 0.7000524401664734,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 11545
+    },
+    {
+      "epoch": 1.3262879788639366,
+      "grad_norm": 0.5763112902641296,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 11546
+    },
+    {
+      "epoch": 1.3264028487737636,
+      "grad_norm": 0.6176822185516357,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 11547
+    },
+    {
+      "epoch": 1.3265177186835908,
+      "grad_norm": 0.5590544939041138,
+      "learning_rate": 0.0001,
+      "loss": 1.3263,
+      "step": 11548
+    },
+    {
+      "epoch": 1.326632588593418,
+      "grad_norm": 0.5652593374252319,
+      "learning_rate": 0.0001,
+      "loss": 1.3331,
+      "step": 11549
+    },
+    {
+      "epoch": 1.326747458503245,
+      "grad_norm": 0.657227635383606,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 11550
+    },
+    {
+      "epoch": 1.3268623284130723,
+      "grad_norm": 0.62965327501297,
+      "learning_rate": 0.0001,
+      "loss": 1.4074,
+      "step": 11551
+    },
+    {
+      "epoch": 1.3269771983228993,
+      "grad_norm": 0.5422511100769043,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 11552
+    },
+    {
+      "epoch": 1.3270920682327265,
+      "grad_norm": 0.5476226806640625,
+      "learning_rate": 0.0001,
+      "loss": 1.4102,
+      "step": 11553
+    },
+    {
+      "epoch": 1.3272069381425535,
+      "grad_norm": 0.6080157160758972,
+      "learning_rate": 0.0001,
+      "loss": 1.4651,
+      "step": 11554
+    },
+    {
+      "epoch": 1.3273218080523808,
+      "grad_norm": 0.6048029661178589,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 11555
+    },
+    {
+      "epoch": 1.3274366779622078,
+      "grad_norm": 0.5942696332931519,
+      "learning_rate": 0.0001,
+      "loss": 1.4198,
+      "step": 11556
+    },
+    {
+      "epoch": 1.327551547872035,
+      "grad_norm": 0.6302083134651184,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 11557
+    },
+    {
+      "epoch": 1.327666417781862,
+      "grad_norm": 0.6311014294624329,
+      "learning_rate": 0.0001,
+      "loss": 1.3774,
+      "step": 11558
+    },
+    {
+      "epoch": 1.3277812876916892,
+      "grad_norm": 0.6695850491523743,
+      "learning_rate": 0.0001,
+      "loss": 1.7421,
+      "step": 11559
+    },
+    {
+      "epoch": 1.3278961576015162,
+      "grad_norm": 0.5893605947494507,
+      "learning_rate": 0.0001,
+      "loss": 1.5078,
+      "step": 11560
+    },
+    {
+      "epoch": 1.3280110275113435,
+      "grad_norm": 0.7168806195259094,
+      "learning_rate": 0.0001,
+      "loss": 1.4677,
+      "step": 11561
+    },
+    {
+      "epoch": 1.3281258974211705,
+      "grad_norm": 0.5717088580131531,
+      "learning_rate": 0.0001,
+      "loss": 1.4689,
+      "step": 11562
+    },
+    {
+      "epoch": 1.3282407673309977,
+      "grad_norm": 0.5938601493835449,
+      "learning_rate": 0.0001,
+      "loss": 1.489,
+      "step": 11563
+    },
+    {
+      "epoch": 1.3283556372408247,
+      "grad_norm": 0.6174132227897644,
+      "learning_rate": 0.0001,
+      "loss": 1.4657,
+      "step": 11564
+    },
+    {
+      "epoch": 1.328470507150652,
+      "grad_norm": 0.6349033713340759,
+      "learning_rate": 0.0001,
+      "loss": 1.4751,
+      "step": 11565
+    },
+    {
+      "epoch": 1.328585377060479,
+      "grad_norm": 0.5308579206466675,
+      "learning_rate": 0.0001,
+      "loss": 1.4448,
+      "step": 11566
+    },
+    {
+      "epoch": 1.3287002469703062,
+      "grad_norm": 0.5638580322265625,
+      "learning_rate": 0.0001,
+      "loss": 1.5103,
+      "step": 11567
+    },
+    {
+      "epoch": 1.3288151168801332,
+      "grad_norm": 0.5843721628189087,
+      "learning_rate": 0.0001,
+      "loss": 1.4057,
+      "step": 11568
+    },
+    {
+      "epoch": 1.3289299867899604,
+      "grad_norm": 0.5811317563056946,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 11569
+    },
+    {
+      "epoch": 1.3290448566997874,
+      "grad_norm": 0.5734536051750183,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 11570
+    },
+    {
+      "epoch": 1.3291597266096147,
+      "grad_norm": 0.5713332295417786,
+      "learning_rate": 0.0001,
+      "loss": 1.3056,
+      "step": 11571
+    },
+    {
+      "epoch": 1.3292745965194417,
+      "grad_norm": 0.570878803730011,
+      "learning_rate": 0.0001,
+      "loss": 1.303,
+      "step": 11572
+    },
+    {
+      "epoch": 1.329389466429269,
+      "grad_norm": 0.5729346871376038,
+      "learning_rate": 0.0001,
+      "loss": 1.2979,
+      "step": 11573
+    },
+    {
+      "epoch": 1.329504336339096,
+      "grad_norm": 0.5870584845542908,
+      "learning_rate": 0.0001,
+      "loss": 1.4395,
+      "step": 11574
+    },
+    {
+      "epoch": 1.3296192062489232,
+      "grad_norm": 0.5481575131416321,
+      "learning_rate": 0.0001,
+      "loss": 1.3214,
+      "step": 11575
+    },
+    {
+      "epoch": 1.3297340761587502,
+      "grad_norm": 0.5822715759277344,
+      "learning_rate": 0.0001,
+      "loss": 1.3497,
+      "step": 11576
+    },
+    {
+      "epoch": 1.3298489460685774,
+      "grad_norm": 0.5590030550956726,
+      "learning_rate": 0.0001,
+      "loss": 1.3659,
+      "step": 11577
+    },
+    {
+      "epoch": 1.3299638159784044,
+      "grad_norm": 0.5739883184432983,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 11578
+    },
+    {
+      "epoch": 1.3300786858882316,
+      "grad_norm": 0.5733195543289185,
+      "learning_rate": 0.0001,
+      "loss": 1.6149,
+      "step": 11579
+    },
+    {
+      "epoch": 1.3301935557980586,
+      "grad_norm": 0.5605318546295166,
+      "learning_rate": 0.0001,
+      "loss": 1.3656,
+      "step": 11580
+    },
+    {
+      "epoch": 1.3303084257078859,
+      "grad_norm": 0.574635922908783,
+      "learning_rate": 0.0001,
+      "loss": 1.3818,
+      "step": 11581
+    },
+    {
+      "epoch": 1.3304232956177129,
+      "grad_norm": 0.5863053202629089,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 11582
+    },
+    {
+      "epoch": 1.3305381655275401,
+      "grad_norm": 0.6203168630599976,
+      "learning_rate": 0.0001,
+      "loss": 1.5662,
+      "step": 11583
+    },
+    {
+      "epoch": 1.3306530354373671,
+      "grad_norm": 0.5553234815597534,
+      "learning_rate": 0.0001,
+      "loss": 1.2708,
+      "step": 11584
+    },
+    {
+      "epoch": 1.3307679053471944,
+      "grad_norm": 0.5634999871253967,
+      "learning_rate": 0.0001,
+      "loss": 1.2638,
+      "step": 11585
+    },
+    {
+      "epoch": 1.3308827752570214,
+      "grad_norm": 0.5584697723388672,
+      "learning_rate": 0.0001,
+      "loss": 1.3846,
+      "step": 11586
+    },
+    {
+      "epoch": 1.3309976451668486,
+      "grad_norm": 0.5859621167182922,
+      "learning_rate": 0.0001,
+      "loss": 1.4662,
+      "step": 11587
+    },
+    {
+      "epoch": 1.3311125150766756,
+      "grad_norm": 0.5909789800643921,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 11588
+    },
+    {
+      "epoch": 1.3312273849865028,
+      "grad_norm": 0.674757182598114,
+      "learning_rate": 0.0001,
+      "loss": 1.6189,
+      "step": 11589
+    },
+    {
+      "epoch": 1.3313422548963298,
+      "grad_norm": 0.5663519501686096,
+      "learning_rate": 0.0001,
+      "loss": 1.3768,
+      "step": 11590
+    },
+    {
+      "epoch": 1.331457124806157,
+      "grad_norm": 0.5656293630599976,
+      "learning_rate": 0.0001,
+      "loss": 1.3759,
+      "step": 11591
+    },
+    {
+      "epoch": 1.331571994715984,
+      "grad_norm": 0.652201771736145,
+      "learning_rate": 0.0001,
+      "loss": 1.6337,
+      "step": 11592
+    },
+    {
+      "epoch": 1.3316868646258113,
+      "grad_norm": 0.6089327931404114,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 11593
+    },
+    {
+      "epoch": 1.3318017345356383,
+      "grad_norm": 0.5640665888786316,
+      "learning_rate": 0.0001,
+      "loss": 1.4744,
+      "step": 11594
+    },
+    {
+      "epoch": 1.3319166044454656,
+      "grad_norm": 0.5926685333251953,
+      "learning_rate": 0.0001,
+      "loss": 1.6095,
+      "step": 11595
+    },
+    {
+      "epoch": 1.3320314743552926,
+      "grad_norm": 0.5491368770599365,
+      "learning_rate": 0.0001,
+      "loss": 1.4922,
+      "step": 11596
+    },
+    {
+      "epoch": 1.3321463442651198,
+      "grad_norm": 0.6609952449798584,
+      "learning_rate": 0.0001,
+      "loss": 1.6714,
+      "step": 11597
+    },
+    {
+      "epoch": 1.3322612141749468,
+      "grad_norm": 0.6160357594490051,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 11598
+    },
+    {
+      "epoch": 1.332376084084774,
+      "grad_norm": 0.5720334649085999,
+      "learning_rate": 0.0001,
+      "loss": 1.4616,
+      "step": 11599
+    },
+    {
+      "epoch": 1.332490953994601,
+      "grad_norm": 0.6039928197860718,
+      "learning_rate": 0.0001,
+      "loss": 1.5472,
+      "step": 11600
+    },
+    {
+      "epoch": 1.3326058239044283,
+      "grad_norm": 0.6017964482307434,
+      "learning_rate": 0.0001,
+      "loss": 1.3536,
+      "step": 11601
+    },
+    {
+      "epoch": 1.3327206938142553,
+      "grad_norm": 0.5808287858963013,
+      "learning_rate": 0.0001,
+      "loss": 1.5072,
+      "step": 11602
+    },
+    {
+      "epoch": 1.3328355637240825,
+      "grad_norm": 0.6469992995262146,
+      "learning_rate": 0.0001,
+      "loss": 1.6845,
+      "step": 11603
+    },
+    {
+      "epoch": 1.3329504336339095,
+      "grad_norm": 0.6133192777633667,
+      "learning_rate": 0.0001,
+      "loss": 1.385,
+      "step": 11604
+    },
+    {
+      "epoch": 1.3330653035437368,
+      "grad_norm": 0.575235903263092,
+      "learning_rate": 0.0001,
+      "loss": 1.465,
+      "step": 11605
+    },
+    {
+      "epoch": 1.3331801734535638,
+      "grad_norm": 0.5712215900421143,
+      "learning_rate": 0.0001,
+      "loss": 1.3618,
+      "step": 11606
+    },
+    {
+      "epoch": 1.333295043363391,
+      "grad_norm": 0.5588753819465637,
+      "learning_rate": 0.0001,
+      "loss": 1.4052,
+      "step": 11607
+    },
+    {
+      "epoch": 1.333409913273218,
+      "grad_norm": 0.5291929244995117,
+      "learning_rate": 0.0001,
+      "loss": 1.3567,
+      "step": 11608
+    },
+    {
+      "epoch": 1.3335247831830452,
+      "grad_norm": 0.5625868439674377,
+      "learning_rate": 0.0001,
+      "loss": 1.3899,
+      "step": 11609
+    },
+    {
+      "epoch": 1.3336396530928722,
+      "grad_norm": 0.5754292607307434,
+      "learning_rate": 0.0001,
+      "loss": 1.3608,
+      "step": 11610
+    },
+    {
+      "epoch": 1.3337545230026995,
+      "grad_norm": 0.6050917506217957,
+      "learning_rate": 0.0001,
+      "loss": 1.4607,
+      "step": 11611
+    },
+    {
+      "epoch": 1.3338693929125265,
+      "grad_norm": 0.5786774754524231,
+      "learning_rate": 0.0001,
+      "loss": 1.6412,
+      "step": 11612
+    },
+    {
+      "epoch": 1.3339842628223537,
+      "grad_norm": 0.6983840465545654,
+      "learning_rate": 0.0001,
+      "loss": 1.7368,
+      "step": 11613
+    },
+    {
+      "epoch": 1.3340991327321807,
+      "grad_norm": 0.5841389298439026,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 11614
+    },
+    {
+      "epoch": 1.334214002642008,
+      "grad_norm": 0.6447943449020386,
+      "learning_rate": 0.0001,
+      "loss": 1.6709,
+      "step": 11615
+    },
+    {
+      "epoch": 1.334328872551835,
+      "grad_norm": 0.5992538332939148,
+      "learning_rate": 0.0001,
+      "loss": 1.4572,
+      "step": 11616
+    },
+    {
+      "epoch": 1.3344437424616622,
+      "grad_norm": 0.6163121461868286,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 11617
+    },
+    {
+      "epoch": 1.3345586123714892,
+      "grad_norm": 0.5886964797973633,
+      "learning_rate": 0.0001,
+      "loss": 1.3598,
+      "step": 11618
+    },
+    {
+      "epoch": 1.3346734822813164,
+      "grad_norm": 0.5401976704597473,
+      "learning_rate": 0.0001,
+      "loss": 1.3276,
+      "step": 11619
+    },
+    {
+      "epoch": 1.3347883521911434,
+      "grad_norm": 0.543461799621582,
+      "learning_rate": 0.0001,
+      "loss": 1.4984,
+      "step": 11620
+    },
+    {
+      "epoch": 1.3349032221009707,
+      "grad_norm": 0.6902713179588318,
+      "learning_rate": 0.0001,
+      "loss": 1.5798,
+      "step": 11621
+    },
+    {
+      "epoch": 1.335018092010798,
+      "grad_norm": 0.6112170219421387,
+      "learning_rate": 0.0001,
+      "loss": 1.4523,
+      "step": 11622
+    },
+    {
+      "epoch": 1.335132961920625,
+      "grad_norm": 0.6658702492713928,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 11623
+    },
+    {
+      "epoch": 1.335247831830452,
+      "grad_norm": 0.5938554406166077,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 11624
+    },
+    {
+      "epoch": 1.3353627017402792,
+      "grad_norm": 0.5592337250709534,
+      "learning_rate": 0.0001,
+      "loss": 1.5078,
+      "step": 11625
+    },
+    {
+      "epoch": 1.3354775716501064,
+      "grad_norm": 0.5694896578788757,
+      "learning_rate": 0.0001,
+      "loss": 1.2984,
+      "step": 11626
+    },
+    {
+      "epoch": 1.3355924415599334,
+      "grad_norm": 0.6331267356872559,
+      "learning_rate": 0.0001,
+      "loss": 1.4092,
+      "step": 11627
+    },
+    {
+      "epoch": 1.3357073114697604,
+      "grad_norm": 0.5441679358482361,
+      "learning_rate": 0.0001,
+      "loss": 1.4533,
+      "step": 11628
+    },
+    {
+      "epoch": 1.3358221813795876,
+      "grad_norm": 0.5393032431602478,
+      "learning_rate": 0.0001,
+      "loss": 1.4713,
+      "step": 11629
+    },
+    {
+      "epoch": 1.3359370512894149,
+      "grad_norm": 0.5676888227462769,
+      "learning_rate": 0.0001,
+      "loss": 1.428,
+      "step": 11630
+    },
+    {
+      "epoch": 1.3360519211992419,
+      "grad_norm": 0.6380576491355896,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 11631
+    },
+    {
+      "epoch": 1.3361667911090689,
+      "grad_norm": 0.6264674663543701,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 11632
+    },
+    {
+      "epoch": 1.3362816610188961,
+      "grad_norm": 0.660614550113678,
+      "learning_rate": 0.0001,
+      "loss": 1.3874,
+      "step": 11633
+    },
+    {
+      "epoch": 1.3363965309287233,
+      "grad_norm": 0.5967418551445007,
+      "learning_rate": 0.0001,
+      "loss": 1.4457,
+      "step": 11634
+    },
+    {
+      "epoch": 1.3365114008385504,
+      "grad_norm": 0.5896640419960022,
+      "learning_rate": 0.0001,
+      "loss": 1.3916,
+      "step": 11635
+    },
+    {
+      "epoch": 1.3366262707483774,
+      "grad_norm": 0.5509446859359741,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 11636
+    },
+    {
+      "epoch": 1.3367411406582046,
+      "grad_norm": 0.5520437955856323,
+      "learning_rate": 0.0001,
+      "loss": 1.4022,
+      "step": 11637
+    },
+    {
+      "epoch": 1.3368560105680318,
+      "grad_norm": 0.6015188694000244,
+      "learning_rate": 0.0001,
+      "loss": 1.4852,
+      "step": 11638
+    },
+    {
+      "epoch": 1.3369708804778588,
+      "grad_norm": 0.6142217516899109,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 11639
+    },
+    {
+      "epoch": 1.3370857503876858,
+      "grad_norm": 0.5661596059799194,
+      "learning_rate": 0.0001,
+      "loss": 1.3181,
+      "step": 11640
+    },
+    {
+      "epoch": 1.337200620297513,
+      "grad_norm": 0.5863280296325684,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 11641
+    },
+    {
+      "epoch": 1.3373154902073403,
+      "grad_norm": 0.6303154826164246,
+      "learning_rate": 0.0001,
+      "loss": 1.437,
+      "step": 11642
+    },
+    {
+      "epoch": 1.3374303601171673,
+      "grad_norm": 0.6611382961273193,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 11643
+    },
+    {
+      "epoch": 1.3375452300269943,
+      "grad_norm": 0.5862720012664795,
+      "learning_rate": 0.0001,
+      "loss": 1.4288,
+      "step": 11644
+    },
+    {
+      "epoch": 1.3376600999368216,
+      "grad_norm": 0.6255212426185608,
+      "learning_rate": 0.0001,
+      "loss": 1.513,
+      "step": 11645
+    },
+    {
+      "epoch": 1.3377749698466488,
+      "grad_norm": 0.6159241199493408,
+      "learning_rate": 0.0001,
+      "loss": 1.5834,
+      "step": 11646
+    },
+    {
+      "epoch": 1.3378898397564758,
+      "grad_norm": 0.6486058235168457,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 11647
+    },
+    {
+      "epoch": 1.3380047096663028,
+      "grad_norm": 0.577648401260376,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 11648
+    },
+    {
+      "epoch": 1.33811957957613,
+      "grad_norm": 0.554535448551178,
+      "learning_rate": 0.0001,
+      "loss": 1.328,
+      "step": 11649
+    },
+    {
+      "epoch": 1.3382344494859573,
+      "grad_norm": 0.6508966088294983,
+      "learning_rate": 0.0001,
+      "loss": 1.475,
+      "step": 11650
+    },
+    {
+      "epoch": 1.3383493193957843,
+      "grad_norm": 0.558700442314148,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 11651
+    },
+    {
+      "epoch": 1.3384641893056113,
+      "grad_norm": 0.5516871809959412,
+      "learning_rate": 0.0001,
+      "loss": 1.3675,
+      "step": 11652
+    },
+    {
+      "epoch": 1.3385790592154385,
+      "grad_norm": 0.5952174663543701,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 11653
+    },
+    {
+      "epoch": 1.3386939291252657,
+      "grad_norm": 0.5594809651374817,
+      "learning_rate": 0.0001,
+      "loss": 1.2648,
+      "step": 11654
+    },
+    {
+      "epoch": 1.3388087990350928,
+      "grad_norm": 0.6401504278182983,
+      "learning_rate": 0.0001,
+      "loss": 1.6669,
+      "step": 11655
+    },
+    {
+      "epoch": 1.3389236689449198,
+      "grad_norm": 0.5477997660636902,
+      "learning_rate": 0.0001,
+      "loss": 1.3875,
+      "step": 11656
+    },
+    {
+      "epoch": 1.339038538854747,
+      "grad_norm": 0.6509205102920532,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 11657
+    },
+    {
+      "epoch": 1.3391534087645742,
+      "grad_norm": 0.6122426986694336,
+      "learning_rate": 0.0001,
+      "loss": 1.4431,
+      "step": 11658
+    },
+    {
+      "epoch": 1.3392682786744012,
+      "grad_norm": 0.6294790506362915,
+      "learning_rate": 0.0001,
+      "loss": 1.3445,
+      "step": 11659
+    },
+    {
+      "epoch": 1.3393831485842282,
+      "grad_norm": 0.6401886940002441,
+      "learning_rate": 0.0001,
+      "loss": 1.5471,
+      "step": 11660
+    },
+    {
+      "epoch": 1.3394980184940555,
+      "grad_norm": 0.6541320085525513,
+      "learning_rate": 0.0001,
+      "loss": 1.7049,
+      "step": 11661
+    },
+    {
+      "epoch": 1.3396128884038827,
+      "grad_norm": 0.6253013014793396,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 11662
+    },
+    {
+      "epoch": 1.3397277583137097,
+      "grad_norm": 0.5853136777877808,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 11663
+    },
+    {
+      "epoch": 1.3398426282235367,
+      "grad_norm": 0.638195276260376,
+      "learning_rate": 0.0001,
+      "loss": 1.4023,
+      "step": 11664
+    },
+    {
+      "epoch": 1.339957498133364,
+      "grad_norm": 0.5563250184059143,
+      "learning_rate": 0.0001,
+      "loss": 1.4894,
+      "step": 11665
+    },
+    {
+      "epoch": 1.3400723680431912,
+      "grad_norm": 0.6137783527374268,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 11666
+    },
+    {
+      "epoch": 1.3401872379530182,
+      "grad_norm": 0.5672292709350586,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 11667
+    },
+    {
+      "epoch": 1.3403021078628452,
+      "grad_norm": 0.6190178394317627,
+      "learning_rate": 0.0001,
+      "loss": 1.3538,
+      "step": 11668
+    },
+    {
+      "epoch": 1.3404169777726724,
+      "grad_norm": 0.5791795253753662,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 11669
+    },
+    {
+      "epoch": 1.3405318476824997,
+      "grad_norm": 0.5504187345504761,
+      "learning_rate": 0.0001,
+      "loss": 1.5983,
+      "step": 11670
+    },
+    {
+      "epoch": 1.3406467175923267,
+      "grad_norm": 0.5945659875869751,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 11671
+    },
+    {
+      "epoch": 1.3407615875021537,
+      "grad_norm": 0.6134802103042603,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 11672
+    },
+    {
+      "epoch": 1.340876457411981,
+      "grad_norm": 0.5988081097602844,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 11673
+    },
+    {
+      "epoch": 1.3409913273218081,
+      "grad_norm": 0.5827198028564453,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 11674
+    },
+    {
+      "epoch": 1.3411061972316352,
+      "grad_norm": 0.5701010227203369,
+      "learning_rate": 0.0001,
+      "loss": 1.3642,
+      "step": 11675
+    },
+    {
+      "epoch": 1.3412210671414622,
+      "grad_norm": 0.6469464898109436,
+      "learning_rate": 0.0001,
+      "loss": 1.5485,
+      "step": 11676
+    },
+    {
+      "epoch": 1.3413359370512894,
+      "grad_norm": 0.5903701782226562,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 11677
+    },
+    {
+      "epoch": 1.3414508069611166,
+      "grad_norm": 0.5560135245323181,
+      "learning_rate": 0.0001,
+      "loss": 1.4042,
+      "step": 11678
+    },
+    {
+      "epoch": 1.3415656768709436,
+      "grad_norm": 0.5839285850524902,
+      "learning_rate": 0.0001,
+      "loss": 1.5252,
+      "step": 11679
+    },
+    {
+      "epoch": 1.3416805467807706,
+      "grad_norm": 0.6008885502815247,
+      "learning_rate": 0.0001,
+      "loss": 1.7412,
+      "step": 11680
+    },
+    {
+      "epoch": 1.3417954166905979,
+      "grad_norm": 0.542387843132019,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 11681
+    },
+    {
+      "epoch": 1.341910286600425,
+      "grad_norm": 0.5818324685096741,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 11682
+    },
+    {
+      "epoch": 1.3420251565102521,
+      "grad_norm": 0.5558339953422546,
+      "learning_rate": 0.0001,
+      "loss": 1.3248,
+      "step": 11683
+    },
+    {
+      "epoch": 1.3421400264200791,
+      "grad_norm": 0.5538620948791504,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 11684
+    },
+    {
+      "epoch": 1.3422548963299064,
+      "grad_norm": 0.5701099038124084,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 11685
+    },
+    {
+      "epoch": 1.3423697662397336,
+      "grad_norm": 0.5815318822860718,
+      "learning_rate": 0.0001,
+      "loss": 1.4229,
+      "step": 11686
+    },
+    {
+      "epoch": 1.3424846361495606,
+      "grad_norm": 0.5730632543563843,
+      "learning_rate": 0.0001,
+      "loss": 1.252,
+      "step": 11687
+    },
+    {
+      "epoch": 1.3425995060593878,
+      "grad_norm": 0.587175726890564,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 11688
+    },
+    {
+      "epoch": 1.3427143759692148,
+      "grad_norm": 0.6182733774185181,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 11689
+    },
+    {
+      "epoch": 1.342829245879042,
+      "grad_norm": 0.636389434337616,
+      "learning_rate": 0.0001,
+      "loss": 1.4714,
+      "step": 11690
+    },
+    {
+      "epoch": 1.342944115788869,
+      "grad_norm": 0.5567188858985901,
+      "learning_rate": 0.0001,
+      "loss": 1.2964,
+      "step": 11691
+    },
+    {
+      "epoch": 1.3430589856986963,
+      "grad_norm": 0.7240939140319824,
+      "learning_rate": 0.0001,
+      "loss": 1.4965,
+      "step": 11692
+    },
+    {
+      "epoch": 1.3431738556085233,
+      "grad_norm": 0.5807350277900696,
+      "learning_rate": 0.0001,
+      "loss": 1.4122,
+      "step": 11693
+    },
+    {
+      "epoch": 1.3432887255183505,
+      "grad_norm": 0.5738909840583801,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 11694
+    },
+    {
+      "epoch": 1.3434035954281776,
+      "grad_norm": 0.6102243661880493,
+      "learning_rate": 0.0001,
+      "loss": 1.501,
+      "step": 11695
+    },
+    {
+      "epoch": 1.3435184653380048,
+      "grad_norm": 0.5855363607406616,
+      "learning_rate": 0.0001,
+      "loss": 1.4738,
+      "step": 11696
+    },
+    {
+      "epoch": 1.3436333352478318,
+      "grad_norm": 0.6184929609298706,
+      "learning_rate": 0.0001,
+      "loss": 1.3653,
+      "step": 11697
+    },
+    {
+      "epoch": 1.343748205157659,
+      "grad_norm": 0.6024360060691833,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 11698
+    },
+    {
+      "epoch": 1.343863075067486,
+      "grad_norm": 0.6664621233940125,
+      "learning_rate": 0.0001,
+      "loss": 1.4674,
+      "step": 11699
+    },
+    {
+      "epoch": 1.3439779449773133,
+      "grad_norm": 0.59235018491745,
+      "learning_rate": 0.0001,
+      "loss": 1.4277,
+      "step": 11700
+    },
+    {
+      "epoch": 1.3440928148871403,
+      "grad_norm": 0.5766885876655579,
+      "learning_rate": 0.0001,
+      "loss": 1.3778,
+      "step": 11701
+    },
+    {
+      "epoch": 1.3442076847969675,
+      "grad_norm": 0.6111820936203003,
+      "learning_rate": 0.0001,
+      "loss": 1.5498,
+      "step": 11702
+    },
+    {
+      "epoch": 1.3443225547067945,
+      "grad_norm": 0.5713093280792236,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 11703
+    },
+    {
+      "epoch": 1.3444374246166217,
+      "grad_norm": 0.5723089575767517,
+      "learning_rate": 0.0001,
+      "loss": 1.4658,
+      "step": 11704
+    },
+    {
+      "epoch": 1.3445522945264488,
+      "grad_norm": 0.5831340551376343,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 11705
+    },
+    {
+      "epoch": 1.344667164436276,
+      "grad_norm": 0.5750026702880859,
+      "learning_rate": 0.0001,
+      "loss": 1.4922,
+      "step": 11706
+    },
+    {
+      "epoch": 1.344782034346103,
+      "grad_norm": 0.6101966500282288,
+      "learning_rate": 0.0001,
+      "loss": 1.3965,
+      "step": 11707
+    },
+    {
+      "epoch": 1.3448969042559302,
+      "grad_norm": 0.5940831303596497,
+      "learning_rate": 0.0001,
+      "loss": 1.3699,
+      "step": 11708
+    },
+    {
+      "epoch": 1.3450117741657572,
+      "grad_norm": 0.5816580057144165,
+      "learning_rate": 0.0001,
+      "loss": 1.2569,
+      "step": 11709
+    },
+    {
+      "epoch": 1.3451266440755845,
+      "grad_norm": 0.651119589805603,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 11710
+    },
+    {
+      "epoch": 1.3452415139854115,
+      "grad_norm": 0.5532911419868469,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 11711
+    },
+    {
+      "epoch": 1.3453563838952387,
+      "grad_norm": 0.5616252422332764,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 11712
+    },
+    {
+      "epoch": 1.3454712538050657,
+      "grad_norm": 0.6075287461280823,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 11713
+    },
+    {
+      "epoch": 1.345586123714893,
+      "grad_norm": 0.5586172938346863,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 11714
+    },
+    {
+      "epoch": 1.34570099362472,
+      "grad_norm": 0.6126081943511963,
+      "learning_rate": 0.0001,
+      "loss": 1.3231,
+      "step": 11715
+    },
+    {
+      "epoch": 1.3458158635345472,
+      "grad_norm": 0.619485080242157,
+      "learning_rate": 0.0001,
+      "loss": 1.5135,
+      "step": 11716
+    },
+    {
+      "epoch": 1.3459307334443742,
+      "grad_norm": 0.5817553997039795,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 11717
+    },
+    {
+      "epoch": 1.3460456033542014,
+      "grad_norm": 0.6091306209564209,
+      "learning_rate": 0.0001,
+      "loss": 1.1376,
+      "step": 11718
+    },
+    {
+      "epoch": 1.3461604732640284,
+      "grad_norm": 0.6346269845962524,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 11719
+    },
+    {
+      "epoch": 1.3462753431738557,
+      "grad_norm": 0.6019834876060486,
+      "learning_rate": 0.0001,
+      "loss": 1.1867,
+      "step": 11720
+    },
+    {
+      "epoch": 1.3463902130836827,
+      "grad_norm": 0.5959239602088928,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 11721
+    },
+    {
+      "epoch": 1.34650508299351,
+      "grad_norm": 0.5814604759216309,
+      "learning_rate": 0.0001,
+      "loss": 1.4918,
+      "step": 11722
+    },
+    {
+      "epoch": 1.346619952903337,
+      "grad_norm": 0.5841962695121765,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 11723
+    },
+    {
+      "epoch": 1.3467348228131641,
+      "grad_norm": 0.626183032989502,
+      "learning_rate": 0.0001,
+      "loss": 1.4611,
+      "step": 11724
+    },
+    {
+      "epoch": 1.3468496927229912,
+      "grad_norm": 0.5390433073043823,
+      "learning_rate": 0.0001,
+      "loss": 1.5152,
+      "step": 11725
+    },
+    {
+      "epoch": 1.3469645626328184,
+      "grad_norm": 0.6256446242332458,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 11726
+    },
+    {
+      "epoch": 1.3470794325426454,
+      "grad_norm": 0.5755500793457031,
+      "learning_rate": 0.0001,
+      "loss": 1.4643,
+      "step": 11727
+    },
+    {
+      "epoch": 1.3471943024524726,
+      "grad_norm": 0.627149224281311,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 11728
+    },
+    {
+      "epoch": 1.3473091723622996,
+      "grad_norm": 0.5608257055282593,
+      "learning_rate": 0.0001,
+      "loss": 1.4637,
+      "step": 11729
+    },
+    {
+      "epoch": 1.3474240422721269,
+      "grad_norm": 0.5481683611869812,
+      "learning_rate": 0.0001,
+      "loss": 1.3497,
+      "step": 11730
+    },
+    {
+      "epoch": 1.3475389121819539,
+      "grad_norm": 0.5451379418373108,
+      "learning_rate": 0.0001,
+      "loss": 1.4324,
+      "step": 11731
+    },
+    {
+      "epoch": 1.347653782091781,
+      "grad_norm": 0.594100832939148,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 11732
+    },
+    {
+      "epoch": 1.3477686520016081,
+      "grad_norm": 0.5474432706832886,
+      "learning_rate": 0.0001,
+      "loss": 1.4132,
+      "step": 11733
+    },
+    {
+      "epoch": 1.3478835219114353,
+      "grad_norm": 0.5873172283172607,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 11734
+    },
+    {
+      "epoch": 1.3479983918212624,
+      "grad_norm": 0.5776464939117432,
+      "learning_rate": 0.0001,
+      "loss": 1.4039,
+      "step": 11735
+    },
+    {
+      "epoch": 1.3481132617310896,
+      "grad_norm": 0.5772274732589722,
+      "learning_rate": 0.0001,
+      "loss": 1.2844,
+      "step": 11736
+    },
+    {
+      "epoch": 1.3482281316409166,
+      "grad_norm": 0.6043945550918579,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 11737
+    },
+    {
+      "epoch": 1.3483430015507438,
+      "grad_norm": 0.5803393125534058,
+      "learning_rate": 0.0001,
+      "loss": 1.3332,
+      "step": 11738
+    },
+    {
+      "epoch": 1.3484578714605708,
+      "grad_norm": 0.5934081673622131,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 11739
+    },
+    {
+      "epoch": 1.348572741370398,
+      "grad_norm": 0.746346652507782,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 11740
+    },
+    {
+      "epoch": 1.348687611280225,
+      "grad_norm": 0.5849841833114624,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 11741
+    },
+    {
+      "epoch": 1.3488024811900523,
+      "grad_norm": 0.54823237657547,
+      "learning_rate": 0.0001,
+      "loss": 1.3391,
+      "step": 11742
+    },
+    {
+      "epoch": 1.3489173510998793,
+      "grad_norm": 0.5797605514526367,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 11743
+    },
+    {
+      "epoch": 1.3490322210097065,
+      "grad_norm": 0.6208351850509644,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 11744
+    },
+    {
+      "epoch": 1.3491470909195336,
+      "grad_norm": 0.599469780921936,
+      "learning_rate": 0.0001,
+      "loss": 1.4858,
+      "step": 11745
+    },
+    {
+      "epoch": 1.3492619608293608,
+      "grad_norm": 0.5936621427536011,
+      "learning_rate": 0.0001,
+      "loss": 1.4009,
+      "step": 11746
+    },
+    {
+      "epoch": 1.3493768307391878,
+      "grad_norm": 0.5813693404197693,
+      "learning_rate": 0.0001,
+      "loss": 1.4898,
+      "step": 11747
+    },
+    {
+      "epoch": 1.349491700649015,
+      "grad_norm": 0.686309278011322,
+      "learning_rate": 0.0001,
+      "loss": 1.4567,
+      "step": 11748
+    },
+    {
+      "epoch": 1.349606570558842,
+      "grad_norm": 0.5535640120506287,
+      "learning_rate": 0.0001,
+      "loss": 1.3404,
+      "step": 11749
+    },
+    {
+      "epoch": 1.3497214404686693,
+      "grad_norm": 0.6416987776756287,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 11750
+    },
+    {
+      "epoch": 1.3498363103784963,
+      "grad_norm": 0.5824869275093079,
+      "learning_rate": 0.0001,
+      "loss": 1.3471,
+      "step": 11751
+    },
+    {
+      "epoch": 1.3499511802883235,
+      "grad_norm": 0.5436984896659851,
+      "learning_rate": 0.0001,
+      "loss": 1.4902,
+      "step": 11752
+    },
+    {
+      "epoch": 1.3500660501981505,
+      "grad_norm": 0.5425761342048645,
+      "learning_rate": 0.0001,
+      "loss": 1.424,
+      "step": 11753
+    },
+    {
+      "epoch": 1.3501809201079777,
+      "grad_norm": 0.5270239114761353,
+      "learning_rate": 0.0001,
+      "loss": 1.1909,
+      "step": 11754
+    },
+    {
+      "epoch": 1.3502957900178048,
+      "grad_norm": 0.5511675477027893,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 11755
+    },
+    {
+      "epoch": 1.350410659927632,
+      "grad_norm": 0.5438040494918823,
+      "learning_rate": 0.0001,
+      "loss": 1.4739,
+      "step": 11756
+    },
+    {
+      "epoch": 1.350525529837459,
+      "grad_norm": 0.5267738699913025,
+      "learning_rate": 0.0001,
+      "loss": 1.4464,
+      "step": 11757
+    },
+    {
+      "epoch": 1.3506403997472862,
+      "grad_norm": 0.5484282374382019,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 11758
+    },
+    {
+      "epoch": 1.3507552696571135,
+      "grad_norm": 0.5937207341194153,
+      "learning_rate": 0.0001,
+      "loss": 1.3918,
+      "step": 11759
+    },
+    {
+      "epoch": 1.3508701395669405,
+      "grad_norm": 0.5717691779136658,
+      "learning_rate": 0.0001,
+      "loss": 1.392,
+      "step": 11760
+    },
+    {
+      "epoch": 1.3509850094767675,
+      "grad_norm": 0.5394260287284851,
+      "learning_rate": 0.0001,
+      "loss": 1.426,
+      "step": 11761
+    },
+    {
+      "epoch": 1.3510998793865947,
+      "grad_norm": 0.6104713082313538,
+      "learning_rate": 0.0001,
+      "loss": 1.4614,
+      "step": 11762
+    },
+    {
+      "epoch": 1.351214749296422,
+      "grad_norm": 0.5929358601570129,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 11763
+    },
+    {
+      "epoch": 1.351329619206249,
+      "grad_norm": 0.581203818321228,
+      "learning_rate": 0.0001,
+      "loss": 1.4168,
+      "step": 11764
+    },
+    {
+      "epoch": 1.351444489116076,
+      "grad_norm": 0.5509240031242371,
+      "learning_rate": 0.0001,
+      "loss": 1.3605,
+      "step": 11765
+    },
+    {
+      "epoch": 1.3515593590259032,
+      "grad_norm": 0.5640619993209839,
+      "learning_rate": 0.0001,
+      "loss": 1.4869,
+      "step": 11766
+    },
+    {
+      "epoch": 1.3516742289357304,
+      "grad_norm": 0.5916575193405151,
+      "learning_rate": 0.0001,
+      "loss": 1.4334,
+      "step": 11767
+    },
+    {
+      "epoch": 1.3517890988455574,
+      "grad_norm": 0.573228120803833,
+      "learning_rate": 0.0001,
+      "loss": 1.4595,
+      "step": 11768
+    },
+    {
+      "epoch": 1.3519039687553844,
+      "grad_norm": 0.662548840045929,
+      "learning_rate": 0.0001,
+      "loss": 1.4985,
+      "step": 11769
+    },
+    {
+      "epoch": 1.3520188386652117,
+      "grad_norm": 0.5832405090332031,
+      "learning_rate": 0.0001,
+      "loss": 1.3511,
+      "step": 11770
+    },
+    {
+      "epoch": 1.352133708575039,
+      "grad_norm": 0.6026256084442139,
+      "learning_rate": 0.0001,
+      "loss": 1.4318,
+      "step": 11771
+    },
+    {
+      "epoch": 1.352248578484866,
+      "grad_norm": 0.5654007792472839,
+      "learning_rate": 0.0001,
+      "loss": 1.3918,
+      "step": 11772
+    },
+    {
+      "epoch": 1.352363448394693,
+      "grad_norm": 0.6701568961143494,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 11773
+    },
+    {
+      "epoch": 1.3524783183045201,
+      "grad_norm": 0.5737709999084473,
+      "learning_rate": 0.0001,
+      "loss": 1.291,
+      "step": 11774
+    },
+    {
+      "epoch": 1.3525931882143474,
+      "grad_norm": 0.6998119950294495,
+      "learning_rate": 0.0001,
+      "loss": 1.7295,
+      "step": 11775
+    },
+    {
+      "epoch": 1.3527080581241744,
+      "grad_norm": 0.645850658416748,
+      "learning_rate": 0.0001,
+      "loss": 1.4265,
+      "step": 11776
+    },
+    {
+      "epoch": 1.3528229280340014,
+      "grad_norm": 0.5715954899787903,
+      "learning_rate": 0.0001,
+      "loss": 1.282,
+      "step": 11777
+    },
+    {
+      "epoch": 1.3529377979438286,
+      "grad_norm": 0.5338268280029297,
+      "learning_rate": 0.0001,
+      "loss": 1.3373,
+      "step": 11778
+    },
+    {
+      "epoch": 1.3530526678536559,
+      "grad_norm": 0.5903444290161133,
+      "learning_rate": 0.0001,
+      "loss": 1.4409,
+      "step": 11779
+    },
+    {
+      "epoch": 1.3531675377634829,
+      "grad_norm": 0.5709140300750732,
+      "learning_rate": 0.0001,
+      "loss": 1.4397,
+      "step": 11780
+    },
+    {
+      "epoch": 1.3532824076733099,
+      "grad_norm": 0.6345384120941162,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 11781
+    },
+    {
+      "epoch": 1.353397277583137,
+      "grad_norm": 0.6136090755462646,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 11782
+    },
+    {
+      "epoch": 1.3535121474929643,
+      "grad_norm": 0.6243228316307068,
+      "learning_rate": 0.0001,
+      "loss": 1.4065,
+      "step": 11783
+    },
+    {
+      "epoch": 1.3536270174027913,
+      "grad_norm": 0.6512452960014343,
+      "learning_rate": 0.0001,
+      "loss": 1.5381,
+      "step": 11784
+    },
+    {
+      "epoch": 1.3537418873126184,
+      "grad_norm": 0.5914151668548584,
+      "learning_rate": 0.0001,
+      "loss": 1.3474,
+      "step": 11785
+    },
+    {
+      "epoch": 1.3538567572224456,
+      "grad_norm": 0.5646949410438538,
+      "learning_rate": 0.0001,
+      "loss": 1.4334,
+      "step": 11786
+    },
+    {
+      "epoch": 1.3539716271322728,
+      "grad_norm": 0.5643793940544128,
+      "learning_rate": 0.0001,
+      "loss": 1.3393,
+      "step": 11787
+    },
+    {
+      "epoch": 1.3540864970420998,
+      "grad_norm": 0.5716279745101929,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 11788
+    },
+    {
+      "epoch": 1.3542013669519268,
+      "grad_norm": 0.5285089612007141,
+      "learning_rate": 0.0001,
+      "loss": 1.2825,
+      "step": 11789
+    },
+    {
+      "epoch": 1.354316236861754,
+      "grad_norm": 0.5500566363334656,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 11790
+    },
+    {
+      "epoch": 1.3544311067715813,
+      "grad_norm": 0.556171715259552,
+      "learning_rate": 0.0001,
+      "loss": 1.4946,
+      "step": 11791
+    },
+    {
+      "epoch": 1.3545459766814083,
+      "grad_norm": 0.62319415807724,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 11792
+    },
+    {
+      "epoch": 1.3546608465912353,
+      "grad_norm": 0.5419564843177795,
+      "learning_rate": 0.0001,
+      "loss": 1.2894,
+      "step": 11793
+    },
+    {
+      "epoch": 1.3547757165010625,
+      "grad_norm": 0.581043541431427,
+      "learning_rate": 0.0001,
+      "loss": 1.4787,
+      "step": 11794
+    },
+    {
+      "epoch": 1.3548905864108898,
+      "grad_norm": 0.5013282895088196,
+      "learning_rate": 0.0001,
+      "loss": 1.1919,
+      "step": 11795
+    },
+    {
+      "epoch": 1.3550054563207168,
+      "grad_norm": 0.6230014562606812,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 11796
+    },
+    {
+      "epoch": 1.3551203262305438,
+      "grad_norm": 0.5700322389602661,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 11797
+    },
+    {
+      "epoch": 1.355235196140371,
+      "grad_norm": 0.5606763362884521,
+      "learning_rate": 0.0001,
+      "loss": 1.4711,
+      "step": 11798
+    },
+    {
+      "epoch": 1.3553500660501983,
+      "grad_norm": 0.5901811122894287,
+      "learning_rate": 0.0001,
+      "loss": 1.7085,
+      "step": 11799
+    },
+    {
+      "epoch": 1.3554649359600253,
+      "grad_norm": 0.5982941389083862,
+      "learning_rate": 0.0001,
+      "loss": 1.4593,
+      "step": 11800
+    },
+    {
+      "epoch": 1.3555798058698523,
+      "grad_norm": 0.5848938822746277,
+      "learning_rate": 0.0001,
+      "loss": 1.466,
+      "step": 11801
+    },
+    {
+      "epoch": 1.3556946757796795,
+      "grad_norm": 0.6245958209037781,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 11802
+    },
+    {
+      "epoch": 1.3558095456895067,
+      "grad_norm": 0.6128213405609131,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 11803
+    },
+    {
+      "epoch": 1.3559244155993337,
+      "grad_norm": 0.5603303909301758,
+      "learning_rate": 0.0001,
+      "loss": 1.3408,
+      "step": 11804
+    },
+    {
+      "epoch": 1.3560392855091608,
+      "grad_norm": 0.6169303059577942,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 11805
+    },
+    {
+      "epoch": 1.356154155418988,
+      "grad_norm": 0.5828307867050171,
+      "learning_rate": 0.0001,
+      "loss": 1.35,
+      "step": 11806
+    },
+    {
+      "epoch": 1.3562690253288152,
+      "grad_norm": 0.6106387972831726,
+      "learning_rate": 0.0001,
+      "loss": 1.4773,
+      "step": 11807
+    },
+    {
+      "epoch": 1.3563838952386422,
+      "grad_norm": 0.6269760131835938,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 11808
+    },
+    {
+      "epoch": 1.3564987651484692,
+      "grad_norm": 0.6034385561943054,
+      "learning_rate": 0.0001,
+      "loss": 1.5904,
+      "step": 11809
+    },
+    {
+      "epoch": 1.3566136350582965,
+      "grad_norm": 0.5654463768005371,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 11810
+    },
+    {
+      "epoch": 1.3567285049681237,
+      "grad_norm": 0.5962846875190735,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 11811
+    },
+    {
+      "epoch": 1.3568433748779507,
+      "grad_norm": 0.6377612352371216,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 11812
+    },
+    {
+      "epoch": 1.3569582447877777,
+      "grad_norm": 0.5528261065483093,
+      "learning_rate": 0.0001,
+      "loss": 1.3851,
+      "step": 11813
+    },
+    {
+      "epoch": 1.357073114697605,
+      "grad_norm": 0.5887748599052429,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 11814
+    },
+    {
+      "epoch": 1.3571879846074322,
+      "grad_norm": 0.6080367565155029,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 11815
+    },
+    {
+      "epoch": 1.3573028545172592,
+      "grad_norm": 0.5874114632606506,
+      "learning_rate": 0.0001,
+      "loss": 1.3564,
+      "step": 11816
+    },
+    {
+      "epoch": 1.3574177244270862,
+      "grad_norm": 0.5622687339782715,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 11817
+    },
+    {
+      "epoch": 1.3575325943369134,
+      "grad_norm": 0.5545808672904968,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 11818
+    },
+    {
+      "epoch": 1.3576474642467407,
+      "grad_norm": 0.5163676738739014,
+      "learning_rate": 0.0001,
+      "loss": 1.3761,
+      "step": 11819
+    },
+    {
+      "epoch": 1.3577623341565677,
+      "grad_norm": 0.5534794330596924,
+      "learning_rate": 0.0001,
+      "loss": 1.4561,
+      "step": 11820
+    },
+    {
+      "epoch": 1.3578772040663947,
+      "grad_norm": 0.582586407661438,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 11821
+    },
+    {
+      "epoch": 1.357992073976222,
+      "grad_norm": 0.5760107040405273,
+      "learning_rate": 0.0001,
+      "loss": 1.329,
+      "step": 11822
+    },
+    {
+      "epoch": 1.3581069438860491,
+      "grad_norm": 0.5939583778381348,
+      "learning_rate": 0.0001,
+      "loss": 1.3137,
+      "step": 11823
+    },
+    {
+      "epoch": 1.3582218137958761,
+      "grad_norm": 0.6262675523757935,
+      "learning_rate": 0.0001,
+      "loss": 1.5452,
+      "step": 11824
+    },
+    {
+      "epoch": 1.3583366837057034,
+      "grad_norm": 0.5676589012145996,
+      "learning_rate": 0.0001,
+      "loss": 1.525,
+      "step": 11825
+    },
+    {
+      "epoch": 1.3584515536155304,
+      "grad_norm": 0.5600330233573914,
+      "learning_rate": 0.0001,
+      "loss": 1.3734,
+      "step": 11826
+    },
+    {
+      "epoch": 1.3585664235253576,
+      "grad_norm": 0.6240646243095398,
+      "learning_rate": 0.0001,
+      "loss": 1.5052,
+      "step": 11827
+    },
+    {
+      "epoch": 1.3586812934351846,
+      "grad_norm": 0.6038419008255005,
+      "learning_rate": 0.0001,
+      "loss": 1.4135,
+      "step": 11828
+    },
+    {
+      "epoch": 1.3587961633450119,
+      "grad_norm": 0.5694374442100525,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 11829
+    },
+    {
+      "epoch": 1.3589110332548389,
+      "grad_norm": 0.5696335434913635,
+      "learning_rate": 0.0001,
+      "loss": 1.3895,
+      "step": 11830
+    },
+    {
+      "epoch": 1.359025903164666,
+      "grad_norm": 0.6053335666656494,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 11831
+    },
+    {
+      "epoch": 1.359140773074493,
+      "grad_norm": 0.5585938096046448,
+      "learning_rate": 0.0001,
+      "loss": 1.4728,
+      "step": 11832
+    },
+    {
+      "epoch": 1.3592556429843203,
+      "grad_norm": 0.5684964060783386,
+      "learning_rate": 0.0001,
+      "loss": 1.4404,
+      "step": 11833
+    },
+    {
+      "epoch": 1.3593705128941473,
+      "grad_norm": 0.5710092782974243,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 11834
+    },
+    {
+      "epoch": 1.3594853828039746,
+      "grad_norm": 0.545071542263031,
+      "learning_rate": 0.0001,
+      "loss": 1.3686,
+      "step": 11835
+    },
+    {
+      "epoch": 1.3596002527138016,
+      "grad_norm": 0.5682846307754517,
+      "learning_rate": 0.0001,
+      "loss": 1.4138,
+      "step": 11836
+    },
+    {
+      "epoch": 1.3597151226236288,
+      "grad_norm": 0.6198849081993103,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 11837
+    },
+    {
+      "epoch": 1.3598299925334558,
+      "grad_norm": 0.6224766969680786,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 11838
+    },
+    {
+      "epoch": 1.359944862443283,
+      "grad_norm": 0.5848830938339233,
+      "learning_rate": 0.0001,
+      "loss": 1.3544,
+      "step": 11839
+    },
+    {
+      "epoch": 1.36005973235311,
+      "grad_norm": 0.6136349439620972,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 11840
+    },
+    {
+      "epoch": 1.3601746022629373,
+      "grad_norm": 0.6168116927146912,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 11841
+    },
+    {
+      "epoch": 1.3602894721727643,
+      "grad_norm": 0.5960310697555542,
+      "learning_rate": 0.0001,
+      "loss": 1.4425,
+      "step": 11842
+    },
+    {
+      "epoch": 1.3604043420825915,
+      "grad_norm": 0.731585681438446,
+      "learning_rate": 0.0001,
+      "loss": 1.6861,
+      "step": 11843
+    },
+    {
+      "epoch": 1.3605192119924185,
+      "grad_norm": 0.6656792759895325,
+      "learning_rate": 0.0001,
+      "loss": 1.5898,
+      "step": 11844
+    },
+    {
+      "epoch": 1.3606340819022458,
+      "grad_norm": 0.5910264849662781,
+      "learning_rate": 0.0001,
+      "loss": 1.3898,
+      "step": 11845
+    },
+    {
+      "epoch": 1.3607489518120728,
+      "grad_norm": 0.6095188856124878,
+      "learning_rate": 0.0001,
+      "loss": 1.5024,
+      "step": 11846
+    },
+    {
+      "epoch": 1.3608638217219,
+      "grad_norm": 0.5570575594902039,
+      "learning_rate": 0.0001,
+      "loss": 1.3486,
+      "step": 11847
+    },
+    {
+      "epoch": 1.360978691631727,
+      "grad_norm": 0.5648567080497742,
+      "learning_rate": 0.0001,
+      "loss": 1.4618,
+      "step": 11848
+    },
+    {
+      "epoch": 1.3610935615415543,
+      "grad_norm": 0.5962167978286743,
+      "learning_rate": 0.0001,
+      "loss": 1.5203,
+      "step": 11849
+    },
+    {
+      "epoch": 1.3612084314513813,
+      "grad_norm": 0.6409988403320312,
+      "learning_rate": 0.0001,
+      "loss": 1.367,
+      "step": 11850
+    },
+    {
+      "epoch": 1.3613233013612085,
+      "grad_norm": 0.6274490356445312,
+      "learning_rate": 0.0001,
+      "loss": 1.4525,
+      "step": 11851
+    },
+    {
+      "epoch": 1.3614381712710355,
+      "grad_norm": 0.541801393032074,
+      "learning_rate": 0.0001,
+      "loss": 1.4423,
+      "step": 11852
+    },
+    {
+      "epoch": 1.3615530411808627,
+      "grad_norm": 0.6285209655761719,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 11853
+    },
+    {
+      "epoch": 1.3616679110906897,
+      "grad_norm": 0.5430086255073547,
+      "learning_rate": 0.0001,
+      "loss": 1.4768,
+      "step": 11854
+    },
+    {
+      "epoch": 1.361782781000517,
+      "grad_norm": 0.5744656324386597,
+      "learning_rate": 0.0001,
+      "loss": 1.3223,
+      "step": 11855
+    },
+    {
+      "epoch": 1.361897650910344,
+      "grad_norm": 0.6095870733261108,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 11856
+    },
+    {
+      "epoch": 1.3620125208201712,
+      "grad_norm": 0.6219245791435242,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 11857
+    },
+    {
+      "epoch": 1.3621273907299982,
+      "grad_norm": 0.5849238038063049,
+      "learning_rate": 0.0001,
+      "loss": 1.5043,
+      "step": 11858
+    },
+    {
+      "epoch": 1.3622422606398255,
+      "grad_norm": 0.6130156517028809,
+      "learning_rate": 0.0001,
+      "loss": 1.3665,
+      "step": 11859
+    },
+    {
+      "epoch": 1.3623571305496525,
+      "grad_norm": 0.6351000070571899,
+      "learning_rate": 0.0001,
+      "loss": 1.5092,
+      "step": 11860
+    },
+    {
+      "epoch": 1.3624720004594797,
+      "grad_norm": 0.5915348529815674,
+      "learning_rate": 0.0001,
+      "loss": 1.4129,
+      "step": 11861
+    },
+    {
+      "epoch": 1.3625868703693067,
+      "grad_norm": 0.6348772644996643,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 11862
+    },
+    {
+      "epoch": 1.362701740279134,
+      "grad_norm": 0.6160379648208618,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 11863
+    },
+    {
+      "epoch": 1.362816610188961,
+      "grad_norm": 0.6016219258308411,
+      "learning_rate": 0.0001,
+      "loss": 1.4674,
+      "step": 11864
+    },
+    {
+      "epoch": 1.3629314800987882,
+      "grad_norm": 0.6062735915184021,
+      "learning_rate": 0.0001,
+      "loss": 1.3464,
+      "step": 11865
+    },
+    {
+      "epoch": 1.3630463500086152,
+      "grad_norm": 0.6335691809654236,
+      "learning_rate": 0.0001,
+      "loss": 1.2892,
+      "step": 11866
+    },
+    {
+      "epoch": 1.3631612199184424,
+      "grad_norm": 0.597547173500061,
+      "learning_rate": 0.0001,
+      "loss": 1.398,
+      "step": 11867
+    },
+    {
+      "epoch": 1.3632760898282694,
+      "grad_norm": 0.6174020767211914,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 11868
+    },
+    {
+      "epoch": 1.3633909597380967,
+      "grad_norm": 0.5814570784568787,
+      "learning_rate": 0.0001,
+      "loss": 1.2317,
+      "step": 11869
+    },
+    {
+      "epoch": 1.3635058296479237,
+      "grad_norm": 0.5660645365715027,
+      "learning_rate": 0.0001,
+      "loss": 1.3388,
+      "step": 11870
+    },
+    {
+      "epoch": 1.363620699557751,
+      "grad_norm": 0.5966171026229858,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 11871
+    },
+    {
+      "epoch": 1.363735569467578,
+      "grad_norm": 0.6484749913215637,
+      "learning_rate": 0.0001,
+      "loss": 1.5419,
+      "step": 11872
+    },
+    {
+      "epoch": 1.3638504393774051,
+      "grad_norm": 0.6023862361907959,
+      "learning_rate": 0.0001,
+      "loss": 1.4739,
+      "step": 11873
+    },
+    {
+      "epoch": 1.3639653092872321,
+      "grad_norm": 0.6137213706970215,
+      "learning_rate": 0.0001,
+      "loss": 1.5059,
+      "step": 11874
+    },
+    {
+      "epoch": 1.3640801791970594,
+      "grad_norm": 0.5899903774261475,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 11875
+    },
+    {
+      "epoch": 1.3641950491068864,
+      "grad_norm": 0.6372878551483154,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 11876
+    },
+    {
+      "epoch": 1.3643099190167136,
+      "grad_norm": 0.6417999863624573,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 11877
+    },
+    {
+      "epoch": 1.3644247889265406,
+      "grad_norm": 0.6401478052139282,
+      "learning_rate": 0.0001,
+      "loss": 1.5189,
+      "step": 11878
+    },
+    {
+      "epoch": 1.3645396588363679,
+      "grad_norm": 0.5522693395614624,
+      "learning_rate": 0.0001,
+      "loss": 1.4322,
+      "step": 11879
+    },
+    {
+      "epoch": 1.3646545287461949,
+      "grad_norm": 0.5718218088150024,
+      "learning_rate": 0.0001,
+      "loss": 1.445,
+      "step": 11880
+    },
+    {
+      "epoch": 1.364769398656022,
+      "grad_norm": 0.5563505291938782,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 11881
+    },
+    {
+      "epoch": 1.364884268565849,
+      "grad_norm": 0.5796937346458435,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 11882
+    },
+    {
+      "epoch": 1.3649991384756763,
+      "grad_norm": 0.5500465035438538,
+      "learning_rate": 0.0001,
+      "loss": 1.3196,
+      "step": 11883
+    },
+    {
+      "epoch": 1.3651140083855033,
+      "grad_norm": 0.5704468488693237,
+      "learning_rate": 0.0001,
+      "loss": 1.4476,
+      "step": 11884
+    },
+    {
+      "epoch": 1.3652288782953306,
+      "grad_norm": 0.6012165546417236,
+      "learning_rate": 0.0001,
+      "loss": 1.3108,
+      "step": 11885
+    },
+    {
+      "epoch": 1.3653437482051576,
+      "grad_norm": 0.5745958089828491,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 11886
+    },
+    {
+      "epoch": 1.3654586181149848,
+      "grad_norm": 0.5989466905593872,
+      "learning_rate": 0.0001,
+      "loss": 1.4294,
+      "step": 11887
+    },
+    {
+      "epoch": 1.3655734880248118,
+      "grad_norm": 0.5610129237174988,
+      "learning_rate": 0.0001,
+      "loss": 1.4047,
+      "step": 11888
+    },
+    {
+      "epoch": 1.365688357934639,
+      "grad_norm": 0.5722390413284302,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 11889
+    },
+    {
+      "epoch": 1.365803227844466,
+      "grad_norm": 0.5391660332679749,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 11890
+    },
+    {
+      "epoch": 1.3659180977542933,
+      "grad_norm": 0.6333369016647339,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 11891
+    },
+    {
+      "epoch": 1.3660329676641203,
+      "grad_norm": 0.5749605298042297,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 11892
+    },
+    {
+      "epoch": 1.3661478375739475,
+      "grad_norm": 0.5995118021965027,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 11893
+    },
+    {
+      "epoch": 1.3662627074837745,
+      "grad_norm": 0.6667972207069397,
+      "learning_rate": 0.0001,
+      "loss": 1.7452,
+      "step": 11894
+    },
+    {
+      "epoch": 1.3663775773936018,
+      "grad_norm": 0.5708993077278137,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 11895
+    },
+    {
+      "epoch": 1.366492447303429,
+      "grad_norm": 0.5501135587692261,
+      "learning_rate": 0.0001,
+      "loss": 1.4129,
+      "step": 11896
+    },
+    {
+      "epoch": 1.366607317213256,
+      "grad_norm": 0.5960387587547302,
+      "learning_rate": 0.0001,
+      "loss": 1.4311,
+      "step": 11897
+    },
+    {
+      "epoch": 1.366722187123083,
+      "grad_norm": 0.6244403123855591,
+      "learning_rate": 0.0001,
+      "loss": 1.5404,
+      "step": 11898
+    },
+    {
+      "epoch": 1.3668370570329103,
+      "grad_norm": 0.5725081562995911,
+      "learning_rate": 0.0001,
+      "loss": 1.4793,
+      "step": 11899
+    },
+    {
+      "epoch": 1.3669519269427375,
+      "grad_norm": 0.5716732144355774,
+      "learning_rate": 0.0001,
+      "loss": 1.2992,
+      "step": 11900
+    },
+    {
+      "epoch": 1.3670667968525645,
+      "grad_norm": 0.6741322875022888,
+      "learning_rate": 0.0001,
+      "loss": 1.6103,
+      "step": 11901
+    },
+    {
+      "epoch": 1.3671816667623915,
+      "grad_norm": 0.5683753490447998,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 11902
+    },
+    {
+      "epoch": 1.3672965366722187,
+      "grad_norm": 0.5858237743377686,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 11903
+    },
+    {
+      "epoch": 1.367411406582046,
+      "grad_norm": 0.6018313765525818,
+      "learning_rate": 0.0001,
+      "loss": 1.3942,
+      "step": 11904
+    },
+    {
+      "epoch": 1.367526276491873,
+      "grad_norm": 0.5790035724639893,
+      "learning_rate": 0.0001,
+      "loss": 1.3989,
+      "step": 11905
+    },
+    {
+      "epoch": 1.3676411464017,
+      "grad_norm": 0.5737671256065369,
+      "learning_rate": 0.0001,
+      "loss": 1.3992,
+      "step": 11906
+    },
+    {
+      "epoch": 1.3677560163115272,
+      "grad_norm": 0.5786615610122681,
+      "learning_rate": 0.0001,
+      "loss": 1.4582,
+      "step": 11907
+    },
+    {
+      "epoch": 1.3678708862213544,
+      "grad_norm": 0.6264104247093201,
+      "learning_rate": 0.0001,
+      "loss": 1.429,
+      "step": 11908
+    },
+    {
+      "epoch": 1.3679857561311815,
+      "grad_norm": 0.5502702593803406,
+      "learning_rate": 0.0001,
+      "loss": 1.4506,
+      "step": 11909
+    },
+    {
+      "epoch": 1.3681006260410085,
+      "grad_norm": 0.5532498359680176,
+      "learning_rate": 0.0001,
+      "loss": 1.3379,
+      "step": 11910
+    },
+    {
+      "epoch": 1.3682154959508357,
+      "grad_norm": 0.5454950332641602,
+      "learning_rate": 0.0001,
+      "loss": 1.3526,
+      "step": 11911
+    },
+    {
+      "epoch": 1.368330365860663,
+      "grad_norm": 0.6295995712280273,
+      "learning_rate": 0.0001,
+      "loss": 1.4195,
+      "step": 11912
+    },
+    {
+      "epoch": 1.36844523577049,
+      "grad_norm": 0.6341611742973328,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 11913
+    },
+    {
+      "epoch": 1.368560105680317,
+      "grad_norm": 0.620488703250885,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 11914
+    },
+    {
+      "epoch": 1.3686749755901442,
+      "grad_norm": 0.5347122550010681,
+      "learning_rate": 0.0001,
+      "loss": 1.46,
+      "step": 11915
+    },
+    {
+      "epoch": 1.3687898454999714,
+      "grad_norm": 0.5783818960189819,
+      "learning_rate": 0.0001,
+      "loss": 1.2346,
+      "step": 11916
+    },
+    {
+      "epoch": 1.3689047154097984,
+      "grad_norm": 0.5772425532341003,
+      "learning_rate": 0.0001,
+      "loss": 1.4085,
+      "step": 11917
+    },
+    {
+      "epoch": 1.3690195853196254,
+      "grad_norm": 0.5653284788131714,
+      "learning_rate": 0.0001,
+      "loss": 1.5086,
+      "step": 11918
+    },
+    {
+      "epoch": 1.3691344552294527,
+      "grad_norm": 0.6505599021911621,
+      "learning_rate": 0.0001,
+      "loss": 1.4088,
+      "step": 11919
+    },
+    {
+      "epoch": 1.3692493251392799,
+      "grad_norm": 0.6298074126243591,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 11920
+    },
+    {
+      "epoch": 1.369364195049107,
+      "grad_norm": 0.5957285761833191,
+      "learning_rate": 0.0001,
+      "loss": 1.4399,
+      "step": 11921
+    },
+    {
+      "epoch": 1.369479064958934,
+      "grad_norm": 0.5449360609054565,
+      "learning_rate": 0.0001,
+      "loss": 1.2528,
+      "step": 11922
+    },
+    {
+      "epoch": 1.3695939348687611,
+      "grad_norm": 0.6365380883216858,
+      "learning_rate": 0.0001,
+      "loss": 1.1535,
+      "step": 11923
+    },
+    {
+      "epoch": 1.3697088047785884,
+      "grad_norm": 0.668751060962677,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 11924
+    },
+    {
+      "epoch": 1.3698236746884154,
+      "grad_norm": 0.631753146648407,
+      "learning_rate": 0.0001,
+      "loss": 1.1487,
+      "step": 11925
+    },
+    {
+      "epoch": 1.3699385445982424,
+      "grad_norm": 0.5405638813972473,
+      "learning_rate": 0.0001,
+      "loss": 1.1319,
+      "step": 11926
+    },
+    {
+      "epoch": 1.3700534145080696,
+      "grad_norm": 0.6844695806503296,
+      "learning_rate": 0.0001,
+      "loss": 1.4908,
+      "step": 11927
+    },
+    {
+      "epoch": 1.3701682844178968,
+      "grad_norm": 0.6165083050727844,
+      "learning_rate": 0.0001,
+      "loss": 1.4513,
+      "step": 11928
+    },
+    {
+      "epoch": 1.3702831543277239,
+      "grad_norm": 0.5696147084236145,
+      "learning_rate": 0.0001,
+      "loss": 1.4323,
+      "step": 11929
+    },
+    {
+      "epoch": 1.3703980242375509,
+      "grad_norm": 0.6744541525840759,
+      "learning_rate": 0.0001,
+      "loss": 1.6197,
+      "step": 11930
+    },
+    {
+      "epoch": 1.370512894147378,
+      "grad_norm": 0.5944900512695312,
+      "learning_rate": 0.0001,
+      "loss": 1.3649,
+      "step": 11931
+    },
+    {
+      "epoch": 1.3706277640572053,
+      "grad_norm": 0.60816490650177,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 11932
+    },
+    {
+      "epoch": 1.3707426339670323,
+      "grad_norm": 0.5886834859848022,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 11933
+    },
+    {
+      "epoch": 1.3708575038768593,
+      "grad_norm": 0.5649728178977966,
+      "learning_rate": 0.0001,
+      "loss": 1.4151,
+      "step": 11934
+    },
+    {
+      "epoch": 1.3709723737866866,
+      "grad_norm": 0.5838742852210999,
+      "learning_rate": 0.0001,
+      "loss": 1.4892,
+      "step": 11935
+    },
+    {
+      "epoch": 1.3710872436965138,
+      "grad_norm": 0.556206464767456,
+      "learning_rate": 0.0001,
+      "loss": 1.4125,
+      "step": 11936
+    },
+    {
+      "epoch": 1.3712021136063408,
+      "grad_norm": 0.5563154816627502,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 11937
+    },
+    {
+      "epoch": 1.3713169835161678,
+      "grad_norm": 0.5660498738288879,
+      "learning_rate": 0.0001,
+      "loss": 1.4787,
+      "step": 11938
+    },
+    {
+      "epoch": 1.371431853425995,
+      "grad_norm": 0.604303777217865,
+      "learning_rate": 0.0001,
+      "loss": 1.5178,
+      "step": 11939
+    },
+    {
+      "epoch": 1.3715467233358223,
+      "grad_norm": 0.624233067035675,
+      "learning_rate": 0.0001,
+      "loss": 1.4388,
+      "step": 11940
+    },
+    {
+      "epoch": 1.3716615932456493,
+      "grad_norm": 0.5636304020881653,
+      "learning_rate": 0.0001,
+      "loss": 1.2747,
+      "step": 11941
+    },
+    {
+      "epoch": 1.3717764631554763,
+      "grad_norm": 0.543554425239563,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 11942
+    },
+    {
+      "epoch": 1.3718913330653035,
+      "grad_norm": 0.5257536768913269,
+      "learning_rate": 0.0001,
+      "loss": 1.1757,
+      "step": 11943
+    },
+    {
+      "epoch": 1.3720062029751308,
+      "grad_norm": 0.6228072047233582,
+      "learning_rate": 0.0001,
+      "loss": 1.3553,
+      "step": 11944
+    },
+    {
+      "epoch": 1.3721210728849578,
+      "grad_norm": 0.7739028334617615,
+      "learning_rate": 0.0001,
+      "loss": 1.355,
+      "step": 11945
+    },
+    {
+      "epoch": 1.3722359427947848,
+      "grad_norm": 0.6044634580612183,
+      "learning_rate": 0.0001,
+      "loss": 1.5408,
+      "step": 11946
+    },
+    {
+      "epoch": 1.372350812704612,
+      "grad_norm": 0.6621956825256348,
+      "learning_rate": 0.0001,
+      "loss": 1.6615,
+      "step": 11947
+    },
+    {
+      "epoch": 1.3724656826144392,
+      "grad_norm": 0.5978637933731079,
+      "learning_rate": 0.0001,
+      "loss": 1.504,
+      "step": 11948
+    },
+    {
+      "epoch": 1.3725805525242663,
+      "grad_norm": 0.5536312460899353,
+      "learning_rate": 0.0001,
+      "loss": 1.4278,
+      "step": 11949
+    },
+    {
+      "epoch": 1.3726954224340933,
+      "grad_norm": 0.5836893320083618,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 11950
+    },
+    {
+      "epoch": 1.3728102923439205,
+      "grad_norm": 0.5607746243476868,
+      "learning_rate": 0.0001,
+      "loss": 1.5454,
+      "step": 11951
+    },
+    {
+      "epoch": 1.3729251622537477,
+      "grad_norm": 0.7326905727386475,
+      "learning_rate": 0.0001,
+      "loss": 1.7962,
+      "step": 11952
+    },
+    {
+      "epoch": 1.3730400321635747,
+      "grad_norm": 0.5599920749664307,
+      "learning_rate": 0.0001,
+      "loss": 1.4866,
+      "step": 11953
+    },
+    {
+      "epoch": 1.3731549020734017,
+      "grad_norm": 0.5913311839103699,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 11954
+    },
+    {
+      "epoch": 1.373269771983229,
+      "grad_norm": 0.596919059753418,
+      "learning_rate": 0.0001,
+      "loss": 1.3861,
+      "step": 11955
+    },
+    {
+      "epoch": 1.3733846418930562,
+      "grad_norm": 0.6204702854156494,
+      "learning_rate": 0.0001,
+      "loss": 1.572,
+      "step": 11956
+    },
+    {
+      "epoch": 1.3734995118028832,
+      "grad_norm": 0.6467247605323792,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 11957
+    },
+    {
+      "epoch": 1.3736143817127102,
+      "grad_norm": 0.6159607768058777,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 11958
+    },
+    {
+      "epoch": 1.3737292516225375,
+      "grad_norm": 0.5385196208953857,
+      "learning_rate": 0.0001,
+      "loss": 1.3476,
+      "step": 11959
+    },
+    {
+      "epoch": 1.3738441215323647,
+      "grad_norm": 0.5562816262245178,
+      "learning_rate": 0.0001,
+      "loss": 1.354,
+      "step": 11960
+    },
+    {
+      "epoch": 1.3739589914421917,
+      "grad_norm": 0.5553545355796814,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 11961
+    },
+    {
+      "epoch": 1.374073861352019,
+      "grad_norm": 0.5768440365791321,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 11962
+    },
+    {
+      "epoch": 1.374188731261846,
+      "grad_norm": 0.6480252146720886,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 11963
+    },
+    {
+      "epoch": 1.3743036011716732,
+      "grad_norm": 0.5714027881622314,
+      "learning_rate": 0.0001,
+      "loss": 1.651,
+      "step": 11964
+    },
+    {
+      "epoch": 1.3744184710815002,
+      "grad_norm": 0.5680522918701172,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 11965
+    },
+    {
+      "epoch": 1.3745333409913274,
+      "grad_norm": 0.6007116436958313,
+      "learning_rate": 0.0001,
+      "loss": 1.4423,
+      "step": 11966
+    },
+    {
+      "epoch": 1.3746482109011544,
+      "grad_norm": 0.5150221586227417,
+      "learning_rate": 0.0001,
+      "loss": 1.3199,
+      "step": 11967
+    },
+    {
+      "epoch": 1.3747630808109816,
+      "grad_norm": 0.5811969041824341,
+      "learning_rate": 0.0001,
+      "loss": 1.3401,
+      "step": 11968
+    },
+    {
+      "epoch": 1.3748779507208087,
+      "grad_norm": 0.5529438853263855,
+      "learning_rate": 0.0001,
+      "loss": 1.445,
+      "step": 11969
+    },
+    {
+      "epoch": 1.3749928206306359,
+      "grad_norm": 0.5678512454032898,
+      "learning_rate": 0.0001,
+      "loss": 1.3625,
+      "step": 11970
+    },
+    {
+      "epoch": 1.375107690540463,
+      "grad_norm": 0.5907145738601685,
+      "learning_rate": 0.0001,
+      "loss": 1.5975,
+      "step": 11971
+    },
+    {
+      "epoch": 1.3752225604502901,
+      "grad_norm": 0.5750979781150818,
+      "learning_rate": 0.0001,
+      "loss": 1.3721,
+      "step": 11972
+    },
+    {
+      "epoch": 1.3753374303601171,
+      "grad_norm": 0.5326782464981079,
+      "learning_rate": 0.0001,
+      "loss": 1.2268,
+      "step": 11973
+    },
+    {
+      "epoch": 1.3754523002699444,
+      "grad_norm": 0.6372673511505127,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 11974
+    },
+    {
+      "epoch": 1.3755671701797714,
+      "grad_norm": 0.702731192111969,
+      "learning_rate": 0.0001,
+      "loss": 1.5823,
+      "step": 11975
+    },
+    {
+      "epoch": 1.3756820400895986,
+      "grad_norm": 0.5595875978469849,
+      "learning_rate": 0.0001,
+      "loss": 1.3401,
+      "step": 11976
+    },
+    {
+      "epoch": 1.3757969099994256,
+      "grad_norm": 0.5655556917190552,
+      "learning_rate": 0.0001,
+      "loss": 1.4331,
+      "step": 11977
+    },
+    {
+      "epoch": 1.3759117799092528,
+      "grad_norm": 0.5677425265312195,
+      "learning_rate": 0.0001,
+      "loss": 1.435,
+      "step": 11978
+    },
+    {
+      "epoch": 1.3760266498190799,
+      "grad_norm": 0.5768133401870728,
+      "learning_rate": 0.0001,
+      "loss": 1.4443,
+      "step": 11979
+    },
+    {
+      "epoch": 1.376141519728907,
+      "grad_norm": 0.6946627497673035,
+      "learning_rate": 0.0001,
+      "loss": 1.437,
+      "step": 11980
+    },
+    {
+      "epoch": 1.376256389638734,
+      "grad_norm": 0.5521852374076843,
+      "learning_rate": 0.0001,
+      "loss": 1.3945,
+      "step": 11981
+    },
+    {
+      "epoch": 1.3763712595485613,
+      "grad_norm": 0.5932571291923523,
+      "learning_rate": 0.0001,
+      "loss": 1.3905,
+      "step": 11982
+    },
+    {
+      "epoch": 1.3764861294583883,
+      "grad_norm": 0.5667235851287842,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 11983
+    },
+    {
+      "epoch": 1.3766009993682156,
+      "grad_norm": 0.6410410404205322,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 11984
+    },
+    {
+      "epoch": 1.3767158692780426,
+      "grad_norm": 0.6235799193382263,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 11985
+    },
+    {
+      "epoch": 1.3768307391878698,
+      "grad_norm": 0.5973227024078369,
+      "learning_rate": 0.0001,
+      "loss": 1.4338,
+      "step": 11986
+    },
+    {
+      "epoch": 1.3769456090976968,
+      "grad_norm": 0.5862446427345276,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 11987
+    },
+    {
+      "epoch": 1.377060479007524,
+      "grad_norm": 0.6002928018569946,
+      "learning_rate": 0.0001,
+      "loss": 1.4927,
+      "step": 11988
+    },
+    {
+      "epoch": 1.377175348917351,
+      "grad_norm": 0.6475585699081421,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 11989
+    },
+    {
+      "epoch": 1.3772902188271783,
+      "grad_norm": 0.6113678216934204,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 11990
+    },
+    {
+      "epoch": 1.3774050887370053,
+      "grad_norm": 0.5376540422439575,
+      "learning_rate": 0.0001,
+      "loss": 1.2808,
+      "step": 11991
+    },
+    {
+      "epoch": 1.3775199586468325,
+      "grad_norm": 0.6065835356712341,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 11992
+    },
+    {
+      "epoch": 1.3776348285566595,
+      "grad_norm": 0.5882073640823364,
+      "learning_rate": 0.0001,
+      "loss": 1.3892,
+      "step": 11993
+    },
+    {
+      "epoch": 1.3777496984664868,
+      "grad_norm": 0.5682032704353333,
+      "learning_rate": 0.0001,
+      "loss": 1.3817,
+      "step": 11994
+    },
+    {
+      "epoch": 1.3778645683763138,
+      "grad_norm": 0.6414339542388916,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 11995
+    },
+    {
+      "epoch": 1.377979438286141,
+      "grad_norm": 0.6156001091003418,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 11996
+    },
+    {
+      "epoch": 1.378094308195968,
+      "grad_norm": 0.5534403324127197,
+      "learning_rate": 0.0001,
+      "loss": 1.2606,
+      "step": 11997
+    },
+    {
+      "epoch": 1.3782091781057952,
+      "grad_norm": 0.5855156779289246,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 11998
+    },
+    {
+      "epoch": 1.3783240480156222,
+      "grad_norm": 0.5981359481811523,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 11999
+    },
+    {
+      "epoch": 1.3784389179254495,
+      "grad_norm": 0.6743270754814148,
+      "learning_rate": 0.0001,
+      "loss": 1.558,
+      "step": 12000
+    },
+    {
+      "epoch": 1.3785537878352765,
+      "grad_norm": 0.6285853981971741,
+      "learning_rate": 0.0001,
+      "loss": 1.7823,
+      "step": 12001
+    },
+    {
+      "epoch": 1.3786686577451037,
+      "grad_norm": 0.6270994544029236,
+      "learning_rate": 0.0001,
+      "loss": 1.3792,
+      "step": 12002
+    },
+    {
+      "epoch": 1.3787835276549307,
+      "grad_norm": 0.5615768432617188,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 12003
+    },
+    {
+      "epoch": 1.378898397564758,
+      "grad_norm": 0.5567690134048462,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 12004
+    },
+    {
+      "epoch": 1.379013267474585,
+      "grad_norm": 0.6072371602058411,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 12005
+    },
+    {
+      "epoch": 1.3791281373844122,
+      "grad_norm": 0.6124165058135986,
+      "learning_rate": 0.0001,
+      "loss": 1.396,
+      "step": 12006
+    },
+    {
+      "epoch": 1.3792430072942392,
+      "grad_norm": 0.6033795475959778,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 12007
+    },
+    {
+      "epoch": 1.3793578772040664,
+      "grad_norm": 0.5871466994285583,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 12008
+    },
+    {
+      "epoch": 1.3794727471138934,
+      "grad_norm": 0.6192461252212524,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 12009
+    },
+    {
+      "epoch": 1.3795876170237207,
+      "grad_norm": 0.5672609806060791,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 12010
+    },
+    {
+      "epoch": 1.3797024869335477,
+      "grad_norm": 0.5758570432662964,
+      "learning_rate": 0.0001,
+      "loss": 1.6809,
+      "step": 12011
+    },
+    {
+      "epoch": 1.379817356843375,
+      "grad_norm": 0.5541750192642212,
+      "learning_rate": 0.0001,
+      "loss": 1.4669,
+      "step": 12012
+    },
+    {
+      "epoch": 1.379932226753202,
+      "grad_norm": 0.652824878692627,
+      "learning_rate": 0.0001,
+      "loss": 1.2463,
+      "step": 12013
+    },
+    {
+      "epoch": 1.3800470966630292,
+      "grad_norm": 0.6101529598236084,
+      "learning_rate": 0.0001,
+      "loss": 1.6506,
+      "step": 12014
+    },
+    {
+      "epoch": 1.3801619665728562,
+      "grad_norm": 0.5852307677268982,
+      "learning_rate": 0.0001,
+      "loss": 1.6664,
+      "step": 12015
+    },
+    {
+      "epoch": 1.3802768364826834,
+      "grad_norm": 0.5421605706214905,
+      "learning_rate": 0.0001,
+      "loss": 1.3547,
+      "step": 12016
+    },
+    {
+      "epoch": 1.3803917063925104,
+      "grad_norm": 0.5122795701026917,
+      "learning_rate": 0.0001,
+      "loss": 1.3986,
+      "step": 12017
+    },
+    {
+      "epoch": 1.3805065763023376,
+      "grad_norm": 0.5709700584411621,
+      "learning_rate": 0.0001,
+      "loss": 1.4771,
+      "step": 12018
+    },
+    {
+      "epoch": 1.3806214462121646,
+      "grad_norm": 0.568841814994812,
+      "learning_rate": 0.0001,
+      "loss": 1.2557,
+      "step": 12019
+    },
+    {
+      "epoch": 1.3807363161219919,
+      "grad_norm": 0.5416236519813538,
+      "learning_rate": 0.0001,
+      "loss": 1.5226,
+      "step": 12020
+    },
+    {
+      "epoch": 1.380851186031819,
+      "grad_norm": 0.5720876455307007,
+      "learning_rate": 0.0001,
+      "loss": 1.4735,
+      "step": 12021
+    },
+    {
+      "epoch": 1.3809660559416461,
+      "grad_norm": 0.5533315539360046,
+      "learning_rate": 0.0001,
+      "loss": 1.2757,
+      "step": 12022
+    },
+    {
+      "epoch": 1.3810809258514731,
+      "grad_norm": 0.6400322914123535,
+      "learning_rate": 0.0001,
+      "loss": 1.3811,
+      "step": 12023
+    },
+    {
+      "epoch": 1.3811957957613004,
+      "grad_norm": 0.6613686680793762,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 12024
+    },
+    {
+      "epoch": 1.3813106656711274,
+      "grad_norm": 0.6082622408866882,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 12025
+    },
+    {
+      "epoch": 1.3814255355809546,
+      "grad_norm": 0.6383165121078491,
+      "learning_rate": 0.0001,
+      "loss": 1.579,
+      "step": 12026
+    },
+    {
+      "epoch": 1.3815404054907816,
+      "grad_norm": 0.5541550517082214,
+      "learning_rate": 0.0001,
+      "loss": 1.4047,
+      "step": 12027
+    },
+    {
+      "epoch": 1.3816552754006088,
+      "grad_norm": 0.5779566168785095,
+      "learning_rate": 0.0001,
+      "loss": 1.3272,
+      "step": 12028
+    },
+    {
+      "epoch": 1.3817701453104358,
+      "grad_norm": 0.599031925201416,
+      "learning_rate": 0.0001,
+      "loss": 1.5207,
+      "step": 12029
+    },
+    {
+      "epoch": 1.381885015220263,
+      "grad_norm": 0.5647594928741455,
+      "learning_rate": 0.0001,
+      "loss": 1.4146,
+      "step": 12030
+    },
+    {
+      "epoch": 1.38199988513009,
+      "grad_norm": 0.6167514324188232,
+      "learning_rate": 0.0001,
+      "loss": 1.5239,
+      "step": 12031
+    },
+    {
+      "epoch": 1.3821147550399173,
+      "grad_norm": 0.5759591460227966,
+      "learning_rate": 0.0001,
+      "loss": 1.3562,
+      "step": 12032
+    },
+    {
+      "epoch": 1.3822296249497446,
+      "grad_norm": 0.6348143815994263,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 12033
+    },
+    {
+      "epoch": 1.3823444948595716,
+      "grad_norm": 0.6396414041519165,
+      "learning_rate": 0.0001,
+      "loss": 1.595,
+      "step": 12034
+    },
+    {
+      "epoch": 1.3824593647693986,
+      "grad_norm": 0.5942586064338684,
+      "learning_rate": 0.0001,
+      "loss": 1.4623,
+      "step": 12035
+    },
+    {
+      "epoch": 1.3825742346792258,
+      "grad_norm": 0.625024676322937,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 12036
+    },
+    {
+      "epoch": 1.382689104589053,
+      "grad_norm": 0.6331309676170349,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 12037
+    },
+    {
+      "epoch": 1.38280397449888,
+      "grad_norm": 0.6286697387695312,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 12038
+    },
+    {
+      "epoch": 1.382918844408707,
+      "grad_norm": 0.6172159910202026,
+      "learning_rate": 0.0001,
+      "loss": 1.3461,
+      "step": 12039
+    },
+    {
+      "epoch": 1.3830337143185343,
+      "grad_norm": 0.5521469712257385,
+      "learning_rate": 0.0001,
+      "loss": 1.5414,
+      "step": 12040
+    },
+    {
+      "epoch": 1.3831485842283615,
+      "grad_norm": 0.5680788159370422,
+      "learning_rate": 0.0001,
+      "loss": 1.392,
+      "step": 12041
+    },
+    {
+      "epoch": 1.3832634541381885,
+      "grad_norm": 0.5680851936340332,
+      "learning_rate": 0.0001,
+      "loss": 1.4833,
+      "step": 12042
+    },
+    {
+      "epoch": 1.3833783240480155,
+      "grad_norm": 0.6148812174797058,
+      "learning_rate": 0.0001,
+      "loss": 1.4922,
+      "step": 12043
+    },
+    {
+      "epoch": 1.3834931939578428,
+      "grad_norm": 0.5677898526191711,
+      "learning_rate": 0.0001,
+      "loss": 1.3504,
+      "step": 12044
+    },
+    {
+      "epoch": 1.38360806386767,
+      "grad_norm": 0.5935441255569458,
+      "learning_rate": 0.0001,
+      "loss": 1.3455,
+      "step": 12045
+    },
+    {
+      "epoch": 1.383722933777497,
+      "grad_norm": 0.5716426968574524,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 12046
+    },
+    {
+      "epoch": 1.383837803687324,
+      "grad_norm": 0.5644375085830688,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 12047
+    },
+    {
+      "epoch": 1.3839526735971512,
+      "grad_norm": 0.5767795443534851,
+      "learning_rate": 0.0001,
+      "loss": 1.5569,
+      "step": 12048
+    },
+    {
+      "epoch": 1.3840675435069785,
+      "grad_norm": 0.5734127759933472,
+      "learning_rate": 0.0001,
+      "loss": 1.3649,
+      "step": 12049
+    },
+    {
+      "epoch": 1.3841824134168055,
+      "grad_norm": 0.628250777721405,
+      "learning_rate": 0.0001,
+      "loss": 1.5152,
+      "step": 12050
+    },
+    {
+      "epoch": 1.3842972833266325,
+      "grad_norm": 0.622211217880249,
+      "learning_rate": 0.0001,
+      "loss": 1.4912,
+      "step": 12051
+    },
+    {
+      "epoch": 1.3844121532364597,
+      "grad_norm": 0.5941185355186462,
+      "learning_rate": 0.0001,
+      "loss": 1.4555,
+      "step": 12052
+    },
+    {
+      "epoch": 1.384527023146287,
+      "grad_norm": 0.6594216227531433,
+      "learning_rate": 0.0001,
+      "loss": 1.4542,
+      "step": 12053
+    },
+    {
+      "epoch": 1.384641893056114,
+      "grad_norm": 0.5944557785987854,
+      "learning_rate": 0.0001,
+      "loss": 1.4187,
+      "step": 12054
+    },
+    {
+      "epoch": 1.384756762965941,
+      "grad_norm": 0.612331748008728,
+      "learning_rate": 0.0001,
+      "loss": 1.4077,
+      "step": 12055
+    },
+    {
+      "epoch": 1.3848716328757682,
+      "grad_norm": 0.5708249807357788,
+      "learning_rate": 0.0001,
+      "loss": 1.3858,
+      "step": 12056
+    },
+    {
+      "epoch": 1.3849865027855954,
+      "grad_norm": 0.7231257557868958,
+      "learning_rate": 0.0001,
+      "loss": 1.5112,
+      "step": 12057
+    },
+    {
+      "epoch": 1.3851013726954224,
+      "grad_norm": 0.5759585499763489,
+      "learning_rate": 0.0001,
+      "loss": 1.4769,
+      "step": 12058
+    },
+    {
+      "epoch": 1.3852162426052494,
+      "grad_norm": 0.5974624752998352,
+      "learning_rate": 0.0001,
+      "loss": 1.5837,
+      "step": 12059
+    },
+    {
+      "epoch": 1.3853311125150767,
+      "grad_norm": 0.5672141313552856,
+      "learning_rate": 0.0001,
+      "loss": 1.3981,
+      "step": 12060
+    },
+    {
+      "epoch": 1.385445982424904,
+      "grad_norm": 0.5883855819702148,
+      "learning_rate": 0.0001,
+      "loss": 1.2928,
+      "step": 12061
+    },
+    {
+      "epoch": 1.385560852334731,
+      "grad_norm": 0.6936026215553284,
+      "learning_rate": 0.0001,
+      "loss": 1.6683,
+      "step": 12062
+    },
+    {
+      "epoch": 1.385675722244558,
+      "grad_norm": 0.5789870619773865,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 12063
+    },
+    {
+      "epoch": 1.3857905921543852,
+      "grad_norm": 0.6692520380020142,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 12064
+    },
+    {
+      "epoch": 1.3859054620642124,
+      "grad_norm": 0.6601521372795105,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 12065
+    },
+    {
+      "epoch": 1.3860203319740394,
+      "grad_norm": 0.6020188927650452,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 12066
+    },
+    {
+      "epoch": 1.3861352018838664,
+      "grad_norm": 0.5850477814674377,
+      "learning_rate": 0.0001,
+      "loss": 1.5276,
+      "step": 12067
+    },
+    {
+      "epoch": 1.3862500717936936,
+      "grad_norm": 0.6538852453231812,
+      "learning_rate": 0.0001,
+      "loss": 1.6425,
+      "step": 12068
+    },
+    {
+      "epoch": 1.3863649417035209,
+      "grad_norm": 0.5990716218948364,
+      "learning_rate": 0.0001,
+      "loss": 1.4191,
+      "step": 12069
+    },
+    {
+      "epoch": 1.3864798116133479,
+      "grad_norm": 0.6036950945854187,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 12070
+    },
+    {
+      "epoch": 1.3865946815231749,
+      "grad_norm": 0.5441368818283081,
+      "learning_rate": 0.0001,
+      "loss": 1.4551,
+      "step": 12071
+    },
+    {
+      "epoch": 1.3867095514330021,
+      "grad_norm": 0.6289240121841431,
+      "learning_rate": 0.0001,
+      "loss": 1.274,
+      "step": 12072
+    },
+    {
+      "epoch": 1.3868244213428293,
+      "grad_norm": 0.5667722225189209,
+      "learning_rate": 0.0001,
+      "loss": 1.3627,
+      "step": 12073
+    },
+    {
+      "epoch": 1.3869392912526564,
+      "grad_norm": 0.5763718485832214,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 12074
+    },
+    {
+      "epoch": 1.3870541611624834,
+      "grad_norm": 0.5512832999229431,
+      "learning_rate": 0.0001,
+      "loss": 1.4434,
+      "step": 12075
+    },
+    {
+      "epoch": 1.3871690310723106,
+      "grad_norm": 0.5643221735954285,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 12076
+    },
+    {
+      "epoch": 1.3872839009821378,
+      "grad_norm": 0.5980443954467773,
+      "learning_rate": 0.0001,
+      "loss": 1.3511,
+      "step": 12077
+    },
+    {
+      "epoch": 1.3873987708919648,
+      "grad_norm": 0.6038445234298706,
+      "learning_rate": 0.0001,
+      "loss": 1.6927,
+      "step": 12078
+    },
+    {
+      "epoch": 1.3875136408017918,
+      "grad_norm": 0.6184426546096802,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 12079
+    },
+    {
+      "epoch": 1.387628510711619,
+      "grad_norm": 0.6515767574310303,
+      "learning_rate": 0.0001,
+      "loss": 1.684,
+      "step": 12080
+    },
+    {
+      "epoch": 1.3877433806214463,
+      "grad_norm": 0.6065769195556641,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 12081
+    },
+    {
+      "epoch": 1.3878582505312733,
+      "grad_norm": 0.6202980875968933,
+      "learning_rate": 0.0001,
+      "loss": 1.3916,
+      "step": 12082
+    },
+    {
+      "epoch": 1.3879731204411003,
+      "grad_norm": 0.6353384852409363,
+      "learning_rate": 0.0001,
+      "loss": 1.3461,
+      "step": 12083
+    },
+    {
+      "epoch": 1.3880879903509276,
+      "grad_norm": 0.6570160984992981,
+      "learning_rate": 0.0001,
+      "loss": 1.3667,
+      "step": 12084
+    },
+    {
+      "epoch": 1.3882028602607548,
+      "grad_norm": 0.5832175612449646,
+      "learning_rate": 0.0001,
+      "loss": 1.3499,
+      "step": 12085
+    },
+    {
+      "epoch": 1.3883177301705818,
+      "grad_norm": 0.5629262924194336,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 12086
+    },
+    {
+      "epoch": 1.3884326000804088,
+      "grad_norm": 0.586079478263855,
+      "learning_rate": 0.0001,
+      "loss": 1.432,
+      "step": 12087
+    },
+    {
+      "epoch": 1.388547469990236,
+      "grad_norm": 0.5707273483276367,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 12088
+    },
+    {
+      "epoch": 1.3886623399000633,
+      "grad_norm": 0.6245285868644714,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 12089
+    },
+    {
+      "epoch": 1.3887772098098903,
+      "grad_norm": 0.6213446855545044,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 12090
+    },
+    {
+      "epoch": 1.3888920797197173,
+      "grad_norm": 0.6027031540870667,
+      "learning_rate": 0.0001,
+      "loss": 1.5837,
+      "step": 12091
+    },
+    {
+      "epoch": 1.3890069496295445,
+      "grad_norm": 0.5713696479797363,
+      "learning_rate": 0.0001,
+      "loss": 1.4307,
+      "step": 12092
+    },
+    {
+      "epoch": 1.3891218195393717,
+      "grad_norm": 0.6017079949378967,
+      "learning_rate": 0.0001,
+      "loss": 1.3979,
+      "step": 12093
+    },
+    {
+      "epoch": 1.3892366894491988,
+      "grad_norm": 0.5776708722114563,
+      "learning_rate": 0.0001,
+      "loss": 1.4307,
+      "step": 12094
+    },
+    {
+      "epoch": 1.3893515593590258,
+      "grad_norm": 0.5756542086601257,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 12095
+    },
+    {
+      "epoch": 1.389466429268853,
+      "grad_norm": 0.6287577152252197,
+      "learning_rate": 0.0001,
+      "loss": 1.3661,
+      "step": 12096
+    },
+    {
+      "epoch": 1.3895812991786802,
+      "grad_norm": 0.5801504850387573,
+      "learning_rate": 0.0001,
+      "loss": 1.4198,
+      "step": 12097
+    },
+    {
+      "epoch": 1.3896961690885072,
+      "grad_norm": 0.6163144111633301,
+      "learning_rate": 0.0001,
+      "loss": 1.3697,
+      "step": 12098
+    },
+    {
+      "epoch": 1.3898110389983345,
+      "grad_norm": 0.623882532119751,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 12099
+    },
+    {
+      "epoch": 1.3899259089081615,
+      "grad_norm": 0.6868638396263123,
+      "learning_rate": 0.0001,
+      "loss": 1.7235,
+      "step": 12100
+    },
+    {
+      "epoch": 1.3900407788179887,
+      "grad_norm": 0.5804413557052612,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 12101
+    },
+    {
+      "epoch": 1.3901556487278157,
+      "grad_norm": 0.6150039434432983,
+      "learning_rate": 0.0001,
+      "loss": 1.4018,
+      "step": 12102
+    },
+    {
+      "epoch": 1.390270518637643,
+      "grad_norm": 0.6017144322395325,
+      "learning_rate": 0.0001,
+      "loss": 1.326,
+      "step": 12103
+    },
+    {
+      "epoch": 1.39038538854747,
+      "grad_norm": 0.5922073721885681,
+      "learning_rate": 0.0001,
+      "loss": 1.4635,
+      "step": 12104
+    },
+    {
+      "epoch": 1.3905002584572972,
+      "grad_norm": 0.6403117775917053,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 12105
+    },
+    {
+      "epoch": 1.3906151283671242,
+      "grad_norm": 0.5838686227798462,
+      "learning_rate": 0.0001,
+      "loss": 1.5511,
+      "step": 12106
+    },
+    {
+      "epoch": 1.3907299982769514,
+      "grad_norm": 0.575851559638977,
+      "learning_rate": 0.0001,
+      "loss": 1.4667,
+      "step": 12107
+    },
+    {
+      "epoch": 1.3908448681867784,
+      "grad_norm": 0.5583367347717285,
+      "learning_rate": 0.0001,
+      "loss": 1.3506,
+      "step": 12108
+    },
+    {
+      "epoch": 1.3909597380966057,
+      "grad_norm": 0.6150028109550476,
+      "learning_rate": 0.0001,
+      "loss": 1.5256,
+      "step": 12109
+    },
+    {
+      "epoch": 1.3910746080064327,
+      "grad_norm": 0.5967298746109009,
+      "learning_rate": 0.0001,
+      "loss": 1.4282,
+      "step": 12110
+    },
+    {
+      "epoch": 1.39118947791626,
+      "grad_norm": 0.6403521299362183,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 12111
+    },
+    {
+      "epoch": 1.391304347826087,
+      "grad_norm": 0.579937219619751,
+      "learning_rate": 0.0001,
+      "loss": 1.4695,
+      "step": 12112
+    },
+    {
+      "epoch": 1.3914192177359141,
+      "grad_norm": 0.5739872455596924,
+      "learning_rate": 0.0001,
+      "loss": 1.6012,
+      "step": 12113
+    },
+    {
+      "epoch": 1.3915340876457412,
+      "grad_norm": 0.5874963402748108,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 12114
+    },
+    {
+      "epoch": 1.3916489575555684,
+      "grad_norm": 0.5656614899635315,
+      "learning_rate": 0.0001,
+      "loss": 1.4357,
+      "step": 12115
+    },
+    {
+      "epoch": 1.3917638274653954,
+      "grad_norm": 0.6302921175956726,
+      "learning_rate": 0.0001,
+      "loss": 1.5072,
+      "step": 12116
+    },
+    {
+      "epoch": 1.3918786973752226,
+      "grad_norm": 0.6151950359344482,
+      "learning_rate": 0.0001,
+      "loss": 1.4652,
+      "step": 12117
+    },
+    {
+      "epoch": 1.3919935672850496,
+      "grad_norm": 0.6326630711555481,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 12118
+    },
+    {
+      "epoch": 1.3921084371948769,
+      "grad_norm": 0.6131033897399902,
+      "learning_rate": 0.0001,
+      "loss": 1.382,
+      "step": 12119
+    },
+    {
+      "epoch": 1.3922233071047039,
+      "grad_norm": 0.6083925366401672,
+      "learning_rate": 0.0001,
+      "loss": 1.47,
+      "step": 12120
+    },
+    {
+      "epoch": 1.392338177014531,
+      "grad_norm": 0.5470343232154846,
+      "learning_rate": 0.0001,
+      "loss": 1.3606,
+      "step": 12121
+    },
+    {
+      "epoch": 1.3924530469243581,
+      "grad_norm": 0.6349021196365356,
+      "learning_rate": 0.0001,
+      "loss": 1.6476,
+      "step": 12122
+    },
+    {
+      "epoch": 1.3925679168341853,
+      "grad_norm": 0.5835825204849243,
+      "learning_rate": 0.0001,
+      "loss": 1.3953,
+      "step": 12123
+    },
+    {
+      "epoch": 1.3926827867440124,
+      "grad_norm": 0.6109292507171631,
+      "learning_rate": 0.0001,
+      "loss": 1.5743,
+      "step": 12124
+    },
+    {
+      "epoch": 1.3927976566538396,
+      "grad_norm": 0.5654585957527161,
+      "learning_rate": 0.0001,
+      "loss": 1.4229,
+      "step": 12125
+    },
+    {
+      "epoch": 1.3929125265636666,
+      "grad_norm": 0.5883094072341919,
+      "learning_rate": 0.0001,
+      "loss": 1.5037,
+      "step": 12126
+    },
+    {
+      "epoch": 1.3930273964734938,
+      "grad_norm": 0.6077958345413208,
+      "learning_rate": 0.0001,
+      "loss": 1.4381,
+      "step": 12127
+    },
+    {
+      "epoch": 1.3931422663833208,
+      "grad_norm": 0.6568667888641357,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 12128
+    },
+    {
+      "epoch": 1.393257136293148,
+      "grad_norm": 0.5563843250274658,
+      "learning_rate": 0.0001,
+      "loss": 1.2562,
+      "step": 12129
+    },
+    {
+      "epoch": 1.393372006202975,
+      "grad_norm": 0.5851126313209534,
+      "learning_rate": 0.0001,
+      "loss": 1.3322,
+      "step": 12130
+    },
+    {
+      "epoch": 1.3934868761128023,
+      "grad_norm": 0.579210102558136,
+      "learning_rate": 0.0001,
+      "loss": 1.3495,
+      "step": 12131
+    },
+    {
+      "epoch": 1.3936017460226293,
+      "grad_norm": 0.6247128844261169,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 12132
+    },
+    {
+      "epoch": 1.3937166159324565,
+      "grad_norm": 0.6491114497184753,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 12133
+    },
+    {
+      "epoch": 1.3938314858422836,
+      "grad_norm": 0.5388944149017334,
+      "learning_rate": 0.0001,
+      "loss": 1.3398,
+      "step": 12134
+    },
+    {
+      "epoch": 1.3939463557521108,
+      "grad_norm": 0.5601046681404114,
+      "learning_rate": 0.0001,
+      "loss": 1.3384,
+      "step": 12135
+    },
+    {
+      "epoch": 1.3940612256619378,
+      "grad_norm": 0.5893126726150513,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 12136
+    },
+    {
+      "epoch": 1.394176095571765,
+      "grad_norm": 0.6289069056510925,
+      "learning_rate": 0.0001,
+      "loss": 1.5355,
+      "step": 12137
+    },
+    {
+      "epoch": 1.394290965481592,
+      "grad_norm": 0.6383712887763977,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 12138
+    },
+    {
+      "epoch": 1.3944058353914193,
+      "grad_norm": 0.6105307340621948,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 12139
+    },
+    {
+      "epoch": 1.3945207053012463,
+      "grad_norm": 0.5872035622596741,
+      "learning_rate": 0.0001,
+      "loss": 1.4613,
+      "step": 12140
+    },
+    {
+      "epoch": 1.3946355752110735,
+      "grad_norm": 0.5394282937049866,
+      "learning_rate": 0.0001,
+      "loss": 1.4145,
+      "step": 12141
+    },
+    {
+      "epoch": 1.3947504451209005,
+      "grad_norm": 0.6004125475883484,
+      "learning_rate": 0.0001,
+      "loss": 1.3922,
+      "step": 12142
+    },
+    {
+      "epoch": 1.3948653150307277,
+      "grad_norm": 0.6273020505905151,
+      "learning_rate": 0.0001,
+      "loss": 1.5747,
+      "step": 12143
+    },
+    {
+      "epoch": 1.3949801849405548,
+      "grad_norm": 0.5593479871749878,
+      "learning_rate": 0.0001,
+      "loss": 1.4331,
+      "step": 12144
+    },
+    {
+      "epoch": 1.395095054850382,
+      "grad_norm": 0.5342775583267212,
+      "learning_rate": 0.0001,
+      "loss": 1.3331,
+      "step": 12145
+    },
+    {
+      "epoch": 1.395209924760209,
+      "grad_norm": 0.5714641213417053,
+      "learning_rate": 0.0001,
+      "loss": 1.1656,
+      "step": 12146
+    },
+    {
+      "epoch": 1.3953247946700362,
+      "grad_norm": 0.6640602350234985,
+      "learning_rate": 0.0001,
+      "loss": 1.4388,
+      "step": 12147
+    },
+    {
+      "epoch": 1.3954396645798632,
+      "grad_norm": 0.5924851298332214,
+      "learning_rate": 0.0001,
+      "loss": 1.2908,
+      "step": 12148
+    },
+    {
+      "epoch": 1.3955545344896905,
+      "grad_norm": 0.6022685170173645,
+      "learning_rate": 0.0001,
+      "loss": 1.4481,
+      "step": 12149
+    },
+    {
+      "epoch": 1.3956694043995175,
+      "grad_norm": 0.6584579348564148,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 12150
+    },
+    {
+      "epoch": 1.3957842743093447,
+      "grad_norm": 0.6461522579193115,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 12151
+    },
+    {
+      "epoch": 1.3958991442191717,
+      "grad_norm": 0.6562781929969788,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 12152
+    },
+    {
+      "epoch": 1.396014014128999,
+      "grad_norm": 0.593710720539093,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 12153
+    },
+    {
+      "epoch": 1.396128884038826,
+      "grad_norm": 0.6154862642288208,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 12154
+    },
+    {
+      "epoch": 1.3962437539486532,
+      "grad_norm": 0.5625523328781128,
+      "learning_rate": 0.0001,
+      "loss": 1.4415,
+      "step": 12155
+    },
+    {
+      "epoch": 1.3963586238584802,
+      "grad_norm": 0.5562409162521362,
+      "learning_rate": 0.0001,
+      "loss": 1.3692,
+      "step": 12156
+    },
+    {
+      "epoch": 1.3964734937683074,
+      "grad_norm": 0.6092026829719543,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 12157
+    },
+    {
+      "epoch": 1.3965883636781344,
+      "grad_norm": 0.5918128490447998,
+      "learning_rate": 0.0001,
+      "loss": 1.4551,
+      "step": 12158
+    },
+    {
+      "epoch": 1.3967032335879617,
+      "grad_norm": 0.6120827794075012,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 12159
+    },
+    {
+      "epoch": 1.3968181034977887,
+      "grad_norm": 0.6138426065444946,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 12160
+    },
+    {
+      "epoch": 1.396932973407616,
+      "grad_norm": 0.641486406326294,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 12161
+    },
+    {
+      "epoch": 1.397047843317443,
+      "grad_norm": 0.6255931258201599,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 12162
+    },
+    {
+      "epoch": 1.3971627132272701,
+      "grad_norm": 0.6870328187942505,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 12163
+    },
+    {
+      "epoch": 1.3972775831370972,
+      "grad_norm": 0.5824881196022034,
+      "learning_rate": 0.0001,
+      "loss": 1.4912,
+      "step": 12164
+    },
+    {
+      "epoch": 1.3973924530469244,
+      "grad_norm": 0.557790994644165,
+      "learning_rate": 0.0001,
+      "loss": 1.4021,
+      "step": 12165
+    },
+    {
+      "epoch": 1.3975073229567514,
+      "grad_norm": 0.6548678874969482,
+      "learning_rate": 0.0001,
+      "loss": 1.3511,
+      "step": 12166
+    },
+    {
+      "epoch": 1.3976221928665786,
+      "grad_norm": 0.5523027777671814,
+      "learning_rate": 0.0001,
+      "loss": 1.1925,
+      "step": 12167
+    },
+    {
+      "epoch": 1.3977370627764056,
+      "grad_norm": 0.5813341736793518,
+      "learning_rate": 0.0001,
+      "loss": 1.477,
+      "step": 12168
+    },
+    {
+      "epoch": 1.3978519326862329,
+      "grad_norm": 0.5378603935241699,
+      "learning_rate": 0.0001,
+      "loss": 1.2486,
+      "step": 12169
+    },
+    {
+      "epoch": 1.39796680259606,
+      "grad_norm": 0.5774914622306824,
+      "learning_rate": 0.0001,
+      "loss": 1.361,
+      "step": 12170
+    },
+    {
+      "epoch": 1.398081672505887,
+      "grad_norm": 0.602354884147644,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 12171
+    },
+    {
+      "epoch": 1.3981965424157141,
+      "grad_norm": 0.559851884841919,
+      "learning_rate": 0.0001,
+      "loss": 1.488,
+      "step": 12172
+    },
+    {
+      "epoch": 1.3983114123255413,
+      "grad_norm": 0.6218519806861877,
+      "learning_rate": 0.0001,
+      "loss": 1.355,
+      "step": 12173
+    },
+    {
+      "epoch": 1.3984262822353686,
+      "grad_norm": 0.6444374918937683,
+      "learning_rate": 0.0001,
+      "loss": 1.5331,
+      "step": 12174
+    },
+    {
+      "epoch": 1.3985411521451956,
+      "grad_norm": 0.5984706282615662,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 12175
+    },
+    {
+      "epoch": 1.3986560220550226,
+      "grad_norm": 0.6040021777153015,
+      "learning_rate": 0.0001,
+      "loss": 1.4337,
+      "step": 12176
+    },
+    {
+      "epoch": 1.3987708919648498,
+      "grad_norm": 0.5464459657669067,
+      "learning_rate": 0.0001,
+      "loss": 1.4417,
+      "step": 12177
+    },
+    {
+      "epoch": 1.398885761874677,
+      "grad_norm": 0.5808142423629761,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 12178
+    },
+    {
+      "epoch": 1.399000631784504,
+      "grad_norm": 0.5869821906089783,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 12179
+    },
+    {
+      "epoch": 1.399115501694331,
+      "grad_norm": 0.5694946050643921,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 12180
+    },
+    {
+      "epoch": 1.3992303716041583,
+      "grad_norm": 0.5913307666778564,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 12181
+    },
+    {
+      "epoch": 1.3993452415139855,
+      "grad_norm": 0.5427361130714417,
+      "learning_rate": 0.0001,
+      "loss": 1.4013,
+      "step": 12182
+    },
+    {
+      "epoch": 1.3994601114238125,
+      "grad_norm": 0.5571943521499634,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 12183
+    },
+    {
+      "epoch": 1.3995749813336396,
+      "grad_norm": 0.5840848684310913,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 12184
+    },
+    {
+      "epoch": 1.3996898512434668,
+      "grad_norm": 0.5792492628097534,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 12185
+    },
+    {
+      "epoch": 1.399804721153294,
+      "grad_norm": 0.5461111068725586,
+      "learning_rate": 0.0001,
+      "loss": 1.4224,
+      "step": 12186
+    },
+    {
+      "epoch": 1.399919591063121,
+      "grad_norm": 0.6092454195022583,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 12187
+    },
+    {
+      "epoch": 1.400034460972948,
+      "grad_norm": 0.5773031711578369,
+      "learning_rate": 0.0001,
+      "loss": 1.4072,
+      "step": 12188
+    },
+    {
+      "epoch": 1.4001493308827753,
+      "grad_norm": 0.6374520063400269,
+      "learning_rate": 0.0001,
+      "loss": 1.5028,
+      "step": 12189
+    },
+    {
+      "epoch": 1.4002642007926025,
+      "grad_norm": 0.5584102272987366,
+      "learning_rate": 0.0001,
+      "loss": 1.2906,
+      "step": 12190
+    },
+    {
+      "epoch": 1.4003790707024295,
+      "grad_norm": 0.6517335772514343,
+      "learning_rate": 0.0001,
+      "loss": 1.7582,
+      "step": 12191
+    },
+    {
+      "epoch": 1.4004939406122565,
+      "grad_norm": 0.5977898240089417,
+      "learning_rate": 0.0001,
+      "loss": 1.517,
+      "step": 12192
+    },
+    {
+      "epoch": 1.4006088105220837,
+      "grad_norm": 0.5880120992660522,
+      "learning_rate": 0.0001,
+      "loss": 1.6576,
+      "step": 12193
+    },
+    {
+      "epoch": 1.400723680431911,
+      "grad_norm": 0.5685825943946838,
+      "learning_rate": 0.0001,
+      "loss": 1.4966,
+      "step": 12194
+    },
+    {
+      "epoch": 1.400838550341738,
+      "grad_norm": 0.5807334780693054,
+      "learning_rate": 0.0001,
+      "loss": 1.3646,
+      "step": 12195
+    },
+    {
+      "epoch": 1.400953420251565,
+      "grad_norm": 0.602315366268158,
+      "learning_rate": 0.0001,
+      "loss": 1.2503,
+      "step": 12196
+    },
+    {
+      "epoch": 1.4010682901613922,
+      "grad_norm": 0.6140511631965637,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 12197
+    },
+    {
+      "epoch": 1.4011831600712195,
+      "grad_norm": 0.6030164361000061,
+      "learning_rate": 0.0001,
+      "loss": 1.4811,
+      "step": 12198
+    },
+    {
+      "epoch": 1.4012980299810465,
+      "grad_norm": 0.6865645051002502,
+      "learning_rate": 0.0001,
+      "loss": 1.5387,
+      "step": 12199
+    },
+    {
+      "epoch": 1.4014128998908735,
+      "grad_norm": 0.5891702175140381,
+      "learning_rate": 0.0001,
+      "loss": 1.3482,
+      "step": 12200
+    },
+    {
+      "epoch": 1.4015277698007007,
+      "grad_norm": 0.6280384063720703,
+      "learning_rate": 0.0001,
+      "loss": 1.4052,
+      "step": 12201
+    },
+    {
+      "epoch": 1.401642639710528,
+      "grad_norm": 0.5294430255889893,
+      "learning_rate": 0.0001,
+      "loss": 1.4011,
+      "step": 12202
+    },
+    {
+      "epoch": 1.401757509620355,
+      "grad_norm": 0.5137497782707214,
+      "learning_rate": 0.0001,
+      "loss": 1.3008,
+      "step": 12203
+    },
+    {
+      "epoch": 1.401872379530182,
+      "grad_norm": 0.6277523636817932,
+      "learning_rate": 0.0001,
+      "loss": 1.3415,
+      "step": 12204
+    },
+    {
+      "epoch": 1.4019872494400092,
+      "grad_norm": 0.5599936842918396,
+      "learning_rate": 0.0001,
+      "loss": 1.524,
+      "step": 12205
+    },
+    {
+      "epoch": 1.4021021193498364,
+      "grad_norm": 0.5398606657981873,
+      "learning_rate": 0.0001,
+      "loss": 1.3835,
+      "step": 12206
+    },
+    {
+      "epoch": 1.4022169892596634,
+      "grad_norm": 0.5670443773269653,
+      "learning_rate": 0.0001,
+      "loss": 1.4954,
+      "step": 12207
+    },
+    {
+      "epoch": 1.4023318591694904,
+      "grad_norm": 0.5825430154800415,
+      "learning_rate": 0.0001,
+      "loss": 1.4757,
+      "step": 12208
+    },
+    {
+      "epoch": 1.4024467290793177,
+      "grad_norm": 0.6514697670936584,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 12209
+    },
+    {
+      "epoch": 1.402561598989145,
+      "grad_norm": 0.5851296782493591,
+      "learning_rate": 0.0001,
+      "loss": 1.4696,
+      "step": 12210
+    },
+    {
+      "epoch": 1.402676468898972,
+      "grad_norm": 0.6211963295936584,
+      "learning_rate": 0.0001,
+      "loss": 1.192,
+      "step": 12211
+    },
+    {
+      "epoch": 1.402791338808799,
+      "grad_norm": 0.5917556881904602,
+      "learning_rate": 0.0001,
+      "loss": 1.3838,
+      "step": 12212
+    },
+    {
+      "epoch": 1.4029062087186261,
+      "grad_norm": 0.637103259563446,
+      "learning_rate": 0.0001,
+      "loss": 1.5705,
+      "step": 12213
+    },
+    {
+      "epoch": 1.4030210786284534,
+      "grad_norm": 0.6678568720817566,
+      "learning_rate": 0.0001,
+      "loss": 1.7563,
+      "step": 12214
+    },
+    {
+      "epoch": 1.4031359485382804,
+      "grad_norm": 0.6561994552612305,
+      "learning_rate": 0.0001,
+      "loss": 1.485,
+      "step": 12215
+    },
+    {
+      "epoch": 1.4032508184481074,
+      "grad_norm": 0.590522825717926,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 12216
+    },
+    {
+      "epoch": 1.4033656883579346,
+      "grad_norm": 0.5511714816093445,
+      "learning_rate": 0.0001,
+      "loss": 1.405,
+      "step": 12217
+    },
+    {
+      "epoch": 1.4034805582677619,
+      "grad_norm": 0.567674458026886,
+      "learning_rate": 0.0001,
+      "loss": 1.2435,
+      "step": 12218
+    },
+    {
+      "epoch": 1.4035954281775889,
+      "grad_norm": 0.5764709711074829,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 12219
+    },
+    {
+      "epoch": 1.4037102980874159,
+      "grad_norm": 0.5814299583435059,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 12220
+    },
+    {
+      "epoch": 1.403825167997243,
+      "grad_norm": 0.6138479709625244,
+      "learning_rate": 0.0001,
+      "loss": 1.3847,
+      "step": 12221
+    },
+    {
+      "epoch": 1.4039400379070703,
+      "grad_norm": 0.577967643737793,
+      "learning_rate": 0.0001,
+      "loss": 1.4667,
+      "step": 12222
+    },
+    {
+      "epoch": 1.4040549078168973,
+      "grad_norm": 0.5953364372253418,
+      "learning_rate": 0.0001,
+      "loss": 1.433,
+      "step": 12223
+    },
+    {
+      "epoch": 1.4041697777267244,
+      "grad_norm": 0.6030754446983337,
+      "learning_rate": 0.0001,
+      "loss": 1.3456,
+      "step": 12224
+    },
+    {
+      "epoch": 1.4042846476365516,
+      "grad_norm": 0.6069220304489136,
+      "learning_rate": 0.0001,
+      "loss": 1.4485,
+      "step": 12225
+    },
+    {
+      "epoch": 1.4043995175463788,
+      "grad_norm": 0.5867301821708679,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 12226
+    },
+    {
+      "epoch": 1.4045143874562058,
+      "grad_norm": 0.6797637343406677,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 12227
+    },
+    {
+      "epoch": 1.4046292573660328,
+      "grad_norm": 0.6498664617538452,
+      "learning_rate": 0.0001,
+      "loss": 1.2396,
+      "step": 12228
+    },
+    {
+      "epoch": 1.40474412727586,
+      "grad_norm": 0.645907461643219,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 12229
+    },
+    {
+      "epoch": 1.4048589971856873,
+      "grad_norm": 0.5839877128601074,
+      "learning_rate": 0.0001,
+      "loss": 1.366,
+      "step": 12230
+    },
+    {
+      "epoch": 1.4049738670955143,
+      "grad_norm": 0.5825426578521729,
+      "learning_rate": 0.0001,
+      "loss": 1.3383,
+      "step": 12231
+    },
+    {
+      "epoch": 1.4050887370053413,
+      "grad_norm": 0.6786919236183167,
+      "learning_rate": 0.0001,
+      "loss": 1.6,
+      "step": 12232
+    },
+    {
+      "epoch": 1.4052036069151685,
+      "grad_norm": 0.572919487953186,
+      "learning_rate": 0.0001,
+      "loss": 1.5307,
+      "step": 12233
+    },
+    {
+      "epoch": 1.4053184768249958,
+      "grad_norm": 0.5826205611228943,
+      "learning_rate": 0.0001,
+      "loss": 1.227,
+      "step": 12234
+    },
+    {
+      "epoch": 1.4054333467348228,
+      "grad_norm": 0.5886760354042053,
+      "learning_rate": 0.0001,
+      "loss": 1.3468,
+      "step": 12235
+    },
+    {
+      "epoch": 1.40554821664465,
+      "grad_norm": 0.579044759273529,
+      "learning_rate": 0.0001,
+      "loss": 1.3405,
+      "step": 12236
+    },
+    {
+      "epoch": 1.405663086554477,
+      "grad_norm": 0.5720400810241699,
+      "learning_rate": 0.0001,
+      "loss": 1.5135,
+      "step": 12237
+    },
+    {
+      "epoch": 1.4057779564643043,
+      "grad_norm": 0.5574225783348083,
+      "learning_rate": 0.0001,
+      "loss": 1.3995,
+      "step": 12238
+    },
+    {
+      "epoch": 1.4058928263741313,
+      "grad_norm": 0.5698866844177246,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 12239
+    },
+    {
+      "epoch": 1.4060076962839585,
+      "grad_norm": 0.5897749066352844,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 12240
+    },
+    {
+      "epoch": 1.4061225661937855,
+      "grad_norm": 0.5804178714752197,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 12241
+    },
+    {
+      "epoch": 1.4062374361036127,
+      "grad_norm": 0.5859702229499817,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 12242
+    },
+    {
+      "epoch": 1.4063523060134397,
+      "grad_norm": 0.5918383598327637,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 12243
+    },
+    {
+      "epoch": 1.406467175923267,
+      "grad_norm": 0.5748845934867859,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 12244
+    },
+    {
+      "epoch": 1.406582045833094,
+      "grad_norm": 0.6131268739700317,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 12245
+    },
+    {
+      "epoch": 1.4066969157429212,
+      "grad_norm": 0.6054122447967529,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 12246
+    },
+    {
+      "epoch": 1.4068117856527482,
+      "grad_norm": 0.6443667411804199,
+      "learning_rate": 0.0001,
+      "loss": 1.5172,
+      "step": 12247
+    },
+    {
+      "epoch": 1.4069266555625755,
+      "grad_norm": 0.6411632895469666,
+      "learning_rate": 0.0001,
+      "loss": 1.4362,
+      "step": 12248
+    },
+    {
+      "epoch": 1.4070415254724025,
+      "grad_norm": 0.5433965921401978,
+      "learning_rate": 0.0001,
+      "loss": 1.4244,
+      "step": 12249
+    },
+    {
+      "epoch": 1.4071563953822297,
+      "grad_norm": 0.6453516483306885,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 12250
+    },
+    {
+      "epoch": 1.4072712652920567,
+      "grad_norm": 0.5545192360877991,
+      "learning_rate": 0.0001,
+      "loss": 1.1791,
+      "step": 12251
+    },
+    {
+      "epoch": 1.407386135201884,
+      "grad_norm": 0.5485440492630005,
+      "learning_rate": 0.0001,
+      "loss": 1.2924,
+      "step": 12252
+    },
+    {
+      "epoch": 1.407501005111711,
+      "grad_norm": 0.614592432975769,
+      "learning_rate": 0.0001,
+      "loss": 1.2284,
+      "step": 12253
+    },
+    {
+      "epoch": 1.4076158750215382,
+      "grad_norm": 0.5941545963287354,
+      "learning_rate": 0.0001,
+      "loss": 1.4719,
+      "step": 12254
+    },
+    {
+      "epoch": 1.4077307449313652,
+      "grad_norm": 0.5332536697387695,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 12255
+    },
+    {
+      "epoch": 1.4078456148411924,
+      "grad_norm": 0.6020111441612244,
+      "learning_rate": 0.0001,
+      "loss": 1.603,
+      "step": 12256
+    },
+    {
+      "epoch": 1.4079604847510194,
+      "grad_norm": 0.6182122826576233,
+      "learning_rate": 0.0001,
+      "loss": 1.4545,
+      "step": 12257
+    },
+    {
+      "epoch": 1.4080753546608467,
+      "grad_norm": 0.5976362824440002,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 12258
+    },
+    {
+      "epoch": 1.4081902245706737,
+      "grad_norm": 0.5798993706703186,
+      "learning_rate": 0.0001,
+      "loss": 1.3952,
+      "step": 12259
+    },
+    {
+      "epoch": 1.408305094480501,
+      "grad_norm": 0.5920160412788391,
+      "learning_rate": 0.0001,
+      "loss": 1.4257,
+      "step": 12260
+    },
+    {
+      "epoch": 1.408419964390328,
+      "grad_norm": 0.6999684572219849,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 12261
+    },
+    {
+      "epoch": 1.4085348343001551,
+      "grad_norm": 0.6646612286567688,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 12262
+    },
+    {
+      "epoch": 1.4086497042099821,
+      "grad_norm": 0.6494892239570618,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 12263
+    },
+    {
+      "epoch": 1.4087645741198094,
+      "grad_norm": 0.6330276727676392,
+      "learning_rate": 0.0001,
+      "loss": 1.5038,
+      "step": 12264
+    },
+    {
+      "epoch": 1.4088794440296364,
+      "grad_norm": 0.6663432717323303,
+      "learning_rate": 0.0001,
+      "loss": 1.3514,
+      "step": 12265
+    },
+    {
+      "epoch": 1.4089943139394636,
+      "grad_norm": 0.6159442067146301,
+      "learning_rate": 0.0001,
+      "loss": 1.5268,
+      "step": 12266
+    },
+    {
+      "epoch": 1.4091091838492906,
+      "grad_norm": 0.5885686278343201,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 12267
+    },
+    {
+      "epoch": 1.4092240537591179,
+      "grad_norm": 0.6478608250617981,
+      "learning_rate": 0.0001,
+      "loss": 1.3103,
+      "step": 12268
+    },
+    {
+      "epoch": 1.4093389236689449,
+      "grad_norm": 0.598627507686615,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 12269
+    },
+    {
+      "epoch": 1.409453793578772,
+      "grad_norm": 0.5440186262130737,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 12270
+    },
+    {
+      "epoch": 1.409568663488599,
+      "grad_norm": 0.5913492441177368,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 12271
+    },
+    {
+      "epoch": 1.4096835333984263,
+      "grad_norm": 0.6720054149627686,
+      "learning_rate": 0.0001,
+      "loss": 1.6878,
+      "step": 12272
+    },
+    {
+      "epoch": 1.4097984033082533,
+      "grad_norm": 0.550347089767456,
+      "learning_rate": 0.0001,
+      "loss": 1.4825,
+      "step": 12273
+    },
+    {
+      "epoch": 1.4099132732180806,
+      "grad_norm": 0.5974934697151184,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 12274
+    },
+    {
+      "epoch": 1.4100281431279076,
+      "grad_norm": 0.568218469619751,
+      "learning_rate": 0.0001,
+      "loss": 1.3921,
+      "step": 12275
+    },
+    {
+      "epoch": 1.4101430130377348,
+      "grad_norm": 0.6132494211196899,
+      "learning_rate": 0.0001,
+      "loss": 1.3798,
+      "step": 12276
+    },
+    {
+      "epoch": 1.4102578829475618,
+      "grad_norm": 0.5562697649002075,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 12277
+    },
+    {
+      "epoch": 1.410372752857389,
+      "grad_norm": 0.5881964564323425,
+      "learning_rate": 0.0001,
+      "loss": 1.4807,
+      "step": 12278
+    },
+    {
+      "epoch": 1.410487622767216,
+      "grad_norm": 0.6052611470222473,
+      "learning_rate": 0.0001,
+      "loss": 1.3232,
+      "step": 12279
+    },
+    {
+      "epoch": 1.4106024926770433,
+      "grad_norm": 0.568783164024353,
+      "learning_rate": 0.0001,
+      "loss": 1.4518,
+      "step": 12280
+    },
+    {
+      "epoch": 1.4107173625868703,
+      "grad_norm": 0.5390822291374207,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 12281
+    },
+    {
+      "epoch": 1.4108322324966975,
+      "grad_norm": 0.6221309900283813,
+      "learning_rate": 0.0001,
+      "loss": 1.4751,
+      "step": 12282
+    },
+    {
+      "epoch": 1.4109471024065245,
+      "grad_norm": 0.6314241886138916,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 12283
+    },
+    {
+      "epoch": 1.4110619723163518,
+      "grad_norm": 0.5789543390274048,
+      "learning_rate": 0.0001,
+      "loss": 1.4302,
+      "step": 12284
+    },
+    {
+      "epoch": 1.4111768422261788,
+      "grad_norm": 0.6005547642707825,
+      "learning_rate": 0.0001,
+      "loss": 1.2483,
+      "step": 12285
+    },
+    {
+      "epoch": 1.411291712136006,
+      "grad_norm": 0.6064243912696838,
+      "learning_rate": 0.0001,
+      "loss": 1.5456,
+      "step": 12286
+    },
+    {
+      "epoch": 1.411406582045833,
+      "grad_norm": 0.6371053457260132,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 12287
+    },
+    {
+      "epoch": 1.4115214519556603,
+      "grad_norm": 0.6137773394584656,
+      "learning_rate": 0.0001,
+      "loss": 1.4473,
+      "step": 12288
+    },
+    {
+      "epoch": 1.4116363218654873,
+      "grad_norm": 0.6478999853134155,
+      "learning_rate": 0.0001,
+      "loss": 1.6225,
+      "step": 12289
+    },
+    {
+      "epoch": 1.4117511917753145,
+      "grad_norm": 0.5467931032180786,
+      "learning_rate": 0.0001,
+      "loss": 1.3421,
+      "step": 12290
+    },
+    {
+      "epoch": 1.4118660616851415,
+      "grad_norm": 0.5850721001625061,
+      "learning_rate": 0.0001,
+      "loss": 1.3787,
+      "step": 12291
+    },
+    {
+      "epoch": 1.4119809315949687,
+      "grad_norm": 0.5739957690238953,
+      "learning_rate": 0.0001,
+      "loss": 1.4527,
+      "step": 12292
+    },
+    {
+      "epoch": 1.4120958015047957,
+      "grad_norm": 0.6541691422462463,
+      "learning_rate": 0.0001,
+      "loss": 1.5428,
+      "step": 12293
+    },
+    {
+      "epoch": 1.412210671414623,
+      "grad_norm": 0.6030059456825256,
+      "learning_rate": 0.0001,
+      "loss": 1.4744,
+      "step": 12294
+    },
+    {
+      "epoch": 1.41232554132445,
+      "grad_norm": 0.5912495255470276,
+      "learning_rate": 0.0001,
+      "loss": 1.3846,
+      "step": 12295
+    },
+    {
+      "epoch": 1.4124404112342772,
+      "grad_norm": 0.6480311751365662,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 12296
+    },
+    {
+      "epoch": 1.4125552811441042,
+      "grad_norm": 0.6374021768569946,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 12297
+    },
+    {
+      "epoch": 1.4126701510539315,
+      "grad_norm": 0.7061396241188049,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 12298
+    },
+    {
+      "epoch": 1.4127850209637585,
+      "grad_norm": 0.5619686841964722,
+      "learning_rate": 0.0001,
+      "loss": 1.4337,
+      "step": 12299
+    },
+    {
+      "epoch": 1.4128998908735857,
+      "grad_norm": 0.5579246282577515,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 12300
+    },
+    {
+      "epoch": 1.4130147607834127,
+      "grad_norm": 0.5994849801063538,
+      "learning_rate": 0.0001,
+      "loss": 1.4444,
+      "step": 12301
+    },
+    {
+      "epoch": 1.41312963069324,
+      "grad_norm": 0.5560767650604248,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 12302
+    },
+    {
+      "epoch": 1.413244500603067,
+      "grad_norm": 0.5769314765930176,
+      "learning_rate": 0.0001,
+      "loss": 1.3982,
+      "step": 12303
+    },
+    {
+      "epoch": 1.4133593705128942,
+      "grad_norm": 0.5469468235969543,
+      "learning_rate": 0.0001,
+      "loss": 1.3255,
+      "step": 12304
+    },
+    {
+      "epoch": 1.4134742404227212,
+      "grad_norm": 0.5225635170936584,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 12305
+    },
+    {
+      "epoch": 1.4135891103325484,
+      "grad_norm": 0.5575945377349854,
+      "learning_rate": 0.0001,
+      "loss": 1.4781,
+      "step": 12306
+    },
+    {
+      "epoch": 1.4137039802423756,
+      "grad_norm": 0.6048114895820618,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 12307
+    },
+    {
+      "epoch": 1.4138188501522027,
+      "grad_norm": 0.560836911201477,
+      "learning_rate": 0.0001,
+      "loss": 1.3568,
+      "step": 12308
+    },
+    {
+      "epoch": 1.4139337200620297,
+      "grad_norm": 0.5700927376747131,
+      "learning_rate": 0.0001,
+      "loss": 1.4198,
+      "step": 12309
+    },
+    {
+      "epoch": 1.414048589971857,
+      "grad_norm": 0.549271285533905,
+      "learning_rate": 0.0001,
+      "loss": 1.4586,
+      "step": 12310
+    },
+    {
+      "epoch": 1.4141634598816841,
+      "grad_norm": 0.6149371266365051,
+      "learning_rate": 0.0001,
+      "loss": 1.4705,
+      "step": 12311
+    },
+    {
+      "epoch": 1.4142783297915111,
+      "grad_norm": 0.6246234178543091,
+      "learning_rate": 0.0001,
+      "loss": 1.427,
+      "step": 12312
+    },
+    {
+      "epoch": 1.4143931997013381,
+      "grad_norm": 0.6859720349311829,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 12313
+    },
+    {
+      "epoch": 1.4145080696111654,
+      "grad_norm": 0.5962328910827637,
+      "learning_rate": 0.0001,
+      "loss": 1.4477,
+      "step": 12314
+    },
+    {
+      "epoch": 1.4146229395209926,
+      "grad_norm": 0.6179618239402771,
+      "learning_rate": 0.0001,
+      "loss": 1.3629,
+      "step": 12315
+    },
+    {
+      "epoch": 1.4147378094308196,
+      "grad_norm": 0.5705560445785522,
+      "learning_rate": 0.0001,
+      "loss": 1.4853,
+      "step": 12316
+    },
+    {
+      "epoch": 1.4148526793406466,
+      "grad_norm": 0.6366815567016602,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 12317
+    },
+    {
+      "epoch": 1.4149675492504739,
+      "grad_norm": 0.664026141166687,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 12318
+    },
+    {
+      "epoch": 1.415082419160301,
+      "grad_norm": 0.6808172464370728,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 12319
+    },
+    {
+      "epoch": 1.415197289070128,
+      "grad_norm": 0.5884132385253906,
+      "learning_rate": 0.0001,
+      "loss": 1.4446,
+      "step": 12320
+    },
+    {
+      "epoch": 1.415312158979955,
+      "grad_norm": 0.6115908026695251,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 12321
+    },
+    {
+      "epoch": 1.4154270288897823,
+      "grad_norm": 0.610191285610199,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 12322
+    },
+    {
+      "epoch": 1.4155418987996096,
+      "grad_norm": 0.5975642800331116,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 12323
+    },
+    {
+      "epoch": 1.4156567687094366,
+      "grad_norm": 0.5843347907066345,
+      "learning_rate": 0.0001,
+      "loss": 1.4948,
+      "step": 12324
+    },
+    {
+      "epoch": 1.4157716386192636,
+      "grad_norm": 0.5770195126533508,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 12325
+    },
+    {
+      "epoch": 1.4158865085290908,
+      "grad_norm": 0.5760985016822815,
+      "learning_rate": 0.0001,
+      "loss": 1.4043,
+      "step": 12326
+    },
+    {
+      "epoch": 1.416001378438918,
+      "grad_norm": 0.5809157490730286,
+      "learning_rate": 0.0001,
+      "loss": 1.5153,
+      "step": 12327
+    },
+    {
+      "epoch": 1.416116248348745,
+      "grad_norm": 0.59894198179245,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 12328
+    },
+    {
+      "epoch": 1.416231118258572,
+      "grad_norm": 0.5635504722595215,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 12329
+    },
+    {
+      "epoch": 1.4163459881683993,
+      "grad_norm": 0.5939011573791504,
+      "learning_rate": 0.0001,
+      "loss": 1.6191,
+      "step": 12330
+    },
+    {
+      "epoch": 1.4164608580782265,
+      "grad_norm": 0.5841618776321411,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 12331
+    },
+    {
+      "epoch": 1.4165757279880535,
+      "grad_norm": 0.5628432631492615,
+      "learning_rate": 0.0001,
+      "loss": 1.3689,
+      "step": 12332
+    },
+    {
+      "epoch": 1.4166905978978805,
+      "grad_norm": 0.5952514410018921,
+      "learning_rate": 0.0001,
+      "loss": 1.3696,
+      "step": 12333
+    },
+    {
+      "epoch": 1.4168054678077078,
+      "grad_norm": 0.5918675065040588,
+      "learning_rate": 0.0001,
+      "loss": 1.4191,
+      "step": 12334
+    },
+    {
+      "epoch": 1.416920337717535,
+      "grad_norm": 0.629119873046875,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 12335
+    },
+    {
+      "epoch": 1.417035207627362,
+      "grad_norm": 0.5868558287620544,
+      "learning_rate": 0.0001,
+      "loss": 1.4134,
+      "step": 12336
+    },
+    {
+      "epoch": 1.417150077537189,
+      "grad_norm": 0.6055565476417542,
+      "learning_rate": 0.0001,
+      "loss": 1.3543,
+      "step": 12337
+    },
+    {
+      "epoch": 1.4172649474470163,
+      "grad_norm": 0.5665161609649658,
+      "learning_rate": 0.0001,
+      "loss": 1.4376,
+      "step": 12338
+    },
+    {
+      "epoch": 1.4173798173568435,
+      "grad_norm": 0.6178751587867737,
+      "learning_rate": 0.0001,
+      "loss": 1.4882,
+      "step": 12339
+    },
+    {
+      "epoch": 1.4174946872666705,
+      "grad_norm": 0.6303858160972595,
+      "learning_rate": 0.0001,
+      "loss": 1.5314,
+      "step": 12340
+    },
+    {
+      "epoch": 1.4176095571764975,
+      "grad_norm": 0.6005488634109497,
+      "learning_rate": 0.0001,
+      "loss": 1.4275,
+      "step": 12341
+    },
+    {
+      "epoch": 1.4177244270863247,
+      "grad_norm": 0.6242131590843201,
+      "learning_rate": 0.0001,
+      "loss": 1.2941,
+      "step": 12342
+    },
+    {
+      "epoch": 1.417839296996152,
+      "grad_norm": 0.6233993768692017,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 12343
+    },
+    {
+      "epoch": 1.417954166905979,
+      "grad_norm": 0.6324300169944763,
+      "learning_rate": 0.0001,
+      "loss": 1.5645,
+      "step": 12344
+    },
+    {
+      "epoch": 1.418069036815806,
+      "grad_norm": 0.6000421047210693,
+      "learning_rate": 0.0001,
+      "loss": 1.4319,
+      "step": 12345
+    },
+    {
+      "epoch": 1.4181839067256332,
+      "grad_norm": 0.6152875423431396,
+      "learning_rate": 0.0001,
+      "loss": 1.3411,
+      "step": 12346
+    },
+    {
+      "epoch": 1.4182987766354604,
+      "grad_norm": 0.6526452898979187,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 12347
+    },
+    {
+      "epoch": 1.4184136465452875,
+      "grad_norm": 0.6107121109962463,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 12348
+    },
+    {
+      "epoch": 1.4185285164551145,
+      "grad_norm": 0.5946810841560364,
+      "learning_rate": 0.0001,
+      "loss": 1.5623,
+      "step": 12349
+    },
+    {
+      "epoch": 1.4186433863649417,
+      "grad_norm": 0.5849339365959167,
+      "learning_rate": 0.0001,
+      "loss": 1.4287,
+      "step": 12350
+    },
+    {
+      "epoch": 1.418758256274769,
+      "grad_norm": 0.6404920816421509,
+      "learning_rate": 0.0001,
+      "loss": 1.596,
+      "step": 12351
+    },
+    {
+      "epoch": 1.418873126184596,
+      "grad_norm": 0.6268463134765625,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 12352
+    },
+    {
+      "epoch": 1.418987996094423,
+      "grad_norm": 0.5831562876701355,
+      "learning_rate": 0.0001,
+      "loss": 1.3506,
+      "step": 12353
+    },
+    {
+      "epoch": 1.4191028660042502,
+      "grad_norm": 0.6019381284713745,
+      "learning_rate": 0.0001,
+      "loss": 1.428,
+      "step": 12354
+    },
+    {
+      "epoch": 1.4192177359140774,
+      "grad_norm": 0.6480855941772461,
+      "learning_rate": 0.0001,
+      "loss": 1.4975,
+      "step": 12355
+    },
+    {
+      "epoch": 1.4193326058239044,
+      "grad_norm": 0.6324148774147034,
+      "learning_rate": 0.0001,
+      "loss": 1.3886,
+      "step": 12356
+    },
+    {
+      "epoch": 1.4194474757337314,
+      "grad_norm": 0.5891407132148743,
+      "learning_rate": 0.0001,
+      "loss": 1.4664,
+      "step": 12357
+    },
+    {
+      "epoch": 1.4195623456435587,
+      "grad_norm": 0.5745696425437927,
+      "learning_rate": 0.0001,
+      "loss": 1.4676,
+      "step": 12358
+    },
+    {
+      "epoch": 1.4196772155533859,
+      "grad_norm": 0.5865878462791443,
+      "learning_rate": 0.0001,
+      "loss": 1.4144,
+      "step": 12359
+    },
+    {
+      "epoch": 1.419792085463213,
+      "grad_norm": 0.5463162064552307,
+      "learning_rate": 0.0001,
+      "loss": 1.2957,
+      "step": 12360
+    },
+    {
+      "epoch": 1.41990695537304,
+      "grad_norm": 0.5702036619186401,
+      "learning_rate": 0.0001,
+      "loss": 1.3271,
+      "step": 12361
+    },
+    {
+      "epoch": 1.4200218252828671,
+      "grad_norm": 0.5767183899879456,
+      "learning_rate": 0.0001,
+      "loss": 1.5055,
+      "step": 12362
+    },
+    {
+      "epoch": 1.4201366951926944,
+      "grad_norm": 0.6075996160507202,
+      "learning_rate": 0.0001,
+      "loss": 1.5536,
+      "step": 12363
+    },
+    {
+      "epoch": 1.4202515651025214,
+      "grad_norm": 0.6220189332962036,
+      "learning_rate": 0.0001,
+      "loss": 1.4745,
+      "step": 12364
+    },
+    {
+      "epoch": 1.4203664350123484,
+      "grad_norm": 0.5823253393173218,
+      "learning_rate": 0.0001,
+      "loss": 1.4564,
+      "step": 12365
+    },
+    {
+      "epoch": 1.4204813049221756,
+      "grad_norm": 0.592589259147644,
+      "learning_rate": 0.0001,
+      "loss": 1.3409,
+      "step": 12366
+    },
+    {
+      "epoch": 1.4205961748320028,
+      "grad_norm": 0.624343991279602,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 12367
+    },
+    {
+      "epoch": 1.4207110447418299,
+      "grad_norm": 0.568588137626648,
+      "learning_rate": 0.0001,
+      "loss": 1.4421,
+      "step": 12368
+    },
+    {
+      "epoch": 1.4208259146516569,
+      "grad_norm": 0.6244280338287354,
+      "learning_rate": 0.0001,
+      "loss": 1.3673,
+      "step": 12369
+    },
+    {
+      "epoch": 1.420940784561484,
+      "grad_norm": 0.5823641419410706,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 12370
+    },
+    {
+      "epoch": 1.4210556544713113,
+      "grad_norm": 0.627984344959259,
+      "learning_rate": 0.0001,
+      "loss": 1.396,
+      "step": 12371
+    },
+    {
+      "epoch": 1.4211705243811383,
+      "grad_norm": 0.6518086194992065,
+      "learning_rate": 0.0001,
+      "loss": 1.1716,
+      "step": 12372
+    },
+    {
+      "epoch": 1.4212853942909656,
+      "grad_norm": 0.6243144273757935,
+      "learning_rate": 0.0001,
+      "loss": 1.5423,
+      "step": 12373
+    },
+    {
+      "epoch": 1.4214002642007926,
+      "grad_norm": 0.6055872440338135,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 12374
+    },
+    {
+      "epoch": 1.4215151341106198,
+      "grad_norm": 0.5985282063484192,
+      "learning_rate": 0.0001,
+      "loss": 1.4185,
+      "step": 12375
+    },
+    {
+      "epoch": 1.4216300040204468,
+      "grad_norm": 0.5503548383712769,
+      "learning_rate": 0.0001,
+      "loss": 1.524,
+      "step": 12376
+    },
+    {
+      "epoch": 1.421744873930274,
+      "grad_norm": 0.5615320205688477,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 12377
+    },
+    {
+      "epoch": 1.421859743840101,
+      "grad_norm": 0.5671427249908447,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 12378
+    },
+    {
+      "epoch": 1.4219746137499283,
+      "grad_norm": 0.6086657643318176,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 12379
+    },
+    {
+      "epoch": 1.4220894836597553,
+      "grad_norm": 0.567254364490509,
+      "learning_rate": 0.0001,
+      "loss": 1.4461,
+      "step": 12380
+    },
+    {
+      "epoch": 1.4222043535695825,
+      "grad_norm": 0.655489444732666,
+      "learning_rate": 0.0001,
+      "loss": 1.4207,
+      "step": 12381
+    },
+    {
+      "epoch": 1.4223192234794095,
+      "grad_norm": 0.5274955630302429,
+      "learning_rate": 0.0001,
+      "loss": 1.2654,
+      "step": 12382
+    },
+    {
+      "epoch": 1.4224340933892368,
+      "grad_norm": 0.574601948261261,
+      "learning_rate": 0.0001,
+      "loss": 1.4474,
+      "step": 12383
+    },
+    {
+      "epoch": 1.4225489632990638,
+      "grad_norm": 0.5798234343528748,
+      "learning_rate": 0.0001,
+      "loss": 1.4091,
+      "step": 12384
+    },
+    {
+      "epoch": 1.422663833208891,
+      "grad_norm": 0.6089479923248291,
+      "learning_rate": 0.0001,
+      "loss": 1.2936,
+      "step": 12385
+    },
+    {
+      "epoch": 1.422778703118718,
+      "grad_norm": 0.5854637026786804,
+      "learning_rate": 0.0001,
+      "loss": 1.4876,
+      "step": 12386
+    },
+    {
+      "epoch": 1.4228935730285452,
+      "grad_norm": 0.681310772895813,
+      "learning_rate": 0.0001,
+      "loss": 1.7594,
+      "step": 12387
+    },
+    {
+      "epoch": 1.4230084429383723,
+      "grad_norm": 0.5869307518005371,
+      "learning_rate": 0.0001,
+      "loss": 1.3828,
+      "step": 12388
+    },
+    {
+      "epoch": 1.4231233128481995,
+      "grad_norm": 0.5680341124534607,
+      "learning_rate": 0.0001,
+      "loss": 1.3795,
+      "step": 12389
+    },
+    {
+      "epoch": 1.4232381827580265,
+      "grad_norm": 0.5726298093795776,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 12390
+    },
+    {
+      "epoch": 1.4233530526678537,
+      "grad_norm": 0.6122094988822937,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 12391
+    },
+    {
+      "epoch": 1.4234679225776807,
+      "grad_norm": 0.6463402509689331,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 12392
+    },
+    {
+      "epoch": 1.423582792487508,
+      "grad_norm": 0.6010882258415222,
+      "learning_rate": 0.0001,
+      "loss": 1.3532,
+      "step": 12393
+    },
+    {
+      "epoch": 1.423697662397335,
+      "grad_norm": 0.5905262231826782,
+      "learning_rate": 0.0001,
+      "loss": 1.2767,
+      "step": 12394
+    },
+    {
+      "epoch": 1.4238125323071622,
+      "grad_norm": 0.5448614358901978,
+      "learning_rate": 0.0001,
+      "loss": 1.4319,
+      "step": 12395
+    },
+    {
+      "epoch": 1.4239274022169892,
+      "grad_norm": 0.5370798707008362,
+      "learning_rate": 0.0001,
+      "loss": 1.4836,
+      "step": 12396
+    },
+    {
+      "epoch": 1.4240422721268164,
+      "grad_norm": 0.5748822689056396,
+      "learning_rate": 0.0001,
+      "loss": 1.4327,
+      "step": 12397
+    },
+    {
+      "epoch": 1.4241571420366435,
+      "grad_norm": 0.6279718279838562,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 12398
+    },
+    {
+      "epoch": 1.4242720119464707,
+      "grad_norm": 0.612967848777771,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 12399
+    },
+    {
+      "epoch": 1.4243868818562977,
+      "grad_norm": 0.5825915336608887,
+      "learning_rate": 0.0001,
+      "loss": 1.4405,
+      "step": 12400
+    },
+    {
+      "epoch": 1.424501751766125,
+      "grad_norm": 0.5820374488830566,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 12401
+    },
+    {
+      "epoch": 1.424616621675952,
+      "grad_norm": 0.5795537829399109,
+      "learning_rate": 0.0001,
+      "loss": 1.3931,
+      "step": 12402
+    },
+    {
+      "epoch": 1.4247314915857792,
+      "grad_norm": 0.5845256447792053,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 12403
+    },
+    {
+      "epoch": 1.4248463614956062,
+      "grad_norm": 0.6226243376731873,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 12404
+    },
+    {
+      "epoch": 1.4249612314054334,
+      "grad_norm": 0.5559086203575134,
+      "learning_rate": 0.0001,
+      "loss": 1.488,
+      "step": 12405
+    },
+    {
+      "epoch": 1.4250761013152604,
+      "grad_norm": 0.5699470043182373,
+      "learning_rate": 0.0001,
+      "loss": 1.4175,
+      "step": 12406
+    },
+    {
+      "epoch": 1.4251909712250876,
+      "grad_norm": 0.6197097301483154,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 12407
+    },
+    {
+      "epoch": 1.4253058411349147,
+      "grad_norm": 0.5672743320465088,
+      "learning_rate": 0.0001,
+      "loss": 1.4876,
+      "step": 12408
+    },
+    {
+      "epoch": 1.4254207110447419,
+      "grad_norm": 0.5639980435371399,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 12409
+    },
+    {
+      "epoch": 1.425535580954569,
+      "grad_norm": 0.6176483631134033,
+      "learning_rate": 0.0001,
+      "loss": 1.6189,
+      "step": 12410
+    },
+    {
+      "epoch": 1.4256504508643961,
+      "grad_norm": 0.599812388420105,
+      "learning_rate": 0.0001,
+      "loss": 1.3885,
+      "step": 12411
+    },
+    {
+      "epoch": 1.4257653207742231,
+      "grad_norm": 0.5804879069328308,
+      "learning_rate": 0.0001,
+      "loss": 1.4213,
+      "step": 12412
+    },
+    {
+      "epoch": 1.4258801906840504,
+      "grad_norm": 0.553686261177063,
+      "learning_rate": 0.0001,
+      "loss": 1.4293,
+      "step": 12413
+    },
+    {
+      "epoch": 1.4259950605938774,
+      "grad_norm": 0.5969316959381104,
+      "learning_rate": 0.0001,
+      "loss": 1.4132,
+      "step": 12414
+    },
+    {
+      "epoch": 1.4261099305037046,
+      "grad_norm": 0.6206128597259521,
+      "learning_rate": 0.0001,
+      "loss": 1.4204,
+      "step": 12415
+    },
+    {
+      "epoch": 1.4262248004135316,
+      "grad_norm": 0.6273772120475769,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 12416
+    },
+    {
+      "epoch": 1.4263396703233588,
+      "grad_norm": 0.5739526748657227,
+      "learning_rate": 0.0001,
+      "loss": 1.5694,
+      "step": 12417
+    },
+    {
+      "epoch": 1.4264545402331859,
+      "grad_norm": 0.5758925080299377,
+      "learning_rate": 0.0001,
+      "loss": 1.3505,
+      "step": 12418
+    },
+    {
+      "epoch": 1.426569410143013,
+      "grad_norm": 0.5880077481269836,
+      "learning_rate": 0.0001,
+      "loss": 1.4535,
+      "step": 12419
+    },
+    {
+      "epoch": 1.42668428005284,
+      "grad_norm": 0.6464236378669739,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 12420
+    },
+    {
+      "epoch": 1.4267991499626673,
+      "grad_norm": 0.6200028657913208,
+      "learning_rate": 0.0001,
+      "loss": 1.4498,
+      "step": 12421
+    },
+    {
+      "epoch": 1.4269140198724943,
+      "grad_norm": 0.5777352452278137,
+      "learning_rate": 0.0001,
+      "loss": 1.3672,
+      "step": 12422
+    },
+    {
+      "epoch": 1.4270288897823216,
+      "grad_norm": 0.6160234808921814,
+      "learning_rate": 0.0001,
+      "loss": 1.6537,
+      "step": 12423
+    },
+    {
+      "epoch": 1.4271437596921486,
+      "grad_norm": 0.5613859295845032,
+      "learning_rate": 0.0001,
+      "loss": 1.1053,
+      "step": 12424
+    },
+    {
+      "epoch": 1.4272586296019758,
+      "grad_norm": 0.5593403577804565,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 12425
+    },
+    {
+      "epoch": 1.4273734995118028,
+      "grad_norm": 0.5781793594360352,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 12426
+    },
+    {
+      "epoch": 1.42748836942163,
+      "grad_norm": 0.5942292809486389,
+      "learning_rate": 0.0001,
+      "loss": 1.4096,
+      "step": 12427
+    },
+    {
+      "epoch": 1.427603239331457,
+      "grad_norm": 0.5752783417701721,
+      "learning_rate": 0.0001,
+      "loss": 1.3625,
+      "step": 12428
+    },
+    {
+      "epoch": 1.4277181092412843,
+      "grad_norm": 0.7537838816642761,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 12429
+    },
+    {
+      "epoch": 1.4278329791511113,
+      "grad_norm": 0.5820350050926208,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 12430
+    },
+    {
+      "epoch": 1.4279478490609385,
+      "grad_norm": 0.5638639330863953,
+      "learning_rate": 0.0001,
+      "loss": 1.4653,
+      "step": 12431
+    },
+    {
+      "epoch": 1.4280627189707655,
+      "grad_norm": 0.6400308012962341,
+      "learning_rate": 0.0001,
+      "loss": 1.6973,
+      "step": 12432
+    },
+    {
+      "epoch": 1.4281775888805928,
+      "grad_norm": 0.5983318090438843,
+      "learning_rate": 0.0001,
+      "loss": 1.4876,
+      "step": 12433
+    },
+    {
+      "epoch": 1.4282924587904198,
+      "grad_norm": 0.6239731311798096,
+      "learning_rate": 0.0001,
+      "loss": 1.7114,
+      "step": 12434
+    },
+    {
+      "epoch": 1.428407328700247,
+      "grad_norm": 0.6159213781356812,
+      "learning_rate": 0.0001,
+      "loss": 1.4131,
+      "step": 12435
+    },
+    {
+      "epoch": 1.428522198610074,
+      "grad_norm": 0.5617234110832214,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 12436
+    },
+    {
+      "epoch": 1.4286370685199012,
+      "grad_norm": 0.5928561687469482,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 12437
+    },
+    {
+      "epoch": 1.4287519384297283,
+      "grad_norm": 0.6398497819900513,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 12438
+    },
+    {
+      "epoch": 1.4288668083395555,
+      "grad_norm": 0.6484475135803223,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 12439
+    },
+    {
+      "epoch": 1.4289816782493825,
+      "grad_norm": 0.6493231058120728,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 12440
+    },
+    {
+      "epoch": 1.4290965481592097,
+      "grad_norm": 0.5870835781097412,
+      "learning_rate": 0.0001,
+      "loss": 1.2109,
+      "step": 12441
+    },
+    {
+      "epoch": 1.4292114180690367,
+      "grad_norm": 0.6100109815597534,
+      "learning_rate": 0.0001,
+      "loss": 1.4725,
+      "step": 12442
+    },
+    {
+      "epoch": 1.429326287978864,
+      "grad_norm": 0.6027024388313293,
+      "learning_rate": 0.0001,
+      "loss": 1.262,
+      "step": 12443
+    },
+    {
+      "epoch": 1.4294411578886912,
+      "grad_norm": 0.635283887386322,
+      "learning_rate": 0.0001,
+      "loss": 1.4945,
+      "step": 12444
+    },
+    {
+      "epoch": 1.4295560277985182,
+      "grad_norm": 0.5338662266731262,
+      "learning_rate": 0.0001,
+      "loss": 1.2908,
+      "step": 12445
+    },
+    {
+      "epoch": 1.4296708977083452,
+      "grad_norm": 0.579879879951477,
+      "learning_rate": 0.0001,
+      "loss": 1.3954,
+      "step": 12446
+    },
+    {
+      "epoch": 1.4297857676181724,
+      "grad_norm": 0.5709049105644226,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 12447
+    },
+    {
+      "epoch": 1.4299006375279997,
+      "grad_norm": 0.5717498660087585,
+      "learning_rate": 0.0001,
+      "loss": 1.4321,
+      "step": 12448
+    },
+    {
+      "epoch": 1.4300155074378267,
+      "grad_norm": 0.6005428433418274,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 12449
+    },
+    {
+      "epoch": 1.4301303773476537,
+      "grad_norm": 0.5593380928039551,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 12450
+    },
+    {
+      "epoch": 1.430245247257481,
+      "grad_norm": 0.5910133719444275,
+      "learning_rate": 0.0001,
+      "loss": 1.3897,
+      "step": 12451
+    },
+    {
+      "epoch": 1.4303601171673082,
+      "grad_norm": 0.6213868856430054,
+      "learning_rate": 0.0001,
+      "loss": 1.4776,
+      "step": 12452
+    },
+    {
+      "epoch": 1.4304749870771352,
+      "grad_norm": 0.552069365978241,
+      "learning_rate": 0.0001,
+      "loss": 1.0546,
+      "step": 12453
+    },
+    {
+      "epoch": 1.4305898569869622,
+      "grad_norm": 0.5821637511253357,
+      "learning_rate": 0.0001,
+      "loss": 1.4505,
+      "step": 12454
+    },
+    {
+      "epoch": 1.4307047268967894,
+      "grad_norm": 0.6346561312675476,
+      "learning_rate": 0.0001,
+      "loss": 1.3883,
+      "step": 12455
+    },
+    {
+      "epoch": 1.4308195968066166,
+      "grad_norm": 0.5623376369476318,
+      "learning_rate": 0.0001,
+      "loss": 1.2515,
+      "step": 12456
+    },
+    {
+      "epoch": 1.4309344667164436,
+      "grad_norm": 0.5908299088478088,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 12457
+    },
+    {
+      "epoch": 1.4310493366262707,
+      "grad_norm": 0.5913065671920776,
+      "learning_rate": 0.0001,
+      "loss": 1.3443,
+      "step": 12458
+    },
+    {
+      "epoch": 1.4311642065360979,
+      "grad_norm": 0.6305943727493286,
+      "learning_rate": 0.0001,
+      "loss": 1.4428,
+      "step": 12459
+    },
+    {
+      "epoch": 1.4312790764459251,
+      "grad_norm": 0.611919105052948,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 12460
+    },
+    {
+      "epoch": 1.4313939463557521,
+      "grad_norm": 0.5658090710639954,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 12461
+    },
+    {
+      "epoch": 1.4315088162655791,
+      "grad_norm": 0.5929034948348999,
+      "learning_rate": 0.0001,
+      "loss": 1.4665,
+      "step": 12462
+    },
+    {
+      "epoch": 1.4316236861754064,
+      "grad_norm": 0.6326093673706055,
+      "learning_rate": 0.0001,
+      "loss": 1.4622,
+      "step": 12463
+    },
+    {
+      "epoch": 1.4317385560852336,
+      "grad_norm": 0.6289833784103394,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 12464
+    },
+    {
+      "epoch": 1.4318534259950606,
+      "grad_norm": 0.6180969476699829,
+      "learning_rate": 0.0001,
+      "loss": 1.5338,
+      "step": 12465
+    },
+    {
+      "epoch": 1.4319682959048876,
+      "grad_norm": 0.5622907876968384,
+      "learning_rate": 0.0001,
+      "loss": 1.4497,
+      "step": 12466
+    },
+    {
+      "epoch": 1.4320831658147148,
+      "grad_norm": 0.5641545653343201,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 12467
+    },
+    {
+      "epoch": 1.432198035724542,
+      "grad_norm": 0.6269649863243103,
+      "learning_rate": 0.0001,
+      "loss": 1.3424,
+      "step": 12468
+    },
+    {
+      "epoch": 1.432312905634369,
+      "grad_norm": 0.5639840960502625,
+      "learning_rate": 0.0001,
+      "loss": 1.4185,
+      "step": 12469
+    },
+    {
+      "epoch": 1.432427775544196,
+      "grad_norm": 0.630839467048645,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 12470
+    },
+    {
+      "epoch": 1.4325426454540233,
+      "grad_norm": 0.5446004867553711,
+      "learning_rate": 0.0001,
+      "loss": 1.2578,
+      "step": 12471
+    },
+    {
+      "epoch": 1.4326575153638506,
+      "grad_norm": 0.5677520036697388,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 12472
+    },
+    {
+      "epoch": 1.4327723852736776,
+      "grad_norm": 0.6485244035720825,
+      "learning_rate": 0.0001,
+      "loss": 1.336,
+      "step": 12473
+    },
+    {
+      "epoch": 1.4328872551835046,
+      "grad_norm": 0.6378061771392822,
+      "learning_rate": 0.0001,
+      "loss": 1.6011,
+      "step": 12474
+    },
+    {
+      "epoch": 1.4330021250933318,
+      "grad_norm": 0.5592242479324341,
+      "learning_rate": 0.0001,
+      "loss": 1.5086,
+      "step": 12475
+    },
+    {
+      "epoch": 1.433116995003159,
+      "grad_norm": 0.5973872542381287,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 12476
+    },
+    {
+      "epoch": 1.433231864912986,
+      "grad_norm": 0.6042687892913818,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 12477
+    },
+    {
+      "epoch": 1.433346734822813,
+      "grad_norm": 0.6249580979347229,
+      "learning_rate": 0.0001,
+      "loss": 1.504,
+      "step": 12478
+    },
+    {
+      "epoch": 1.4334616047326403,
+      "grad_norm": 0.5797646045684814,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 12479
+    },
+    {
+      "epoch": 1.4335764746424675,
+      "grad_norm": 0.616908609867096,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 12480
+    },
+    {
+      "epoch": 1.4336913445522945,
+      "grad_norm": 0.5572378039360046,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 12481
+    },
+    {
+      "epoch": 1.4338062144621215,
+      "grad_norm": 0.6019822955131531,
+      "learning_rate": 0.0001,
+      "loss": 1.2772,
+      "step": 12482
+    },
+    {
+      "epoch": 1.4339210843719488,
+      "grad_norm": 0.5784163475036621,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 12483
+    },
+    {
+      "epoch": 1.434035954281776,
+      "grad_norm": 0.6143441796302795,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 12484
+    },
+    {
+      "epoch": 1.434150824191603,
+      "grad_norm": 0.6093927621841431,
+      "learning_rate": 0.0001,
+      "loss": 1.3583,
+      "step": 12485
+    },
+    {
+      "epoch": 1.43426569410143,
+      "grad_norm": 0.6261629462242126,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 12486
+    },
+    {
+      "epoch": 1.4343805640112572,
+      "grad_norm": 0.5928403735160828,
+      "learning_rate": 0.0001,
+      "loss": 1.4865,
+      "step": 12487
+    },
+    {
+      "epoch": 1.4344954339210845,
+      "grad_norm": 0.5965852737426758,
+      "learning_rate": 0.0001,
+      "loss": 1.5387,
+      "step": 12488
+    },
+    {
+      "epoch": 1.4346103038309115,
+      "grad_norm": 0.6321581602096558,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 12489
+    },
+    {
+      "epoch": 1.4347251737407385,
+      "grad_norm": 0.5445297956466675,
+      "learning_rate": 0.0001,
+      "loss": 1.2522,
+      "step": 12490
+    },
+    {
+      "epoch": 1.4348400436505657,
+      "grad_norm": 0.5668053030967712,
+      "learning_rate": 0.0001,
+      "loss": 1.2608,
+      "step": 12491
+    },
+    {
+      "epoch": 1.434954913560393,
+      "grad_norm": 0.606621265411377,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 12492
+    },
+    {
+      "epoch": 1.43506978347022,
+      "grad_norm": 0.5955449938774109,
+      "learning_rate": 0.0001,
+      "loss": 1.2963,
+      "step": 12493
+    },
+    {
+      "epoch": 1.435184653380047,
+      "grad_norm": 0.6068549156188965,
+      "learning_rate": 0.0001,
+      "loss": 1.6448,
+      "step": 12494
+    },
+    {
+      "epoch": 1.4352995232898742,
+      "grad_norm": 0.5606309175491333,
+      "learning_rate": 0.0001,
+      "loss": 1.3839,
+      "step": 12495
+    },
+    {
+      "epoch": 1.4354143931997014,
+      "grad_norm": 0.58128821849823,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 12496
+    },
+    {
+      "epoch": 1.4355292631095284,
+      "grad_norm": 0.5731803774833679,
+      "learning_rate": 0.0001,
+      "loss": 1.3732,
+      "step": 12497
+    },
+    {
+      "epoch": 1.4356441330193555,
+      "grad_norm": 0.5677974224090576,
+      "learning_rate": 0.0001,
+      "loss": 1.4033,
+      "step": 12498
+    },
+    {
+      "epoch": 1.4357590029291827,
+      "grad_norm": 0.6500546336174011,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 12499
+    },
+    {
+      "epoch": 1.43587387283901,
+      "grad_norm": 0.5387493968009949,
+      "learning_rate": 0.0001,
+      "loss": 1.3859,
+      "step": 12500
+    },
+    {
+      "epoch": 1.435988742748837,
+      "grad_norm": 0.630449652671814,
+      "learning_rate": 0.0001,
+      "loss": 1.5172,
+      "step": 12501
+    },
+    {
+      "epoch": 1.436103612658664,
+      "grad_norm": 0.5695238709449768,
+      "learning_rate": 0.0001,
+      "loss": 1.2597,
+      "step": 12502
+    },
+    {
+      "epoch": 1.4362184825684912,
+      "grad_norm": 0.611565113067627,
+      "learning_rate": 0.0001,
+      "loss": 1.4379,
+      "step": 12503
+    },
+    {
+      "epoch": 1.4363333524783184,
+      "grad_norm": 0.5593751668930054,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 12504
+    },
+    {
+      "epoch": 1.4364482223881454,
+      "grad_norm": 0.5889508724212646,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 12505
+    },
+    {
+      "epoch": 1.4365630922979724,
+      "grad_norm": 0.6173695921897888,
+      "learning_rate": 0.0001,
+      "loss": 1.5322,
+      "step": 12506
+    },
+    {
+      "epoch": 1.4366779622077996,
+      "grad_norm": 0.591052770614624,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 12507
+    },
+    {
+      "epoch": 1.4367928321176269,
+      "grad_norm": 0.5677117109298706,
+      "learning_rate": 0.0001,
+      "loss": 1.4339,
+      "step": 12508
+    },
+    {
+      "epoch": 1.4369077020274539,
+      "grad_norm": 0.5549595355987549,
+      "learning_rate": 0.0001,
+      "loss": 1.2642,
+      "step": 12509
+    },
+    {
+      "epoch": 1.4370225719372811,
+      "grad_norm": 0.5533115267753601,
+      "learning_rate": 0.0001,
+      "loss": 1.4143,
+      "step": 12510
+    },
+    {
+      "epoch": 1.4371374418471081,
+      "grad_norm": 0.5983604788780212,
+      "learning_rate": 0.0001,
+      "loss": 1.3387,
+      "step": 12511
+    },
+    {
+      "epoch": 1.4372523117569354,
+      "grad_norm": 0.6087487936019897,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 12512
+    },
+    {
+      "epoch": 1.4373671816667624,
+      "grad_norm": 0.629578709602356,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 12513
+    },
+    {
+      "epoch": 1.4374820515765896,
+      "grad_norm": 0.6290776133537292,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 12514
+    },
+    {
+      "epoch": 1.4375969214864166,
+      "grad_norm": 0.6354580521583557,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 12515
+    },
+    {
+      "epoch": 1.4377117913962438,
+      "grad_norm": 0.5979134440422058,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 12516
+    },
+    {
+      "epoch": 1.4378266613060708,
+      "grad_norm": 0.5574628114700317,
+      "learning_rate": 0.0001,
+      "loss": 1.2554,
+      "step": 12517
+    },
+    {
+      "epoch": 1.437941531215898,
+      "grad_norm": 0.5842434763908386,
+      "learning_rate": 0.0001,
+      "loss": 1.3712,
+      "step": 12518
+    },
+    {
+      "epoch": 1.438056401125725,
+      "grad_norm": 0.6743392944335938,
+      "learning_rate": 0.0001,
+      "loss": 1.7602,
+      "step": 12519
+    },
+    {
+      "epoch": 1.4381712710355523,
+      "grad_norm": 0.576398491859436,
+      "learning_rate": 0.0001,
+      "loss": 1.3749,
+      "step": 12520
+    },
+    {
+      "epoch": 1.4382861409453793,
+      "grad_norm": 0.5756230354309082,
+      "learning_rate": 0.0001,
+      "loss": 1.4367,
+      "step": 12521
+    },
+    {
+      "epoch": 1.4384010108552066,
+      "grad_norm": 0.6155824065208435,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 12522
+    },
+    {
+      "epoch": 1.4385158807650336,
+      "grad_norm": 0.6590107679367065,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 12523
+    },
+    {
+      "epoch": 1.4386307506748608,
+      "grad_norm": 0.6111704707145691,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 12524
+    },
+    {
+      "epoch": 1.4387456205846878,
+      "grad_norm": 0.6221973299980164,
+      "learning_rate": 0.0001,
+      "loss": 1.4813,
+      "step": 12525
+    },
+    {
+      "epoch": 1.438860490494515,
+      "grad_norm": 0.5589838624000549,
+      "learning_rate": 0.0001,
+      "loss": 1.2726,
+      "step": 12526
+    },
+    {
+      "epoch": 1.438975360404342,
+      "grad_norm": 0.6924046874046326,
+      "learning_rate": 0.0001,
+      "loss": 1.1983,
+      "step": 12527
+    },
+    {
+      "epoch": 1.4390902303141693,
+      "grad_norm": 0.5677695870399475,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 12528
+    },
+    {
+      "epoch": 1.4392051002239963,
+      "grad_norm": 0.6168467402458191,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 12529
+    },
+    {
+      "epoch": 1.4393199701338235,
+      "grad_norm": 0.5841947197914124,
+      "learning_rate": 0.0001,
+      "loss": 1.5834,
+      "step": 12530
+    },
+    {
+      "epoch": 1.4394348400436505,
+      "grad_norm": 0.6030665636062622,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 12531
+    },
+    {
+      "epoch": 1.4395497099534778,
+      "grad_norm": 0.6445423364639282,
+      "learning_rate": 0.0001,
+      "loss": 1.6561,
+      "step": 12532
+    },
+    {
+      "epoch": 1.4396645798633048,
+      "grad_norm": 0.5743504166603088,
+      "learning_rate": 0.0001,
+      "loss": 1.3297,
+      "step": 12533
+    },
+    {
+      "epoch": 1.439779449773132,
+      "grad_norm": 0.6461590528488159,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 12534
+    },
+    {
+      "epoch": 1.439894319682959,
+      "grad_norm": 0.535851001739502,
+      "learning_rate": 0.0001,
+      "loss": 1.343,
+      "step": 12535
+    },
+    {
+      "epoch": 1.4400091895927862,
+      "grad_norm": 0.6115723252296448,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 12536
+    },
+    {
+      "epoch": 1.4401240595026132,
+      "grad_norm": 0.5796070694923401,
+      "learning_rate": 0.0001,
+      "loss": 1.4744,
+      "step": 12537
+    },
+    {
+      "epoch": 1.4402389294124405,
+      "grad_norm": 0.6212785243988037,
+      "learning_rate": 0.0001,
+      "loss": 1.4577,
+      "step": 12538
+    },
+    {
+      "epoch": 1.4403537993222675,
+      "grad_norm": 0.6039834022521973,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 12539
+    },
+    {
+      "epoch": 1.4404686692320947,
+      "grad_norm": 0.6413995027542114,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 12540
+    },
+    {
+      "epoch": 1.4405835391419217,
+      "grad_norm": 0.6055036187171936,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 12541
+    },
+    {
+      "epoch": 1.440698409051749,
+      "grad_norm": 0.5628538727760315,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 12542
+    },
+    {
+      "epoch": 1.440813278961576,
+      "grad_norm": 0.6067841649055481,
+      "learning_rate": 0.0001,
+      "loss": 1.5706,
+      "step": 12543
+    },
+    {
+      "epoch": 1.4409281488714032,
+      "grad_norm": 0.5712527632713318,
+      "learning_rate": 0.0001,
+      "loss": 1.421,
+      "step": 12544
+    },
+    {
+      "epoch": 1.4410430187812302,
+      "grad_norm": 0.5913591384887695,
+      "learning_rate": 0.0001,
+      "loss": 1.469,
+      "step": 12545
+    },
+    {
+      "epoch": 1.4411578886910574,
+      "grad_norm": 0.5872635841369629,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 12546
+    },
+    {
+      "epoch": 1.4412727586008844,
+      "grad_norm": 0.6234807372093201,
+      "learning_rate": 0.0001,
+      "loss": 1.4239,
+      "step": 12547
+    },
+    {
+      "epoch": 1.4413876285107117,
+      "grad_norm": 0.6241154074668884,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 12548
+    },
+    {
+      "epoch": 1.4415024984205387,
+      "grad_norm": 0.5826482772827148,
+      "learning_rate": 0.0001,
+      "loss": 1.3162,
+      "step": 12549
+    },
+    {
+      "epoch": 1.441617368330366,
+      "grad_norm": 0.5742958784103394,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 12550
+    },
+    {
+      "epoch": 1.441732238240193,
+      "grad_norm": 0.5949505567550659,
+      "learning_rate": 0.0001,
+      "loss": 1.3672,
+      "step": 12551
+    },
+    {
+      "epoch": 1.4418471081500202,
+      "grad_norm": 0.6136342287063599,
+      "learning_rate": 0.0001,
+      "loss": 1.4474,
+      "step": 12552
+    },
+    {
+      "epoch": 1.4419619780598472,
+      "grad_norm": 0.5760694146156311,
+      "learning_rate": 0.0001,
+      "loss": 1.3999,
+      "step": 12553
+    },
+    {
+      "epoch": 1.4420768479696744,
+      "grad_norm": 0.6037359833717346,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 12554
+    },
+    {
+      "epoch": 1.4421917178795014,
+      "grad_norm": 0.5782321095466614,
+      "learning_rate": 0.0001,
+      "loss": 1.3474,
+      "step": 12555
+    },
+    {
+      "epoch": 1.4423065877893286,
+      "grad_norm": 0.5973963737487793,
+      "learning_rate": 0.0001,
+      "loss": 1.5053,
+      "step": 12556
+    },
+    {
+      "epoch": 1.4424214576991556,
+      "grad_norm": 0.588340699672699,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 12557
+    },
+    {
+      "epoch": 1.4425363276089829,
+      "grad_norm": 0.5831438302993774,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 12558
+    },
+    {
+      "epoch": 1.4426511975188099,
+      "grad_norm": 0.6051815748214722,
+      "learning_rate": 0.0001,
+      "loss": 1.466,
+      "step": 12559
+    },
+    {
+      "epoch": 1.4427660674286371,
+      "grad_norm": 0.5373536348342896,
+      "learning_rate": 0.0001,
+      "loss": 1.049,
+      "step": 12560
+    },
+    {
+      "epoch": 1.4428809373384641,
+      "grad_norm": 0.6360757350921631,
+      "learning_rate": 0.0001,
+      "loss": 1.497,
+      "step": 12561
+    },
+    {
+      "epoch": 1.4429958072482914,
+      "grad_norm": 0.6303163170814514,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 12562
+    },
+    {
+      "epoch": 1.4431106771581184,
+      "grad_norm": 0.6021014451980591,
+      "learning_rate": 0.0001,
+      "loss": 1.3477,
+      "step": 12563
+    },
+    {
+      "epoch": 1.4432255470679456,
+      "grad_norm": 0.6191940307617188,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 12564
+    },
+    {
+      "epoch": 1.4433404169777726,
+      "grad_norm": 0.6756370067596436,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 12565
+    },
+    {
+      "epoch": 1.4434552868875998,
+      "grad_norm": 0.7018874883651733,
+      "learning_rate": 0.0001,
+      "loss": 1.863,
+      "step": 12566
+    },
+    {
+      "epoch": 1.4435701567974268,
+      "grad_norm": 0.5474680066108704,
+      "learning_rate": 0.0001,
+      "loss": 1.3566,
+      "step": 12567
+    },
+    {
+      "epoch": 1.443685026707254,
+      "grad_norm": 0.6140058040618896,
+      "learning_rate": 0.0001,
+      "loss": 1.2638,
+      "step": 12568
+    },
+    {
+      "epoch": 1.443799896617081,
+      "grad_norm": 0.5624337196350098,
+      "learning_rate": 0.0001,
+      "loss": 1.4342,
+      "step": 12569
+    },
+    {
+      "epoch": 1.4439147665269083,
+      "grad_norm": 0.6037470698356628,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 12570
+    },
+    {
+      "epoch": 1.4440296364367353,
+      "grad_norm": 0.594399094581604,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 12571
+    },
+    {
+      "epoch": 1.4441445063465626,
+      "grad_norm": 0.5606778264045715,
+      "learning_rate": 0.0001,
+      "loss": 1.3396,
+      "step": 12572
+    },
+    {
+      "epoch": 1.4442593762563896,
+      "grad_norm": 0.6168720126152039,
+      "learning_rate": 0.0001,
+      "loss": 1.4917,
+      "step": 12573
+    },
+    {
+      "epoch": 1.4443742461662168,
+      "grad_norm": 0.5519090294837952,
+      "learning_rate": 0.0001,
+      "loss": 1.5099,
+      "step": 12574
+    },
+    {
+      "epoch": 1.4444891160760438,
+      "grad_norm": 0.5982587933540344,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 12575
+    },
+    {
+      "epoch": 1.444603985985871,
+      "grad_norm": 0.6236387491226196,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 12576
+    },
+    {
+      "epoch": 1.444718855895698,
+      "grad_norm": 0.6040252447128296,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 12577
+    },
+    {
+      "epoch": 1.4448337258055253,
+      "grad_norm": 0.5864963531494141,
+      "learning_rate": 0.0001,
+      "loss": 1.4526,
+      "step": 12578
+    },
+    {
+      "epoch": 1.4449485957153523,
+      "grad_norm": 0.5800042152404785,
+      "learning_rate": 0.0001,
+      "loss": 1.3706,
+      "step": 12579
+    },
+    {
+      "epoch": 1.4450634656251795,
+      "grad_norm": 0.6318053007125854,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 12580
+    },
+    {
+      "epoch": 1.4451783355350067,
+      "grad_norm": 0.6059226989746094,
+      "learning_rate": 0.0001,
+      "loss": 1.3472,
+      "step": 12581
+    },
+    {
+      "epoch": 1.4452932054448338,
+      "grad_norm": 0.5987119078636169,
+      "learning_rate": 0.0001,
+      "loss": 1.4228,
+      "step": 12582
+    },
+    {
+      "epoch": 1.4454080753546608,
+      "grad_norm": 0.6017065048217773,
+      "learning_rate": 0.0001,
+      "loss": 1.3683,
+      "step": 12583
+    },
+    {
+      "epoch": 1.445522945264488,
+      "grad_norm": 0.5962862968444824,
+      "learning_rate": 0.0001,
+      "loss": 1.5466,
+      "step": 12584
+    },
+    {
+      "epoch": 1.4456378151743152,
+      "grad_norm": 0.6121211647987366,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 12585
+    },
+    {
+      "epoch": 1.4457526850841422,
+      "grad_norm": 0.5887848138809204,
+      "learning_rate": 0.0001,
+      "loss": 1.3906,
+      "step": 12586
+    },
+    {
+      "epoch": 1.4458675549939692,
+      "grad_norm": 0.5756409168243408,
+      "learning_rate": 0.0001,
+      "loss": 1.345,
+      "step": 12587
+    },
+    {
+      "epoch": 1.4459824249037965,
+      "grad_norm": 0.6187815070152283,
+      "learning_rate": 0.0001,
+      "loss": 1.2619,
+      "step": 12588
+    },
+    {
+      "epoch": 1.4460972948136237,
+      "grad_norm": 0.6031472086906433,
+      "learning_rate": 0.0001,
+      "loss": 1.4761,
+      "step": 12589
+    },
+    {
+      "epoch": 1.4462121647234507,
+      "grad_norm": 0.5487967729568481,
+      "learning_rate": 0.0001,
+      "loss": 1.258,
+      "step": 12590
+    },
+    {
+      "epoch": 1.4463270346332777,
+      "grad_norm": 0.5621272921562195,
+      "learning_rate": 0.0001,
+      "loss": 1.0927,
+      "step": 12591
+    },
+    {
+      "epoch": 1.446441904543105,
+      "grad_norm": 0.598263144493103,
+      "learning_rate": 0.0001,
+      "loss": 1.462,
+      "step": 12592
+    },
+    {
+      "epoch": 1.4465567744529322,
+      "grad_norm": 0.5940863490104675,
+      "learning_rate": 0.0001,
+      "loss": 1.426,
+      "step": 12593
+    },
+    {
+      "epoch": 1.4466716443627592,
+      "grad_norm": 0.6629210114479065,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 12594
+    },
+    {
+      "epoch": 1.4467865142725862,
+      "grad_norm": 0.5858151912689209,
+      "learning_rate": 0.0001,
+      "loss": 1.3906,
+      "step": 12595
+    },
+    {
+      "epoch": 1.4469013841824134,
+      "grad_norm": 0.6476069092750549,
+      "learning_rate": 0.0001,
+      "loss": 1.3094,
+      "step": 12596
+    },
+    {
+      "epoch": 1.4470162540922407,
+      "grad_norm": 0.6032461524009705,
+      "learning_rate": 0.0001,
+      "loss": 1.3134,
+      "step": 12597
+    },
+    {
+      "epoch": 1.4471311240020677,
+      "grad_norm": 0.591644823551178,
+      "learning_rate": 0.0001,
+      "loss": 1.3682,
+      "step": 12598
+    },
+    {
+      "epoch": 1.4472459939118947,
+      "grad_norm": 0.656358540058136,
+      "learning_rate": 0.0001,
+      "loss": 1.1411,
+      "step": 12599
+    },
+    {
+      "epoch": 1.447360863821722,
+      "grad_norm": 0.6357599496841431,
+      "learning_rate": 0.0001,
+      "loss": 1.5471,
+      "step": 12600
+    },
+    {
+      "epoch": 1.4474757337315491,
+      "grad_norm": 0.5842412710189819,
+      "learning_rate": 0.0001,
+      "loss": 1.4944,
+      "step": 12601
+    },
+    {
+      "epoch": 1.4475906036413762,
+      "grad_norm": 0.580599308013916,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 12602
+    },
+    {
+      "epoch": 1.4477054735512032,
+      "grad_norm": 0.6119430661201477,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 12603
+    },
+    {
+      "epoch": 1.4478203434610304,
+      "grad_norm": 0.5994784832000732,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 12604
+    },
+    {
+      "epoch": 1.4479352133708576,
+      "grad_norm": 0.5609249472618103,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 12605
+    },
+    {
+      "epoch": 1.4480500832806846,
+      "grad_norm": 0.5915836095809937,
+      "learning_rate": 0.0001,
+      "loss": 1.3597,
+      "step": 12606
+    },
+    {
+      "epoch": 1.4481649531905116,
+      "grad_norm": 0.6126052141189575,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 12607
+    },
+    {
+      "epoch": 1.4482798231003389,
+      "grad_norm": 0.6269632577896118,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 12608
+    },
+    {
+      "epoch": 1.448394693010166,
+      "grad_norm": 0.5238544344902039,
+      "learning_rate": 0.0001,
+      "loss": 1.4742,
+      "step": 12609
+    },
+    {
+      "epoch": 1.448509562919993,
+      "grad_norm": 0.607177197933197,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 12610
+    },
+    {
+      "epoch": 1.4486244328298201,
+      "grad_norm": 0.5543795228004456,
+      "learning_rate": 0.0001,
+      "loss": 1.4268,
+      "step": 12611
+    },
+    {
+      "epoch": 1.4487393027396473,
+      "grad_norm": 0.6204040050506592,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 12612
+    },
+    {
+      "epoch": 1.4488541726494746,
+      "grad_norm": 0.5601338148117065,
+      "learning_rate": 0.0001,
+      "loss": 1.5038,
+      "step": 12613
+    },
+    {
+      "epoch": 1.4489690425593016,
+      "grad_norm": 0.5878952145576477,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 12614
+    },
+    {
+      "epoch": 1.4490839124691286,
+      "grad_norm": 0.5514255166053772,
+      "learning_rate": 0.0001,
+      "loss": 1.3956,
+      "step": 12615
+    },
+    {
+      "epoch": 1.4491987823789558,
+      "grad_norm": 0.6429269313812256,
+      "learning_rate": 0.0001,
+      "loss": 1.3746,
+      "step": 12616
+    },
+    {
+      "epoch": 1.449313652288783,
+      "grad_norm": 0.5910865068435669,
+      "learning_rate": 0.0001,
+      "loss": 1.4062,
+      "step": 12617
+    },
+    {
+      "epoch": 1.44942852219861,
+      "grad_norm": 0.6003382205963135,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 12618
+    },
+    {
+      "epoch": 1.449543392108437,
+      "grad_norm": 0.5780314803123474,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 12619
+    },
+    {
+      "epoch": 1.4496582620182643,
+      "grad_norm": 0.5906615257263184,
+      "learning_rate": 0.0001,
+      "loss": 1.414,
+      "step": 12620
+    },
+    {
+      "epoch": 1.4497731319280915,
+      "grad_norm": 0.6168799996376038,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 12621
+    },
+    {
+      "epoch": 1.4498880018379185,
+      "grad_norm": 0.5736768245697021,
+      "learning_rate": 0.0001,
+      "loss": 1.5504,
+      "step": 12622
+    },
+    {
+      "epoch": 1.4500028717477456,
+      "grad_norm": 0.5871389508247375,
+      "learning_rate": 0.0001,
+      "loss": 1.488,
+      "step": 12623
+    },
+    {
+      "epoch": 1.4501177416575728,
+      "grad_norm": 0.6376186013221741,
+      "learning_rate": 0.0001,
+      "loss": 1.4979,
+      "step": 12624
+    },
+    {
+      "epoch": 1.4502326115674,
+      "grad_norm": 0.5403526425361633,
+      "learning_rate": 0.0001,
+      "loss": 1.2595,
+      "step": 12625
+    },
+    {
+      "epoch": 1.450347481477227,
+      "grad_norm": 0.5680791139602661,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 12626
+    },
+    {
+      "epoch": 1.450462351387054,
+      "grad_norm": 0.5831477642059326,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 12627
+    },
+    {
+      "epoch": 1.4505772212968813,
+      "grad_norm": 0.5702657103538513,
+      "learning_rate": 0.0001,
+      "loss": 1.5092,
+      "step": 12628
+    },
+    {
+      "epoch": 1.4506920912067085,
+      "grad_norm": 0.5878103971481323,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 12629
+    },
+    {
+      "epoch": 1.4508069611165355,
+      "grad_norm": 0.6267426609992981,
+      "learning_rate": 0.0001,
+      "loss": 1.3794,
+      "step": 12630
+    },
+    {
+      "epoch": 1.4509218310263625,
+      "grad_norm": 0.578369140625,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 12631
+    },
+    {
+      "epoch": 1.4510367009361897,
+      "grad_norm": 0.601699709892273,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 12632
+    },
+    {
+      "epoch": 1.451151570846017,
+      "grad_norm": 0.5736028552055359,
+      "learning_rate": 0.0001,
+      "loss": 1.3779,
+      "step": 12633
+    },
+    {
+      "epoch": 1.451266440755844,
+      "grad_norm": 0.5805512070655823,
+      "learning_rate": 0.0001,
+      "loss": 1.4821,
+      "step": 12634
+    },
+    {
+      "epoch": 1.451381310665671,
+      "grad_norm": 0.6317859292030334,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 12635
+    },
+    {
+      "epoch": 1.4514961805754982,
+      "grad_norm": 0.5671908259391785,
+      "learning_rate": 0.0001,
+      "loss": 1.3133,
+      "step": 12636
+    },
+    {
+      "epoch": 1.4516110504853255,
+      "grad_norm": 0.5456831455230713,
+      "learning_rate": 0.0001,
+      "loss": 1.463,
+      "step": 12637
+    },
+    {
+      "epoch": 1.4517259203951525,
+      "grad_norm": 0.6048300266265869,
+      "learning_rate": 0.0001,
+      "loss": 1.4596,
+      "step": 12638
+    },
+    {
+      "epoch": 1.4518407903049795,
+      "grad_norm": 0.5492627024650574,
+      "learning_rate": 0.0001,
+      "loss": 1.372,
+      "step": 12639
+    },
+    {
+      "epoch": 1.4519556602148067,
+      "grad_norm": 0.6605719923973083,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 12640
+    },
+    {
+      "epoch": 1.452070530124634,
+      "grad_norm": 0.6105278730392456,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 12641
+    },
+    {
+      "epoch": 1.452185400034461,
+      "grad_norm": 0.6340276598930359,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 12642
+    },
+    {
+      "epoch": 1.452300269944288,
+      "grad_norm": 0.5779246687889099,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 12643
+    },
+    {
+      "epoch": 1.4524151398541152,
+      "grad_norm": 0.5638899803161621,
+      "learning_rate": 0.0001,
+      "loss": 1.3403,
+      "step": 12644
+    },
+    {
+      "epoch": 1.4525300097639424,
+      "grad_norm": 0.5835622549057007,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 12645
+    },
+    {
+      "epoch": 1.4526448796737694,
+      "grad_norm": 0.5531151294708252,
+      "learning_rate": 0.0001,
+      "loss": 1.3624,
+      "step": 12646
+    },
+    {
+      "epoch": 1.4527597495835967,
+      "grad_norm": 0.6040196418762207,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 12647
+    },
+    {
+      "epoch": 1.4528746194934237,
+      "grad_norm": 0.6672126054763794,
+      "learning_rate": 0.0001,
+      "loss": 1.3766,
+      "step": 12648
+    },
+    {
+      "epoch": 1.452989489403251,
+      "grad_norm": 0.5610442757606506,
+      "learning_rate": 0.0001,
+      "loss": 1.3557,
+      "step": 12649
+    },
+    {
+      "epoch": 1.453104359313078,
+      "grad_norm": 0.5616607666015625,
+      "learning_rate": 0.0001,
+      "loss": 1.3134,
+      "step": 12650
+    },
+    {
+      "epoch": 1.4532192292229051,
+      "grad_norm": 0.5792796015739441,
+      "learning_rate": 0.0001,
+      "loss": 1.7026,
+      "step": 12651
+    },
+    {
+      "epoch": 1.4533340991327321,
+      "grad_norm": 0.5875088572502136,
+      "learning_rate": 0.0001,
+      "loss": 1.6033,
+      "step": 12652
+    },
+    {
+      "epoch": 1.4534489690425594,
+      "grad_norm": 0.5907067060470581,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 12653
+    },
+    {
+      "epoch": 1.4535638389523864,
+      "grad_norm": 0.5939152836799622,
+      "learning_rate": 0.0001,
+      "loss": 1.4968,
+      "step": 12654
+    },
+    {
+      "epoch": 1.4536787088622136,
+      "grad_norm": 0.585138201713562,
+      "learning_rate": 0.0001,
+      "loss": 1.4367,
+      "step": 12655
+    },
+    {
+      "epoch": 1.4537935787720406,
+      "grad_norm": 0.634059727191925,
+      "learning_rate": 0.0001,
+      "loss": 1.3877,
+      "step": 12656
+    },
+    {
+      "epoch": 1.4539084486818679,
+      "grad_norm": 0.5916883945465088,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 12657
+    },
+    {
+      "epoch": 1.4540233185916949,
+      "grad_norm": 0.6057246923446655,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 12658
+    },
+    {
+      "epoch": 1.454138188501522,
+      "grad_norm": 0.5644753575325012,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 12659
+    },
+    {
+      "epoch": 1.454253058411349,
+      "grad_norm": 0.6254203915596008,
+      "learning_rate": 0.0001,
+      "loss": 1.501,
+      "step": 12660
+    },
+    {
+      "epoch": 1.4543679283211763,
+      "grad_norm": 0.6106442213058472,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 12661
+    },
+    {
+      "epoch": 1.4544827982310033,
+      "grad_norm": 0.5752881765365601,
+      "learning_rate": 0.0001,
+      "loss": 1.4354,
+      "step": 12662
+    },
+    {
+      "epoch": 1.4545976681408306,
+      "grad_norm": 0.6398709416389465,
+      "learning_rate": 0.0001,
+      "loss": 1.3857,
+      "step": 12663
+    },
+    {
+      "epoch": 1.4547125380506576,
+      "grad_norm": 0.6222221255302429,
+      "learning_rate": 0.0001,
+      "loss": 1.3953,
+      "step": 12664
+    },
+    {
+      "epoch": 1.4548274079604848,
+      "grad_norm": 0.5810672044754028,
+      "learning_rate": 0.0001,
+      "loss": 1.4166,
+      "step": 12665
+    },
+    {
+      "epoch": 1.4549422778703118,
+      "grad_norm": 0.5670533180236816,
+      "learning_rate": 0.0001,
+      "loss": 1.3454,
+      "step": 12666
+    },
+    {
+      "epoch": 1.455057147780139,
+      "grad_norm": 0.6999189853668213,
+      "learning_rate": 0.0001,
+      "loss": 1.6628,
+      "step": 12667
+    },
+    {
+      "epoch": 1.455172017689966,
+      "grad_norm": 0.6700528860092163,
+      "learning_rate": 0.0001,
+      "loss": 1.3558,
+      "step": 12668
+    },
+    {
+      "epoch": 1.4552868875997933,
+      "grad_norm": 0.6250571012496948,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 12669
+    },
+    {
+      "epoch": 1.4554017575096203,
+      "grad_norm": 0.573645293712616,
+      "learning_rate": 0.0001,
+      "loss": 1.4697,
+      "step": 12670
+    },
+    {
+      "epoch": 1.4555166274194475,
+      "grad_norm": 0.5824663043022156,
+      "learning_rate": 0.0001,
+      "loss": 1.3154,
+      "step": 12671
+    },
+    {
+      "epoch": 1.4556314973292745,
+      "grad_norm": 0.5397822260856628,
+      "learning_rate": 0.0001,
+      "loss": 1.3799,
+      "step": 12672
+    },
+    {
+      "epoch": 1.4557463672391018,
+      "grad_norm": 0.6227380633354187,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 12673
+    },
+    {
+      "epoch": 1.4558612371489288,
+      "grad_norm": 0.5699251294136047,
+      "learning_rate": 0.0001,
+      "loss": 1.272,
+      "step": 12674
+    },
+    {
+      "epoch": 1.455976107058756,
+      "grad_norm": 0.58524489402771,
+      "learning_rate": 0.0001,
+      "loss": 1.3419,
+      "step": 12675
+    },
+    {
+      "epoch": 1.456090976968583,
+      "grad_norm": 0.6593015789985657,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 12676
+    },
+    {
+      "epoch": 1.4562058468784103,
+      "grad_norm": 0.6179425120353699,
+      "learning_rate": 0.0001,
+      "loss": 1.3337,
+      "step": 12677
+    },
+    {
+      "epoch": 1.4563207167882373,
+      "grad_norm": 0.591990053653717,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 12678
+    },
+    {
+      "epoch": 1.4564355866980645,
+      "grad_norm": 0.5718491673469543,
+      "learning_rate": 0.0001,
+      "loss": 1.5063,
+      "step": 12679
+    },
+    {
+      "epoch": 1.4565504566078915,
+      "grad_norm": 0.609443187713623,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 12680
+    },
+    {
+      "epoch": 1.4566653265177187,
+      "grad_norm": 0.5620328187942505,
+      "learning_rate": 0.0001,
+      "loss": 1.4505,
+      "step": 12681
+    },
+    {
+      "epoch": 1.4567801964275457,
+      "grad_norm": 0.6245093941688538,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 12682
+    },
+    {
+      "epoch": 1.456895066337373,
+      "grad_norm": 0.5638138651847839,
+      "learning_rate": 0.0001,
+      "loss": 1.3709,
+      "step": 12683
+    },
+    {
+      "epoch": 1.4570099362472,
+      "grad_norm": 0.6003065705299377,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 12684
+    },
+    {
+      "epoch": 1.4571248061570272,
+      "grad_norm": 0.6570206880569458,
+      "learning_rate": 0.0001,
+      "loss": 1.3938,
+      "step": 12685
+    },
+    {
+      "epoch": 1.4572396760668542,
+      "grad_norm": 0.6485840678215027,
+      "learning_rate": 0.0001,
+      "loss": 1.3326,
+      "step": 12686
+    },
+    {
+      "epoch": 1.4573545459766815,
+      "grad_norm": 0.603420615196228,
+      "learning_rate": 0.0001,
+      "loss": 1.4235,
+      "step": 12687
+    },
+    {
+      "epoch": 1.4574694158865085,
+      "grad_norm": 0.6239102482795715,
+      "learning_rate": 0.0001,
+      "loss": 1.3894,
+      "step": 12688
+    },
+    {
+      "epoch": 1.4575842857963357,
+      "grad_norm": 0.6117807626724243,
+      "learning_rate": 0.0001,
+      "loss": 1.333,
+      "step": 12689
+    },
+    {
+      "epoch": 1.4576991557061627,
+      "grad_norm": 0.6549608707427979,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 12690
+    },
+    {
+      "epoch": 1.45781402561599,
+      "grad_norm": 0.6886609196662903,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 12691
+    },
+    {
+      "epoch": 1.457928895525817,
+      "grad_norm": 0.6125257611274719,
+      "learning_rate": 0.0001,
+      "loss": 1.3924,
+      "step": 12692
+    },
+    {
+      "epoch": 1.4580437654356442,
+      "grad_norm": 0.5616730451583862,
+      "learning_rate": 0.0001,
+      "loss": 1.4452,
+      "step": 12693
+    },
+    {
+      "epoch": 1.4581586353454712,
+      "grad_norm": 0.6304099559783936,
+      "learning_rate": 0.0001,
+      "loss": 1.556,
+      "step": 12694
+    },
+    {
+      "epoch": 1.4582735052552984,
+      "grad_norm": 0.6536027789115906,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 12695
+    },
+    {
+      "epoch": 1.4583883751651254,
+      "grad_norm": 0.6076638698577881,
+      "learning_rate": 0.0001,
+      "loss": 1.4965,
+      "step": 12696
+    },
+    {
+      "epoch": 1.4585032450749527,
+      "grad_norm": 0.6095083355903625,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 12697
+    },
+    {
+      "epoch": 1.4586181149847797,
+      "grad_norm": 0.5856631994247437,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 12698
+    },
+    {
+      "epoch": 1.458732984894607,
+      "grad_norm": 0.582143247127533,
+      "learning_rate": 0.0001,
+      "loss": 1.4553,
+      "step": 12699
+    },
+    {
+      "epoch": 1.458847854804434,
+      "grad_norm": 0.6474102735519409,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 12700
+    },
+    {
+      "epoch": 1.4589627247142611,
+      "grad_norm": 0.6195456981658936,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 12701
+    },
+    {
+      "epoch": 1.4590775946240881,
+      "grad_norm": 0.5719879269599915,
+      "learning_rate": 0.0001,
+      "loss": 1.3453,
+      "step": 12702
+    },
+    {
+      "epoch": 1.4591924645339154,
+      "grad_norm": 0.5978543162345886,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 12703
+    },
+    {
+      "epoch": 1.4593073344437424,
+      "grad_norm": 0.5765978097915649,
+      "learning_rate": 0.0001,
+      "loss": 1.2654,
+      "step": 12704
+    },
+    {
+      "epoch": 1.4594222043535696,
+      "grad_norm": 0.6159289479255676,
+      "learning_rate": 0.0001,
+      "loss": 1.3155,
+      "step": 12705
+    },
+    {
+      "epoch": 1.4595370742633966,
+      "grad_norm": 0.6160826683044434,
+      "learning_rate": 0.0001,
+      "loss": 1.4765,
+      "step": 12706
+    },
+    {
+      "epoch": 1.4596519441732239,
+      "grad_norm": 0.5803123712539673,
+      "learning_rate": 0.0001,
+      "loss": 1.3856,
+      "step": 12707
+    },
+    {
+      "epoch": 1.4597668140830509,
+      "grad_norm": 0.6179252862930298,
+      "learning_rate": 0.0001,
+      "loss": 1.3149,
+      "step": 12708
+    },
+    {
+      "epoch": 1.459881683992878,
+      "grad_norm": 0.6384448409080505,
+      "learning_rate": 0.0001,
+      "loss": 1.2182,
+      "step": 12709
+    },
+    {
+      "epoch": 1.459996553902705,
+      "grad_norm": 0.5776406526565552,
+      "learning_rate": 0.0001,
+      "loss": 1.3238,
+      "step": 12710
+    },
+    {
+      "epoch": 1.4601114238125323,
+      "grad_norm": 0.6758487820625305,
+      "learning_rate": 0.0001,
+      "loss": 1.4016,
+      "step": 12711
+    },
+    {
+      "epoch": 1.4602262937223593,
+      "grad_norm": 0.7246870398521423,
+      "learning_rate": 0.0001,
+      "loss": 1.6412,
+      "step": 12712
+    },
+    {
+      "epoch": 1.4603411636321866,
+      "grad_norm": 0.6004265546798706,
+      "learning_rate": 0.0001,
+      "loss": 1.2532,
+      "step": 12713
+    },
+    {
+      "epoch": 1.4604560335420136,
+      "grad_norm": 0.6100044846534729,
+      "learning_rate": 0.0001,
+      "loss": 1.4896,
+      "step": 12714
+    },
+    {
+      "epoch": 1.4605709034518408,
+      "grad_norm": 0.6924980282783508,
+      "learning_rate": 0.0001,
+      "loss": 1.7668,
+      "step": 12715
+    },
+    {
+      "epoch": 1.4606857733616678,
+      "grad_norm": 0.5812683701515198,
+      "learning_rate": 0.0001,
+      "loss": 1.4722,
+      "step": 12716
+    },
+    {
+      "epoch": 1.460800643271495,
+      "grad_norm": 0.6266067624092102,
+      "learning_rate": 0.0001,
+      "loss": 1.4486,
+      "step": 12717
+    },
+    {
+      "epoch": 1.4609155131813223,
+      "grad_norm": 0.5701542496681213,
+      "learning_rate": 0.0001,
+      "loss": 1.4599,
+      "step": 12718
+    },
+    {
+      "epoch": 1.4610303830911493,
+      "grad_norm": 0.6322028636932373,
+      "learning_rate": 0.0001,
+      "loss": 1.4134,
+      "step": 12719
+    },
+    {
+      "epoch": 1.4611452530009763,
+      "grad_norm": 0.6089679598808289,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 12720
+    },
+    {
+      "epoch": 1.4612601229108035,
+      "grad_norm": 0.5674778819084167,
+      "learning_rate": 0.0001,
+      "loss": 1.4329,
+      "step": 12721
+    },
+    {
+      "epoch": 1.4613749928206308,
+      "grad_norm": 0.5603479146957397,
+      "learning_rate": 0.0001,
+      "loss": 1.4397,
+      "step": 12722
+    },
+    {
+      "epoch": 1.4614898627304578,
+      "grad_norm": 0.5788118839263916,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 12723
+    },
+    {
+      "epoch": 1.4616047326402848,
+      "grad_norm": 0.5956910848617554,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 12724
+    },
+    {
+      "epoch": 1.461719602550112,
+      "grad_norm": 0.5729474425315857,
+      "learning_rate": 0.0001,
+      "loss": 1.4532,
+      "step": 12725
+    },
+    {
+      "epoch": 1.4618344724599392,
+      "grad_norm": 0.6295613646507263,
+      "learning_rate": 0.0001,
+      "loss": 1.6566,
+      "step": 12726
+    },
+    {
+      "epoch": 1.4619493423697663,
+      "grad_norm": 0.5930306911468506,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 12727
+    },
+    {
+      "epoch": 1.4620642122795933,
+      "grad_norm": 0.6040239930152893,
+      "learning_rate": 0.0001,
+      "loss": 1.3305,
+      "step": 12728
+    },
+    {
+      "epoch": 1.4621790821894205,
+      "grad_norm": 0.5989452004432678,
+      "learning_rate": 0.0001,
+      "loss": 1.4624,
+      "step": 12729
+    },
+    {
+      "epoch": 1.4622939520992477,
+      "grad_norm": 0.5778357982635498,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 12730
+    },
+    {
+      "epoch": 1.4624088220090747,
+      "grad_norm": 0.6063084602355957,
+      "learning_rate": 0.0001,
+      "loss": 1.3575,
+      "step": 12731
+    },
+    {
+      "epoch": 1.4625236919189017,
+      "grad_norm": 0.6218013763427734,
+      "learning_rate": 0.0001,
+      "loss": 1.2905,
+      "step": 12732
+    },
+    {
+      "epoch": 1.462638561828729,
+      "grad_norm": 0.6372556090354919,
+      "learning_rate": 0.0001,
+      "loss": 1.3767,
+      "step": 12733
+    },
+    {
+      "epoch": 1.4627534317385562,
+      "grad_norm": 0.624243438243866,
+      "learning_rate": 0.0001,
+      "loss": 1.3002,
+      "step": 12734
+    },
+    {
+      "epoch": 1.4628683016483832,
+      "grad_norm": 0.6607589721679688,
+      "learning_rate": 0.0001,
+      "loss": 1.3411,
+      "step": 12735
+    },
+    {
+      "epoch": 1.4629831715582102,
+      "grad_norm": 0.5998088121414185,
+      "learning_rate": 0.0001,
+      "loss": 1.3656,
+      "step": 12736
+    },
+    {
+      "epoch": 1.4630980414680375,
+      "grad_norm": 0.5681934356689453,
+      "learning_rate": 0.0001,
+      "loss": 1.3278,
+      "step": 12737
+    },
+    {
+      "epoch": 1.4632129113778647,
+      "grad_norm": 0.6242961883544922,
+      "learning_rate": 0.0001,
+      "loss": 1.7247,
+      "step": 12738
+    },
+    {
+      "epoch": 1.4633277812876917,
+      "grad_norm": 0.5514942407608032,
+      "learning_rate": 0.0001,
+      "loss": 1.6772,
+      "step": 12739
+    },
+    {
+      "epoch": 1.4634426511975187,
+      "grad_norm": 0.5704249739646912,
+      "learning_rate": 0.0001,
+      "loss": 1.2248,
+      "step": 12740
+    },
+    {
+      "epoch": 1.463557521107346,
+      "grad_norm": 0.6051421165466309,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 12741
+    },
+    {
+      "epoch": 1.4636723910171732,
+      "grad_norm": 0.59198397397995,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 12742
+    },
+    {
+      "epoch": 1.4637872609270002,
+      "grad_norm": 0.5700596570968628,
+      "learning_rate": 0.0001,
+      "loss": 1.2688,
+      "step": 12743
+    },
+    {
+      "epoch": 1.4639021308368272,
+      "grad_norm": 0.613609254360199,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 12744
+    },
+    {
+      "epoch": 1.4640170007466544,
+      "grad_norm": 0.6272452473640442,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 12745
+    },
+    {
+      "epoch": 1.4641318706564816,
+      "grad_norm": 0.5880221128463745,
+      "learning_rate": 0.0001,
+      "loss": 1.4157,
+      "step": 12746
+    },
+    {
+      "epoch": 1.4642467405663087,
+      "grad_norm": 0.6015984416007996,
+      "learning_rate": 0.0001,
+      "loss": 1.3987,
+      "step": 12747
+    },
+    {
+      "epoch": 1.4643616104761357,
+      "grad_norm": 0.6657059192657471,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 12748
+    },
+    {
+      "epoch": 1.464476480385963,
+      "grad_norm": 0.6863231062889099,
+      "learning_rate": 0.0001,
+      "loss": 1.3948,
+      "step": 12749
+    },
+    {
+      "epoch": 1.4645913502957901,
+      "grad_norm": 0.5913318395614624,
+      "learning_rate": 0.0001,
+      "loss": 1.412,
+      "step": 12750
+    },
+    {
+      "epoch": 1.4647062202056171,
+      "grad_norm": 0.5482379794120789,
+      "learning_rate": 0.0001,
+      "loss": 1.5154,
+      "step": 12751
+    },
+    {
+      "epoch": 1.4648210901154441,
+      "grad_norm": 0.5808722972869873,
+      "learning_rate": 0.0001,
+      "loss": 1.5977,
+      "step": 12752
+    },
+    {
+      "epoch": 1.4649359600252714,
+      "grad_norm": 0.553925633430481,
+      "learning_rate": 0.0001,
+      "loss": 1.2278,
+      "step": 12753
+    },
+    {
+      "epoch": 1.4650508299350986,
+      "grad_norm": 0.5533438920974731,
+      "learning_rate": 0.0001,
+      "loss": 1.401,
+      "step": 12754
+    },
+    {
+      "epoch": 1.4651656998449256,
+      "grad_norm": 0.6688501834869385,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 12755
+    },
+    {
+      "epoch": 1.4652805697547526,
+      "grad_norm": 0.6241793036460876,
+      "learning_rate": 0.0001,
+      "loss": 1.5635,
+      "step": 12756
+    },
+    {
+      "epoch": 1.4653954396645799,
+      "grad_norm": 0.6105805039405823,
+      "learning_rate": 0.0001,
+      "loss": 1.3178,
+      "step": 12757
+    },
+    {
+      "epoch": 1.465510309574407,
+      "grad_norm": 0.621698260307312,
+      "learning_rate": 0.0001,
+      "loss": 1.369,
+      "step": 12758
+    },
+    {
+      "epoch": 1.465625179484234,
+      "grad_norm": 0.6072282791137695,
+      "learning_rate": 0.0001,
+      "loss": 1.3693,
+      "step": 12759
+    },
+    {
+      "epoch": 1.465740049394061,
+      "grad_norm": 0.5882850885391235,
+      "learning_rate": 0.0001,
+      "loss": 1.4573,
+      "step": 12760
+    },
+    {
+      "epoch": 1.4658549193038883,
+      "grad_norm": 0.5509545207023621,
+      "learning_rate": 0.0001,
+      "loss": 1.4878,
+      "step": 12761
+    },
+    {
+      "epoch": 1.4659697892137156,
+      "grad_norm": 0.6017540693283081,
+      "learning_rate": 0.0001,
+      "loss": 1.2881,
+      "step": 12762
+    },
+    {
+      "epoch": 1.4660846591235426,
+      "grad_norm": 0.6178678870201111,
+      "learning_rate": 0.0001,
+      "loss": 1.4212,
+      "step": 12763
+    },
+    {
+      "epoch": 1.4661995290333696,
+      "grad_norm": 0.5732308626174927,
+      "learning_rate": 0.0001,
+      "loss": 1.4519,
+      "step": 12764
+    },
+    {
+      "epoch": 1.4663143989431968,
+      "grad_norm": 0.5819944739341736,
+      "learning_rate": 0.0001,
+      "loss": 1.4526,
+      "step": 12765
+    },
+    {
+      "epoch": 1.466429268853024,
+      "grad_norm": 0.6570119857788086,
+      "learning_rate": 0.0001,
+      "loss": 1.68,
+      "step": 12766
+    },
+    {
+      "epoch": 1.466544138762851,
+      "grad_norm": 0.6044562458992004,
+      "learning_rate": 0.0001,
+      "loss": 1.4818,
+      "step": 12767
+    },
+    {
+      "epoch": 1.466659008672678,
+      "grad_norm": 0.6516227126121521,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 12768
+    },
+    {
+      "epoch": 1.4667738785825053,
+      "grad_norm": 0.5833653211593628,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 12769
+    },
+    {
+      "epoch": 1.4668887484923325,
+      "grad_norm": 0.6204677820205688,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 12770
+    },
+    {
+      "epoch": 1.4670036184021595,
+      "grad_norm": 0.5666163563728333,
+      "learning_rate": 0.0001,
+      "loss": 1.4902,
+      "step": 12771
+    },
+    {
+      "epoch": 1.4671184883119865,
+      "grad_norm": 0.5615315437316895,
+      "learning_rate": 0.0001,
+      "loss": 1.3985,
+      "step": 12772
+    },
+    {
+      "epoch": 1.4672333582218138,
+      "grad_norm": 0.5774404406547546,
+      "learning_rate": 0.0001,
+      "loss": 1.2866,
+      "step": 12773
+    },
+    {
+      "epoch": 1.467348228131641,
+      "grad_norm": 0.6420813798904419,
+      "learning_rate": 0.0001,
+      "loss": 1.2587,
+      "step": 12774
+    },
+    {
+      "epoch": 1.467463098041468,
+      "grad_norm": 0.5721668601036072,
+      "learning_rate": 0.0001,
+      "loss": 1.4001,
+      "step": 12775
+    },
+    {
+      "epoch": 1.467577967951295,
+      "grad_norm": 0.5667369961738586,
+      "learning_rate": 0.0001,
+      "loss": 1.4447,
+      "step": 12776
+    },
+    {
+      "epoch": 1.4676928378611223,
+      "grad_norm": 0.5806662440299988,
+      "learning_rate": 0.0001,
+      "loss": 1.4093,
+      "step": 12777
+    },
+    {
+      "epoch": 1.4678077077709495,
+      "grad_norm": 0.5815864205360413,
+      "learning_rate": 0.0001,
+      "loss": 1.4261,
+      "step": 12778
+    },
+    {
+      "epoch": 1.4679225776807765,
+      "grad_norm": 0.5797282457351685,
+      "learning_rate": 0.0001,
+      "loss": 1.3684,
+      "step": 12779
+    },
+    {
+      "epoch": 1.4680374475906035,
+      "grad_norm": 0.686774492263794,
+      "learning_rate": 0.0001,
+      "loss": 1.6016,
+      "step": 12780
+    },
+    {
+      "epoch": 1.4681523175004307,
+      "grad_norm": 0.570772647857666,
+      "learning_rate": 0.0001,
+      "loss": 1.3664,
+      "step": 12781
+    },
+    {
+      "epoch": 1.468267187410258,
+      "grad_norm": 0.6441384553909302,
+      "learning_rate": 0.0001,
+      "loss": 1.5228,
+      "step": 12782
+    },
+    {
+      "epoch": 1.468382057320085,
+      "grad_norm": 0.6247745156288147,
+      "learning_rate": 0.0001,
+      "loss": 1.3402,
+      "step": 12783
+    },
+    {
+      "epoch": 1.468496927229912,
+      "grad_norm": 0.5949456095695496,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 12784
+    },
+    {
+      "epoch": 1.4686117971397392,
+      "grad_norm": 0.5714166760444641,
+      "learning_rate": 0.0001,
+      "loss": 1.4388,
+      "step": 12785
+    },
+    {
+      "epoch": 1.4687266670495664,
+      "grad_norm": 0.5946084856987,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 12786
+    },
+    {
+      "epoch": 1.4688415369593935,
+      "grad_norm": 0.607331395149231,
+      "learning_rate": 0.0001,
+      "loss": 1.3089,
+      "step": 12787
+    },
+    {
+      "epoch": 1.4689564068692207,
+      "grad_norm": 0.5990446209907532,
+      "learning_rate": 0.0001,
+      "loss": 1.4873,
+      "step": 12788
+    },
+    {
+      "epoch": 1.4690712767790477,
+      "grad_norm": 0.6370789408683777,
+      "learning_rate": 0.0001,
+      "loss": 1.5659,
+      "step": 12789
+    },
+    {
+      "epoch": 1.469186146688875,
+      "grad_norm": 0.6360231041908264,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 12790
+    },
+    {
+      "epoch": 1.469301016598702,
+      "grad_norm": 0.6127180457115173,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 12791
+    },
+    {
+      "epoch": 1.4694158865085292,
+      "grad_norm": 0.5970214009284973,
+      "learning_rate": 0.0001,
+      "loss": 1.2743,
+      "step": 12792
+    },
+    {
+      "epoch": 1.4695307564183562,
+      "grad_norm": 0.6505621075630188,
+      "learning_rate": 0.0001,
+      "loss": 1.7703,
+      "step": 12793
+    },
+    {
+      "epoch": 1.4696456263281834,
+      "grad_norm": 0.5832335352897644,
+      "learning_rate": 0.0001,
+      "loss": 1.3529,
+      "step": 12794
+    },
+    {
+      "epoch": 1.4697604962380104,
+      "grad_norm": 0.61517733335495,
+      "learning_rate": 0.0001,
+      "loss": 1.6536,
+      "step": 12795
+    },
+    {
+      "epoch": 1.4698753661478376,
+      "grad_norm": 0.5566610097885132,
+      "learning_rate": 0.0001,
+      "loss": 1.269,
+      "step": 12796
+    },
+    {
+      "epoch": 1.4699902360576647,
+      "grad_norm": 0.5564651489257812,
+      "learning_rate": 0.0001,
+      "loss": 1.2414,
+      "step": 12797
+    },
+    {
+      "epoch": 1.4701051059674919,
+      "grad_norm": 0.5723354816436768,
+      "learning_rate": 0.0001,
+      "loss": 1.4185,
+      "step": 12798
+    },
+    {
+      "epoch": 1.470219975877319,
+      "grad_norm": 0.6019946932792664,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 12799
+    },
+    {
+      "epoch": 1.4703348457871461,
+      "grad_norm": 0.6336812973022461,
+      "learning_rate": 0.0001,
+      "loss": 1.3159,
+      "step": 12800
+    },
+    {
+      "epoch": 1.4704497156969731,
+      "grad_norm": 0.602342426776886,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 12801
+    },
+    {
+      "epoch": 1.4705645856068004,
+      "grad_norm": 0.6818926334381104,
+      "learning_rate": 0.0001,
+      "loss": 1.4655,
+      "step": 12802
+    },
+    {
+      "epoch": 1.4706794555166274,
+      "grad_norm": 0.6180601716041565,
+      "learning_rate": 0.0001,
+      "loss": 1.457,
+      "step": 12803
+    },
+    {
+      "epoch": 1.4707943254264546,
+      "grad_norm": 0.6229212284088135,
+      "learning_rate": 0.0001,
+      "loss": 1.5556,
+      "step": 12804
+    },
+    {
+      "epoch": 1.4709091953362816,
+      "grad_norm": 0.6223541498184204,
+      "learning_rate": 0.0001,
+      "loss": 1.4241,
+      "step": 12805
+    },
+    {
+      "epoch": 1.4710240652461088,
+      "grad_norm": 0.5592056512832642,
+      "learning_rate": 0.0001,
+      "loss": 1.3271,
+      "step": 12806
+    },
+    {
+      "epoch": 1.4711389351559359,
+      "grad_norm": 0.569546639919281,
+      "learning_rate": 0.0001,
+      "loss": 1.4232,
+      "step": 12807
+    },
+    {
+      "epoch": 1.471253805065763,
+      "grad_norm": 0.5602309107780457,
+      "learning_rate": 0.0001,
+      "loss": 1.543,
+      "step": 12808
+    },
+    {
+      "epoch": 1.47136867497559,
+      "grad_norm": 0.7094202637672424,
+      "learning_rate": 0.0001,
+      "loss": 1.6989,
+      "step": 12809
+    },
+    {
+      "epoch": 1.4714835448854173,
+      "grad_norm": 0.6181465983390808,
+      "learning_rate": 0.0001,
+      "loss": 1.6225,
+      "step": 12810
+    },
+    {
+      "epoch": 1.4715984147952443,
+      "grad_norm": 0.5809864401817322,
+      "learning_rate": 0.0001,
+      "loss": 1.3377,
+      "step": 12811
+    },
+    {
+      "epoch": 1.4717132847050716,
+      "grad_norm": 0.5605827569961548,
+      "learning_rate": 0.0001,
+      "loss": 1.4811,
+      "step": 12812
+    },
+    {
+      "epoch": 1.4718281546148986,
+      "grad_norm": 0.5921470522880554,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 12813
+    },
+    {
+      "epoch": 1.4719430245247258,
+      "grad_norm": 0.5679218173027039,
+      "learning_rate": 0.0001,
+      "loss": 1.2956,
+      "step": 12814
+    },
+    {
+      "epoch": 1.4720578944345528,
+      "grad_norm": 0.6096241474151611,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 12815
+    },
+    {
+      "epoch": 1.47217276434438,
+      "grad_norm": 0.5679489970207214,
+      "learning_rate": 0.0001,
+      "loss": 1.4101,
+      "step": 12816
+    },
+    {
+      "epoch": 1.472287634254207,
+      "grad_norm": 0.5891135931015015,
+      "learning_rate": 0.0001,
+      "loss": 1.4284,
+      "step": 12817
+    },
+    {
+      "epoch": 1.4724025041640343,
+      "grad_norm": 0.6344767212867737,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 12818
+    },
+    {
+      "epoch": 1.4725173740738613,
+      "grad_norm": 0.5692608952522278,
+      "learning_rate": 0.0001,
+      "loss": 1.3893,
+      "step": 12819
+    },
+    {
+      "epoch": 1.4726322439836885,
+      "grad_norm": 0.6075809001922607,
+      "learning_rate": 0.0001,
+      "loss": 1.391,
+      "step": 12820
+    },
+    {
+      "epoch": 1.4727471138935155,
+      "grad_norm": 0.5786142349243164,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 12821
+    },
+    {
+      "epoch": 1.4728619838033428,
+      "grad_norm": 0.5931715965270996,
+      "learning_rate": 0.0001,
+      "loss": 1.3525,
+      "step": 12822
+    },
+    {
+      "epoch": 1.4729768537131698,
+      "grad_norm": 0.5575855374336243,
+      "learning_rate": 0.0001,
+      "loss": 1.4526,
+      "step": 12823
+    },
+    {
+      "epoch": 1.473091723622997,
+      "grad_norm": 0.5985252857208252,
+      "learning_rate": 0.0001,
+      "loss": 1.4732,
+      "step": 12824
+    },
+    {
+      "epoch": 1.473206593532824,
+      "grad_norm": 0.5632007718086243,
+      "learning_rate": 0.0001,
+      "loss": 1.5949,
+      "step": 12825
+    },
+    {
+      "epoch": 1.4733214634426512,
+      "grad_norm": 0.5793201923370361,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 12826
+    },
+    {
+      "epoch": 1.4734363333524783,
+      "grad_norm": 0.5845635533332825,
+      "learning_rate": 0.0001,
+      "loss": 1.4624,
+      "step": 12827
+    },
+    {
+      "epoch": 1.4735512032623055,
+      "grad_norm": 0.6314181089401245,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 12828
+    },
+    {
+      "epoch": 1.4736660731721325,
+      "grad_norm": 0.6269012093544006,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 12829
+    },
+    {
+      "epoch": 1.4737809430819597,
+      "grad_norm": 0.6091486215591431,
+      "learning_rate": 0.0001,
+      "loss": 1.3224,
+      "step": 12830
+    },
+    {
+      "epoch": 1.4738958129917867,
+      "grad_norm": 0.6564414501190186,
+      "learning_rate": 0.0001,
+      "loss": 1.482,
+      "step": 12831
+    },
+    {
+      "epoch": 1.474010682901614,
+      "grad_norm": 0.5696612000465393,
+      "learning_rate": 0.0001,
+      "loss": 1.403,
+      "step": 12832
+    },
+    {
+      "epoch": 1.474125552811441,
+      "grad_norm": 0.612686276435852,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 12833
+    },
+    {
+      "epoch": 1.4742404227212682,
+      "grad_norm": 0.5725769996643066,
+      "learning_rate": 0.0001,
+      "loss": 1.474,
+      "step": 12834
+    },
+    {
+      "epoch": 1.4743552926310952,
+      "grad_norm": 0.6267582774162292,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 12835
+    },
+    {
+      "epoch": 1.4744701625409224,
+      "grad_norm": 0.6302061676979065,
+      "learning_rate": 0.0001,
+      "loss": 1.4283,
+      "step": 12836
+    },
+    {
+      "epoch": 1.4745850324507495,
+      "grad_norm": 0.5919556617736816,
+      "learning_rate": 0.0001,
+      "loss": 1.4403,
+      "step": 12837
+    },
+    {
+      "epoch": 1.4746999023605767,
+      "grad_norm": 0.6446247100830078,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 12838
+    },
+    {
+      "epoch": 1.4748147722704037,
+      "grad_norm": 0.6327805519104004,
+      "learning_rate": 0.0001,
+      "loss": 1.4543,
+      "step": 12839
+    },
+    {
+      "epoch": 1.474929642180231,
+      "grad_norm": 0.6169029474258423,
+      "learning_rate": 0.0001,
+      "loss": 1.5194,
+      "step": 12840
+    },
+    {
+      "epoch": 1.475044512090058,
+      "grad_norm": 0.576792299747467,
+      "learning_rate": 0.0001,
+      "loss": 1.2857,
+      "step": 12841
+    },
+    {
+      "epoch": 1.4751593819998852,
+      "grad_norm": 0.6225866675376892,
+      "learning_rate": 0.0001,
+      "loss": 1.5189,
+      "step": 12842
+    },
+    {
+      "epoch": 1.4752742519097122,
+      "grad_norm": 0.5920522212982178,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 12843
+    },
+    {
+      "epoch": 1.4753891218195394,
+      "grad_norm": 0.5702757835388184,
+      "learning_rate": 0.0001,
+      "loss": 1.4564,
+      "step": 12844
+    },
+    {
+      "epoch": 1.4755039917293664,
+      "grad_norm": 0.5631170272827148,
+      "learning_rate": 0.0001,
+      "loss": 1.3581,
+      "step": 12845
+    },
+    {
+      "epoch": 1.4756188616391936,
+      "grad_norm": 0.5859529376029968,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 12846
+    },
+    {
+      "epoch": 1.4757337315490207,
+      "grad_norm": 0.5922108888626099,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 12847
+    },
+    {
+      "epoch": 1.4758486014588479,
+      "grad_norm": 0.5865465998649597,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 12848
+    },
+    {
+      "epoch": 1.475963471368675,
+      "grad_norm": 0.5963805317878723,
+      "learning_rate": 0.0001,
+      "loss": 1.3117,
+      "step": 12849
+    },
+    {
+      "epoch": 1.4760783412785021,
+      "grad_norm": 0.6101550459861755,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 12850
+    },
+    {
+      "epoch": 1.4761932111883291,
+      "grad_norm": 0.5703292489051819,
+      "learning_rate": 0.0001,
+      "loss": 1.4169,
+      "step": 12851
+    },
+    {
+      "epoch": 1.4763080810981564,
+      "grad_norm": 0.726719856262207,
+      "learning_rate": 0.0001,
+      "loss": 1.3441,
+      "step": 12852
+    },
+    {
+      "epoch": 1.4764229510079834,
+      "grad_norm": 0.5899834036827087,
+      "learning_rate": 0.0001,
+      "loss": 1.3448,
+      "step": 12853
+    },
+    {
+      "epoch": 1.4765378209178106,
+      "grad_norm": 0.6098437905311584,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 12854
+    },
+    {
+      "epoch": 1.4766526908276378,
+      "grad_norm": 0.5887052416801453,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 12855
+    },
+    {
+      "epoch": 1.4767675607374648,
+      "grad_norm": 0.6153585910797119,
+      "learning_rate": 0.0001,
+      "loss": 1.4361,
+      "step": 12856
+    },
+    {
+      "epoch": 1.4768824306472919,
+      "grad_norm": 0.6315271854400635,
+      "learning_rate": 0.0001,
+      "loss": 1.3454,
+      "step": 12857
+    },
+    {
+      "epoch": 1.476997300557119,
+      "grad_norm": 0.6163341403007507,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 12858
+    },
+    {
+      "epoch": 1.4771121704669463,
+      "grad_norm": 0.6163727045059204,
+      "learning_rate": 0.0001,
+      "loss": 1.4261,
+      "step": 12859
+    },
+    {
+      "epoch": 1.4772270403767733,
+      "grad_norm": 0.5802782773971558,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 12860
+    },
+    {
+      "epoch": 1.4773419102866003,
+      "grad_norm": 0.6903051733970642,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 12861
+    },
+    {
+      "epoch": 1.4774567801964276,
+      "grad_norm": 0.5828868746757507,
+      "learning_rate": 0.0001,
+      "loss": 1.381,
+      "step": 12862
+    },
+    {
+      "epoch": 1.4775716501062548,
+      "grad_norm": 0.5923581719398499,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 12863
+    },
+    {
+      "epoch": 1.4776865200160818,
+      "grad_norm": 0.606850802898407,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 12864
+    },
+    {
+      "epoch": 1.4778013899259088,
+      "grad_norm": 0.5898141264915466,
+      "learning_rate": 0.0001,
+      "loss": 1.2978,
+      "step": 12865
+    },
+    {
+      "epoch": 1.477916259835736,
+      "grad_norm": 0.5947832465171814,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 12866
+    },
+    {
+      "epoch": 1.4780311297455633,
+      "grad_norm": 0.5990186929702759,
+      "learning_rate": 0.0001,
+      "loss": 1.4618,
+      "step": 12867
+    },
+    {
+      "epoch": 1.4781459996553903,
+      "grad_norm": 0.6089809536933899,
+      "learning_rate": 0.0001,
+      "loss": 1.5126,
+      "step": 12868
+    },
+    {
+      "epoch": 1.4782608695652173,
+      "grad_norm": 0.5526347756385803,
+      "learning_rate": 0.0001,
+      "loss": 1.3861,
+      "step": 12869
+    },
+    {
+      "epoch": 1.4783757394750445,
+      "grad_norm": 0.631928026676178,
+      "learning_rate": 0.0001,
+      "loss": 1.4951,
+      "step": 12870
+    },
+    {
+      "epoch": 1.4784906093848718,
+      "grad_norm": 0.6184859275817871,
+      "learning_rate": 0.0001,
+      "loss": 1.3549,
+      "step": 12871
+    },
+    {
+      "epoch": 1.4786054792946988,
+      "grad_norm": 0.5659068822860718,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 12872
+    },
+    {
+      "epoch": 1.4787203492045258,
+      "grad_norm": 0.5731703639030457,
+      "learning_rate": 0.0001,
+      "loss": 1.1864,
+      "step": 12873
+    },
+    {
+      "epoch": 1.478835219114353,
+      "grad_norm": 0.5723096132278442,
+      "learning_rate": 0.0001,
+      "loss": 1.5536,
+      "step": 12874
+    },
+    {
+      "epoch": 1.4789500890241802,
+      "grad_norm": 0.5880361199378967,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 12875
+    },
+    {
+      "epoch": 1.4790649589340072,
+      "grad_norm": 0.6019386053085327,
+      "learning_rate": 0.0001,
+      "loss": 1.4355,
+      "step": 12876
+    },
+    {
+      "epoch": 1.4791798288438343,
+      "grad_norm": 0.6749945282936096,
+      "learning_rate": 0.0001,
+      "loss": 1.1855,
+      "step": 12877
+    },
+    {
+      "epoch": 1.4792946987536615,
+      "grad_norm": 0.5709156394004822,
+      "learning_rate": 0.0001,
+      "loss": 1.4813,
+      "step": 12878
+    },
+    {
+      "epoch": 1.4794095686634887,
+      "grad_norm": 0.604844331741333,
+      "learning_rate": 0.0001,
+      "loss": 1.487,
+      "step": 12879
+    },
+    {
+      "epoch": 1.4795244385733157,
+      "grad_norm": 0.608182966709137,
+      "learning_rate": 0.0001,
+      "loss": 1.3741,
+      "step": 12880
+    },
+    {
+      "epoch": 1.4796393084831427,
+      "grad_norm": 0.5870488882064819,
+      "learning_rate": 0.0001,
+      "loss": 1.3744,
+      "step": 12881
+    },
+    {
+      "epoch": 1.47975417839297,
+      "grad_norm": 0.6059988737106323,
+      "learning_rate": 0.0001,
+      "loss": 1.3412,
+      "step": 12882
+    },
+    {
+      "epoch": 1.4798690483027972,
+      "grad_norm": 0.6506321430206299,
+      "learning_rate": 0.0001,
+      "loss": 1.6666,
+      "step": 12883
+    },
+    {
+      "epoch": 1.4799839182126242,
+      "grad_norm": 0.5870694518089294,
+      "learning_rate": 0.0001,
+      "loss": 1.3905,
+      "step": 12884
+    },
+    {
+      "epoch": 1.4800987881224512,
+      "grad_norm": 0.5936076641082764,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 12885
+    },
+    {
+      "epoch": 1.4802136580322784,
+      "grad_norm": 0.6202870607376099,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 12886
+    },
+    {
+      "epoch": 1.4803285279421057,
+      "grad_norm": 0.5941158533096313,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 12887
+    },
+    {
+      "epoch": 1.4804433978519327,
+      "grad_norm": 0.5812159180641174,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 12888
+    },
+    {
+      "epoch": 1.4805582677617597,
+      "grad_norm": 0.6922297477722168,
+      "learning_rate": 0.0001,
+      "loss": 1.6634,
+      "step": 12889
+    },
+    {
+      "epoch": 1.480673137671587,
+      "grad_norm": 0.6124345064163208,
+      "learning_rate": 0.0001,
+      "loss": 1.3621,
+      "step": 12890
+    },
+    {
+      "epoch": 1.4807880075814142,
+      "grad_norm": 0.5976356267929077,
+      "learning_rate": 0.0001,
+      "loss": 1.4017,
+      "step": 12891
+    },
+    {
+      "epoch": 1.4809028774912412,
+      "grad_norm": 0.6526122689247131,
+      "learning_rate": 0.0001,
+      "loss": 1.2287,
+      "step": 12892
+    },
+    {
+      "epoch": 1.4810177474010682,
+      "grad_norm": 0.5740790367126465,
+      "learning_rate": 0.0001,
+      "loss": 1.4163,
+      "step": 12893
+    },
+    {
+      "epoch": 1.4811326173108954,
+      "grad_norm": 0.6110181212425232,
+      "learning_rate": 0.0001,
+      "loss": 1.3558,
+      "step": 12894
+    },
+    {
+      "epoch": 1.4812474872207226,
+      "grad_norm": 0.62778240442276,
+      "learning_rate": 0.0001,
+      "loss": 1.4462,
+      "step": 12895
+    },
+    {
+      "epoch": 1.4813623571305496,
+      "grad_norm": 0.6331201195716858,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 12896
+    },
+    {
+      "epoch": 1.4814772270403767,
+      "grad_norm": 0.647627055644989,
+      "learning_rate": 0.0001,
+      "loss": 1.3772,
+      "step": 12897
+    },
+    {
+      "epoch": 1.4815920969502039,
+      "grad_norm": 0.5829721093177795,
+      "learning_rate": 0.0001,
+      "loss": 1.4,
+      "step": 12898
+    },
+    {
+      "epoch": 1.4817069668600311,
+      "grad_norm": 0.6361407041549683,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 12899
+    },
+    {
+      "epoch": 1.4818218367698581,
+      "grad_norm": 0.5852049589157104,
+      "learning_rate": 0.0001,
+      "loss": 1.3346,
+      "step": 12900
+    },
+    {
+      "epoch": 1.4819367066796851,
+      "grad_norm": 0.5944021344184875,
+      "learning_rate": 0.0001,
+      "loss": 1.5299,
+      "step": 12901
+    },
+    {
+      "epoch": 1.4820515765895124,
+      "grad_norm": 0.5802842378616333,
+      "learning_rate": 0.0001,
+      "loss": 1.4209,
+      "step": 12902
+    },
+    {
+      "epoch": 1.4821664464993396,
+      "grad_norm": 0.582327663898468,
+      "learning_rate": 0.0001,
+      "loss": 1.2857,
+      "step": 12903
+    },
+    {
+      "epoch": 1.4822813164091666,
+      "grad_norm": 0.6188213229179382,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 12904
+    },
+    {
+      "epoch": 1.4823961863189936,
+      "grad_norm": 0.6501442790031433,
+      "learning_rate": 0.0001,
+      "loss": 1.3503,
+      "step": 12905
+    },
+    {
+      "epoch": 1.4825110562288208,
+      "grad_norm": 0.6042030453681946,
+      "learning_rate": 0.0001,
+      "loss": 1.3655,
+      "step": 12906
+    },
+    {
+      "epoch": 1.482625926138648,
+      "grad_norm": 0.6173478960990906,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 12907
+    },
+    {
+      "epoch": 1.482740796048475,
+      "grad_norm": 0.6603444814682007,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 12908
+    },
+    {
+      "epoch": 1.482855665958302,
+      "grad_norm": 0.5978038311004639,
+      "learning_rate": 0.0001,
+      "loss": 1.2935,
+      "step": 12909
+    },
+    {
+      "epoch": 1.4829705358681293,
+      "grad_norm": 0.6178515553474426,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 12910
+    },
+    {
+      "epoch": 1.4830854057779566,
+      "grad_norm": 0.6077051758766174,
+      "learning_rate": 0.0001,
+      "loss": 1.4124,
+      "step": 12911
+    },
+    {
+      "epoch": 1.4832002756877836,
+      "grad_norm": 0.6282074451446533,
+      "learning_rate": 0.0001,
+      "loss": 1.5461,
+      "step": 12912
+    },
+    {
+      "epoch": 1.4833151455976106,
+      "grad_norm": 0.5881144404411316,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 12913
+    },
+    {
+      "epoch": 1.4834300155074378,
+      "grad_norm": 0.5621463656425476,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 12914
+    },
+    {
+      "epoch": 1.483544885417265,
+      "grad_norm": 0.6100106239318848,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 12915
+    },
+    {
+      "epoch": 1.483659755327092,
+      "grad_norm": 0.5708780288696289,
+      "learning_rate": 0.0001,
+      "loss": 1.3741,
+      "step": 12916
+    },
+    {
+      "epoch": 1.483774625236919,
+      "grad_norm": 0.6632856726646423,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 12917
+    },
+    {
+      "epoch": 1.4838894951467463,
+      "grad_norm": 0.598359227180481,
+      "learning_rate": 0.0001,
+      "loss": 1.3337,
+      "step": 12918
+    },
+    {
+      "epoch": 1.4840043650565735,
+      "grad_norm": 0.6106839179992676,
+      "learning_rate": 0.0001,
+      "loss": 1.4881,
+      "step": 12919
+    },
+    {
+      "epoch": 1.4841192349664005,
+      "grad_norm": 0.6464507579803467,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 12920
+    },
+    {
+      "epoch": 1.4842341048762275,
+      "grad_norm": 0.6744695901870728,
+      "learning_rate": 0.0001,
+      "loss": 1.6106,
+      "step": 12921
+    },
+    {
+      "epoch": 1.4843489747860548,
+      "grad_norm": 0.6888026595115662,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 12922
+    },
+    {
+      "epoch": 1.484463844695882,
+      "grad_norm": 0.5786687731742859,
+      "learning_rate": 0.0001,
+      "loss": 1.4022,
+      "step": 12923
+    },
+    {
+      "epoch": 1.484578714605709,
+      "grad_norm": 0.5795042514801025,
+      "learning_rate": 0.0001,
+      "loss": 1.4461,
+      "step": 12924
+    },
+    {
+      "epoch": 1.4846935845155362,
+      "grad_norm": 0.5988849997520447,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 12925
+    },
+    {
+      "epoch": 1.4848084544253632,
+      "grad_norm": 0.5869229435920715,
+      "learning_rate": 0.0001,
+      "loss": 1.3634,
+      "step": 12926
+    },
+    {
+      "epoch": 1.4849233243351905,
+      "grad_norm": 0.6069849133491516,
+      "learning_rate": 0.0001,
+      "loss": 1.4881,
+      "step": 12927
+    },
+    {
+      "epoch": 1.4850381942450175,
+      "grad_norm": 0.6400530338287354,
+      "learning_rate": 0.0001,
+      "loss": 1.6889,
+      "step": 12928
+    },
+    {
+      "epoch": 1.4851530641548447,
+      "grad_norm": 0.605957567691803,
+      "learning_rate": 0.0001,
+      "loss": 1.3198,
+      "step": 12929
+    },
+    {
+      "epoch": 1.4852679340646717,
+      "grad_norm": 0.6827652454376221,
+      "learning_rate": 0.0001,
+      "loss": 1.6951,
+      "step": 12930
+    },
+    {
+      "epoch": 1.485382803974499,
+      "grad_norm": 0.6778180003166199,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 12931
+    },
+    {
+      "epoch": 1.485497673884326,
+      "grad_norm": 0.5894115567207336,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 12932
+    },
+    {
+      "epoch": 1.4856125437941532,
+      "grad_norm": 0.5633677244186401,
+      "learning_rate": 0.0001,
+      "loss": 1.4671,
+      "step": 12933
+    },
+    {
+      "epoch": 1.4857274137039802,
+      "grad_norm": 0.5746277570724487,
+      "learning_rate": 0.0001,
+      "loss": 1.3921,
+      "step": 12934
+    },
+    {
+      "epoch": 1.4858422836138074,
+      "grad_norm": 0.5946472883224487,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 12935
+    },
+    {
+      "epoch": 1.4859571535236344,
+      "grad_norm": 0.6017822027206421,
+      "learning_rate": 0.0001,
+      "loss": 1.3947,
+      "step": 12936
+    },
+    {
+      "epoch": 1.4860720234334617,
+      "grad_norm": 0.5574742555618286,
+      "learning_rate": 0.0001,
+      "loss": 1.3781,
+      "step": 12937
+    },
+    {
+      "epoch": 1.4861868933432887,
+      "grad_norm": 0.6060687899589539,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 12938
+    },
+    {
+      "epoch": 1.486301763253116,
+      "grad_norm": 0.5431711673736572,
+      "learning_rate": 0.0001,
+      "loss": 1.4442,
+      "step": 12939
+    },
+    {
+      "epoch": 1.486416633162943,
+      "grad_norm": 0.5980778336524963,
+      "learning_rate": 0.0001,
+      "loss": 1.4474,
+      "step": 12940
+    },
+    {
+      "epoch": 1.4865315030727702,
+      "grad_norm": 0.5813719034194946,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 12941
+    },
+    {
+      "epoch": 1.4866463729825972,
+      "grad_norm": 0.5789249539375305,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 12942
+    },
+    {
+      "epoch": 1.4867612428924244,
+      "grad_norm": 0.592308759689331,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 12943
+    },
+    {
+      "epoch": 1.4868761128022514,
+      "grad_norm": 0.5930988192558289,
+      "learning_rate": 0.0001,
+      "loss": 1.3819,
+      "step": 12944
+    },
+    {
+      "epoch": 1.4869909827120786,
+      "grad_norm": 0.5978574752807617,
+      "learning_rate": 0.0001,
+      "loss": 1.3716,
+      "step": 12945
+    },
+    {
+      "epoch": 1.4871058526219056,
+      "grad_norm": 0.554966151714325,
+      "learning_rate": 0.0001,
+      "loss": 1.2418,
+      "step": 12946
+    },
+    {
+      "epoch": 1.4872207225317329,
+      "grad_norm": 0.6341855525970459,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 12947
+    },
+    {
+      "epoch": 1.4873355924415599,
+      "grad_norm": 0.5823864340782166,
+      "learning_rate": 0.0001,
+      "loss": 1.461,
+      "step": 12948
+    },
+    {
+      "epoch": 1.4874504623513871,
+      "grad_norm": 0.5773583650588989,
+      "learning_rate": 0.0001,
+      "loss": 1.4707,
+      "step": 12949
+    },
+    {
+      "epoch": 1.4875653322612141,
+      "grad_norm": 0.6104335188865662,
+      "learning_rate": 0.0001,
+      "loss": 1.2777,
+      "step": 12950
+    },
+    {
+      "epoch": 1.4876802021710414,
+      "grad_norm": 0.5461379885673523,
+      "learning_rate": 0.0001,
+      "loss": 1.3399,
+      "step": 12951
+    },
+    {
+      "epoch": 1.4877950720808684,
+      "grad_norm": 0.5723500847816467,
+      "learning_rate": 0.0001,
+      "loss": 1.3066,
+      "step": 12952
+    },
+    {
+      "epoch": 1.4879099419906956,
+      "grad_norm": 0.6174570918083191,
+      "learning_rate": 0.0001,
+      "loss": 1.6148,
+      "step": 12953
+    },
+    {
+      "epoch": 1.4880248119005226,
+      "grad_norm": 0.5926865935325623,
+      "learning_rate": 0.0001,
+      "loss": 1.5758,
+      "step": 12954
+    },
+    {
+      "epoch": 1.4881396818103498,
+      "grad_norm": 0.593763530254364,
+      "learning_rate": 0.0001,
+      "loss": 1.4821,
+      "step": 12955
+    },
+    {
+      "epoch": 1.4882545517201768,
+      "grad_norm": 0.6111153364181519,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 12956
+    },
+    {
+      "epoch": 1.488369421630004,
+      "grad_norm": 0.6312853693962097,
+      "learning_rate": 0.0001,
+      "loss": 1.4476,
+      "step": 12957
+    },
+    {
+      "epoch": 1.488484291539831,
+      "grad_norm": 0.6247866153717041,
+      "learning_rate": 0.0001,
+      "loss": 1.3772,
+      "step": 12958
+    },
+    {
+      "epoch": 1.4885991614496583,
+      "grad_norm": 0.5472710728645325,
+      "learning_rate": 0.0001,
+      "loss": 1.1579,
+      "step": 12959
+    },
+    {
+      "epoch": 1.4887140313594853,
+      "grad_norm": 0.5936057567596436,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 12960
+    },
+    {
+      "epoch": 1.4888289012693126,
+      "grad_norm": 0.5668746829032898,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 12961
+    },
+    {
+      "epoch": 1.4889437711791396,
+      "grad_norm": 0.605015218257904,
+      "learning_rate": 0.0001,
+      "loss": 1.7382,
+      "step": 12962
+    },
+    {
+      "epoch": 1.4890586410889668,
+      "grad_norm": 0.6417834162712097,
+      "learning_rate": 0.0001,
+      "loss": 1.5262,
+      "step": 12963
+    },
+    {
+      "epoch": 1.4891735109987938,
+      "grad_norm": 0.6083729267120361,
+      "learning_rate": 0.0001,
+      "loss": 1.4273,
+      "step": 12964
+    },
+    {
+      "epoch": 1.489288380908621,
+      "grad_norm": 0.6126262545585632,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 12965
+    },
+    {
+      "epoch": 1.489403250818448,
+      "grad_norm": 0.6185763478279114,
+      "learning_rate": 0.0001,
+      "loss": 1.3363,
+      "step": 12966
+    },
+    {
+      "epoch": 1.4895181207282753,
+      "grad_norm": 0.6177452802658081,
+      "learning_rate": 0.0001,
+      "loss": 1.2466,
+      "step": 12967
+    },
+    {
+      "epoch": 1.4896329906381023,
+      "grad_norm": 0.6014622449874878,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 12968
+    },
+    {
+      "epoch": 1.4897478605479295,
+      "grad_norm": 0.6131553053855896,
+      "learning_rate": 0.0001,
+      "loss": 1.4818,
+      "step": 12969
+    },
+    {
+      "epoch": 1.4898627304577565,
+      "grad_norm": 0.6212412118911743,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 12970
+    },
+    {
+      "epoch": 1.4899776003675838,
+      "grad_norm": 0.595173180103302,
+      "learning_rate": 0.0001,
+      "loss": 1.4124,
+      "step": 12971
+    },
+    {
+      "epoch": 1.4900924702774108,
+      "grad_norm": 0.623760998249054,
+      "learning_rate": 0.0001,
+      "loss": 1.4035,
+      "step": 12972
+    },
+    {
+      "epoch": 1.490207340187238,
+      "grad_norm": 0.5903745889663696,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 12973
+    },
+    {
+      "epoch": 1.490322210097065,
+      "grad_norm": 0.5988485813140869,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 12974
+    },
+    {
+      "epoch": 1.4904370800068922,
+      "grad_norm": 0.592363178730011,
+      "learning_rate": 0.0001,
+      "loss": 1.3643,
+      "step": 12975
+    },
+    {
+      "epoch": 1.4905519499167192,
+      "grad_norm": 0.6512883901596069,
+      "learning_rate": 0.0001,
+      "loss": 1.3463,
+      "step": 12976
+    },
+    {
+      "epoch": 1.4906668198265465,
+      "grad_norm": 0.5865827798843384,
+      "learning_rate": 0.0001,
+      "loss": 1.3116,
+      "step": 12977
+    },
+    {
+      "epoch": 1.4907816897363735,
+      "grad_norm": 0.5729326605796814,
+      "learning_rate": 0.0001,
+      "loss": 1.3622,
+      "step": 12978
+    },
+    {
+      "epoch": 1.4908965596462007,
+      "grad_norm": 0.599138617515564,
+      "learning_rate": 0.0001,
+      "loss": 1.5096,
+      "step": 12979
+    },
+    {
+      "epoch": 1.4910114295560277,
+      "grad_norm": 0.6441994905471802,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 12980
+    },
+    {
+      "epoch": 1.491126299465855,
+      "grad_norm": 0.5867522358894348,
+      "learning_rate": 0.0001,
+      "loss": 1.4342,
+      "step": 12981
+    },
+    {
+      "epoch": 1.491241169375682,
+      "grad_norm": 0.6497024297714233,
+      "learning_rate": 0.0001,
+      "loss": 1.5281,
+      "step": 12982
+    },
+    {
+      "epoch": 1.4913560392855092,
+      "grad_norm": 0.6006242632865906,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 12983
+    },
+    {
+      "epoch": 1.4914709091953362,
+      "grad_norm": 0.5676483511924744,
+      "learning_rate": 0.0001,
+      "loss": 1.2727,
+      "step": 12984
+    },
+    {
+      "epoch": 1.4915857791051634,
+      "grad_norm": 0.6318432092666626,
+      "learning_rate": 0.0001,
+      "loss": 1.56,
+      "step": 12985
+    },
+    {
+      "epoch": 1.4917006490149904,
+      "grad_norm": 0.5977509021759033,
+      "learning_rate": 0.0001,
+      "loss": 1.289,
+      "step": 12986
+    },
+    {
+      "epoch": 1.4918155189248177,
+      "grad_norm": 0.6188353896141052,
+      "learning_rate": 0.0001,
+      "loss": 1.3374,
+      "step": 12987
+    },
+    {
+      "epoch": 1.4919303888346447,
+      "grad_norm": 0.5568304657936096,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 12988
+    },
+    {
+      "epoch": 1.492045258744472,
+      "grad_norm": 0.5960107445716858,
+      "learning_rate": 0.0001,
+      "loss": 1.6009,
+      "step": 12989
+    },
+    {
+      "epoch": 1.492160128654299,
+      "grad_norm": 0.5843378305435181,
+      "learning_rate": 0.0001,
+      "loss": 1.4434,
+      "step": 12990
+    },
+    {
+      "epoch": 1.4922749985641262,
+      "grad_norm": 0.5970795154571533,
+      "learning_rate": 0.0001,
+      "loss": 1.6927,
+      "step": 12991
+    },
+    {
+      "epoch": 1.4923898684739534,
+      "grad_norm": 0.5971647500991821,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 12992
+    },
+    {
+      "epoch": 1.4925047383837804,
+      "grad_norm": 0.5765950679779053,
+      "learning_rate": 0.0001,
+      "loss": 1.4302,
+      "step": 12993
+    },
+    {
+      "epoch": 1.4926196082936074,
+      "grad_norm": 0.6342061758041382,
+      "learning_rate": 0.0001,
+      "loss": 1.4708,
+      "step": 12994
+    },
+    {
+      "epoch": 1.4927344782034346,
+      "grad_norm": 0.5808407664299011,
+      "learning_rate": 0.0001,
+      "loss": 1.4504,
+      "step": 12995
+    },
+    {
+      "epoch": 1.4928493481132619,
+      "grad_norm": 0.7014811635017395,
+      "learning_rate": 0.0001,
+      "loss": 1.6261,
+      "step": 12996
+    },
+    {
+      "epoch": 1.4929642180230889,
+      "grad_norm": 0.6212725639343262,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 12997
+    },
+    {
+      "epoch": 1.4930790879329159,
+      "grad_norm": 0.5805021524429321,
+      "learning_rate": 0.0001,
+      "loss": 1.4603,
+      "step": 12998
+    },
+    {
+      "epoch": 1.4931939578427431,
+      "grad_norm": 0.5734280943870544,
+      "learning_rate": 0.0001,
+      "loss": 1.5504,
+      "step": 12999
+    },
+    {
+      "epoch": 1.4933088277525703,
+      "grad_norm": 0.5856978297233582,
+      "learning_rate": 0.0001,
+      "loss": 1.3655,
+      "step": 13000
+    },
+    {
+      "epoch": 1.4934236976623974,
+      "grad_norm": 0.5353971123695374,
+      "learning_rate": 0.0001,
+      "loss": 1.2464,
+      "step": 13001
+    },
+    {
+      "epoch": 1.4935385675722244,
+      "grad_norm": 0.6052659749984741,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 13002
+    },
+    {
+      "epoch": 1.4936534374820516,
+      "grad_norm": 0.5781797170639038,
+      "learning_rate": 0.0001,
+      "loss": 1.3597,
+      "step": 13003
+    },
+    {
+      "epoch": 1.4937683073918788,
+      "grad_norm": 0.576927125453949,
+      "learning_rate": 0.0001,
+      "loss": 1.4228,
+      "step": 13004
+    },
+    {
+      "epoch": 1.4938831773017058,
+      "grad_norm": 0.6039324998855591,
+      "learning_rate": 0.0001,
+      "loss": 1.4246,
+      "step": 13005
+    },
+    {
+      "epoch": 1.4939980472115328,
+      "grad_norm": 0.5810524821281433,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 13006
+    },
+    {
+      "epoch": 1.49411291712136,
+      "grad_norm": 0.5944836735725403,
+      "learning_rate": 0.0001,
+      "loss": 1.4696,
+      "step": 13007
+    },
+    {
+      "epoch": 1.4942277870311873,
+      "grad_norm": 0.6197353601455688,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 13008
+    },
+    {
+      "epoch": 1.4943426569410143,
+      "grad_norm": 0.5848913788795471,
+      "learning_rate": 0.0001,
+      "loss": 1.4212,
+      "step": 13009
+    },
+    {
+      "epoch": 1.4944575268508413,
+      "grad_norm": 0.6071828603744507,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 13010
+    },
+    {
+      "epoch": 1.4945723967606686,
+      "grad_norm": 0.6387497186660767,
+      "learning_rate": 0.0001,
+      "loss": 1.5511,
+      "step": 13011
+    },
+    {
+      "epoch": 1.4946872666704958,
+      "grad_norm": 0.5986767411231995,
+      "learning_rate": 0.0001,
+      "loss": 1.461,
+      "step": 13012
+    },
+    {
+      "epoch": 1.4948021365803228,
+      "grad_norm": 0.6346409916877747,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 13013
+    },
+    {
+      "epoch": 1.4949170064901498,
+      "grad_norm": 0.6802158951759338,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 13014
+    },
+    {
+      "epoch": 1.495031876399977,
+      "grad_norm": 0.6537509560585022,
+      "learning_rate": 0.0001,
+      "loss": 1.1075,
+      "step": 13015
+    },
+    {
+      "epoch": 1.4951467463098043,
+      "grad_norm": 0.5992099642753601,
+      "learning_rate": 0.0001,
+      "loss": 1.3288,
+      "step": 13016
+    },
+    {
+      "epoch": 1.4952616162196313,
+      "grad_norm": 0.6179826855659485,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 13017
+    },
+    {
+      "epoch": 1.4953764861294583,
+      "grad_norm": 0.6495833992958069,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 13018
+    },
+    {
+      "epoch": 1.4954913560392855,
+      "grad_norm": 0.6056380271911621,
+      "learning_rate": 0.0001,
+      "loss": 1.4111,
+      "step": 13019
+    },
+    {
+      "epoch": 1.4956062259491127,
+      "grad_norm": 0.7197732925415039,
+      "learning_rate": 0.0001,
+      "loss": 1.8647,
+      "step": 13020
+    },
+    {
+      "epoch": 1.4957210958589398,
+      "grad_norm": 0.63360196352005,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 13021
+    },
+    {
+      "epoch": 1.4958359657687668,
+      "grad_norm": 0.6025167107582092,
+      "learning_rate": 0.0001,
+      "loss": 1.425,
+      "step": 13022
+    },
+    {
+      "epoch": 1.495950835678594,
+      "grad_norm": 0.596898078918457,
+      "learning_rate": 0.0001,
+      "loss": 1.6411,
+      "step": 13023
+    },
+    {
+      "epoch": 1.4960657055884212,
+      "grad_norm": 0.5755943059921265,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 13024
+    },
+    {
+      "epoch": 1.4961805754982482,
+      "grad_norm": 0.5723254084587097,
+      "learning_rate": 0.0001,
+      "loss": 1.5093,
+      "step": 13025
+    },
+    {
+      "epoch": 1.4962954454080752,
+      "grad_norm": 0.5714227557182312,
+      "learning_rate": 0.0001,
+      "loss": 1.3679,
+      "step": 13026
+    },
+    {
+      "epoch": 1.4964103153179025,
+      "grad_norm": 0.6034862399101257,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 13027
+    },
+    {
+      "epoch": 1.4965251852277297,
+      "grad_norm": 0.6068670749664307,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 13028
+    },
+    {
+      "epoch": 1.4966400551375567,
+      "grad_norm": 0.5679345726966858,
+      "learning_rate": 0.0001,
+      "loss": 1.236,
+      "step": 13029
+    },
+    {
+      "epoch": 1.4967549250473837,
+      "grad_norm": 0.56649249792099,
+      "learning_rate": 0.0001,
+      "loss": 1.3815,
+      "step": 13030
+    },
+    {
+      "epoch": 1.496869794957211,
+      "grad_norm": 0.617186963558197,
+      "learning_rate": 0.0001,
+      "loss": 1.4181,
+      "step": 13031
+    },
+    {
+      "epoch": 1.4969846648670382,
+      "grad_norm": 0.5779293179512024,
+      "learning_rate": 0.0001,
+      "loss": 1.4227,
+      "step": 13032
+    },
+    {
+      "epoch": 1.4970995347768652,
+      "grad_norm": 0.6814537644386292,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 13033
+    },
+    {
+      "epoch": 1.4972144046866922,
+      "grad_norm": 0.6161949038505554,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 13034
+    },
+    {
+      "epoch": 1.4973292745965194,
+      "grad_norm": 0.6186149716377258,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 13035
+    },
+    {
+      "epoch": 1.4974441445063467,
+      "grad_norm": 0.6470484137535095,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 13036
+    },
+    {
+      "epoch": 1.4975590144161737,
+      "grad_norm": 0.6378332376480103,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 13037
+    },
+    {
+      "epoch": 1.4976738843260007,
+      "grad_norm": 0.6506776809692383,
+      "learning_rate": 0.0001,
+      "loss": 1.6681,
+      "step": 13038
+    },
+    {
+      "epoch": 1.497788754235828,
+      "grad_norm": 0.6596044301986694,
+      "learning_rate": 0.0001,
+      "loss": 1.634,
+      "step": 13039
+    },
+    {
+      "epoch": 1.4979036241456551,
+      "grad_norm": 0.5984453558921814,
+      "learning_rate": 0.0001,
+      "loss": 1.3366,
+      "step": 13040
+    },
+    {
+      "epoch": 1.4980184940554822,
+      "grad_norm": 0.6259294748306274,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 13041
+    },
+    {
+      "epoch": 1.4981333639653092,
+      "grad_norm": 0.6096507906913757,
+      "learning_rate": 0.0001,
+      "loss": 1.4598,
+      "step": 13042
+    },
+    {
+      "epoch": 1.4982482338751364,
+      "grad_norm": 0.6044954657554626,
+      "learning_rate": 0.0001,
+      "loss": 1.4842,
+      "step": 13043
+    },
+    {
+      "epoch": 1.4983631037849636,
+      "grad_norm": 0.5996847748756409,
+      "learning_rate": 0.0001,
+      "loss": 1.4205,
+      "step": 13044
+    },
+    {
+      "epoch": 1.4984779736947906,
+      "grad_norm": 0.5740419030189514,
+      "learning_rate": 0.0001,
+      "loss": 1.3223,
+      "step": 13045
+    },
+    {
+      "epoch": 1.4985928436046176,
+      "grad_norm": 0.6097593903541565,
+      "learning_rate": 0.0001,
+      "loss": 1.4634,
+      "step": 13046
+    },
+    {
+      "epoch": 1.4987077135144449,
+      "grad_norm": 0.6650487780570984,
+      "learning_rate": 0.0001,
+      "loss": 1.3505,
+      "step": 13047
+    },
+    {
+      "epoch": 1.498822583424272,
+      "grad_norm": 0.5681633949279785,
+      "learning_rate": 0.0001,
+      "loss": 1.3723,
+      "step": 13048
+    },
+    {
+      "epoch": 1.4989374533340991,
+      "grad_norm": 0.6020470857620239,
+      "learning_rate": 0.0001,
+      "loss": 1.4584,
+      "step": 13049
+    },
+    {
+      "epoch": 1.4990523232439261,
+      "grad_norm": 0.6068710088729858,
+      "learning_rate": 0.0001,
+      "loss": 1.2623,
+      "step": 13050
+    },
+    {
+      "epoch": 1.4991671931537534,
+      "grad_norm": 0.6180760860443115,
+      "learning_rate": 0.0001,
+      "loss": 1.4976,
+      "step": 13051
+    },
+    {
+      "epoch": 1.4992820630635806,
+      "grad_norm": 0.6852160096168518,
+      "learning_rate": 0.0001,
+      "loss": 1.3216,
+      "step": 13052
+    },
+    {
+      "epoch": 1.4993969329734076,
+      "grad_norm": 0.6118769645690918,
+      "learning_rate": 0.0001,
+      "loss": 1.4306,
+      "step": 13053
+    },
+    {
+      "epoch": 1.4995118028832346,
+      "grad_norm": 0.5923478007316589,
+      "learning_rate": 0.0001,
+      "loss": 1.3799,
+      "step": 13054
+    },
+    {
+      "epoch": 1.4996266727930618,
+      "grad_norm": 0.5937912464141846,
+      "learning_rate": 0.0001,
+      "loss": 1.2611,
+      "step": 13055
+    },
+    {
+      "epoch": 1.499741542702889,
+      "grad_norm": 0.5981078147888184,
+      "learning_rate": 0.0001,
+      "loss": 1.363,
+      "step": 13056
+    },
+    {
+      "epoch": 1.499856412612716,
+      "grad_norm": 0.5924525260925293,
+      "learning_rate": 0.0001,
+      "loss": 1.2294,
+      "step": 13057
+    },
+    {
+      "epoch": 1.499971282522543,
+      "grad_norm": 0.6042497754096985,
+      "learning_rate": 0.0001,
+      "loss": 1.3568,
+      "step": 13058
+    },
+    {
+      "epoch": 1.5000861524323703,
+      "grad_norm": 0.6669368147850037,
+      "learning_rate": 0.0001,
+      "loss": 1.3832,
+      "step": 13059
+    },
+    {
+      "epoch": 1.5002010223421975,
+      "grad_norm": 0.6114850044250488,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 13060
+    },
+    {
+      "epoch": 1.5003158922520246,
+      "grad_norm": 0.6326112151145935,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 13061
+    },
+    {
+      "epoch": 1.5004307621618516,
+      "grad_norm": 0.648796796798706,
+      "learning_rate": 0.0001,
+      "loss": 1.5702,
+      "step": 13062
+    },
+    {
+      "epoch": 1.5005456320716788,
+      "grad_norm": 0.618161141872406,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 13063
+    },
+    {
+      "epoch": 1.500660501981506,
+      "grad_norm": 0.5921330451965332,
+      "learning_rate": 0.0001,
+      "loss": 1.3333,
+      "step": 13064
+    },
+    {
+      "epoch": 1.500775371891333,
+      "grad_norm": 0.6372246146202087,
+      "learning_rate": 0.0001,
+      "loss": 1.7721,
+      "step": 13065
+    },
+    {
+      "epoch": 1.50089024180116,
+      "grad_norm": 0.5890547037124634,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 13066
+    },
+    {
+      "epoch": 1.5010051117109873,
+      "grad_norm": 0.5621899366378784,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 13067
+    },
+    {
+      "epoch": 1.5011199816208145,
+      "grad_norm": 0.6191715598106384,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 13068
+    },
+    {
+      "epoch": 1.5012348515306415,
+      "grad_norm": 0.601288914680481,
+      "learning_rate": 0.0001,
+      "loss": 1.4675,
+      "step": 13069
+    },
+    {
+      "epoch": 1.5013497214404685,
+      "grad_norm": 0.5595393180847168,
+      "learning_rate": 0.0001,
+      "loss": 1.4565,
+      "step": 13070
+    },
+    {
+      "epoch": 1.5014645913502958,
+      "grad_norm": 0.6976054906845093,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 13071
+    },
+    {
+      "epoch": 1.501579461260123,
+      "grad_norm": 0.5838610529899597,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 13072
+    },
+    {
+      "epoch": 1.50169433116995,
+      "grad_norm": 0.6203344464302063,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 13073
+    },
+    {
+      "epoch": 1.501809201079777,
+      "grad_norm": 0.5664482712745667,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 13074
+    },
+    {
+      "epoch": 1.5019240709896042,
+      "grad_norm": 0.5609536170959473,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 13075
+    },
+    {
+      "epoch": 1.5020389408994315,
+      "grad_norm": 0.6260766386985779,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 13076
+    },
+    {
+      "epoch": 1.5021538108092585,
+      "grad_norm": 0.6110790967941284,
+      "learning_rate": 0.0001,
+      "loss": 1.5518,
+      "step": 13077
+    },
+    {
+      "epoch": 1.5022686807190855,
+      "grad_norm": 0.590969443321228,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 13078
+    },
+    {
+      "epoch": 1.5023835506289127,
+      "grad_norm": 0.5963428616523743,
+      "learning_rate": 0.0001,
+      "loss": 1.4506,
+      "step": 13079
+    },
+    {
+      "epoch": 1.50249842053874,
+      "grad_norm": 0.649579644203186,
+      "learning_rate": 0.0001,
+      "loss": 1.4977,
+      "step": 13080
+    },
+    {
+      "epoch": 1.502613290448567,
+      "grad_norm": 0.5847672820091248,
+      "learning_rate": 0.0001,
+      "loss": 1.4921,
+      "step": 13081
+    },
+    {
+      "epoch": 1.502728160358394,
+      "grad_norm": 0.6328948736190796,
+      "learning_rate": 0.0001,
+      "loss": 1.188,
+      "step": 13082
+    },
+    {
+      "epoch": 1.5028430302682212,
+      "grad_norm": 0.5965337753295898,
+      "learning_rate": 0.0001,
+      "loss": 1.413,
+      "step": 13083
+    },
+    {
+      "epoch": 1.5029579001780484,
+      "grad_norm": 0.6650028824806213,
+      "learning_rate": 0.0001,
+      "loss": 1.4197,
+      "step": 13084
+    },
+    {
+      "epoch": 1.5030727700878754,
+      "grad_norm": 0.6406525373458862,
+      "learning_rate": 0.0001,
+      "loss": 1.6736,
+      "step": 13085
+    },
+    {
+      "epoch": 1.5031876399977024,
+      "grad_norm": 0.5682651400566101,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 13086
+    },
+    {
+      "epoch": 1.5033025099075297,
+      "grad_norm": 0.6791881322860718,
+      "learning_rate": 0.0001,
+      "loss": 1.4621,
+      "step": 13087
+    },
+    {
+      "epoch": 1.503417379817357,
+      "grad_norm": 0.6566523909568787,
+      "learning_rate": 0.0001,
+      "loss": 1.469,
+      "step": 13088
+    },
+    {
+      "epoch": 1.503532249727184,
+      "grad_norm": 0.6123066544532776,
+      "learning_rate": 0.0001,
+      "loss": 1.3571,
+      "step": 13089
+    },
+    {
+      "epoch": 1.503647119637011,
+      "grad_norm": 0.6262611746788025,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 13090
+    },
+    {
+      "epoch": 1.5037619895468382,
+      "grad_norm": 0.5882257223129272,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 13091
+    },
+    {
+      "epoch": 1.5038768594566654,
+      "grad_norm": 0.6783078908920288,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 13092
+    },
+    {
+      "epoch": 1.5039917293664926,
+      "grad_norm": 0.6360931992530823,
+      "learning_rate": 0.0001,
+      "loss": 1.3838,
+      "step": 13093
+    },
+    {
+      "epoch": 1.5041065992763196,
+      "grad_norm": 0.5902203917503357,
+      "learning_rate": 0.0001,
+      "loss": 1.3392,
+      "step": 13094
+    },
+    {
+      "epoch": 1.5042214691861466,
+      "grad_norm": 0.5917932987213135,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 13095
+    },
+    {
+      "epoch": 1.5043363390959739,
+      "grad_norm": 0.5984362959861755,
+      "learning_rate": 0.0001,
+      "loss": 1.4222,
+      "step": 13096
+    },
+    {
+      "epoch": 1.504451209005801,
+      "grad_norm": 0.5655670166015625,
+      "learning_rate": 0.0001,
+      "loss": 1.3595,
+      "step": 13097
+    },
+    {
+      "epoch": 1.504566078915628,
+      "grad_norm": 0.5472081899642944,
+      "learning_rate": 0.0001,
+      "loss": 1.3729,
+      "step": 13098
+    },
+    {
+      "epoch": 1.5046809488254551,
+      "grad_norm": 0.6080209016799927,
+      "learning_rate": 0.0001,
+      "loss": 1.3025,
+      "step": 13099
+    },
+    {
+      "epoch": 1.5047958187352823,
+      "grad_norm": 0.6341947913169861,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 13100
+    },
+    {
+      "epoch": 1.5049106886451096,
+      "grad_norm": 0.6027140617370605,
+      "learning_rate": 0.0001,
+      "loss": 1.4337,
+      "step": 13101
+    },
+    {
+      "epoch": 1.5050255585549366,
+      "grad_norm": 0.6108810901641846,
+      "learning_rate": 0.0001,
+      "loss": 1.1485,
+      "step": 13102
+    },
+    {
+      "epoch": 1.5051404284647636,
+      "grad_norm": 0.6450017094612122,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 13103
+    },
+    {
+      "epoch": 1.5052552983745908,
+      "grad_norm": 0.612722635269165,
+      "learning_rate": 0.0001,
+      "loss": 1.4161,
+      "step": 13104
+    },
+    {
+      "epoch": 1.505370168284418,
+      "grad_norm": 0.6334837675094604,
+      "learning_rate": 0.0001,
+      "loss": 1.3841,
+      "step": 13105
+    },
+    {
+      "epoch": 1.505485038194245,
+      "grad_norm": 0.5804214477539062,
+      "learning_rate": 0.0001,
+      "loss": 1.4075,
+      "step": 13106
+    },
+    {
+      "epoch": 1.505599908104072,
+      "grad_norm": 0.590505063533783,
+      "learning_rate": 0.0001,
+      "loss": 1.3492,
+      "step": 13107
+    },
+    {
+      "epoch": 1.5057147780138993,
+      "grad_norm": 0.5730953216552734,
+      "learning_rate": 0.0001,
+      "loss": 1.6026,
+      "step": 13108
+    },
+    {
+      "epoch": 1.5058296479237265,
+      "grad_norm": 0.6718409061431885,
+      "learning_rate": 0.0001,
+      "loss": 1.4637,
+      "step": 13109
+    },
+    {
+      "epoch": 1.5059445178335535,
+      "grad_norm": 0.5778510570526123,
+      "learning_rate": 0.0001,
+      "loss": 1.419,
+      "step": 13110
+    },
+    {
+      "epoch": 1.5060593877433806,
+      "grad_norm": 0.5760291218757629,
+      "learning_rate": 0.0001,
+      "loss": 1.4774,
+      "step": 13111
+    },
+    {
+      "epoch": 1.5061742576532078,
+      "grad_norm": 0.6402103900909424,
+      "learning_rate": 0.0001,
+      "loss": 1.4435,
+      "step": 13112
+    },
+    {
+      "epoch": 1.506289127563035,
+      "grad_norm": 0.6180935502052307,
+      "learning_rate": 0.0001,
+      "loss": 1.5586,
+      "step": 13113
+    },
+    {
+      "epoch": 1.506403997472862,
+      "grad_norm": 0.5968201756477356,
+      "learning_rate": 0.0001,
+      "loss": 1.2183,
+      "step": 13114
+    },
+    {
+      "epoch": 1.506518867382689,
+      "grad_norm": 0.5823055505752563,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 13115
+    },
+    {
+      "epoch": 1.5066337372925163,
+      "grad_norm": 0.6111396551132202,
+      "learning_rate": 0.0001,
+      "loss": 1.4014,
+      "step": 13116
+    },
+    {
+      "epoch": 1.5067486072023435,
+      "grad_norm": 0.6117756366729736,
+      "learning_rate": 0.0001,
+      "loss": 1.3061,
+      "step": 13117
+    },
+    {
+      "epoch": 1.5068634771121705,
+      "grad_norm": 0.6762123703956604,
+      "learning_rate": 0.0001,
+      "loss": 1.6192,
+      "step": 13118
+    },
+    {
+      "epoch": 1.5069783470219975,
+      "grad_norm": 0.6174688339233398,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 13119
+    },
+    {
+      "epoch": 1.5070932169318247,
+      "grad_norm": 0.6349688172340393,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 13120
+    },
+    {
+      "epoch": 1.507208086841652,
+      "grad_norm": 0.6256809830665588,
+      "learning_rate": 0.0001,
+      "loss": 1.3266,
+      "step": 13121
+    },
+    {
+      "epoch": 1.507322956751479,
+      "grad_norm": 0.6662157773971558,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 13122
+    },
+    {
+      "epoch": 1.507437826661306,
+      "grad_norm": 0.6114436984062195,
+      "learning_rate": 0.0001,
+      "loss": 1.4598,
+      "step": 13123
+    },
+    {
+      "epoch": 1.5075526965711332,
+      "grad_norm": 0.6376405358314514,
+      "learning_rate": 0.0001,
+      "loss": 1.4969,
+      "step": 13124
+    },
+    {
+      "epoch": 1.5076675664809605,
+      "grad_norm": 0.6283373832702637,
+      "learning_rate": 0.0001,
+      "loss": 1.4329,
+      "step": 13125
+    },
+    {
+      "epoch": 1.5077824363907875,
+      "grad_norm": 0.6334078311920166,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 13126
+    },
+    {
+      "epoch": 1.5078973063006145,
+      "grad_norm": 0.606662929058075,
+      "learning_rate": 0.0001,
+      "loss": 1.3271,
+      "step": 13127
+    },
+    {
+      "epoch": 1.5080121762104417,
+      "grad_norm": 0.6204425096511841,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 13128
+    },
+    {
+      "epoch": 1.508127046120269,
+      "grad_norm": 0.6233339905738831,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 13129
+    },
+    {
+      "epoch": 1.508241916030096,
+      "grad_norm": 0.5537604093551636,
+      "learning_rate": 0.0001,
+      "loss": 1.4306,
+      "step": 13130
+    },
+    {
+      "epoch": 1.508356785939923,
+      "grad_norm": 0.5744999051094055,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 13131
+    },
+    {
+      "epoch": 1.5084716558497502,
+      "grad_norm": 0.5970391035079956,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 13132
+    },
+    {
+      "epoch": 1.5085865257595774,
+      "grad_norm": 0.5616945028305054,
+      "learning_rate": 0.0001,
+      "loss": 1.2999,
+      "step": 13133
+    },
+    {
+      "epoch": 1.5087013956694044,
+      "grad_norm": 0.566780686378479,
+      "learning_rate": 0.0001,
+      "loss": 1.3291,
+      "step": 13134
+    },
+    {
+      "epoch": 1.5088162655792314,
+      "grad_norm": 0.6132379174232483,
+      "learning_rate": 0.0001,
+      "loss": 1.3745,
+      "step": 13135
+    },
+    {
+      "epoch": 1.5089311354890587,
+      "grad_norm": 0.6889569163322449,
+      "learning_rate": 0.0001,
+      "loss": 1.4654,
+      "step": 13136
+    },
+    {
+      "epoch": 1.509046005398886,
+      "grad_norm": 0.5599787831306458,
+      "learning_rate": 0.0001,
+      "loss": 1.3133,
+      "step": 13137
+    },
+    {
+      "epoch": 1.509160875308713,
+      "grad_norm": 0.5910610556602478,
+      "learning_rate": 0.0001,
+      "loss": 1.4698,
+      "step": 13138
+    },
+    {
+      "epoch": 1.50927574521854,
+      "grad_norm": 0.6421205401420593,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 13139
+    },
+    {
+      "epoch": 1.5093906151283671,
+      "grad_norm": 0.6151227355003357,
+      "learning_rate": 0.0001,
+      "loss": 1.3632,
+      "step": 13140
+    },
+    {
+      "epoch": 1.5095054850381944,
+      "grad_norm": 0.569758951663971,
+      "learning_rate": 0.0001,
+      "loss": 1.3274,
+      "step": 13141
+    },
+    {
+      "epoch": 1.5096203549480214,
+      "grad_norm": 0.6561845541000366,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 13142
+    },
+    {
+      "epoch": 1.5097352248578484,
+      "grad_norm": 0.6225590705871582,
+      "learning_rate": 0.0001,
+      "loss": 1.6304,
+      "step": 13143
+    },
+    {
+      "epoch": 1.5098500947676756,
+      "grad_norm": 0.6006361842155457,
+      "learning_rate": 0.0001,
+      "loss": 1.4622,
+      "step": 13144
+    },
+    {
+      "epoch": 1.5099649646775029,
+      "grad_norm": 0.7036729454994202,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 13145
+    },
+    {
+      "epoch": 1.5100798345873299,
+      "grad_norm": 0.6505088806152344,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 13146
+    },
+    {
+      "epoch": 1.5101947044971569,
+      "grad_norm": 0.6297341585159302,
+      "learning_rate": 0.0001,
+      "loss": 1.5059,
+      "step": 13147
+    },
+    {
+      "epoch": 1.510309574406984,
+      "grad_norm": 0.6522087454795837,
+      "learning_rate": 0.0001,
+      "loss": 1.4154,
+      "step": 13148
+    },
+    {
+      "epoch": 1.5104244443168113,
+      "grad_norm": 0.6192685961723328,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 13149
+    },
+    {
+      "epoch": 1.5105393142266383,
+      "grad_norm": 0.6250871419906616,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 13150
+    },
+    {
+      "epoch": 1.5106541841364654,
+      "grad_norm": 0.5791242122650146,
+      "learning_rate": 0.0001,
+      "loss": 1.4324,
+      "step": 13151
+    },
+    {
+      "epoch": 1.5107690540462926,
+      "grad_norm": 0.6033053994178772,
+      "learning_rate": 0.0001,
+      "loss": 1.441,
+      "step": 13152
+    },
+    {
+      "epoch": 1.5108839239561198,
+      "grad_norm": 0.5738871693611145,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 13153
+    },
+    {
+      "epoch": 1.5109987938659468,
+      "grad_norm": 0.6556652188301086,
+      "learning_rate": 0.0001,
+      "loss": 1.6179,
+      "step": 13154
+    },
+    {
+      "epoch": 1.5111136637757738,
+      "grad_norm": 0.6059442758560181,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 13155
+    },
+    {
+      "epoch": 1.511228533685601,
+      "grad_norm": 0.6075924038887024,
+      "learning_rate": 0.0001,
+      "loss": 1.3265,
+      "step": 13156
+    },
+    {
+      "epoch": 1.5113434035954283,
+      "grad_norm": 0.5933572053909302,
+      "learning_rate": 0.0001,
+      "loss": 1.5268,
+      "step": 13157
+    },
+    {
+      "epoch": 1.5114582735052553,
+      "grad_norm": 0.6054621934890747,
+      "learning_rate": 0.0001,
+      "loss": 1.66,
+      "step": 13158
+    },
+    {
+      "epoch": 1.5115731434150823,
+      "grad_norm": 0.5910248756408691,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 13159
+    },
+    {
+      "epoch": 1.5116880133249095,
+      "grad_norm": 0.6314256191253662,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 13160
+    },
+    {
+      "epoch": 1.5118028832347368,
+      "grad_norm": 0.5735428929328918,
+      "learning_rate": 0.0001,
+      "loss": 1.3867,
+      "step": 13161
+    },
+    {
+      "epoch": 1.5119177531445638,
+      "grad_norm": 0.5875412821769714,
+      "learning_rate": 0.0001,
+      "loss": 1.2807,
+      "step": 13162
+    },
+    {
+      "epoch": 1.5120326230543908,
+      "grad_norm": 0.5720763206481934,
+      "learning_rate": 0.0001,
+      "loss": 1.3435,
+      "step": 13163
+    },
+    {
+      "epoch": 1.512147492964218,
+      "grad_norm": 0.5954453349113464,
+      "learning_rate": 0.0001,
+      "loss": 1.3863,
+      "step": 13164
+    },
+    {
+      "epoch": 1.5122623628740453,
+      "grad_norm": 0.577720582485199,
+      "learning_rate": 0.0001,
+      "loss": 1.2367,
+      "step": 13165
+    },
+    {
+      "epoch": 1.5123772327838723,
+      "grad_norm": 0.5905912518501282,
+      "learning_rate": 0.0001,
+      "loss": 1.3191,
+      "step": 13166
+    },
+    {
+      "epoch": 1.5124921026936993,
+      "grad_norm": 0.5706838369369507,
+      "learning_rate": 0.0001,
+      "loss": 1.2935,
+      "step": 13167
+    },
+    {
+      "epoch": 1.5126069726035265,
+      "grad_norm": 0.6354959011077881,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 13168
+    },
+    {
+      "epoch": 1.5127218425133537,
+      "grad_norm": 0.6335857510566711,
+      "learning_rate": 0.0001,
+      "loss": 1.4592,
+      "step": 13169
+    },
+    {
+      "epoch": 1.5128367124231807,
+      "grad_norm": 0.6097238063812256,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 13170
+    },
+    {
+      "epoch": 1.5129515823330077,
+      "grad_norm": 0.5822547674179077,
+      "learning_rate": 0.0001,
+      "loss": 1.41,
+      "step": 13171
+    },
+    {
+      "epoch": 1.513066452242835,
+      "grad_norm": 0.6244901418685913,
+      "learning_rate": 0.0001,
+      "loss": 1.4444,
+      "step": 13172
+    },
+    {
+      "epoch": 1.5131813221526622,
+      "grad_norm": 0.5513818860054016,
+      "learning_rate": 0.0001,
+      "loss": 1.3421,
+      "step": 13173
+    },
+    {
+      "epoch": 1.5132961920624892,
+      "grad_norm": 0.582343578338623,
+      "learning_rate": 0.0001,
+      "loss": 1.317,
+      "step": 13174
+    },
+    {
+      "epoch": 1.5134110619723162,
+      "grad_norm": 0.6151304841041565,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 13175
+    },
+    {
+      "epoch": 1.5135259318821435,
+      "grad_norm": 0.5942151546478271,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 13176
+    },
+    {
+      "epoch": 1.5136408017919707,
+      "grad_norm": 0.589524507522583,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 13177
+    },
+    {
+      "epoch": 1.5137556717017977,
+      "grad_norm": 0.5877421498298645,
+      "learning_rate": 0.0001,
+      "loss": 1.4425,
+      "step": 13178
+    },
+    {
+      "epoch": 1.5138705416116247,
+      "grad_norm": 0.5954683423042297,
+      "learning_rate": 0.0001,
+      "loss": 1.5715,
+      "step": 13179
+    },
+    {
+      "epoch": 1.513985411521452,
+      "grad_norm": 0.5939478874206543,
+      "learning_rate": 0.0001,
+      "loss": 1.4957,
+      "step": 13180
+    },
+    {
+      "epoch": 1.5141002814312792,
+      "grad_norm": 0.6020224690437317,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 13181
+    },
+    {
+      "epoch": 1.5142151513411062,
+      "grad_norm": 0.6681519746780396,
+      "learning_rate": 0.0001,
+      "loss": 1.2384,
+      "step": 13182
+    },
+    {
+      "epoch": 1.5143300212509332,
+      "grad_norm": 0.622969925403595,
+      "learning_rate": 0.0001,
+      "loss": 1.423,
+      "step": 13183
+    },
+    {
+      "epoch": 1.5144448911607604,
+      "grad_norm": 0.5562835931777954,
+      "learning_rate": 0.0001,
+      "loss": 1.3621,
+      "step": 13184
+    },
+    {
+      "epoch": 1.5145597610705877,
+      "grad_norm": 0.6770687103271484,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 13185
+    },
+    {
+      "epoch": 1.5146746309804147,
+      "grad_norm": 0.6299651861190796,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 13186
+    },
+    {
+      "epoch": 1.5147895008902417,
+      "grad_norm": 0.566347599029541,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 13187
+    },
+    {
+      "epoch": 1.514904370800069,
+      "grad_norm": 0.5551129579544067,
+      "learning_rate": 0.0001,
+      "loss": 1.4055,
+      "step": 13188
+    },
+    {
+      "epoch": 1.5150192407098961,
+      "grad_norm": 0.5839925408363342,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 13189
+    },
+    {
+      "epoch": 1.5151341106197231,
+      "grad_norm": 0.5784244537353516,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 13190
+    },
+    {
+      "epoch": 1.5152489805295501,
+      "grad_norm": 0.5645365118980408,
+      "learning_rate": 0.0001,
+      "loss": 1.342,
+      "step": 13191
+    },
+    {
+      "epoch": 1.5153638504393774,
+      "grad_norm": 0.6406526565551758,
+      "learning_rate": 0.0001,
+      "loss": 1.3734,
+      "step": 13192
+    },
+    {
+      "epoch": 1.5154787203492046,
+      "grad_norm": 0.6612091064453125,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 13193
+    },
+    {
+      "epoch": 1.5155935902590316,
+      "grad_norm": 0.582217812538147,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 13194
+    },
+    {
+      "epoch": 1.5157084601688586,
+      "grad_norm": 0.6395711302757263,
+      "learning_rate": 0.0001,
+      "loss": 1.4942,
+      "step": 13195
+    },
+    {
+      "epoch": 1.5158233300786859,
+      "grad_norm": 0.6635745763778687,
+      "learning_rate": 0.0001,
+      "loss": 1.4659,
+      "step": 13196
+    },
+    {
+      "epoch": 1.515938199988513,
+      "grad_norm": 0.615152895450592,
+      "learning_rate": 0.0001,
+      "loss": 1.3869,
+      "step": 13197
+    },
+    {
+      "epoch": 1.51605306989834,
+      "grad_norm": 0.6119014620780945,
+      "learning_rate": 0.0001,
+      "loss": 1.5678,
+      "step": 13198
+    },
+    {
+      "epoch": 1.516167939808167,
+      "grad_norm": 0.6164930462837219,
+      "learning_rate": 0.0001,
+      "loss": 1.5306,
+      "step": 13199
+    },
+    {
+      "epoch": 1.5162828097179943,
+      "grad_norm": 0.5925434231758118,
+      "learning_rate": 0.0001,
+      "loss": 1.6357,
+      "step": 13200
+    },
+    {
+      "epoch": 1.5163976796278216,
+      "grad_norm": 0.6095196008682251,
+      "learning_rate": 0.0001,
+      "loss": 1.5921,
+      "step": 13201
+    },
+    {
+      "epoch": 1.5165125495376486,
+      "grad_norm": 0.6188748478889465,
+      "learning_rate": 0.0001,
+      "loss": 1.6986,
+      "step": 13202
+    },
+    {
+      "epoch": 1.5166274194474756,
+      "grad_norm": 0.5641734004020691,
+      "learning_rate": 0.0001,
+      "loss": 1.516,
+      "step": 13203
+    },
+    {
+      "epoch": 1.5167422893573028,
+      "grad_norm": 0.7535280585289001,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 13204
+    },
+    {
+      "epoch": 1.51685715926713,
+      "grad_norm": 0.6287049055099487,
+      "learning_rate": 0.0001,
+      "loss": 1.53,
+      "step": 13205
+    },
+    {
+      "epoch": 1.516972029176957,
+      "grad_norm": 0.6373167037963867,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 13206
+    },
+    {
+      "epoch": 1.517086899086784,
+      "grad_norm": 0.5664235949516296,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 13207
+    },
+    {
+      "epoch": 1.5172017689966113,
+      "grad_norm": 0.5645571351051331,
+      "learning_rate": 0.0001,
+      "loss": 1.2935,
+      "step": 13208
+    },
+    {
+      "epoch": 1.5173166389064385,
+      "grad_norm": 0.6005182862281799,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 13209
+    },
+    {
+      "epoch": 1.5174315088162655,
+      "grad_norm": 0.595731794834137,
+      "learning_rate": 0.0001,
+      "loss": 1.342,
+      "step": 13210
+    },
+    {
+      "epoch": 1.5175463787260925,
+      "grad_norm": 0.6699162721633911,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 13211
+    },
+    {
+      "epoch": 1.5176612486359198,
+      "grad_norm": 0.5972051620483398,
+      "learning_rate": 0.0001,
+      "loss": 1.438,
+      "step": 13212
+    },
+    {
+      "epoch": 1.517776118545747,
+      "grad_norm": 0.6613103747367859,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 13213
+    },
+    {
+      "epoch": 1.517890988455574,
+      "grad_norm": 0.5806509852409363,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 13214
+    },
+    {
+      "epoch": 1.518005858365401,
+      "grad_norm": 0.6001474857330322,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 13215
+    },
+    {
+      "epoch": 1.5181207282752283,
+      "grad_norm": 0.6385689377784729,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 13216
+    },
+    {
+      "epoch": 1.5182355981850555,
+      "grad_norm": 0.6373685002326965,
+      "learning_rate": 0.0001,
+      "loss": 1.478,
+      "step": 13217
+    },
+    {
+      "epoch": 1.5183504680948825,
+      "grad_norm": 0.5941466689109802,
+      "learning_rate": 0.0001,
+      "loss": 1.384,
+      "step": 13218
+    },
+    {
+      "epoch": 1.5184653380047095,
+      "grad_norm": 0.5440075993537903,
+      "learning_rate": 0.0001,
+      "loss": 1.388,
+      "step": 13219
+    },
+    {
+      "epoch": 1.5185802079145367,
+      "grad_norm": 0.5798068046569824,
+      "learning_rate": 0.0001,
+      "loss": 1.4018,
+      "step": 13220
+    },
+    {
+      "epoch": 1.518695077824364,
+      "grad_norm": 0.6054274439811707,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 13221
+    },
+    {
+      "epoch": 1.518809947734191,
+      "grad_norm": 0.623768150806427,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 13222
+    },
+    {
+      "epoch": 1.518924817644018,
+      "grad_norm": 0.5744849443435669,
+      "learning_rate": 0.0001,
+      "loss": 1.3231,
+      "step": 13223
+    },
+    {
+      "epoch": 1.5190396875538452,
+      "grad_norm": 0.6288904547691345,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 13224
+    },
+    {
+      "epoch": 1.5191545574636725,
+      "grad_norm": 0.58237624168396,
+      "learning_rate": 0.0001,
+      "loss": 1.2721,
+      "step": 13225
+    },
+    {
+      "epoch": 1.5192694273734995,
+      "grad_norm": 0.5864928364753723,
+      "learning_rate": 0.0001,
+      "loss": 1.4331,
+      "step": 13226
+    },
+    {
+      "epoch": 1.5193842972833265,
+      "grad_norm": 0.5528786778450012,
+      "learning_rate": 0.0001,
+      "loss": 1.4004,
+      "step": 13227
+    },
+    {
+      "epoch": 1.5194991671931537,
+      "grad_norm": 0.639892578125,
+      "learning_rate": 0.0001,
+      "loss": 1.3726,
+      "step": 13228
+    },
+    {
+      "epoch": 1.519614037102981,
+      "grad_norm": 0.5863984227180481,
+      "learning_rate": 0.0001,
+      "loss": 1.4963,
+      "step": 13229
+    },
+    {
+      "epoch": 1.5197289070128082,
+      "grad_norm": 0.6669512391090393,
+      "learning_rate": 0.0001,
+      "loss": 1.5823,
+      "step": 13230
+    },
+    {
+      "epoch": 1.5198437769226352,
+      "grad_norm": 0.628280520439148,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 13231
+    },
+    {
+      "epoch": 1.5199586468324622,
+      "grad_norm": 0.6582902669906616,
+      "learning_rate": 0.0001,
+      "loss": 1.4258,
+      "step": 13232
+    },
+    {
+      "epoch": 1.5200735167422894,
+      "grad_norm": 0.5654552578926086,
+      "learning_rate": 0.0001,
+      "loss": 1.3741,
+      "step": 13233
+    },
+    {
+      "epoch": 1.5201883866521166,
+      "grad_norm": 0.6104324460029602,
+      "learning_rate": 0.0001,
+      "loss": 1.6584,
+      "step": 13234
+    },
+    {
+      "epoch": 1.5203032565619437,
+      "grad_norm": 0.6294476389884949,
+      "learning_rate": 0.0001,
+      "loss": 1.5948,
+      "step": 13235
+    },
+    {
+      "epoch": 1.5204181264717707,
+      "grad_norm": 0.6104316115379333,
+      "learning_rate": 0.0001,
+      "loss": 1.3727,
+      "step": 13236
+    },
+    {
+      "epoch": 1.520532996381598,
+      "grad_norm": 0.6369021534919739,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 13237
+    },
+    {
+      "epoch": 1.5206478662914251,
+      "grad_norm": 0.5884815454483032,
+      "learning_rate": 0.0001,
+      "loss": 1.5229,
+      "step": 13238
+    },
+    {
+      "epoch": 1.5207627362012521,
+      "grad_norm": 0.6035114526748657,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 13239
+    },
+    {
+      "epoch": 1.5208776061110791,
+      "grad_norm": 0.5970353484153748,
+      "learning_rate": 0.0001,
+      "loss": 1.4528,
+      "step": 13240
+    },
+    {
+      "epoch": 1.5209924760209064,
+      "grad_norm": 0.6134264469146729,
+      "learning_rate": 0.0001,
+      "loss": 1.5959,
+      "step": 13241
+    },
+    {
+      "epoch": 1.5211073459307336,
+      "grad_norm": 0.6086723208427429,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 13242
+    },
+    {
+      "epoch": 1.5212222158405606,
+      "grad_norm": 0.6190268397331238,
+      "learning_rate": 0.0001,
+      "loss": 1.4347,
+      "step": 13243
+    },
+    {
+      "epoch": 1.5213370857503876,
+      "grad_norm": 0.5801183581352234,
+      "learning_rate": 0.0001,
+      "loss": 1.4204,
+      "step": 13244
+    },
+    {
+      "epoch": 1.5214519556602148,
+      "grad_norm": 0.5259165167808533,
+      "learning_rate": 0.0001,
+      "loss": 1.2951,
+      "step": 13245
+    },
+    {
+      "epoch": 1.521566825570042,
+      "grad_norm": 0.6810782551765442,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 13246
+    },
+    {
+      "epoch": 1.521681695479869,
+      "grad_norm": 0.6331126093864441,
+      "learning_rate": 0.0001,
+      "loss": 1.4176,
+      "step": 13247
+    },
+    {
+      "epoch": 1.521796565389696,
+      "grad_norm": 0.561797022819519,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 13248
+    },
+    {
+      "epoch": 1.5219114352995233,
+      "grad_norm": 0.6428273320198059,
+      "learning_rate": 0.0001,
+      "loss": 1.2756,
+      "step": 13249
+    },
+    {
+      "epoch": 1.5220263052093506,
+      "grad_norm": 0.5670008659362793,
+      "learning_rate": 0.0001,
+      "loss": 1.4275,
+      "step": 13250
+    },
+    {
+      "epoch": 1.5221411751191776,
+      "grad_norm": 0.6418376564979553,
+      "learning_rate": 0.0001,
+      "loss": 1.5854,
+      "step": 13251
+    },
+    {
+      "epoch": 1.5222560450290046,
+      "grad_norm": 0.6684396266937256,
+      "learning_rate": 0.0001,
+      "loss": 1.4903,
+      "step": 13252
+    },
+    {
+      "epoch": 1.5223709149388318,
+      "grad_norm": 0.5767911672592163,
+      "learning_rate": 0.0001,
+      "loss": 1.2792,
+      "step": 13253
+    },
+    {
+      "epoch": 1.522485784848659,
+      "grad_norm": 0.620569109916687,
+      "learning_rate": 0.0001,
+      "loss": 1.4702,
+      "step": 13254
+    },
+    {
+      "epoch": 1.522600654758486,
+      "grad_norm": 0.6131866574287415,
+      "learning_rate": 0.0001,
+      "loss": 1.4637,
+      "step": 13255
+    },
+    {
+      "epoch": 1.522715524668313,
+      "grad_norm": 0.6050616502761841,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 13256
+    },
+    {
+      "epoch": 1.5228303945781403,
+      "grad_norm": 0.6381790637969971,
+      "learning_rate": 0.0001,
+      "loss": 1.5207,
+      "step": 13257
+    },
+    {
+      "epoch": 1.5229452644879675,
+      "grad_norm": 0.6472856998443604,
+      "learning_rate": 0.0001,
+      "loss": 1.3212,
+      "step": 13258
+    },
+    {
+      "epoch": 1.5230601343977945,
+      "grad_norm": 0.5918861031532288,
+      "learning_rate": 0.0001,
+      "loss": 1.4754,
+      "step": 13259
+    },
+    {
+      "epoch": 1.5231750043076215,
+      "grad_norm": 0.6339123249053955,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 13260
+    },
+    {
+      "epoch": 1.5232898742174488,
+      "grad_norm": 0.5977100133895874,
+      "learning_rate": 0.0001,
+      "loss": 1.3237,
+      "step": 13261
+    },
+    {
+      "epoch": 1.523404744127276,
+      "grad_norm": 0.6018535494804382,
+      "learning_rate": 0.0001,
+      "loss": 1.354,
+      "step": 13262
+    },
+    {
+      "epoch": 1.523519614037103,
+      "grad_norm": 0.5871580243110657,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 13263
+    },
+    {
+      "epoch": 1.52363448394693,
+      "grad_norm": 0.6061954498291016,
+      "learning_rate": 0.0001,
+      "loss": 1.3217,
+      "step": 13264
+    },
+    {
+      "epoch": 1.5237493538567572,
+      "grad_norm": 0.637395977973938,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 13265
+    },
+    {
+      "epoch": 1.5238642237665845,
+      "grad_norm": 0.7003656029701233,
+      "learning_rate": 0.0001,
+      "loss": 1.7488,
+      "step": 13266
+    },
+    {
+      "epoch": 1.5239790936764115,
+      "grad_norm": 0.6120108366012573,
+      "learning_rate": 0.0001,
+      "loss": 1.4945,
+      "step": 13267
+    },
+    {
+      "epoch": 1.5240939635862385,
+      "grad_norm": 0.5899313688278198,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 13268
+    },
+    {
+      "epoch": 1.5242088334960657,
+      "grad_norm": 0.6798549890518188,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 13269
+    },
+    {
+      "epoch": 1.524323703405893,
+      "grad_norm": 0.6134473085403442,
+      "learning_rate": 0.0001,
+      "loss": 1.4846,
+      "step": 13270
+    },
+    {
+      "epoch": 1.52443857331572,
+      "grad_norm": 0.5412236452102661,
+      "learning_rate": 0.0001,
+      "loss": 1.5211,
+      "step": 13271
+    },
+    {
+      "epoch": 1.524553443225547,
+      "grad_norm": 0.5635609030723572,
+      "learning_rate": 0.0001,
+      "loss": 1.3889,
+      "step": 13272
+    },
+    {
+      "epoch": 1.5246683131353742,
+      "grad_norm": 0.5932890772819519,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 13273
+    },
+    {
+      "epoch": 1.5247831830452014,
+      "grad_norm": 0.5923974514007568,
+      "learning_rate": 0.0001,
+      "loss": 1.6642,
+      "step": 13274
+    },
+    {
+      "epoch": 1.5248980529550284,
+      "grad_norm": 0.6257017254829407,
+      "learning_rate": 0.0001,
+      "loss": 1.599,
+      "step": 13275
+    },
+    {
+      "epoch": 1.5250129228648555,
+      "grad_norm": 0.5884189605712891,
+      "learning_rate": 0.0001,
+      "loss": 1.5512,
+      "step": 13276
+    },
+    {
+      "epoch": 1.5251277927746827,
+      "grad_norm": 0.5920361876487732,
+      "learning_rate": 0.0001,
+      "loss": 1.5368,
+      "step": 13277
+    },
+    {
+      "epoch": 1.52524266268451,
+      "grad_norm": 0.6029490232467651,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 13278
+    },
+    {
+      "epoch": 1.525357532594337,
+      "grad_norm": 0.5848401784896851,
+      "learning_rate": 0.0001,
+      "loss": 1.4009,
+      "step": 13279
+    },
+    {
+      "epoch": 1.525472402504164,
+      "grad_norm": 0.6184453964233398,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 13280
+    },
+    {
+      "epoch": 1.5255872724139912,
+      "grad_norm": 0.7568985223770142,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 13281
+    },
+    {
+      "epoch": 1.5257021423238184,
+      "grad_norm": 0.5871482491493225,
+      "learning_rate": 0.0001,
+      "loss": 1.3935,
+      "step": 13282
+    },
+    {
+      "epoch": 1.5258170122336454,
+      "grad_norm": 0.6043814420700073,
+      "learning_rate": 0.0001,
+      "loss": 1.2961,
+      "step": 13283
+    },
+    {
+      "epoch": 1.5259318821434724,
+      "grad_norm": 0.5470103621482849,
+      "learning_rate": 0.0001,
+      "loss": 1.3294,
+      "step": 13284
+    },
+    {
+      "epoch": 1.5260467520532996,
+      "grad_norm": 0.5336215496063232,
+      "learning_rate": 0.0001,
+      "loss": 1.1915,
+      "step": 13285
+    },
+    {
+      "epoch": 1.5261616219631269,
+      "grad_norm": 0.6141431331634521,
+      "learning_rate": 0.0001,
+      "loss": 1.469,
+      "step": 13286
+    },
+    {
+      "epoch": 1.5262764918729539,
+      "grad_norm": 0.5851207375526428,
+      "learning_rate": 0.0001,
+      "loss": 1.424,
+      "step": 13287
+    },
+    {
+      "epoch": 1.526391361782781,
+      "grad_norm": 0.5804892182350159,
+      "learning_rate": 0.0001,
+      "loss": 1.3296,
+      "step": 13288
+    },
+    {
+      "epoch": 1.5265062316926081,
+      "grad_norm": 0.6511141061782837,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 13289
+    },
+    {
+      "epoch": 1.5266211016024354,
+      "grad_norm": 0.6962987184524536,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 13290
+    },
+    {
+      "epoch": 1.5267359715122624,
+      "grad_norm": 0.5867136120796204,
+      "learning_rate": 0.0001,
+      "loss": 1.2938,
+      "step": 13291
+    },
+    {
+      "epoch": 1.5268508414220894,
+      "grad_norm": 0.6161892414093018,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 13292
+    },
+    {
+      "epoch": 1.5269657113319166,
+      "grad_norm": 0.5810278058052063,
+      "learning_rate": 0.0001,
+      "loss": 1.197,
+      "step": 13293
+    },
+    {
+      "epoch": 1.5270805812417438,
+      "grad_norm": 0.6159192323684692,
+      "learning_rate": 0.0001,
+      "loss": 1.3235,
+      "step": 13294
+    },
+    {
+      "epoch": 1.5271954511515708,
+      "grad_norm": 0.6228445768356323,
+      "learning_rate": 0.0001,
+      "loss": 1.3081,
+      "step": 13295
+    },
+    {
+      "epoch": 1.5273103210613979,
+      "grad_norm": 0.6097662448883057,
+      "learning_rate": 0.0001,
+      "loss": 1.468,
+      "step": 13296
+    },
+    {
+      "epoch": 1.527425190971225,
+      "grad_norm": 0.6655218005180359,
+      "learning_rate": 0.0001,
+      "loss": 1.4785,
+      "step": 13297
+    },
+    {
+      "epoch": 1.5275400608810523,
+      "grad_norm": 0.5990704298019409,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 13298
+    },
+    {
+      "epoch": 1.5276549307908793,
+      "grad_norm": 0.6513421535491943,
+      "learning_rate": 0.0001,
+      "loss": 1.6009,
+      "step": 13299
+    },
+    {
+      "epoch": 1.5277698007007063,
+      "grad_norm": 0.6456665992736816,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 13300
+    },
+    {
+      "epoch": 1.5278846706105336,
+      "grad_norm": 0.6521137952804565,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 13301
+    },
+    {
+      "epoch": 1.5279995405203608,
+      "grad_norm": 0.5967424511909485,
+      "learning_rate": 0.0001,
+      "loss": 1.4281,
+      "step": 13302
+    },
+    {
+      "epoch": 1.5281144104301878,
+      "grad_norm": 0.6816303730010986,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 13303
+    },
+    {
+      "epoch": 1.5282292803400148,
+      "grad_norm": 0.5861914753913879,
+      "learning_rate": 0.0001,
+      "loss": 1.4389,
+      "step": 13304
+    },
+    {
+      "epoch": 1.528344150249842,
+      "grad_norm": 0.6135704517364502,
+      "learning_rate": 0.0001,
+      "loss": 1.3657,
+      "step": 13305
+    },
+    {
+      "epoch": 1.5284590201596693,
+      "grad_norm": 0.6036303043365479,
+      "learning_rate": 0.0001,
+      "loss": 1.4097,
+      "step": 13306
+    },
+    {
+      "epoch": 1.5285738900694963,
+      "grad_norm": 0.5533670783042908,
+      "learning_rate": 0.0001,
+      "loss": 1.3476,
+      "step": 13307
+    },
+    {
+      "epoch": 1.5286887599793233,
+      "grad_norm": 0.5623646974563599,
+      "learning_rate": 0.0001,
+      "loss": 1.5436,
+      "step": 13308
+    },
+    {
+      "epoch": 1.5288036298891505,
+      "grad_norm": 0.5962461829185486,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 13309
+    },
+    {
+      "epoch": 1.5289184997989778,
+      "grad_norm": 0.626592218875885,
+      "learning_rate": 0.0001,
+      "loss": 1.4646,
+      "step": 13310
+    },
+    {
+      "epoch": 1.5290333697088048,
+      "grad_norm": 0.6010507941246033,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 13311
+    },
+    {
+      "epoch": 1.5291482396186318,
+      "grad_norm": 0.5912574529647827,
+      "learning_rate": 0.0001,
+      "loss": 1.4557,
+      "step": 13312
+    },
+    {
+      "epoch": 1.529263109528459,
+      "grad_norm": 0.6870297789573669,
+      "learning_rate": 0.0001,
+      "loss": 1.7032,
+      "step": 13313
+    },
+    {
+      "epoch": 1.5293779794382862,
+      "grad_norm": 0.5987645983695984,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 13314
+    },
+    {
+      "epoch": 1.5294928493481132,
+      "grad_norm": 0.6424558162689209,
+      "learning_rate": 0.0001,
+      "loss": 1.2024,
+      "step": 13315
+    },
+    {
+      "epoch": 1.5296077192579403,
+      "grad_norm": 0.6005625128746033,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 13316
+    },
+    {
+      "epoch": 1.5297225891677675,
+      "grad_norm": 0.6098487973213196,
+      "learning_rate": 0.0001,
+      "loss": 1.3273,
+      "step": 13317
+    },
+    {
+      "epoch": 1.5298374590775947,
+      "grad_norm": 0.625301718711853,
+      "learning_rate": 0.0001,
+      "loss": 1.3964,
+      "step": 13318
+    },
+    {
+      "epoch": 1.5299523289874217,
+      "grad_norm": 0.6038113236427307,
+      "learning_rate": 0.0001,
+      "loss": 1.4294,
+      "step": 13319
+    },
+    {
+      "epoch": 1.5300671988972487,
+      "grad_norm": 0.5868397951126099,
+      "learning_rate": 0.0001,
+      "loss": 1.4582,
+      "step": 13320
+    },
+    {
+      "epoch": 1.530182068807076,
+      "grad_norm": 0.6464180946350098,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 13321
+    },
+    {
+      "epoch": 1.5302969387169032,
+      "grad_norm": 0.604279637336731,
+      "learning_rate": 0.0001,
+      "loss": 1.5115,
+      "step": 13322
+    },
+    {
+      "epoch": 1.5304118086267302,
+      "grad_norm": 0.5744287371635437,
+      "learning_rate": 0.0001,
+      "loss": 1.3912,
+      "step": 13323
+    },
+    {
+      "epoch": 1.5305266785365572,
+      "grad_norm": 0.6124972105026245,
+      "learning_rate": 0.0001,
+      "loss": 1.432,
+      "step": 13324
+    },
+    {
+      "epoch": 1.5306415484463844,
+      "grad_norm": 0.5574883818626404,
+      "learning_rate": 0.0001,
+      "loss": 1.3208,
+      "step": 13325
+    },
+    {
+      "epoch": 1.5307564183562117,
+      "grad_norm": 0.6026033759117126,
+      "learning_rate": 0.0001,
+      "loss": 1.3873,
+      "step": 13326
+    },
+    {
+      "epoch": 1.5308712882660387,
+      "grad_norm": 0.6555309295654297,
+      "learning_rate": 0.0001,
+      "loss": 1.4002,
+      "step": 13327
+    },
+    {
+      "epoch": 1.5309861581758657,
+      "grad_norm": 0.5827085375785828,
+      "learning_rate": 0.0001,
+      "loss": 1.3754,
+      "step": 13328
+    },
+    {
+      "epoch": 1.531101028085693,
+      "grad_norm": 0.6004881262779236,
+      "learning_rate": 0.0001,
+      "loss": 1.2212,
+      "step": 13329
+    },
+    {
+      "epoch": 1.5312158979955202,
+      "grad_norm": 0.7817880511283875,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 13330
+    },
+    {
+      "epoch": 1.5313307679053472,
+      "grad_norm": 0.6626024842262268,
+      "learning_rate": 0.0001,
+      "loss": 1.4952,
+      "step": 13331
+    },
+    {
+      "epoch": 1.5314456378151742,
+      "grad_norm": 0.6087819337844849,
+      "learning_rate": 0.0001,
+      "loss": 1.4887,
+      "step": 13332
+    },
+    {
+      "epoch": 1.5315605077250014,
+      "grad_norm": 0.7612007260322571,
+      "learning_rate": 0.0001,
+      "loss": 1.4343,
+      "step": 13333
+    },
+    {
+      "epoch": 1.5316753776348286,
+      "grad_norm": 0.5998708009719849,
+      "learning_rate": 0.0001,
+      "loss": 1.3433,
+      "step": 13334
+    },
+    {
+      "epoch": 1.5317902475446556,
+      "grad_norm": 0.5892654061317444,
+      "learning_rate": 0.0001,
+      "loss": 1.5178,
+      "step": 13335
+    },
+    {
+      "epoch": 1.5319051174544827,
+      "grad_norm": 0.6381953954696655,
+      "learning_rate": 0.0001,
+      "loss": 1.5053,
+      "step": 13336
+    },
+    {
+      "epoch": 1.5320199873643099,
+      "grad_norm": 0.6319192051887512,
+      "learning_rate": 0.0001,
+      "loss": 1.2256,
+      "step": 13337
+    },
+    {
+      "epoch": 1.5321348572741371,
+      "grad_norm": 0.6061521172523499,
+      "learning_rate": 0.0001,
+      "loss": 1.3842,
+      "step": 13338
+    },
+    {
+      "epoch": 1.5322497271839641,
+      "grad_norm": 0.6164060235023499,
+      "learning_rate": 0.0001,
+      "loss": 1.5006,
+      "step": 13339
+    },
+    {
+      "epoch": 1.5323645970937911,
+      "grad_norm": 0.5808977484703064,
+      "learning_rate": 0.0001,
+      "loss": 1.3016,
+      "step": 13340
+    },
+    {
+      "epoch": 1.5324794670036184,
+      "grad_norm": 0.6272817254066467,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 13341
+    },
+    {
+      "epoch": 1.5325943369134456,
+      "grad_norm": 0.5705950260162354,
+      "learning_rate": 0.0001,
+      "loss": 1.4289,
+      "step": 13342
+    },
+    {
+      "epoch": 1.5327092068232726,
+      "grad_norm": 0.5681813359260559,
+      "learning_rate": 0.0001,
+      "loss": 1.3662,
+      "step": 13343
+    },
+    {
+      "epoch": 1.5328240767330996,
+      "grad_norm": 0.591742753982544,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 13344
+    },
+    {
+      "epoch": 1.5329389466429268,
+      "grad_norm": 0.6885697841644287,
+      "learning_rate": 0.0001,
+      "loss": 1.4915,
+      "step": 13345
+    },
+    {
+      "epoch": 1.533053816552754,
+      "grad_norm": 0.6182805895805359,
+      "learning_rate": 0.0001,
+      "loss": 1.3183,
+      "step": 13346
+    },
+    {
+      "epoch": 1.533168686462581,
+      "grad_norm": 0.630598247051239,
+      "learning_rate": 0.0001,
+      "loss": 1.399,
+      "step": 13347
+    },
+    {
+      "epoch": 1.533283556372408,
+      "grad_norm": 0.5968937277793884,
+      "learning_rate": 0.0001,
+      "loss": 1.3648,
+      "step": 13348
+    },
+    {
+      "epoch": 1.5333984262822353,
+      "grad_norm": 0.6997851133346558,
+      "learning_rate": 0.0001,
+      "loss": 1.5873,
+      "step": 13349
+    },
+    {
+      "epoch": 1.5335132961920626,
+      "grad_norm": 0.615630030632019,
+      "learning_rate": 0.0001,
+      "loss": 1.3916,
+      "step": 13350
+    },
+    {
+      "epoch": 1.5336281661018896,
+      "grad_norm": 0.6529400944709778,
+      "learning_rate": 0.0001,
+      "loss": 1.4654,
+      "step": 13351
+    },
+    {
+      "epoch": 1.5337430360117166,
+      "grad_norm": 0.6269962191581726,
+      "learning_rate": 0.0001,
+      "loss": 1.3357,
+      "step": 13352
+    },
+    {
+      "epoch": 1.5338579059215438,
+      "grad_norm": 0.613075852394104,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 13353
+    },
+    {
+      "epoch": 1.533972775831371,
+      "grad_norm": 0.6532135605812073,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 13354
+    },
+    {
+      "epoch": 1.534087645741198,
+      "grad_norm": 0.6102808117866516,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 13355
+    },
+    {
+      "epoch": 1.534202515651025,
+      "grad_norm": 0.6019791960716248,
+      "learning_rate": 0.0001,
+      "loss": 1.3113,
+      "step": 13356
+    },
+    {
+      "epoch": 1.5343173855608523,
+      "grad_norm": 0.6130610108375549,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 13357
+    },
+    {
+      "epoch": 1.5344322554706795,
+      "grad_norm": 0.5954639911651611,
+      "learning_rate": 0.0001,
+      "loss": 1.5233,
+      "step": 13358
+    },
+    {
+      "epoch": 1.5345471253805065,
+      "grad_norm": 0.6210713982582092,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 13359
+    },
+    {
+      "epoch": 1.5346619952903335,
+      "grad_norm": 0.5767752528190613,
+      "learning_rate": 0.0001,
+      "loss": 1.2935,
+      "step": 13360
+    },
+    {
+      "epoch": 1.5347768652001608,
+      "grad_norm": 0.5763528943061829,
+      "learning_rate": 0.0001,
+      "loss": 1.3317,
+      "step": 13361
+    },
+    {
+      "epoch": 1.534891735109988,
+      "grad_norm": 0.6128670573234558,
+      "learning_rate": 0.0001,
+      "loss": 1.4817,
+      "step": 13362
+    },
+    {
+      "epoch": 1.535006605019815,
+      "grad_norm": 0.6389924883842468,
+      "learning_rate": 0.0001,
+      "loss": 1.4359,
+      "step": 13363
+    },
+    {
+      "epoch": 1.535121474929642,
+      "grad_norm": 0.6564880013465881,
+      "learning_rate": 0.0001,
+      "loss": 1.4894,
+      "step": 13364
+    },
+    {
+      "epoch": 1.5352363448394692,
+      "grad_norm": 0.6105384826660156,
+      "learning_rate": 0.0001,
+      "loss": 1.3985,
+      "step": 13365
+    },
+    {
+      "epoch": 1.5353512147492965,
+      "grad_norm": 0.7015275359153748,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 13366
+    },
+    {
+      "epoch": 1.5354660846591237,
+      "grad_norm": 0.6538365483283997,
+      "learning_rate": 0.0001,
+      "loss": 1.5722,
+      "step": 13367
+    },
+    {
+      "epoch": 1.5355809545689507,
+      "grad_norm": 0.5591805577278137,
+      "learning_rate": 0.0001,
+      "loss": 1.306,
+      "step": 13368
+    },
+    {
+      "epoch": 1.5356958244787777,
+      "grad_norm": 0.5733972787857056,
+      "learning_rate": 0.0001,
+      "loss": 1.2758,
+      "step": 13369
+    },
+    {
+      "epoch": 1.535810694388605,
+      "grad_norm": 0.6370871067047119,
+      "learning_rate": 0.0001,
+      "loss": 1.3703,
+      "step": 13370
+    },
+    {
+      "epoch": 1.5359255642984322,
+      "grad_norm": 0.5944746136665344,
+      "learning_rate": 0.0001,
+      "loss": 1.461,
+      "step": 13371
+    },
+    {
+      "epoch": 1.5360404342082592,
+      "grad_norm": 0.5778318047523499,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 13372
+    },
+    {
+      "epoch": 1.5361553041180862,
+      "grad_norm": 0.6232753992080688,
+      "learning_rate": 0.0001,
+      "loss": 1.5324,
+      "step": 13373
+    },
+    {
+      "epoch": 1.5362701740279134,
+      "grad_norm": 0.6197500228881836,
+      "learning_rate": 0.0001,
+      "loss": 1.437,
+      "step": 13374
+    },
+    {
+      "epoch": 1.5363850439377407,
+      "grad_norm": 0.6072701811790466,
+      "learning_rate": 0.0001,
+      "loss": 1.4755,
+      "step": 13375
+    },
+    {
+      "epoch": 1.5364999138475677,
+      "grad_norm": 0.602379560470581,
+      "learning_rate": 0.0001,
+      "loss": 1.4888,
+      "step": 13376
+    },
+    {
+      "epoch": 1.5366147837573947,
+      "grad_norm": 0.5745888352394104,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 13377
+    },
+    {
+      "epoch": 1.536729653667222,
+      "grad_norm": 0.615543782711029,
+      "learning_rate": 0.0001,
+      "loss": 1.2821,
+      "step": 13378
+    },
+    {
+      "epoch": 1.5368445235770491,
+      "grad_norm": 0.5586651563644409,
+      "learning_rate": 0.0001,
+      "loss": 1.29,
+      "step": 13379
+    },
+    {
+      "epoch": 1.5369593934868762,
+      "grad_norm": 0.5711967945098877,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 13380
+    },
+    {
+      "epoch": 1.5370742633967032,
+      "grad_norm": 0.5827999711036682,
+      "learning_rate": 0.0001,
+      "loss": 1.2618,
+      "step": 13381
+    },
+    {
+      "epoch": 1.5371891333065304,
+      "grad_norm": 0.5403852462768555,
+      "learning_rate": 0.0001,
+      "loss": 1.2958,
+      "step": 13382
+    },
+    {
+      "epoch": 1.5373040032163576,
+      "grad_norm": 0.6437862515449524,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 13383
+    },
+    {
+      "epoch": 1.5374188731261846,
+      "grad_norm": 0.5628765225410461,
+      "learning_rate": 0.0001,
+      "loss": 1.4437,
+      "step": 13384
+    },
+    {
+      "epoch": 1.5375337430360116,
+      "grad_norm": 0.6011320948600769,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 13385
+    },
+    {
+      "epoch": 1.5376486129458389,
+      "grad_norm": 0.5917825102806091,
+      "learning_rate": 0.0001,
+      "loss": 1.4444,
+      "step": 13386
+    },
+    {
+      "epoch": 1.537763482855666,
+      "grad_norm": 0.5793378353118896,
+      "learning_rate": 0.0001,
+      "loss": 1.3797,
+      "step": 13387
+    },
+    {
+      "epoch": 1.5378783527654931,
+      "grad_norm": 0.6069167852401733,
+      "learning_rate": 0.0001,
+      "loss": 1.4303,
+      "step": 13388
+    },
+    {
+      "epoch": 1.5379932226753201,
+      "grad_norm": 0.6057384014129639,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 13389
+    },
+    {
+      "epoch": 1.5381080925851474,
+      "grad_norm": 0.5952368974685669,
+      "learning_rate": 0.0001,
+      "loss": 1.4508,
+      "step": 13390
+    },
+    {
+      "epoch": 1.5382229624949746,
+      "grad_norm": 0.58933025598526,
+      "learning_rate": 0.0001,
+      "loss": 1.4134,
+      "step": 13391
+    },
+    {
+      "epoch": 1.5383378324048016,
+      "grad_norm": 0.5574020743370056,
+      "learning_rate": 0.0001,
+      "loss": 1.2945,
+      "step": 13392
+    },
+    {
+      "epoch": 1.5384527023146286,
+      "grad_norm": 0.5393736958503723,
+      "learning_rate": 0.0001,
+      "loss": 1.3008,
+      "step": 13393
+    },
+    {
+      "epoch": 1.5385675722244558,
+      "grad_norm": 0.5925777554512024,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 13394
+    },
+    {
+      "epoch": 1.538682442134283,
+      "grad_norm": 0.6799826622009277,
+      "learning_rate": 0.0001,
+      "loss": 1.3157,
+      "step": 13395
+    },
+    {
+      "epoch": 1.53879731204411,
+      "grad_norm": 0.6065730452537537,
+      "learning_rate": 0.0001,
+      "loss": 1.3707,
+      "step": 13396
+    },
+    {
+      "epoch": 1.538912181953937,
+      "grad_norm": 0.6125333905220032,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 13397
+    },
+    {
+      "epoch": 1.5390270518637643,
+      "grad_norm": 0.5929028391838074,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 13398
+    },
+    {
+      "epoch": 1.5391419217735915,
+      "grad_norm": 0.6272832751274109,
+      "learning_rate": 0.0001,
+      "loss": 1.3904,
+      "step": 13399
+    },
+    {
+      "epoch": 1.5392567916834186,
+      "grad_norm": 0.5839588642120361,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 13400
+    },
+    {
+      "epoch": 1.5393716615932456,
+      "grad_norm": 0.6149510145187378,
+      "learning_rate": 0.0001,
+      "loss": 1.4483,
+      "step": 13401
+    },
+    {
+      "epoch": 1.5394865315030728,
+      "grad_norm": 0.5844335556030273,
+      "learning_rate": 0.0001,
+      "loss": 1.4239,
+      "step": 13402
+    },
+    {
+      "epoch": 1.5396014014129,
+      "grad_norm": 0.5901253819465637,
+      "learning_rate": 0.0001,
+      "loss": 1.3002,
+      "step": 13403
+    },
+    {
+      "epoch": 1.539716271322727,
+      "grad_norm": 0.6704103946685791,
+      "learning_rate": 0.0001,
+      "loss": 1.5101,
+      "step": 13404
+    },
+    {
+      "epoch": 1.539831141232554,
+      "grad_norm": 0.5861269235610962,
+      "learning_rate": 0.0001,
+      "loss": 1.5826,
+      "step": 13405
+    },
+    {
+      "epoch": 1.5399460111423813,
+      "grad_norm": 0.6078442335128784,
+      "learning_rate": 0.0001,
+      "loss": 1.5645,
+      "step": 13406
+    },
+    {
+      "epoch": 1.5400608810522085,
+      "grad_norm": 0.5604826211929321,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 13407
+    },
+    {
+      "epoch": 1.5401757509620355,
+      "grad_norm": 0.5622230172157288,
+      "learning_rate": 0.0001,
+      "loss": 1.5088,
+      "step": 13408
+    },
+    {
+      "epoch": 1.5402906208718625,
+      "grad_norm": 0.6920380592346191,
+      "learning_rate": 0.0001,
+      "loss": 1.3965,
+      "step": 13409
+    },
+    {
+      "epoch": 1.5404054907816898,
+      "grad_norm": 0.5716781616210938,
+      "learning_rate": 0.0001,
+      "loss": 1.6094,
+      "step": 13410
+    },
+    {
+      "epoch": 1.540520360691517,
+      "grad_norm": 0.5778867602348328,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 13411
+    },
+    {
+      "epoch": 1.540635230601344,
+      "grad_norm": 0.564794659614563,
+      "learning_rate": 0.0001,
+      "loss": 1.4669,
+      "step": 13412
+    },
+    {
+      "epoch": 1.540750100511171,
+      "grad_norm": 0.5998444557189941,
+      "learning_rate": 0.0001,
+      "loss": 1.3983,
+      "step": 13413
+    },
+    {
+      "epoch": 1.5408649704209982,
+      "grad_norm": 0.6303174495697021,
+      "learning_rate": 0.0001,
+      "loss": 1.6117,
+      "step": 13414
+    },
+    {
+      "epoch": 1.5409798403308255,
+      "grad_norm": 0.6035495400428772,
+      "learning_rate": 0.0001,
+      "loss": 1.4046,
+      "step": 13415
+    },
+    {
+      "epoch": 1.5410947102406525,
+      "grad_norm": 0.6213974952697754,
+      "learning_rate": 0.0001,
+      "loss": 1.6235,
+      "step": 13416
+    },
+    {
+      "epoch": 1.5412095801504795,
+      "grad_norm": 0.6108240485191345,
+      "learning_rate": 0.0001,
+      "loss": 1.5057,
+      "step": 13417
+    },
+    {
+      "epoch": 1.5413244500603067,
+      "grad_norm": 0.5867082476615906,
+      "learning_rate": 0.0001,
+      "loss": 1.4281,
+      "step": 13418
+    },
+    {
+      "epoch": 1.541439319970134,
+      "grad_norm": 0.6038832664489746,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 13419
+    },
+    {
+      "epoch": 1.541554189879961,
+      "grad_norm": 0.6144563555717468,
+      "learning_rate": 0.0001,
+      "loss": 1.68,
+      "step": 13420
+    },
+    {
+      "epoch": 1.541669059789788,
+      "grad_norm": 0.568112850189209,
+      "learning_rate": 0.0001,
+      "loss": 1.4584,
+      "step": 13421
+    },
+    {
+      "epoch": 1.5417839296996152,
+      "grad_norm": 0.6124117374420166,
+      "learning_rate": 0.0001,
+      "loss": 1.6241,
+      "step": 13422
+    },
+    {
+      "epoch": 1.5418987996094424,
+      "grad_norm": 0.5999955534934998,
+      "learning_rate": 0.0001,
+      "loss": 1.4416,
+      "step": 13423
+    },
+    {
+      "epoch": 1.5420136695192694,
+      "grad_norm": 0.6205251812934875,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 13424
+    },
+    {
+      "epoch": 1.5421285394290964,
+      "grad_norm": 0.5759456753730774,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 13425
+    },
+    {
+      "epoch": 1.5422434093389237,
+      "grad_norm": 0.6211003661155701,
+      "learning_rate": 0.0001,
+      "loss": 1.5235,
+      "step": 13426
+    },
+    {
+      "epoch": 1.542358279248751,
+      "grad_norm": 0.5795714259147644,
+      "learning_rate": 0.0001,
+      "loss": 1.44,
+      "step": 13427
+    },
+    {
+      "epoch": 1.542473149158578,
+      "grad_norm": 0.5453798770904541,
+      "learning_rate": 0.0001,
+      "loss": 1.3127,
+      "step": 13428
+    },
+    {
+      "epoch": 1.542588019068405,
+      "grad_norm": 0.6491063833236694,
+      "learning_rate": 0.0001,
+      "loss": 1.4514,
+      "step": 13429
+    },
+    {
+      "epoch": 1.5427028889782322,
+      "grad_norm": 0.5597581267356873,
+      "learning_rate": 0.0001,
+      "loss": 1.4427,
+      "step": 13430
+    },
+    {
+      "epoch": 1.5428177588880594,
+      "grad_norm": 0.6014596819877625,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 13431
+    },
+    {
+      "epoch": 1.5429326287978864,
+      "grad_norm": 0.5784545540809631,
+      "learning_rate": 0.0001,
+      "loss": 1.4709,
+      "step": 13432
+    },
+    {
+      "epoch": 1.5430474987077134,
+      "grad_norm": 0.5702968835830688,
+      "learning_rate": 0.0001,
+      "loss": 1.4526,
+      "step": 13433
+    },
+    {
+      "epoch": 1.5431623686175406,
+      "grad_norm": 0.5658290386199951,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 13434
+    },
+    {
+      "epoch": 1.5432772385273679,
+      "grad_norm": 0.598718523979187,
+      "learning_rate": 0.0001,
+      "loss": 1.3576,
+      "step": 13435
+    },
+    {
+      "epoch": 1.5433921084371949,
+      "grad_norm": 0.6943349242210388,
+      "learning_rate": 0.0001,
+      "loss": 1.7476,
+      "step": 13436
+    },
+    {
+      "epoch": 1.5435069783470219,
+      "grad_norm": 0.5755090117454529,
+      "learning_rate": 0.0001,
+      "loss": 1.3345,
+      "step": 13437
+    },
+    {
+      "epoch": 1.5436218482568491,
+      "grad_norm": 0.6275411248207092,
+      "learning_rate": 0.0001,
+      "loss": 1.497,
+      "step": 13438
+    },
+    {
+      "epoch": 1.5437367181666763,
+      "grad_norm": 0.5974034667015076,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 13439
+    },
+    {
+      "epoch": 1.5438515880765034,
+      "grad_norm": 0.5933687090873718,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 13440
+    },
+    {
+      "epoch": 1.5439664579863304,
+      "grad_norm": 0.7212704420089722,
+      "learning_rate": 0.0001,
+      "loss": 1.562,
+      "step": 13441
+    },
+    {
+      "epoch": 1.5440813278961576,
+      "grad_norm": 0.5720301866531372,
+      "learning_rate": 0.0001,
+      "loss": 1.3696,
+      "step": 13442
+    },
+    {
+      "epoch": 1.5441961978059848,
+      "grad_norm": 0.6417856812477112,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 13443
+    },
+    {
+      "epoch": 1.5443110677158118,
+      "grad_norm": 0.6052496433258057,
+      "learning_rate": 0.0001,
+      "loss": 1.2494,
+      "step": 13444
+    },
+    {
+      "epoch": 1.5444259376256388,
+      "grad_norm": 0.6466429829597473,
+      "learning_rate": 0.0001,
+      "loss": 1.483,
+      "step": 13445
+    },
+    {
+      "epoch": 1.544540807535466,
+      "grad_norm": 0.6234352588653564,
+      "learning_rate": 0.0001,
+      "loss": 1.5925,
+      "step": 13446
+    },
+    {
+      "epoch": 1.5446556774452933,
+      "grad_norm": 0.6527000069618225,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 13447
+    },
+    {
+      "epoch": 1.5447705473551203,
+      "grad_norm": 0.5902075171470642,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 13448
+    },
+    {
+      "epoch": 1.5448854172649473,
+      "grad_norm": 0.5839083790779114,
+      "learning_rate": 0.0001,
+      "loss": 1.3611,
+      "step": 13449
+    },
+    {
+      "epoch": 1.5450002871747746,
+      "grad_norm": 0.6253820061683655,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 13450
+    },
+    {
+      "epoch": 1.5451151570846018,
+      "grad_norm": 0.5373639464378357,
+      "learning_rate": 0.0001,
+      "loss": 1.3338,
+      "step": 13451
+    },
+    {
+      "epoch": 1.5452300269944288,
+      "grad_norm": 0.5663428902626038,
+      "learning_rate": 0.0001,
+      "loss": 1.3177,
+      "step": 13452
+    },
+    {
+      "epoch": 1.5453448969042558,
+      "grad_norm": 0.6304009556770325,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 13453
+    },
+    {
+      "epoch": 1.545459766814083,
+      "grad_norm": 0.6668345332145691,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 13454
+    },
+    {
+      "epoch": 1.5455746367239103,
+      "grad_norm": 0.5478126406669617,
+      "learning_rate": 0.0001,
+      "loss": 1.4109,
+      "step": 13455
+    },
+    {
+      "epoch": 1.5456895066337373,
+      "grad_norm": 0.6082738637924194,
+      "learning_rate": 0.0001,
+      "loss": 1.4676,
+      "step": 13456
+    },
+    {
+      "epoch": 1.5458043765435643,
+      "grad_norm": 0.6476391553878784,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 13457
+    },
+    {
+      "epoch": 1.5459192464533915,
+      "grad_norm": 0.6350123286247253,
+      "learning_rate": 0.0001,
+      "loss": 1.4257,
+      "step": 13458
+    },
+    {
+      "epoch": 1.5460341163632187,
+      "grad_norm": 0.621774435043335,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 13459
+    },
+    {
+      "epoch": 1.5461489862730458,
+      "grad_norm": 0.6082205176353455,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 13460
+    },
+    {
+      "epoch": 1.5462638561828728,
+      "grad_norm": 0.5338382124900818,
+      "learning_rate": 0.0001,
+      "loss": 1.2507,
+      "step": 13461
+    },
+    {
+      "epoch": 1.5463787260927,
+      "grad_norm": 0.6611436605453491,
+      "learning_rate": 0.0001,
+      "loss": 1.5725,
+      "step": 13462
+    },
+    {
+      "epoch": 1.5464935960025272,
+      "grad_norm": 0.6387352347373962,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 13463
+    },
+    {
+      "epoch": 1.5466084659123542,
+      "grad_norm": 0.6265326142311096,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 13464
+    },
+    {
+      "epoch": 1.5467233358221812,
+      "grad_norm": 0.5728742480278015,
+      "learning_rate": 0.0001,
+      "loss": 1.3898,
+      "step": 13465
+    },
+    {
+      "epoch": 1.5468382057320085,
+      "grad_norm": 0.5909122228622437,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 13466
+    },
+    {
+      "epoch": 1.5469530756418357,
+      "grad_norm": 0.5926406979560852,
+      "learning_rate": 0.0001,
+      "loss": 1.3875,
+      "step": 13467
+    },
+    {
+      "epoch": 1.5470679455516627,
+      "grad_norm": 0.6227415204048157,
+      "learning_rate": 0.0001,
+      "loss": 1.3836,
+      "step": 13468
+    },
+    {
+      "epoch": 1.5471828154614897,
+      "grad_norm": 0.6383469104766846,
+      "learning_rate": 0.0001,
+      "loss": 1.4241,
+      "step": 13469
+    },
+    {
+      "epoch": 1.547297685371317,
+      "grad_norm": 0.606417715549469,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 13470
+    },
+    {
+      "epoch": 1.5474125552811442,
+      "grad_norm": 0.5884244441986084,
+      "learning_rate": 0.0001,
+      "loss": 1.2275,
+      "step": 13471
+    },
+    {
+      "epoch": 1.5475274251909712,
+      "grad_norm": 0.6884910464286804,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 13472
+    },
+    {
+      "epoch": 1.5476422951007982,
+      "grad_norm": 0.6169816851615906,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 13473
+    },
+    {
+      "epoch": 1.5477571650106254,
+      "grad_norm": 0.5996578335762024,
+      "learning_rate": 0.0001,
+      "loss": 1.3095,
+      "step": 13474
+    },
+    {
+      "epoch": 1.5478720349204527,
+      "grad_norm": 0.6751422882080078,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 13475
+    },
+    {
+      "epoch": 1.5479869048302797,
+      "grad_norm": 0.6498405337333679,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 13476
+    },
+    {
+      "epoch": 1.5481017747401067,
+      "grad_norm": 0.6042332053184509,
+      "learning_rate": 0.0001,
+      "loss": 1.5048,
+      "step": 13477
+    },
+    {
+      "epoch": 1.548216644649934,
+      "grad_norm": 0.6130797863006592,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 13478
+    },
+    {
+      "epoch": 1.5483315145597611,
+      "grad_norm": 0.6066043972969055,
+      "learning_rate": 0.0001,
+      "loss": 1.4461,
+      "step": 13479
+    },
+    {
+      "epoch": 1.5484463844695882,
+      "grad_norm": 0.5683161616325378,
+      "learning_rate": 0.0001,
+      "loss": 1.3388,
+      "step": 13480
+    },
+    {
+      "epoch": 1.5485612543794152,
+      "grad_norm": 0.5794828534126282,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 13481
+    },
+    {
+      "epoch": 1.5486761242892424,
+      "grad_norm": 0.5909842848777771,
+      "learning_rate": 0.0001,
+      "loss": 1.3772,
+      "step": 13482
+    },
+    {
+      "epoch": 1.5487909941990696,
+      "grad_norm": 0.6028822660446167,
+      "learning_rate": 0.0001,
+      "loss": 1.461,
+      "step": 13483
+    },
+    {
+      "epoch": 1.5489058641088966,
+      "grad_norm": 0.6567312479019165,
+      "learning_rate": 0.0001,
+      "loss": 1.4616,
+      "step": 13484
+    },
+    {
+      "epoch": 1.5490207340187236,
+      "grad_norm": 0.6116561889648438,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 13485
+    },
+    {
+      "epoch": 1.5491356039285509,
+      "grad_norm": 0.638192892074585,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 13486
+    },
+    {
+      "epoch": 1.549250473838378,
+      "grad_norm": 0.6210533380508423,
+      "learning_rate": 0.0001,
+      "loss": 1.4859,
+      "step": 13487
+    },
+    {
+      "epoch": 1.5493653437482051,
+      "grad_norm": 0.6891507506370544,
+      "learning_rate": 0.0001,
+      "loss": 1.332,
+      "step": 13488
+    },
+    {
+      "epoch": 1.5494802136580321,
+      "grad_norm": 0.5958152413368225,
+      "learning_rate": 0.0001,
+      "loss": 1.3574,
+      "step": 13489
+    },
+    {
+      "epoch": 1.5495950835678594,
+      "grad_norm": 0.5857873558998108,
+      "learning_rate": 0.0001,
+      "loss": 1.3016,
+      "step": 13490
+    },
+    {
+      "epoch": 1.5497099534776866,
+      "grad_norm": 0.5769304037094116,
+      "learning_rate": 0.0001,
+      "loss": 1.3742,
+      "step": 13491
+    },
+    {
+      "epoch": 1.5498248233875136,
+      "grad_norm": 0.5681525468826294,
+      "learning_rate": 0.0001,
+      "loss": 1.2591,
+      "step": 13492
+    },
+    {
+      "epoch": 1.5499396932973406,
+      "grad_norm": 0.6011735796928406,
+      "learning_rate": 0.0001,
+      "loss": 1.3664,
+      "step": 13493
+    },
+    {
+      "epoch": 1.5500545632071678,
+      "grad_norm": 0.5927474498748779,
+      "learning_rate": 0.0001,
+      "loss": 1.4265,
+      "step": 13494
+    },
+    {
+      "epoch": 1.550169433116995,
+      "grad_norm": 0.6287341713905334,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 13495
+    },
+    {
+      "epoch": 1.550284303026822,
+      "grad_norm": 0.6226458549499512,
+      "learning_rate": 0.0001,
+      "loss": 1.3742,
+      "step": 13496
+    },
+    {
+      "epoch": 1.550399172936649,
+      "grad_norm": 0.6283382773399353,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 13497
+    },
+    {
+      "epoch": 1.5505140428464763,
+      "grad_norm": 0.6250648498535156,
+      "learning_rate": 0.0001,
+      "loss": 1.4687,
+      "step": 13498
+    },
+    {
+      "epoch": 1.5506289127563035,
+      "grad_norm": 0.6573746204376221,
+      "learning_rate": 0.0001,
+      "loss": 1.4691,
+      "step": 13499
+    },
+    {
+      "epoch": 1.5507437826661306,
+      "grad_norm": 0.6034247279167175,
+      "learning_rate": 0.0001,
+      "loss": 1.4162,
+      "step": 13500
+    },
+    {
+      "epoch": 1.5508586525759576,
+      "grad_norm": 0.5965235829353333,
+      "learning_rate": 0.0001,
+      "loss": 1.3892,
+      "step": 13501
+    },
+    {
+      "epoch": 1.5509735224857848,
+      "grad_norm": 0.6325914263725281,
+      "learning_rate": 0.0001,
+      "loss": 1.3612,
+      "step": 13502
+    },
+    {
+      "epoch": 1.551088392395612,
+      "grad_norm": 0.6564382910728455,
+      "learning_rate": 0.0001,
+      "loss": 1.5547,
+      "step": 13503
+    },
+    {
+      "epoch": 1.5512032623054393,
+      "grad_norm": 0.6648733019828796,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 13504
+    },
+    {
+      "epoch": 1.5513181322152663,
+      "grad_norm": 0.6087038516998291,
+      "learning_rate": 0.0001,
+      "loss": 1.2969,
+      "step": 13505
+    },
+    {
+      "epoch": 1.5514330021250933,
+      "grad_norm": 0.5483207106590271,
+      "learning_rate": 0.0001,
+      "loss": 1.2729,
+      "step": 13506
+    },
+    {
+      "epoch": 1.5515478720349205,
+      "grad_norm": 0.5833482146263123,
+      "learning_rate": 0.0001,
+      "loss": 1.2962,
+      "step": 13507
+    },
+    {
+      "epoch": 1.5516627419447477,
+      "grad_norm": 0.6498992443084717,
+      "learning_rate": 0.0001,
+      "loss": 1.4952,
+      "step": 13508
+    },
+    {
+      "epoch": 1.5517776118545747,
+      "grad_norm": 0.5728626847267151,
+      "learning_rate": 0.0001,
+      "loss": 1.3803,
+      "step": 13509
+    },
+    {
+      "epoch": 1.5518924817644018,
+      "grad_norm": 0.5859959125518799,
+      "learning_rate": 0.0001,
+      "loss": 1.4502,
+      "step": 13510
+    },
+    {
+      "epoch": 1.552007351674229,
+      "grad_norm": 0.6737546324729919,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 13511
+    },
+    {
+      "epoch": 1.5521222215840562,
+      "grad_norm": 0.6166685819625854,
+      "learning_rate": 0.0001,
+      "loss": 1.4754,
+      "step": 13512
+    },
+    {
+      "epoch": 1.5522370914938832,
+      "grad_norm": 0.7040491104125977,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 13513
+    },
+    {
+      "epoch": 1.5523519614037102,
+      "grad_norm": 0.6027595400810242,
+      "learning_rate": 0.0001,
+      "loss": 1.4652,
+      "step": 13514
+    },
+    {
+      "epoch": 1.5524668313135375,
+      "grad_norm": 0.5595097541809082,
+      "learning_rate": 0.0001,
+      "loss": 1.3197,
+      "step": 13515
+    },
+    {
+      "epoch": 1.5525817012233647,
+      "grad_norm": 0.5787383317947388,
+      "learning_rate": 0.0001,
+      "loss": 1.3246,
+      "step": 13516
+    },
+    {
+      "epoch": 1.5526965711331917,
+      "grad_norm": 0.64249187707901,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 13517
+    },
+    {
+      "epoch": 1.5528114410430187,
+      "grad_norm": 0.5845315456390381,
+      "learning_rate": 0.0001,
+      "loss": 1.2307,
+      "step": 13518
+    },
+    {
+      "epoch": 1.552926310952846,
+      "grad_norm": 0.6052901744842529,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 13519
+    },
+    {
+      "epoch": 1.5530411808626732,
+      "grad_norm": 0.5542384386062622,
+      "learning_rate": 0.0001,
+      "loss": 1.3201,
+      "step": 13520
+    },
+    {
+      "epoch": 1.5531560507725002,
+      "grad_norm": 0.5924617648124695,
+      "learning_rate": 0.0001,
+      "loss": 1.4083,
+      "step": 13521
+    },
+    {
+      "epoch": 1.5532709206823272,
+      "grad_norm": 0.5823348760604858,
+      "learning_rate": 0.0001,
+      "loss": 1.433,
+      "step": 13522
+    },
+    {
+      "epoch": 1.5533857905921544,
+      "grad_norm": 0.5733996033668518,
+      "learning_rate": 0.0001,
+      "loss": 1.3618,
+      "step": 13523
+    },
+    {
+      "epoch": 1.5535006605019817,
+      "grad_norm": 0.5738714933395386,
+      "learning_rate": 0.0001,
+      "loss": 1.3387,
+      "step": 13524
+    },
+    {
+      "epoch": 1.5536155304118087,
+      "grad_norm": 0.5945634245872498,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 13525
+    },
+    {
+      "epoch": 1.5537304003216357,
+      "grad_norm": 0.6478724479675293,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 13526
+    },
+    {
+      "epoch": 1.553845270231463,
+      "grad_norm": 0.6466807126998901,
+      "learning_rate": 0.0001,
+      "loss": 1.3919,
+      "step": 13527
+    },
+    {
+      "epoch": 1.5539601401412901,
+      "grad_norm": 0.5651605725288391,
+      "learning_rate": 0.0001,
+      "loss": 1.3325,
+      "step": 13528
+    },
+    {
+      "epoch": 1.5540750100511171,
+      "grad_norm": 0.6229522228240967,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 13529
+    },
+    {
+      "epoch": 1.5541898799609442,
+      "grad_norm": 0.6317934393882751,
+      "learning_rate": 0.0001,
+      "loss": 1.524,
+      "step": 13530
+    },
+    {
+      "epoch": 1.5543047498707714,
+      "grad_norm": 0.5871884822845459,
+      "learning_rate": 0.0001,
+      "loss": 1.4019,
+      "step": 13531
+    },
+    {
+      "epoch": 1.5544196197805986,
+      "grad_norm": 0.5875590443611145,
+      "learning_rate": 0.0001,
+      "loss": 1.3716,
+      "step": 13532
+    },
+    {
+      "epoch": 1.5545344896904256,
+      "grad_norm": 0.5747091174125671,
+      "learning_rate": 0.0001,
+      "loss": 1.2858,
+      "step": 13533
+    },
+    {
+      "epoch": 1.5546493596002526,
+      "grad_norm": 0.5531101226806641,
+      "learning_rate": 0.0001,
+      "loss": 1.394,
+      "step": 13534
+    },
+    {
+      "epoch": 1.5547642295100799,
+      "grad_norm": 0.7090408205986023,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 13535
+    },
+    {
+      "epoch": 1.554879099419907,
+      "grad_norm": 0.6552852392196655,
+      "learning_rate": 0.0001,
+      "loss": 1.6069,
+      "step": 13536
+    },
+    {
+      "epoch": 1.554993969329734,
+      "grad_norm": 0.6876298189163208,
+      "learning_rate": 0.0001,
+      "loss": 1.6896,
+      "step": 13537
+    },
+    {
+      "epoch": 1.5551088392395611,
+      "grad_norm": 0.5838922262191772,
+      "learning_rate": 0.0001,
+      "loss": 1.4006,
+      "step": 13538
+    },
+    {
+      "epoch": 1.5552237091493883,
+      "grad_norm": 0.6967250108718872,
+      "learning_rate": 0.0001,
+      "loss": 1.3594,
+      "step": 13539
+    },
+    {
+      "epoch": 1.5553385790592156,
+      "grad_norm": 0.6267087459564209,
+      "learning_rate": 0.0001,
+      "loss": 1.1573,
+      "step": 13540
+    },
+    {
+      "epoch": 1.5554534489690426,
+      "grad_norm": 0.5778437852859497,
+      "learning_rate": 0.0001,
+      "loss": 1.4125,
+      "step": 13541
+    },
+    {
+      "epoch": 1.5555683188788696,
+      "grad_norm": 0.5467836260795593,
+      "learning_rate": 0.0001,
+      "loss": 1.1714,
+      "step": 13542
+    },
+    {
+      "epoch": 1.5556831887886968,
+      "grad_norm": 0.5664315819740295,
+      "learning_rate": 0.0001,
+      "loss": 1.3083,
+      "step": 13543
+    },
+    {
+      "epoch": 1.555798058698524,
+      "grad_norm": 0.6014490127563477,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 13544
+    },
+    {
+      "epoch": 1.555912928608351,
+      "grad_norm": 0.6335508823394775,
+      "learning_rate": 0.0001,
+      "loss": 1.236,
+      "step": 13545
+    },
+    {
+      "epoch": 1.556027798518178,
+      "grad_norm": 0.6208090782165527,
+      "learning_rate": 0.0001,
+      "loss": 1.3677,
+      "step": 13546
+    },
+    {
+      "epoch": 1.5561426684280053,
+      "grad_norm": 0.61933833360672,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 13547
+    },
+    {
+      "epoch": 1.5562575383378325,
+      "grad_norm": 0.6805666089057922,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 13548
+    },
+    {
+      "epoch": 1.5563724082476595,
+      "grad_norm": 0.7156261205673218,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 13549
+    },
+    {
+      "epoch": 1.5564872781574866,
+      "grad_norm": 0.6893269419670105,
+      "learning_rate": 0.0001,
+      "loss": 1.4658,
+      "step": 13550
+    },
+    {
+      "epoch": 1.5566021480673138,
+      "grad_norm": 0.6271970868110657,
+      "learning_rate": 0.0001,
+      "loss": 1.5036,
+      "step": 13551
+    },
+    {
+      "epoch": 1.556717017977141,
+      "grad_norm": 0.6317713856697083,
+      "learning_rate": 0.0001,
+      "loss": 1.5933,
+      "step": 13552
+    },
+    {
+      "epoch": 1.556831887886968,
+      "grad_norm": 0.6471002697944641,
+      "learning_rate": 0.0001,
+      "loss": 1.4808,
+      "step": 13553
+    },
+    {
+      "epoch": 1.556946757796795,
+      "grad_norm": 0.6045787930488586,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 13554
+    },
+    {
+      "epoch": 1.5570616277066223,
+      "grad_norm": 0.5843409299850464,
+      "learning_rate": 0.0001,
+      "loss": 1.3993,
+      "step": 13555
+    },
+    {
+      "epoch": 1.5571764976164495,
+      "grad_norm": 0.5926359295845032,
+      "learning_rate": 0.0001,
+      "loss": 1.2548,
+      "step": 13556
+    },
+    {
+      "epoch": 1.5572913675262765,
+      "grad_norm": 0.5961912274360657,
+      "learning_rate": 0.0001,
+      "loss": 1.677,
+      "step": 13557
+    },
+    {
+      "epoch": 1.5574062374361035,
+      "grad_norm": 0.7315113544464111,
+      "learning_rate": 0.0001,
+      "loss": 1.5993,
+      "step": 13558
+    },
+    {
+      "epoch": 1.5575211073459307,
+      "grad_norm": 0.576897919178009,
+      "learning_rate": 0.0001,
+      "loss": 1.4251,
+      "step": 13559
+    },
+    {
+      "epoch": 1.557635977255758,
+      "grad_norm": 0.5911585092544556,
+      "learning_rate": 0.0001,
+      "loss": 1.2031,
+      "step": 13560
+    },
+    {
+      "epoch": 1.557750847165585,
+      "grad_norm": 0.6434839963912964,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 13561
+    },
+    {
+      "epoch": 1.557865717075412,
+      "grad_norm": 0.5712766647338867,
+      "learning_rate": 0.0001,
+      "loss": 1.1872,
+      "step": 13562
+    },
+    {
+      "epoch": 1.5579805869852392,
+      "grad_norm": 0.6965934634208679,
+      "learning_rate": 0.0001,
+      "loss": 1.3617,
+      "step": 13563
+    },
+    {
+      "epoch": 1.5580954568950665,
+      "grad_norm": 0.5756476521492004,
+      "learning_rate": 0.0001,
+      "loss": 1.424,
+      "step": 13564
+    },
+    {
+      "epoch": 1.5582103268048935,
+      "grad_norm": 0.6307820081710815,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 13565
+    },
+    {
+      "epoch": 1.5583251967147205,
+      "grad_norm": 0.5934085845947266,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 13566
+    },
+    {
+      "epoch": 1.5584400666245477,
+      "grad_norm": 0.6400713324546814,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 13567
+    },
+    {
+      "epoch": 1.558554936534375,
+      "grad_norm": 0.6499615907669067,
+      "learning_rate": 0.0001,
+      "loss": 1.5194,
+      "step": 13568
+    },
+    {
+      "epoch": 1.558669806444202,
+      "grad_norm": 0.6657858490943909,
+      "learning_rate": 0.0001,
+      "loss": 1.5722,
+      "step": 13569
+    },
+    {
+      "epoch": 1.558784676354029,
+      "grad_norm": 0.5988503098487854,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 13570
+    },
+    {
+      "epoch": 1.5588995462638562,
+      "grad_norm": 0.6103773713111877,
+      "learning_rate": 0.0001,
+      "loss": 1.5287,
+      "step": 13571
+    },
+    {
+      "epoch": 1.5590144161736834,
+      "grad_norm": 0.5706753134727478,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 13572
+    },
+    {
+      "epoch": 1.5591292860835104,
+      "grad_norm": 0.5638065338134766,
+      "learning_rate": 0.0001,
+      "loss": 1.3968,
+      "step": 13573
+    },
+    {
+      "epoch": 1.5592441559933374,
+      "grad_norm": 0.5787845849990845,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 13574
+    },
+    {
+      "epoch": 1.5593590259031647,
+      "grad_norm": 0.5700173377990723,
+      "learning_rate": 0.0001,
+      "loss": 1.3468,
+      "step": 13575
+    },
+    {
+      "epoch": 1.559473895812992,
+      "grad_norm": 0.5860016345977783,
+      "learning_rate": 0.0001,
+      "loss": 1.4658,
+      "step": 13576
+    },
+    {
+      "epoch": 1.559588765722819,
+      "grad_norm": 0.604753851890564,
+      "learning_rate": 0.0001,
+      "loss": 1.4624,
+      "step": 13577
+    },
+    {
+      "epoch": 1.559703635632646,
+      "grad_norm": 0.600627601146698,
+      "learning_rate": 0.0001,
+      "loss": 1.5398,
+      "step": 13578
+    },
+    {
+      "epoch": 1.5598185055424731,
+      "grad_norm": 0.6065698266029358,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 13579
+    },
+    {
+      "epoch": 1.5599333754523004,
+      "grad_norm": 0.5824136137962341,
+      "learning_rate": 0.0001,
+      "loss": 1.4065,
+      "step": 13580
+    },
+    {
+      "epoch": 1.5600482453621274,
+      "grad_norm": 0.6180105209350586,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 13581
+    },
+    {
+      "epoch": 1.5601631152719544,
+      "grad_norm": 0.6440869569778442,
+      "learning_rate": 0.0001,
+      "loss": 1.592,
+      "step": 13582
+    },
+    {
+      "epoch": 1.5602779851817816,
+      "grad_norm": 0.6405513286590576,
+      "learning_rate": 0.0001,
+      "loss": 1.3444,
+      "step": 13583
+    },
+    {
+      "epoch": 1.5603928550916089,
+      "grad_norm": 0.6168532371520996,
+      "learning_rate": 0.0001,
+      "loss": 1.2609,
+      "step": 13584
+    },
+    {
+      "epoch": 1.5605077250014359,
+      "grad_norm": 0.5789191722869873,
+      "learning_rate": 0.0001,
+      "loss": 1.2043,
+      "step": 13585
+    },
+    {
+      "epoch": 1.5606225949112629,
+      "grad_norm": 0.6007379293441772,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 13586
+    },
+    {
+      "epoch": 1.56073746482109,
+      "grad_norm": 0.5817810297012329,
+      "learning_rate": 0.0001,
+      "loss": 1.3975,
+      "step": 13587
+    },
+    {
+      "epoch": 1.5608523347309173,
+      "grad_norm": 0.5913118720054626,
+      "learning_rate": 0.0001,
+      "loss": 1.3675,
+      "step": 13588
+    },
+    {
+      "epoch": 1.5609672046407443,
+      "grad_norm": 0.5883614420890808,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 13589
+    },
+    {
+      "epoch": 1.5610820745505714,
+      "grad_norm": 0.6002376079559326,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 13590
+    },
+    {
+      "epoch": 1.5611969444603986,
+      "grad_norm": 0.6250750422477722,
+      "learning_rate": 0.0001,
+      "loss": 1.5516,
+      "step": 13591
+    },
+    {
+      "epoch": 1.5613118143702258,
+      "grad_norm": 0.611410915851593,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 13592
+    },
+    {
+      "epoch": 1.5614266842800528,
+      "grad_norm": 0.5881221890449524,
+      "learning_rate": 0.0001,
+      "loss": 1.5605,
+      "step": 13593
+    },
+    {
+      "epoch": 1.5615415541898798,
+      "grad_norm": 0.5879666805267334,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 13594
+    },
+    {
+      "epoch": 1.561656424099707,
+      "grad_norm": 0.5731499791145325,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 13595
+    },
+    {
+      "epoch": 1.5617712940095343,
+      "grad_norm": 0.6417600512504578,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 13596
+    },
+    {
+      "epoch": 1.5618861639193613,
+      "grad_norm": 0.5718529224395752,
+      "learning_rate": 0.0001,
+      "loss": 1.3185,
+      "step": 13597
+    },
+    {
+      "epoch": 1.5620010338291883,
+      "grad_norm": 0.6178721189498901,
+      "learning_rate": 0.0001,
+      "loss": 1.311,
+      "step": 13598
+    },
+    {
+      "epoch": 1.5621159037390155,
+      "grad_norm": 0.5907917618751526,
+      "learning_rate": 0.0001,
+      "loss": 1.4209,
+      "step": 13599
+    },
+    {
+      "epoch": 1.5622307736488428,
+      "grad_norm": 0.6152604818344116,
+      "learning_rate": 0.0001,
+      "loss": 1.4549,
+      "step": 13600
+    },
+    {
+      "epoch": 1.5623456435586698,
+      "grad_norm": 0.6977373361587524,
+      "learning_rate": 0.0001,
+      "loss": 1.7487,
+      "step": 13601
+    },
+    {
+      "epoch": 1.5624605134684968,
+      "grad_norm": 0.6169999241828918,
+      "learning_rate": 0.0001,
+      "loss": 1.5021,
+      "step": 13602
+    },
+    {
+      "epoch": 1.562575383378324,
+      "grad_norm": 0.5735710859298706,
+      "learning_rate": 0.0001,
+      "loss": 1.3729,
+      "step": 13603
+    },
+    {
+      "epoch": 1.5626902532881513,
+      "grad_norm": 0.5788447856903076,
+      "learning_rate": 0.0001,
+      "loss": 1.2963,
+      "step": 13604
+    },
+    {
+      "epoch": 1.5628051231979783,
+      "grad_norm": 0.6432527303695679,
+      "learning_rate": 0.0001,
+      "loss": 1.4393,
+      "step": 13605
+    },
+    {
+      "epoch": 1.5629199931078053,
+      "grad_norm": 0.6032426357269287,
+      "learning_rate": 0.0001,
+      "loss": 1.5303,
+      "step": 13606
+    },
+    {
+      "epoch": 1.5630348630176325,
+      "grad_norm": 0.6332036852836609,
+      "learning_rate": 0.0001,
+      "loss": 1.5395,
+      "step": 13607
+    },
+    {
+      "epoch": 1.5631497329274597,
+      "grad_norm": 0.5610132217407227,
+      "learning_rate": 0.0001,
+      "loss": 1.4589,
+      "step": 13608
+    },
+    {
+      "epoch": 1.5632646028372867,
+      "grad_norm": 0.5653334259986877,
+      "learning_rate": 0.0001,
+      "loss": 1.2573,
+      "step": 13609
+    },
+    {
+      "epoch": 1.5633794727471138,
+      "grad_norm": 0.5841345191001892,
+      "learning_rate": 0.0001,
+      "loss": 1.4447,
+      "step": 13610
+    },
+    {
+      "epoch": 1.563494342656941,
+      "grad_norm": 0.6390504240989685,
+      "learning_rate": 0.0001,
+      "loss": 1.5338,
+      "step": 13611
+    },
+    {
+      "epoch": 1.5636092125667682,
+      "grad_norm": 0.6101760268211365,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 13612
+    },
+    {
+      "epoch": 1.5637240824765952,
+      "grad_norm": 0.6613549590110779,
+      "learning_rate": 0.0001,
+      "loss": 1.4788,
+      "step": 13613
+    },
+    {
+      "epoch": 1.5638389523864222,
+      "grad_norm": 0.6057902574539185,
+      "learning_rate": 0.0001,
+      "loss": 1.3687,
+      "step": 13614
+    },
+    {
+      "epoch": 1.5639538222962495,
+      "grad_norm": 0.6379287838935852,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 13615
+    },
+    {
+      "epoch": 1.5640686922060767,
+      "grad_norm": 0.597686231136322,
+      "learning_rate": 0.0001,
+      "loss": 1.3398,
+      "step": 13616
+    },
+    {
+      "epoch": 1.5641835621159037,
+      "grad_norm": 0.5676443576812744,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 13617
+    },
+    {
+      "epoch": 1.5642984320257307,
+      "grad_norm": 0.6300573348999023,
+      "learning_rate": 0.0001,
+      "loss": 1.6266,
+      "step": 13618
+    },
+    {
+      "epoch": 1.564413301935558,
+      "grad_norm": 0.5722482204437256,
+      "learning_rate": 0.0001,
+      "loss": 1.4044,
+      "step": 13619
+    },
+    {
+      "epoch": 1.5645281718453852,
+      "grad_norm": 0.6020122170448303,
+      "learning_rate": 0.0001,
+      "loss": 1.4094,
+      "step": 13620
+    },
+    {
+      "epoch": 1.5646430417552122,
+      "grad_norm": 0.6561688780784607,
+      "learning_rate": 0.0001,
+      "loss": 1.5187,
+      "step": 13621
+    },
+    {
+      "epoch": 1.5647579116650392,
+      "grad_norm": 0.589190661907196,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 13622
+    },
+    {
+      "epoch": 1.5648727815748664,
+      "grad_norm": 0.593706488609314,
+      "learning_rate": 0.0001,
+      "loss": 1.5398,
+      "step": 13623
+    },
+    {
+      "epoch": 1.5649876514846937,
+      "grad_norm": 0.6446245312690735,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 13624
+    },
+    {
+      "epoch": 1.5651025213945207,
+      "grad_norm": 0.6118418574333191,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 13625
+    },
+    {
+      "epoch": 1.5652173913043477,
+      "grad_norm": 0.6244831681251526,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 13626
+    },
+    {
+      "epoch": 1.565332261214175,
+      "grad_norm": 0.6229195594787598,
+      "learning_rate": 0.0001,
+      "loss": 1.4809,
+      "step": 13627
+    },
+    {
+      "epoch": 1.5654471311240021,
+      "grad_norm": 0.5555204153060913,
+      "learning_rate": 0.0001,
+      "loss": 1.2535,
+      "step": 13628
+    },
+    {
+      "epoch": 1.5655620010338291,
+      "grad_norm": 0.6049181222915649,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 13629
+    },
+    {
+      "epoch": 1.5656768709436562,
+      "grad_norm": 0.5908517837524414,
+      "learning_rate": 0.0001,
+      "loss": 1.3301,
+      "step": 13630
+    },
+    {
+      "epoch": 1.5657917408534834,
+      "grad_norm": 0.6150537133216858,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 13631
+    },
+    {
+      "epoch": 1.5659066107633106,
+      "grad_norm": 0.5594236850738525,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 13632
+    },
+    {
+      "epoch": 1.5660214806731376,
+      "grad_norm": 0.6012204885482788,
+      "learning_rate": 0.0001,
+      "loss": 1.4467,
+      "step": 13633
+    },
+    {
+      "epoch": 1.5661363505829646,
+      "grad_norm": 0.6006706357002258,
+      "learning_rate": 0.0001,
+      "loss": 1.4424,
+      "step": 13634
+    },
+    {
+      "epoch": 1.5662512204927919,
+      "grad_norm": 0.5952029824256897,
+      "learning_rate": 0.0001,
+      "loss": 1.3238,
+      "step": 13635
+    },
+    {
+      "epoch": 1.566366090402619,
+      "grad_norm": 0.5738430619239807,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 13636
+    },
+    {
+      "epoch": 1.566480960312446,
+      "grad_norm": 0.6680176258087158,
+      "learning_rate": 0.0001,
+      "loss": 1.453,
+      "step": 13637
+    },
+    {
+      "epoch": 1.5665958302222731,
+      "grad_norm": 0.6369667649269104,
+      "learning_rate": 0.0001,
+      "loss": 1.476,
+      "step": 13638
+    },
+    {
+      "epoch": 1.5667107001321003,
+      "grad_norm": 0.6867667436599731,
+      "learning_rate": 0.0001,
+      "loss": 1.491,
+      "step": 13639
+    },
+    {
+      "epoch": 1.5668255700419276,
+      "grad_norm": 0.5914698243141174,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 13640
+    },
+    {
+      "epoch": 1.5669404399517548,
+      "grad_norm": 0.5792491436004639,
+      "learning_rate": 0.0001,
+      "loss": 1.4495,
+      "step": 13641
+    },
+    {
+      "epoch": 1.5670553098615818,
+      "grad_norm": 0.6214803457260132,
+      "learning_rate": 0.0001,
+      "loss": 1.4099,
+      "step": 13642
+    },
+    {
+      "epoch": 1.5671701797714088,
+      "grad_norm": 0.5663346648216248,
+      "learning_rate": 0.0001,
+      "loss": 1.3742,
+      "step": 13643
+    },
+    {
+      "epoch": 1.567285049681236,
+      "grad_norm": 0.5920063257217407,
+      "learning_rate": 0.0001,
+      "loss": 1.4131,
+      "step": 13644
+    },
+    {
+      "epoch": 1.5673999195910633,
+      "grad_norm": 0.6462366580963135,
+      "learning_rate": 0.0001,
+      "loss": 1.422,
+      "step": 13645
+    },
+    {
+      "epoch": 1.5675147895008903,
+      "grad_norm": 0.6337886452674866,
+      "learning_rate": 0.0001,
+      "loss": 1.475,
+      "step": 13646
+    },
+    {
+      "epoch": 1.5676296594107173,
+      "grad_norm": 0.6497291326522827,
+      "learning_rate": 0.0001,
+      "loss": 1.3826,
+      "step": 13647
+    },
+    {
+      "epoch": 1.5677445293205445,
+      "grad_norm": 0.6257938146591187,
+      "learning_rate": 0.0001,
+      "loss": 1.4504,
+      "step": 13648
+    },
+    {
+      "epoch": 1.5678593992303718,
+      "grad_norm": 0.6551761031150818,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 13649
+    },
+    {
+      "epoch": 1.5679742691401988,
+      "grad_norm": 0.6262896656990051,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 13650
+    },
+    {
+      "epoch": 1.5680891390500258,
+      "grad_norm": 0.5628766417503357,
+      "learning_rate": 0.0001,
+      "loss": 1.3154,
+      "step": 13651
+    },
+    {
+      "epoch": 1.568204008959853,
+      "grad_norm": 0.6413166522979736,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 13652
+    },
+    {
+      "epoch": 1.5683188788696802,
+      "grad_norm": 0.568213164806366,
+      "learning_rate": 0.0001,
+      "loss": 1.3689,
+      "step": 13653
+    },
+    {
+      "epoch": 1.5684337487795073,
+      "grad_norm": 0.6713234186172485,
+      "learning_rate": 0.0001,
+      "loss": 1.4422,
+      "step": 13654
+    },
+    {
+      "epoch": 1.5685486186893343,
+      "grad_norm": 0.5839575529098511,
+      "learning_rate": 0.0001,
+      "loss": 1.3933,
+      "step": 13655
+    },
+    {
+      "epoch": 1.5686634885991615,
+      "grad_norm": 0.596601128578186,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 13656
+    },
+    {
+      "epoch": 1.5687783585089887,
+      "grad_norm": 0.5839077830314636,
+      "learning_rate": 0.0001,
+      "loss": 1.3594,
+      "step": 13657
+    },
+    {
+      "epoch": 1.5688932284188157,
+      "grad_norm": 0.6276641488075256,
+      "learning_rate": 0.0001,
+      "loss": 1.4735,
+      "step": 13658
+    },
+    {
+      "epoch": 1.5690080983286427,
+      "grad_norm": 0.6786941289901733,
+      "learning_rate": 0.0001,
+      "loss": 1.4608,
+      "step": 13659
+    },
+    {
+      "epoch": 1.56912296823847,
+      "grad_norm": 0.7449260354042053,
+      "learning_rate": 0.0001,
+      "loss": 1.6882,
+      "step": 13660
+    },
+    {
+      "epoch": 1.5692378381482972,
+      "grad_norm": 0.6506140828132629,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 13661
+    },
+    {
+      "epoch": 1.5693527080581242,
+      "grad_norm": 0.5764363408088684,
+      "learning_rate": 0.0001,
+      "loss": 1.396,
+      "step": 13662
+    },
+    {
+      "epoch": 1.5694675779679512,
+      "grad_norm": 0.6293970346450806,
+      "learning_rate": 0.0001,
+      "loss": 1.3189,
+      "step": 13663
+    },
+    {
+      "epoch": 1.5695824478777785,
+      "grad_norm": 0.5982877016067505,
+      "learning_rate": 0.0001,
+      "loss": 1.3611,
+      "step": 13664
+    },
+    {
+      "epoch": 1.5696973177876057,
+      "grad_norm": 0.7153692841529846,
+      "learning_rate": 0.0001,
+      "loss": 1.5204,
+      "step": 13665
+    },
+    {
+      "epoch": 1.5698121876974327,
+      "grad_norm": 0.556322455406189,
+      "learning_rate": 0.0001,
+      "loss": 1.1547,
+      "step": 13666
+    },
+    {
+      "epoch": 1.5699270576072597,
+      "grad_norm": 0.5707471370697021,
+      "learning_rate": 0.0001,
+      "loss": 1.4342,
+      "step": 13667
+    },
+    {
+      "epoch": 1.570041927517087,
+      "grad_norm": 0.6068229079246521,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 13668
+    },
+    {
+      "epoch": 1.5701567974269142,
+      "grad_norm": 0.5805314183235168,
+      "learning_rate": 0.0001,
+      "loss": 1.3441,
+      "step": 13669
+    },
+    {
+      "epoch": 1.5702716673367412,
+      "grad_norm": 0.5970333814620972,
+      "learning_rate": 0.0001,
+      "loss": 1.4048,
+      "step": 13670
+    },
+    {
+      "epoch": 1.5703865372465682,
+      "grad_norm": 0.6311222314834595,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 13671
+    },
+    {
+      "epoch": 1.5705014071563954,
+      "grad_norm": 0.6435375213623047,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 13672
+    },
+    {
+      "epoch": 1.5706162770662226,
+      "grad_norm": 0.62371826171875,
+      "learning_rate": 0.0001,
+      "loss": 1.4818,
+      "step": 13673
+    },
+    {
+      "epoch": 1.5707311469760497,
+      "grad_norm": 0.6377353072166443,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 13674
+    },
+    {
+      "epoch": 1.5708460168858767,
+      "grad_norm": 0.630799412727356,
+      "learning_rate": 0.0001,
+      "loss": 1.4142,
+      "step": 13675
+    },
+    {
+      "epoch": 1.570960886795704,
+      "grad_norm": 0.6089138388633728,
+      "learning_rate": 0.0001,
+      "loss": 1.344,
+      "step": 13676
+    },
+    {
+      "epoch": 1.5710757567055311,
+      "grad_norm": 0.6124714612960815,
+      "learning_rate": 0.0001,
+      "loss": 1.3856,
+      "step": 13677
+    },
+    {
+      "epoch": 1.5711906266153581,
+      "grad_norm": 0.592974066734314,
+      "learning_rate": 0.0001,
+      "loss": 1.3503,
+      "step": 13678
+    },
+    {
+      "epoch": 1.5713054965251851,
+      "grad_norm": 0.662214994430542,
+      "learning_rate": 0.0001,
+      "loss": 1.4621,
+      "step": 13679
+    },
+    {
+      "epoch": 1.5714203664350124,
+      "grad_norm": 0.5967538952827454,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 13680
+    },
+    {
+      "epoch": 1.5715352363448396,
+      "grad_norm": 0.6241264939308167,
+      "learning_rate": 0.0001,
+      "loss": 1.4568,
+      "step": 13681
+    },
+    {
+      "epoch": 1.5716501062546666,
+      "grad_norm": 0.6036410927772522,
+      "learning_rate": 0.0001,
+      "loss": 1.5817,
+      "step": 13682
+    },
+    {
+      "epoch": 1.5717649761644936,
+      "grad_norm": 0.5998589992523193,
+      "learning_rate": 0.0001,
+      "loss": 1.2894,
+      "step": 13683
+    },
+    {
+      "epoch": 1.5718798460743209,
+      "grad_norm": 0.6088207960128784,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 13684
+    },
+    {
+      "epoch": 1.571994715984148,
+      "grad_norm": 0.6031522750854492,
+      "learning_rate": 0.0001,
+      "loss": 1.4404,
+      "step": 13685
+    },
+    {
+      "epoch": 1.572109585893975,
+      "grad_norm": 0.5819301605224609,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 13686
+    },
+    {
+      "epoch": 1.572224455803802,
+      "grad_norm": 0.5897693634033203,
+      "learning_rate": 0.0001,
+      "loss": 1.306,
+      "step": 13687
+    },
+    {
+      "epoch": 1.5723393257136293,
+      "grad_norm": 0.6375783681869507,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 13688
+    },
+    {
+      "epoch": 1.5724541956234566,
+      "grad_norm": 0.6380894184112549,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 13689
+    },
+    {
+      "epoch": 1.5725690655332836,
+      "grad_norm": 0.6246376633644104,
+      "learning_rate": 0.0001,
+      "loss": 1.3728,
+      "step": 13690
+    },
+    {
+      "epoch": 1.5726839354431106,
+      "grad_norm": 0.6492366790771484,
+      "learning_rate": 0.0001,
+      "loss": 1.3662,
+      "step": 13691
+    },
+    {
+      "epoch": 1.5727988053529378,
+      "grad_norm": 0.6497386693954468,
+      "learning_rate": 0.0001,
+      "loss": 1.4163,
+      "step": 13692
+    },
+    {
+      "epoch": 1.572913675262765,
+      "grad_norm": 0.5843161940574646,
+      "learning_rate": 0.0001,
+      "loss": 1.3522,
+      "step": 13693
+    },
+    {
+      "epoch": 1.573028545172592,
+      "grad_norm": 0.6169954538345337,
+      "learning_rate": 0.0001,
+      "loss": 1.4708,
+      "step": 13694
+    },
+    {
+      "epoch": 1.573143415082419,
+      "grad_norm": 0.5712270140647888,
+      "learning_rate": 0.0001,
+      "loss": 1.2669,
+      "step": 13695
+    },
+    {
+      "epoch": 1.5732582849922463,
+      "grad_norm": 0.6111646294593811,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 13696
+    },
+    {
+      "epoch": 1.5733731549020735,
+      "grad_norm": 0.5616782307624817,
+      "learning_rate": 0.0001,
+      "loss": 1.3518,
+      "step": 13697
+    },
+    {
+      "epoch": 1.5734880248119005,
+      "grad_norm": 0.6015630960464478,
+      "learning_rate": 0.0001,
+      "loss": 1.2921,
+      "step": 13698
+    },
+    {
+      "epoch": 1.5736028947217275,
+      "grad_norm": 0.5990334749221802,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 13699
+    },
+    {
+      "epoch": 1.5737177646315548,
+      "grad_norm": 0.5758739709854126,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 13700
+    },
+    {
+      "epoch": 1.573832634541382,
+      "grad_norm": 0.5954878330230713,
+      "learning_rate": 0.0001,
+      "loss": 1.3584,
+      "step": 13701
+    },
+    {
+      "epoch": 1.573947504451209,
+      "grad_norm": 0.6391401290893555,
+      "learning_rate": 0.0001,
+      "loss": 1.4215,
+      "step": 13702
+    },
+    {
+      "epoch": 1.574062374361036,
+      "grad_norm": 0.7178584337234497,
+      "learning_rate": 0.0001,
+      "loss": 1.3619,
+      "step": 13703
+    },
+    {
+      "epoch": 1.5741772442708633,
+      "grad_norm": 0.6165657639503479,
+      "learning_rate": 0.0001,
+      "loss": 1.2028,
+      "step": 13704
+    },
+    {
+      "epoch": 1.5742921141806905,
+      "grad_norm": 0.5853223204612732,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 13705
+    },
+    {
+      "epoch": 1.5744069840905175,
+      "grad_norm": 0.6249833703041077,
+      "learning_rate": 0.0001,
+      "loss": 1.1105,
+      "step": 13706
+    },
+    {
+      "epoch": 1.5745218540003445,
+      "grad_norm": 0.5916066765785217,
+      "learning_rate": 0.0001,
+      "loss": 1.4113,
+      "step": 13707
+    },
+    {
+      "epoch": 1.5746367239101717,
+      "grad_norm": 0.5843729972839355,
+      "learning_rate": 0.0001,
+      "loss": 1.3678,
+      "step": 13708
+    },
+    {
+      "epoch": 1.574751593819999,
+      "grad_norm": 0.5756968259811401,
+      "learning_rate": 0.0001,
+      "loss": 1.3407,
+      "step": 13709
+    },
+    {
+      "epoch": 1.574866463729826,
+      "grad_norm": 0.6385331749916077,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 13710
+    },
+    {
+      "epoch": 1.574981333639653,
+      "grad_norm": 0.6959099769592285,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 13711
+    },
+    {
+      "epoch": 1.5750962035494802,
+      "grad_norm": 0.5969239473342896,
+      "learning_rate": 0.0001,
+      "loss": 1.4625,
+      "step": 13712
+    },
+    {
+      "epoch": 1.5752110734593074,
+      "grad_norm": 0.5907700061798096,
+      "learning_rate": 0.0001,
+      "loss": 1.4731,
+      "step": 13713
+    },
+    {
+      "epoch": 1.5753259433691345,
+      "grad_norm": 0.5716517567634583,
+      "learning_rate": 0.0001,
+      "loss": 1.0672,
+      "step": 13714
+    },
+    {
+      "epoch": 1.5754408132789615,
+      "grad_norm": 0.6416873931884766,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 13715
+    },
+    {
+      "epoch": 1.5755556831887887,
+      "grad_norm": 0.6345185041427612,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 13716
+    },
+    {
+      "epoch": 1.575670553098616,
+      "grad_norm": 0.63191157579422,
+      "learning_rate": 0.0001,
+      "loss": 1.6803,
+      "step": 13717
+    },
+    {
+      "epoch": 1.575785423008443,
+      "grad_norm": 0.5994105339050293,
+      "learning_rate": 0.0001,
+      "loss": 1.4558,
+      "step": 13718
+    },
+    {
+      "epoch": 1.57590029291827,
+      "grad_norm": 0.657799243927002,
+      "learning_rate": 0.0001,
+      "loss": 1.4011,
+      "step": 13719
+    },
+    {
+      "epoch": 1.5760151628280972,
+      "grad_norm": 0.6098163723945618,
+      "learning_rate": 0.0001,
+      "loss": 1.4295,
+      "step": 13720
+    },
+    {
+      "epoch": 1.5761300327379244,
+      "grad_norm": 0.5927136540412903,
+      "learning_rate": 0.0001,
+      "loss": 1.4307,
+      "step": 13721
+    },
+    {
+      "epoch": 1.5762449026477514,
+      "grad_norm": 0.5886332392692566,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 13722
+    },
+    {
+      "epoch": 1.5763597725575784,
+      "grad_norm": 0.589408814907074,
+      "learning_rate": 0.0001,
+      "loss": 1.3345,
+      "step": 13723
+    },
+    {
+      "epoch": 1.5764746424674057,
+      "grad_norm": 0.6005495190620422,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 13724
+    },
+    {
+      "epoch": 1.5765895123772329,
+      "grad_norm": 0.5907677412033081,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 13725
+    },
+    {
+      "epoch": 1.57670438228706,
+      "grad_norm": 0.6294686794281006,
+      "learning_rate": 0.0001,
+      "loss": 1.4602,
+      "step": 13726
+    },
+    {
+      "epoch": 1.576819252196887,
+      "grad_norm": 0.5746269822120667,
+      "learning_rate": 0.0001,
+      "loss": 1.4667,
+      "step": 13727
+    },
+    {
+      "epoch": 1.5769341221067141,
+      "grad_norm": 0.6354400515556335,
+      "learning_rate": 0.0001,
+      "loss": 1.3841,
+      "step": 13728
+    },
+    {
+      "epoch": 1.5770489920165414,
+      "grad_norm": 0.5498471856117249,
+      "learning_rate": 0.0001,
+      "loss": 1.4757,
+      "step": 13729
+    },
+    {
+      "epoch": 1.5771638619263684,
+      "grad_norm": 0.6611149907112122,
+      "learning_rate": 0.0001,
+      "loss": 1.4133,
+      "step": 13730
+    },
+    {
+      "epoch": 1.5772787318361954,
+      "grad_norm": 0.5527372360229492,
+      "learning_rate": 0.0001,
+      "loss": 1.4975,
+      "step": 13731
+    },
+    {
+      "epoch": 1.5773936017460226,
+      "grad_norm": 0.6152689456939697,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 13732
+    },
+    {
+      "epoch": 1.5775084716558498,
+      "grad_norm": 0.6419612169265747,
+      "learning_rate": 0.0001,
+      "loss": 1.6189,
+      "step": 13733
+    },
+    {
+      "epoch": 1.5776233415656769,
+      "grad_norm": 0.6237484216690063,
+      "learning_rate": 0.0001,
+      "loss": 1.3205,
+      "step": 13734
+    },
+    {
+      "epoch": 1.5777382114755039,
+      "grad_norm": 0.6146785616874695,
+      "learning_rate": 0.0001,
+      "loss": 1.3854,
+      "step": 13735
+    },
+    {
+      "epoch": 1.577853081385331,
+      "grad_norm": 0.5459988117218018,
+      "learning_rate": 0.0001,
+      "loss": 1.2364,
+      "step": 13736
+    },
+    {
+      "epoch": 1.5779679512951583,
+      "grad_norm": 0.5892395377159119,
+      "learning_rate": 0.0001,
+      "loss": 1.4168,
+      "step": 13737
+    },
+    {
+      "epoch": 1.5780828212049853,
+      "grad_norm": 0.6266677975654602,
+      "learning_rate": 0.0001,
+      "loss": 1.4853,
+      "step": 13738
+    },
+    {
+      "epoch": 1.5781976911148123,
+      "grad_norm": 0.6316624283790588,
+      "learning_rate": 0.0001,
+      "loss": 1.3561,
+      "step": 13739
+    },
+    {
+      "epoch": 1.5783125610246396,
+      "grad_norm": 0.5511997938156128,
+      "learning_rate": 0.0001,
+      "loss": 1.4032,
+      "step": 13740
+    },
+    {
+      "epoch": 1.5784274309344668,
+      "grad_norm": 0.6501776576042175,
+      "learning_rate": 0.0001,
+      "loss": 1.4935,
+      "step": 13741
+    },
+    {
+      "epoch": 1.5785423008442938,
+      "grad_norm": 0.6260613203048706,
+      "learning_rate": 0.0001,
+      "loss": 1.4332,
+      "step": 13742
+    },
+    {
+      "epoch": 1.5786571707541208,
+      "grad_norm": 0.6105442047119141,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 13743
+    },
+    {
+      "epoch": 1.578772040663948,
+      "grad_norm": 0.5777299404144287,
+      "learning_rate": 0.0001,
+      "loss": 1.4765,
+      "step": 13744
+    },
+    {
+      "epoch": 1.5788869105737753,
+      "grad_norm": 0.6278029680252075,
+      "learning_rate": 0.0001,
+      "loss": 1.4338,
+      "step": 13745
+    },
+    {
+      "epoch": 1.5790017804836023,
+      "grad_norm": 0.6350043416023254,
+      "learning_rate": 0.0001,
+      "loss": 1.2099,
+      "step": 13746
+    },
+    {
+      "epoch": 1.5791166503934293,
+      "grad_norm": 0.7006800770759583,
+      "learning_rate": 0.0001,
+      "loss": 1.6148,
+      "step": 13747
+    },
+    {
+      "epoch": 1.5792315203032565,
+      "grad_norm": 0.7125732898712158,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 13748
+    },
+    {
+      "epoch": 1.5793463902130838,
+      "grad_norm": 0.595615565776825,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 13749
+    },
+    {
+      "epoch": 1.5794612601229108,
+      "grad_norm": 0.5776601433753967,
+      "learning_rate": 0.0001,
+      "loss": 1.2368,
+      "step": 13750
+    },
+    {
+      "epoch": 1.5795761300327378,
+      "grad_norm": 0.6074332594871521,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 13751
+    },
+    {
+      "epoch": 1.579690999942565,
+      "grad_norm": 0.5884713530540466,
+      "learning_rate": 0.0001,
+      "loss": 1.2479,
+      "step": 13752
+    },
+    {
+      "epoch": 1.5798058698523922,
+      "grad_norm": 0.5512914657592773,
+      "learning_rate": 0.0001,
+      "loss": 1.3712,
+      "step": 13753
+    },
+    {
+      "epoch": 1.5799207397622193,
+      "grad_norm": 0.5844442844390869,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 13754
+    },
+    {
+      "epoch": 1.5800356096720463,
+      "grad_norm": 0.6031823754310608,
+      "learning_rate": 0.0001,
+      "loss": 1.295,
+      "step": 13755
+    },
+    {
+      "epoch": 1.5801504795818735,
+      "grad_norm": 0.6166372895240784,
+      "learning_rate": 0.0001,
+      "loss": 1.3007,
+      "step": 13756
+    },
+    {
+      "epoch": 1.5802653494917007,
+      "grad_norm": 0.5954715013504028,
+      "learning_rate": 0.0001,
+      "loss": 1.4218,
+      "step": 13757
+    },
+    {
+      "epoch": 1.5803802194015277,
+      "grad_norm": 0.6521285176277161,
+      "learning_rate": 0.0001,
+      "loss": 1.4628,
+      "step": 13758
+    },
+    {
+      "epoch": 1.5804950893113547,
+      "grad_norm": 0.6663768887519836,
+      "learning_rate": 0.0001,
+      "loss": 1.6063,
+      "step": 13759
+    },
+    {
+      "epoch": 1.580609959221182,
+      "grad_norm": 0.577915370464325,
+      "learning_rate": 0.0001,
+      "loss": 1.4309,
+      "step": 13760
+    },
+    {
+      "epoch": 1.5807248291310092,
+      "grad_norm": 0.5663382411003113,
+      "learning_rate": 0.0001,
+      "loss": 1.4019,
+      "step": 13761
+    },
+    {
+      "epoch": 1.5808396990408362,
+      "grad_norm": 0.5602874159812927,
+      "learning_rate": 0.0001,
+      "loss": 1.4436,
+      "step": 13762
+    },
+    {
+      "epoch": 1.5809545689506632,
+      "grad_norm": 0.6626793146133423,
+      "learning_rate": 0.0001,
+      "loss": 1.4416,
+      "step": 13763
+    },
+    {
+      "epoch": 1.5810694388604905,
+      "grad_norm": 0.6153534650802612,
+      "learning_rate": 0.0001,
+      "loss": 1.3675,
+      "step": 13764
+    },
+    {
+      "epoch": 1.5811843087703177,
+      "grad_norm": 0.5600067377090454,
+      "learning_rate": 0.0001,
+      "loss": 1.2818,
+      "step": 13765
+    },
+    {
+      "epoch": 1.5812991786801447,
+      "grad_norm": 0.6040723919868469,
+      "learning_rate": 0.0001,
+      "loss": 1.4141,
+      "step": 13766
+    },
+    {
+      "epoch": 1.5814140485899717,
+      "grad_norm": 0.5831596255302429,
+      "learning_rate": 0.0001,
+      "loss": 1.412,
+      "step": 13767
+    },
+    {
+      "epoch": 1.581528918499799,
+      "grad_norm": 0.5813616514205933,
+      "learning_rate": 0.0001,
+      "loss": 1.3028,
+      "step": 13768
+    },
+    {
+      "epoch": 1.5816437884096262,
+      "grad_norm": 0.5949644446372986,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 13769
+    },
+    {
+      "epoch": 1.5817586583194532,
+      "grad_norm": 0.6183162927627563,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 13770
+    },
+    {
+      "epoch": 1.5818735282292802,
+      "grad_norm": 0.5976161360740662,
+      "learning_rate": 0.0001,
+      "loss": 1.5491,
+      "step": 13771
+    },
+    {
+      "epoch": 1.5819883981391074,
+      "grad_norm": 0.6509799957275391,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 13772
+    },
+    {
+      "epoch": 1.5821032680489346,
+      "grad_norm": 0.6055750250816345,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 13773
+    },
+    {
+      "epoch": 1.5822181379587617,
+      "grad_norm": 0.6353389024734497,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 13774
+    },
+    {
+      "epoch": 1.5823330078685887,
+      "grad_norm": 0.645206093788147,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 13775
+    },
+    {
+      "epoch": 1.582447877778416,
+      "grad_norm": 0.6120302081108093,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 13776
+    },
+    {
+      "epoch": 1.5825627476882431,
+      "grad_norm": 0.5803287625312805,
+      "learning_rate": 0.0001,
+      "loss": 1.2655,
+      "step": 13777
+    },
+    {
+      "epoch": 1.5826776175980704,
+      "grad_norm": 0.5672709345817566,
+      "learning_rate": 0.0001,
+      "loss": 1.5853,
+      "step": 13778
+    },
+    {
+      "epoch": 1.5827924875078974,
+      "grad_norm": 0.593664288520813,
+      "learning_rate": 0.0001,
+      "loss": 1.4554,
+      "step": 13779
+    },
+    {
+      "epoch": 1.5829073574177244,
+      "grad_norm": 0.6163424849510193,
+      "learning_rate": 0.0001,
+      "loss": 1.4778,
+      "step": 13780
+    },
+    {
+      "epoch": 1.5830222273275516,
+      "grad_norm": 0.6263569593429565,
+      "learning_rate": 0.0001,
+      "loss": 1.6402,
+      "step": 13781
+    },
+    {
+      "epoch": 1.5831370972373788,
+      "grad_norm": 0.6205965876579285,
+      "learning_rate": 0.0001,
+      "loss": 1.6084,
+      "step": 13782
+    },
+    {
+      "epoch": 1.5832519671472058,
+      "grad_norm": 0.6791396141052246,
+      "learning_rate": 0.0001,
+      "loss": 1.7392,
+      "step": 13783
+    },
+    {
+      "epoch": 1.5833668370570328,
+      "grad_norm": 0.6714447736740112,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 13784
+    },
+    {
+      "epoch": 1.58348170696686,
+      "grad_norm": 0.5691717267036438,
+      "learning_rate": 0.0001,
+      "loss": 1.4059,
+      "step": 13785
+    },
+    {
+      "epoch": 1.5835965768766873,
+      "grad_norm": 0.636568009853363,
+      "learning_rate": 0.0001,
+      "loss": 1.4756,
+      "step": 13786
+    },
+    {
+      "epoch": 1.5837114467865143,
+      "grad_norm": 0.6351962685585022,
+      "learning_rate": 0.0001,
+      "loss": 1.4316,
+      "step": 13787
+    },
+    {
+      "epoch": 1.5838263166963413,
+      "grad_norm": 0.6552459001541138,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 13788
+    },
+    {
+      "epoch": 1.5839411866061686,
+      "grad_norm": 0.6077340841293335,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 13789
+    },
+    {
+      "epoch": 1.5840560565159958,
+      "grad_norm": 0.5727121829986572,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 13790
+    },
+    {
+      "epoch": 1.5841709264258228,
+      "grad_norm": 0.6179769039154053,
+      "learning_rate": 0.0001,
+      "loss": 1.3921,
+      "step": 13791
+    },
+    {
+      "epoch": 1.5842857963356498,
+      "grad_norm": 0.5873903036117554,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 13792
+    },
+    {
+      "epoch": 1.584400666245477,
+      "grad_norm": 0.6171455979347229,
+      "learning_rate": 0.0001,
+      "loss": 1.3184,
+      "step": 13793
+    },
+    {
+      "epoch": 1.5845155361553043,
+      "grad_norm": 0.7028810977935791,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 13794
+    },
+    {
+      "epoch": 1.5846304060651313,
+      "grad_norm": 0.5513997077941895,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 13795
+    },
+    {
+      "epoch": 1.5847452759749583,
+      "grad_norm": 0.5477231740951538,
+      "learning_rate": 0.0001,
+      "loss": 1.3617,
+      "step": 13796
+    },
+    {
+      "epoch": 1.5848601458847855,
+      "grad_norm": 0.6108036041259766,
+      "learning_rate": 0.0001,
+      "loss": 1.3043,
+      "step": 13797
+    },
+    {
+      "epoch": 1.5849750157946128,
+      "grad_norm": 0.559474527835846,
+      "learning_rate": 0.0001,
+      "loss": 1.2964,
+      "step": 13798
+    },
+    {
+      "epoch": 1.5850898857044398,
+      "grad_norm": 0.6026455760002136,
+      "learning_rate": 0.0001,
+      "loss": 1.4424,
+      "step": 13799
+    },
+    {
+      "epoch": 1.5852047556142668,
+      "grad_norm": 0.7762643098831177,
+      "learning_rate": 0.0001,
+      "loss": 1.7879,
+      "step": 13800
+    },
+    {
+      "epoch": 1.585319625524094,
+      "grad_norm": 0.6184485554695129,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 13801
+    },
+    {
+      "epoch": 1.5854344954339212,
+      "grad_norm": 0.6571382880210876,
+      "learning_rate": 0.0001,
+      "loss": 1.6504,
+      "step": 13802
+    },
+    {
+      "epoch": 1.5855493653437482,
+      "grad_norm": 0.6091337203979492,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 13803
+    },
+    {
+      "epoch": 1.5856642352535752,
+      "grad_norm": 0.6771681308746338,
+      "learning_rate": 0.0001,
+      "loss": 1.3213,
+      "step": 13804
+    },
+    {
+      "epoch": 1.5857791051634025,
+      "grad_norm": 0.6501324772834778,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 13805
+    },
+    {
+      "epoch": 1.5858939750732297,
+      "grad_norm": 0.5912746787071228,
+      "learning_rate": 0.0001,
+      "loss": 1.3764,
+      "step": 13806
+    },
+    {
+      "epoch": 1.5860088449830567,
+      "grad_norm": 0.6037330627441406,
+      "learning_rate": 0.0001,
+      "loss": 1.4944,
+      "step": 13807
+    },
+    {
+      "epoch": 1.5861237148928837,
+      "grad_norm": 0.5597751140594482,
+      "learning_rate": 0.0001,
+      "loss": 1.3488,
+      "step": 13808
+    },
+    {
+      "epoch": 1.586238584802711,
+      "grad_norm": 0.5525902509689331,
+      "learning_rate": 0.0001,
+      "loss": 1.4154,
+      "step": 13809
+    },
+    {
+      "epoch": 1.5863534547125382,
+      "grad_norm": 0.6479353308677673,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 13810
+    },
+    {
+      "epoch": 1.5864683246223652,
+      "grad_norm": 0.6290416121482849,
+      "learning_rate": 0.0001,
+      "loss": 1.4498,
+      "step": 13811
+    },
+    {
+      "epoch": 1.5865831945321922,
+      "grad_norm": 0.6369213461875916,
+      "learning_rate": 0.0001,
+      "loss": 1.3789,
+      "step": 13812
+    },
+    {
+      "epoch": 1.5866980644420194,
+      "grad_norm": 0.6600077152252197,
+      "learning_rate": 0.0001,
+      "loss": 1.3561,
+      "step": 13813
+    },
+    {
+      "epoch": 1.5868129343518467,
+      "grad_norm": 0.6449211239814758,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 13814
+    },
+    {
+      "epoch": 1.5869278042616737,
+      "grad_norm": 0.596219003200531,
+      "learning_rate": 0.0001,
+      "loss": 1.4306,
+      "step": 13815
+    },
+    {
+      "epoch": 1.5870426741715007,
+      "grad_norm": 0.6226633191108704,
+      "learning_rate": 0.0001,
+      "loss": 1.4792,
+      "step": 13816
+    },
+    {
+      "epoch": 1.587157544081328,
+      "grad_norm": 0.614152193069458,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 13817
+    },
+    {
+      "epoch": 1.5872724139911552,
+      "grad_norm": 0.6192676424980164,
+      "learning_rate": 0.0001,
+      "loss": 1.3306,
+      "step": 13818
+    },
+    {
+      "epoch": 1.5873872839009822,
+      "grad_norm": 0.6133370995521545,
+      "learning_rate": 0.0001,
+      "loss": 1.2503,
+      "step": 13819
+    },
+    {
+      "epoch": 1.5875021538108092,
+      "grad_norm": 0.6042430996894836,
+      "learning_rate": 0.0001,
+      "loss": 1.4175,
+      "step": 13820
+    },
+    {
+      "epoch": 1.5876170237206364,
+      "grad_norm": 0.7672133445739746,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 13821
+    },
+    {
+      "epoch": 1.5877318936304636,
+      "grad_norm": 0.662254273891449,
+      "learning_rate": 0.0001,
+      "loss": 1.4117,
+      "step": 13822
+    },
+    {
+      "epoch": 1.5878467635402906,
+      "grad_norm": 0.6693131327629089,
+      "learning_rate": 0.0001,
+      "loss": 1.5383,
+      "step": 13823
+    },
+    {
+      "epoch": 1.5879616334501176,
+      "grad_norm": 0.636415421962738,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 13824
+    },
+    {
+      "epoch": 1.5880765033599449,
+      "grad_norm": 0.579950213432312,
+      "learning_rate": 0.0001,
+      "loss": 1.4417,
+      "step": 13825
+    },
+    {
+      "epoch": 1.588191373269772,
+      "grad_norm": 0.6041045188903809,
+      "learning_rate": 0.0001,
+      "loss": 1.3729,
+      "step": 13826
+    },
+    {
+      "epoch": 1.5883062431795991,
+      "grad_norm": 0.638687252998352,
+      "learning_rate": 0.0001,
+      "loss": 1.4003,
+      "step": 13827
+    },
+    {
+      "epoch": 1.5884211130894261,
+      "grad_norm": 0.5657978057861328,
+      "learning_rate": 0.0001,
+      "loss": 1.3076,
+      "step": 13828
+    },
+    {
+      "epoch": 1.5885359829992534,
+      "grad_norm": 0.5743250250816345,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 13829
+    },
+    {
+      "epoch": 1.5886508529090806,
+      "grad_norm": 0.6413049101829529,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 13830
+    },
+    {
+      "epoch": 1.5887657228189076,
+      "grad_norm": 0.5768105983734131,
+      "learning_rate": 0.0001,
+      "loss": 1.4486,
+      "step": 13831
+    },
+    {
+      "epoch": 1.5888805927287346,
+      "grad_norm": 0.6100025773048401,
+      "learning_rate": 0.0001,
+      "loss": 1.5592,
+      "step": 13832
+    },
+    {
+      "epoch": 1.5889954626385618,
+      "grad_norm": 0.6086409091949463,
+      "learning_rate": 0.0001,
+      "loss": 1.3241,
+      "step": 13833
+    },
+    {
+      "epoch": 1.589110332548389,
+      "grad_norm": 0.6724202632904053,
+      "learning_rate": 0.0001,
+      "loss": 1.4075,
+      "step": 13834
+    },
+    {
+      "epoch": 1.589225202458216,
+      "grad_norm": 0.6264391541481018,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 13835
+    },
+    {
+      "epoch": 1.589340072368043,
+      "grad_norm": 0.6160328388214111,
+      "learning_rate": 0.0001,
+      "loss": 1.4792,
+      "step": 13836
+    },
+    {
+      "epoch": 1.5894549422778703,
+      "grad_norm": 0.6358749270439148,
+      "learning_rate": 0.0001,
+      "loss": 1.465,
+      "step": 13837
+    },
+    {
+      "epoch": 1.5895698121876976,
+      "grad_norm": 0.6735308766365051,
+      "learning_rate": 0.0001,
+      "loss": 1.4201,
+      "step": 13838
+    },
+    {
+      "epoch": 1.5896846820975246,
+      "grad_norm": 0.6253379583358765,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 13839
+    },
+    {
+      "epoch": 1.5897995520073516,
+      "grad_norm": 0.6210620999336243,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 13840
+    },
+    {
+      "epoch": 1.5899144219171788,
+      "grad_norm": 0.6258774995803833,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 13841
+    },
+    {
+      "epoch": 1.590029291827006,
+      "grad_norm": 0.6079967021942139,
+      "learning_rate": 0.0001,
+      "loss": 1.4229,
+      "step": 13842
+    },
+    {
+      "epoch": 1.590144161736833,
+      "grad_norm": 0.5905248522758484,
+      "learning_rate": 0.0001,
+      "loss": 1.436,
+      "step": 13843
+    },
+    {
+      "epoch": 1.59025903164666,
+      "grad_norm": 0.623208224773407,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 13844
+    },
+    {
+      "epoch": 1.5903739015564873,
+      "grad_norm": 0.6524278521537781,
+      "learning_rate": 0.0001,
+      "loss": 1.526,
+      "step": 13845
+    },
+    {
+      "epoch": 1.5904887714663145,
+      "grad_norm": 0.6335569024085999,
+      "learning_rate": 0.0001,
+      "loss": 1.4586,
+      "step": 13846
+    },
+    {
+      "epoch": 1.5906036413761415,
+      "grad_norm": 0.5839788317680359,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 13847
+    },
+    {
+      "epoch": 1.5907185112859685,
+      "grad_norm": 0.6016788482666016,
+      "learning_rate": 0.0001,
+      "loss": 1.6573,
+      "step": 13848
+    },
+    {
+      "epoch": 1.5908333811957958,
+      "grad_norm": 0.5814146399497986,
+      "learning_rate": 0.0001,
+      "loss": 1.5764,
+      "step": 13849
+    },
+    {
+      "epoch": 1.590948251105623,
+      "grad_norm": 0.6370062828063965,
+      "learning_rate": 0.0001,
+      "loss": 1.6166,
+      "step": 13850
+    },
+    {
+      "epoch": 1.59106312101545,
+      "grad_norm": 0.5697435736656189,
+      "learning_rate": 0.0001,
+      "loss": 1.5031,
+      "step": 13851
+    },
+    {
+      "epoch": 1.591177990925277,
+      "grad_norm": 0.5778668522834778,
+      "learning_rate": 0.0001,
+      "loss": 1.2866,
+      "step": 13852
+    },
+    {
+      "epoch": 1.5912928608351042,
+      "grad_norm": 0.5944350957870483,
+      "learning_rate": 0.0001,
+      "loss": 1.4024,
+      "step": 13853
+    },
+    {
+      "epoch": 1.5914077307449315,
+      "grad_norm": 0.713706374168396,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 13854
+    },
+    {
+      "epoch": 1.5915226006547585,
+      "grad_norm": 0.5639752149581909,
+      "learning_rate": 0.0001,
+      "loss": 1.4652,
+      "step": 13855
+    },
+    {
+      "epoch": 1.5916374705645855,
+      "grad_norm": 0.6923316717147827,
+      "learning_rate": 0.0001,
+      "loss": 1.529,
+      "step": 13856
+    },
+    {
+      "epoch": 1.5917523404744127,
+      "grad_norm": 0.5926253795623779,
+      "learning_rate": 0.0001,
+      "loss": 1.4854,
+      "step": 13857
+    },
+    {
+      "epoch": 1.59186721038424,
+      "grad_norm": 0.6101776361465454,
+      "learning_rate": 0.0001,
+      "loss": 1.4077,
+      "step": 13858
+    },
+    {
+      "epoch": 1.591982080294067,
+      "grad_norm": 0.6200652122497559,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 13859
+    },
+    {
+      "epoch": 1.592096950203894,
+      "grad_norm": 0.7140838503837585,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 13860
+    },
+    {
+      "epoch": 1.5922118201137212,
+      "grad_norm": 0.6783391237258911,
+      "learning_rate": 0.0001,
+      "loss": 1.6397,
+      "step": 13861
+    },
+    {
+      "epoch": 1.5923266900235484,
+      "grad_norm": 0.6628894805908203,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 13862
+    },
+    {
+      "epoch": 1.5924415599333754,
+      "grad_norm": 0.59294593334198,
+      "learning_rate": 0.0001,
+      "loss": 1.3666,
+      "step": 13863
+    },
+    {
+      "epoch": 1.5925564298432024,
+      "grad_norm": 0.5863275527954102,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 13864
+    },
+    {
+      "epoch": 1.5926712997530297,
+      "grad_norm": 0.5980870723724365,
+      "learning_rate": 0.0001,
+      "loss": 1.4679,
+      "step": 13865
+    },
+    {
+      "epoch": 1.592786169662857,
+      "grad_norm": 0.6048779487609863,
+      "learning_rate": 0.0001,
+      "loss": 1.4745,
+      "step": 13866
+    },
+    {
+      "epoch": 1.592901039572684,
+      "grad_norm": 0.5743735432624817,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 13867
+    },
+    {
+      "epoch": 1.593015909482511,
+      "grad_norm": 0.5809198021888733,
+      "learning_rate": 0.0001,
+      "loss": 1.4706,
+      "step": 13868
+    },
+    {
+      "epoch": 1.5931307793923382,
+      "grad_norm": 0.6342722773551941,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 13869
+    },
+    {
+      "epoch": 1.5932456493021654,
+      "grad_norm": 0.6230271458625793,
+      "learning_rate": 0.0001,
+      "loss": 1.3428,
+      "step": 13870
+    },
+    {
+      "epoch": 1.5933605192119924,
+      "grad_norm": 0.6740293502807617,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 13871
+    },
+    {
+      "epoch": 1.5934753891218194,
+      "grad_norm": 0.7214620113372803,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 13872
+    },
+    {
+      "epoch": 1.5935902590316466,
+      "grad_norm": 0.6256888508796692,
+      "learning_rate": 0.0001,
+      "loss": 1.7852,
+      "step": 13873
+    },
+    {
+      "epoch": 1.5937051289414739,
+      "grad_norm": 0.5903255939483643,
+      "learning_rate": 0.0001,
+      "loss": 1.4396,
+      "step": 13874
+    },
+    {
+      "epoch": 1.5938199988513009,
+      "grad_norm": 0.572929859161377,
+      "learning_rate": 0.0001,
+      "loss": 1.4534,
+      "step": 13875
+    },
+    {
+      "epoch": 1.5939348687611279,
+      "grad_norm": 0.6389855146408081,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 13876
+    },
+    {
+      "epoch": 1.5940497386709551,
+      "grad_norm": 0.614324688911438,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 13877
+    },
+    {
+      "epoch": 1.5941646085807823,
+      "grad_norm": 0.6262052655220032,
+      "learning_rate": 0.0001,
+      "loss": 1.48,
+      "step": 13878
+    },
+    {
+      "epoch": 1.5942794784906094,
+      "grad_norm": 0.6147677898406982,
+      "learning_rate": 0.0001,
+      "loss": 1.4761,
+      "step": 13879
+    },
+    {
+      "epoch": 1.5943943484004364,
+      "grad_norm": 0.635215699672699,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 13880
+    },
+    {
+      "epoch": 1.5945092183102636,
+      "grad_norm": 0.5967612266540527,
+      "learning_rate": 0.0001,
+      "loss": 1.4983,
+      "step": 13881
+    },
+    {
+      "epoch": 1.5946240882200908,
+      "grad_norm": 0.5582137703895569,
+      "learning_rate": 0.0001,
+      "loss": 1.374,
+      "step": 13882
+    },
+    {
+      "epoch": 1.5947389581299178,
+      "grad_norm": 0.6301720142364502,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 13883
+    },
+    {
+      "epoch": 1.5948538280397448,
+      "grad_norm": 0.5917264819145203,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 13884
+    },
+    {
+      "epoch": 1.594968697949572,
+      "grad_norm": 0.6191073656082153,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 13885
+    },
+    {
+      "epoch": 1.5950835678593993,
+      "grad_norm": 0.5992122292518616,
+      "learning_rate": 0.0001,
+      "loss": 1.3994,
+      "step": 13886
+    },
+    {
+      "epoch": 1.5951984377692263,
+      "grad_norm": 0.6241499781608582,
+      "learning_rate": 0.0001,
+      "loss": 1.5801,
+      "step": 13887
+    },
+    {
+      "epoch": 1.5953133076790533,
+      "grad_norm": 0.5631673336029053,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 13888
+    },
+    {
+      "epoch": 1.5954281775888806,
+      "grad_norm": 0.6660466194152832,
+      "learning_rate": 0.0001,
+      "loss": 1.7669,
+      "step": 13889
+    },
+    {
+      "epoch": 1.5955430474987078,
+      "grad_norm": 0.676660418510437,
+      "learning_rate": 0.0001,
+      "loss": 1.1202,
+      "step": 13890
+    },
+    {
+      "epoch": 1.5956579174085348,
+      "grad_norm": 0.6641074419021606,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 13891
+    },
+    {
+      "epoch": 1.5957727873183618,
+      "grad_norm": 0.6057578325271606,
+      "learning_rate": 0.0001,
+      "loss": 1.4497,
+      "step": 13892
+    },
+    {
+      "epoch": 1.595887657228189,
+      "grad_norm": 0.5668378472328186,
+      "learning_rate": 0.0001,
+      "loss": 1.3555,
+      "step": 13893
+    },
+    {
+      "epoch": 1.5960025271380163,
+      "grad_norm": 0.6328227519989014,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 13894
+    },
+    {
+      "epoch": 1.5961173970478433,
+      "grad_norm": 0.605475664138794,
+      "learning_rate": 0.0001,
+      "loss": 1.4586,
+      "step": 13895
+    },
+    {
+      "epoch": 1.5962322669576703,
+      "grad_norm": 0.5906625986099243,
+      "learning_rate": 0.0001,
+      "loss": 1.4653,
+      "step": 13896
+    },
+    {
+      "epoch": 1.5963471368674975,
+      "grad_norm": 0.613008439540863,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 13897
+    },
+    {
+      "epoch": 1.5964620067773247,
+      "grad_norm": 0.6402007937431335,
+      "learning_rate": 0.0001,
+      "loss": 1.3678,
+      "step": 13898
+    },
+    {
+      "epoch": 1.5965768766871518,
+      "grad_norm": 0.6814921498298645,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 13899
+    },
+    {
+      "epoch": 1.5966917465969788,
+      "grad_norm": 0.5518124103546143,
+      "learning_rate": 0.0001,
+      "loss": 1.3886,
+      "step": 13900
+    },
+    {
+      "epoch": 1.596806616506806,
+      "grad_norm": 0.5962539315223694,
+      "learning_rate": 0.0001,
+      "loss": 1.412,
+      "step": 13901
+    },
+    {
+      "epoch": 1.5969214864166332,
+      "grad_norm": 0.5778492093086243,
+      "learning_rate": 0.0001,
+      "loss": 1.3721,
+      "step": 13902
+    },
+    {
+      "epoch": 1.5970363563264602,
+      "grad_norm": 0.6201198101043701,
+      "learning_rate": 0.0001,
+      "loss": 1.3933,
+      "step": 13903
+    },
+    {
+      "epoch": 1.5971512262362872,
+      "grad_norm": 0.6029123663902283,
+      "learning_rate": 0.0001,
+      "loss": 1.3812,
+      "step": 13904
+    },
+    {
+      "epoch": 1.5972660961461145,
+      "grad_norm": 0.6379988789558411,
+      "learning_rate": 0.0001,
+      "loss": 1.2687,
+      "step": 13905
+    },
+    {
+      "epoch": 1.5973809660559417,
+      "grad_norm": 0.5439411997795105,
+      "learning_rate": 0.0001,
+      "loss": 1.3471,
+      "step": 13906
+    },
+    {
+      "epoch": 1.5974958359657687,
+      "grad_norm": 0.6145894527435303,
+      "learning_rate": 0.0001,
+      "loss": 1.3119,
+      "step": 13907
+    },
+    {
+      "epoch": 1.5976107058755957,
+      "grad_norm": 0.6268411874771118,
+      "learning_rate": 0.0001,
+      "loss": 1.4461,
+      "step": 13908
+    },
+    {
+      "epoch": 1.597725575785423,
+      "grad_norm": 0.5672012567520142,
+      "learning_rate": 0.0001,
+      "loss": 1.4152,
+      "step": 13909
+    },
+    {
+      "epoch": 1.5978404456952502,
+      "grad_norm": 0.5719462633132935,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 13910
+    },
+    {
+      "epoch": 1.5979553156050772,
+      "grad_norm": 0.564241349697113,
+      "learning_rate": 0.0001,
+      "loss": 1.2821,
+      "step": 13911
+    },
+    {
+      "epoch": 1.5980701855149042,
+      "grad_norm": 0.6079102158546448,
+      "learning_rate": 0.0001,
+      "loss": 1.3764,
+      "step": 13912
+    },
+    {
+      "epoch": 1.5981850554247314,
+      "grad_norm": 0.605634868144989,
+      "learning_rate": 0.0001,
+      "loss": 1.4621,
+      "step": 13913
+    },
+    {
+      "epoch": 1.5982999253345587,
+      "grad_norm": 0.599432110786438,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 13914
+    },
+    {
+      "epoch": 1.5984147952443857,
+      "grad_norm": 0.6091422438621521,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 13915
+    },
+    {
+      "epoch": 1.598529665154213,
+      "grad_norm": 0.6520305275917053,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 13916
+    },
+    {
+      "epoch": 1.59864453506404,
+      "grad_norm": 0.6225545406341553,
+      "learning_rate": 0.0001,
+      "loss": 1.4667,
+      "step": 13917
+    },
+    {
+      "epoch": 1.5987594049738671,
+      "grad_norm": 0.5742337107658386,
+      "learning_rate": 0.0001,
+      "loss": 1.3543,
+      "step": 13918
+    },
+    {
+      "epoch": 1.5988742748836944,
+      "grad_norm": 0.6456047892570496,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 13919
+    },
+    {
+      "epoch": 1.5989891447935214,
+      "grad_norm": 0.637758195400238,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 13920
+    },
+    {
+      "epoch": 1.5991040147033484,
+      "grad_norm": 0.6224943399429321,
+      "learning_rate": 0.0001,
+      "loss": 1.2575,
+      "step": 13921
+    },
+    {
+      "epoch": 1.5992188846131756,
+      "grad_norm": 0.6721301078796387,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 13922
+    },
+    {
+      "epoch": 1.5993337545230029,
+      "grad_norm": 0.5861270427703857,
+      "learning_rate": 0.0001,
+      "loss": 1.2367,
+      "step": 13923
+    },
+    {
+      "epoch": 1.5994486244328299,
+      "grad_norm": 0.6392655968666077,
+      "learning_rate": 0.0001,
+      "loss": 1.4703,
+      "step": 13924
+    },
+    {
+      "epoch": 1.5995634943426569,
+      "grad_norm": 0.6457921862602234,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 13925
+    },
+    {
+      "epoch": 1.599678364252484,
+      "grad_norm": 0.6379731297492981,
+      "learning_rate": 0.0001,
+      "loss": 1.3977,
+      "step": 13926
+    },
+    {
+      "epoch": 1.5997932341623113,
+      "grad_norm": 0.5802931785583496,
+      "learning_rate": 0.0001,
+      "loss": 1.3204,
+      "step": 13927
+    },
+    {
+      "epoch": 1.5999081040721383,
+      "grad_norm": 0.6664165258407593,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 13928
+    },
+    {
+      "epoch": 1.6000229739819654,
+      "grad_norm": 0.6241664290428162,
+      "learning_rate": 0.0001,
+      "loss": 1.6584,
+      "step": 13929
+    },
+    {
+      "epoch": 1.6001378438917926,
+      "grad_norm": 0.6621495485305786,
+      "learning_rate": 0.0001,
+      "loss": 1.3809,
+      "step": 13930
+    },
+    {
+      "epoch": 1.6002527138016198,
+      "grad_norm": 0.6004970073699951,
+      "learning_rate": 0.0001,
+      "loss": 1.4092,
+      "step": 13931
+    },
+    {
+      "epoch": 1.6003675837114468,
+      "grad_norm": 0.6193183064460754,
+      "learning_rate": 0.0001,
+      "loss": 1.4335,
+      "step": 13932
+    },
+    {
+      "epoch": 1.6004824536212738,
+      "grad_norm": 0.6352545619010925,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 13933
+    },
+    {
+      "epoch": 1.600597323531101,
+      "grad_norm": 0.6299646496772766,
+      "learning_rate": 0.0001,
+      "loss": 1.4958,
+      "step": 13934
+    },
+    {
+      "epoch": 1.6007121934409283,
+      "grad_norm": 0.5891308784484863,
+      "learning_rate": 0.0001,
+      "loss": 1.3664,
+      "step": 13935
+    },
+    {
+      "epoch": 1.6008270633507553,
+      "grad_norm": 0.655714750289917,
+      "learning_rate": 0.0001,
+      "loss": 1.3667,
+      "step": 13936
+    },
+    {
+      "epoch": 1.6009419332605823,
+      "grad_norm": 0.6111890077590942,
+      "learning_rate": 0.0001,
+      "loss": 1.4353,
+      "step": 13937
+    },
+    {
+      "epoch": 1.6010568031704095,
+      "grad_norm": 0.5762326717376709,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 13938
+    },
+    {
+      "epoch": 1.6011716730802368,
+      "grad_norm": 0.5921722650527954,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 13939
+    },
+    {
+      "epoch": 1.6012865429900638,
+      "grad_norm": 0.6111255884170532,
+      "learning_rate": 0.0001,
+      "loss": 1.722,
+      "step": 13940
+    },
+    {
+      "epoch": 1.6014014128998908,
+      "grad_norm": 0.5806509256362915,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 13941
+    },
+    {
+      "epoch": 1.601516282809718,
+      "grad_norm": 0.5703766942024231,
+      "learning_rate": 0.0001,
+      "loss": 1.303,
+      "step": 13942
+    },
+    {
+      "epoch": 1.6016311527195453,
+      "grad_norm": 0.5971461534500122,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 13943
+    },
+    {
+      "epoch": 1.6017460226293723,
+      "grad_norm": 0.6908469796180725,
+      "learning_rate": 0.0001,
+      "loss": 1.4302,
+      "step": 13944
+    },
+    {
+      "epoch": 1.6018608925391993,
+      "grad_norm": 0.5854578614234924,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 13945
+    },
+    {
+      "epoch": 1.6019757624490265,
+      "grad_norm": 0.6084689497947693,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 13946
+    },
+    {
+      "epoch": 1.6020906323588537,
+      "grad_norm": 0.6091856956481934,
+      "learning_rate": 0.0001,
+      "loss": 1.3299,
+      "step": 13947
+    },
+    {
+      "epoch": 1.6022055022686807,
+      "grad_norm": 0.6106900572776794,
+      "learning_rate": 0.0001,
+      "loss": 1.4301,
+      "step": 13948
+    },
+    {
+      "epoch": 1.6023203721785078,
+      "grad_norm": 0.6288970708847046,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 13949
+    },
+    {
+      "epoch": 1.602435242088335,
+      "grad_norm": 0.6874493956565857,
+      "learning_rate": 0.0001,
+      "loss": 1.4993,
+      "step": 13950
+    },
+    {
+      "epoch": 1.6025501119981622,
+      "grad_norm": 0.5642193555831909,
+      "learning_rate": 0.0001,
+      "loss": 1.2891,
+      "step": 13951
+    },
+    {
+      "epoch": 1.6026649819079892,
+      "grad_norm": 0.6149432063102722,
+      "learning_rate": 0.0001,
+      "loss": 1.346,
+      "step": 13952
+    },
+    {
+      "epoch": 1.6027798518178162,
+      "grad_norm": 0.6034269332885742,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 13953
+    },
+    {
+      "epoch": 1.6028947217276435,
+      "grad_norm": 0.5566948056221008,
+      "learning_rate": 0.0001,
+      "loss": 1.3367,
+      "step": 13954
+    },
+    {
+      "epoch": 1.6030095916374707,
+      "grad_norm": 0.6086257100105286,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 13955
+    },
+    {
+      "epoch": 1.6031244615472977,
+      "grad_norm": 0.6284582614898682,
+      "learning_rate": 0.0001,
+      "loss": 1.4821,
+      "step": 13956
+    },
+    {
+      "epoch": 1.6032393314571247,
+      "grad_norm": 0.6378557085990906,
+      "learning_rate": 0.0001,
+      "loss": 1.4118,
+      "step": 13957
+    },
+    {
+      "epoch": 1.603354201366952,
+      "grad_norm": 0.6031190156936646,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 13958
+    },
+    {
+      "epoch": 1.6034690712767792,
+      "grad_norm": 0.6202352046966553,
+      "learning_rate": 0.0001,
+      "loss": 1.3672,
+      "step": 13959
+    },
+    {
+      "epoch": 1.6035839411866062,
+      "grad_norm": 0.6928132772445679,
+      "learning_rate": 0.0001,
+      "loss": 1.2899,
+      "step": 13960
+    },
+    {
+      "epoch": 1.6036988110964332,
+      "grad_norm": 0.6329846978187561,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 13961
+    },
+    {
+      "epoch": 1.6038136810062604,
+      "grad_norm": 0.5797714591026306,
+      "learning_rate": 0.0001,
+      "loss": 1.3294,
+      "step": 13962
+    },
+    {
+      "epoch": 1.6039285509160877,
+      "grad_norm": 0.6104891896247864,
+      "learning_rate": 0.0001,
+      "loss": 1.3985,
+      "step": 13963
+    },
+    {
+      "epoch": 1.6040434208259147,
+      "grad_norm": 0.6021250486373901,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 13964
+    },
+    {
+      "epoch": 1.6041582907357417,
+      "grad_norm": 0.6344681978225708,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 13965
+    },
+    {
+      "epoch": 1.604273160645569,
+      "grad_norm": 0.6106261014938354,
+      "learning_rate": 0.0001,
+      "loss": 1.4263,
+      "step": 13966
+    },
+    {
+      "epoch": 1.6043880305553961,
+      "grad_norm": 0.598883330821991,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 13967
+    },
+    {
+      "epoch": 1.6045029004652231,
+      "grad_norm": 0.6441910862922668,
+      "learning_rate": 0.0001,
+      "loss": 1.3875,
+      "step": 13968
+    },
+    {
+      "epoch": 1.6046177703750502,
+      "grad_norm": 0.6011143326759338,
+      "learning_rate": 0.0001,
+      "loss": 1.4894,
+      "step": 13969
+    },
+    {
+      "epoch": 1.6047326402848774,
+      "grad_norm": 0.6068170070648193,
+      "learning_rate": 0.0001,
+      "loss": 1.2947,
+      "step": 13970
+    },
+    {
+      "epoch": 1.6048475101947046,
+      "grad_norm": 0.6693038940429688,
+      "learning_rate": 0.0001,
+      "loss": 1.4515,
+      "step": 13971
+    },
+    {
+      "epoch": 1.6049623801045316,
+      "grad_norm": 0.6221956610679626,
+      "learning_rate": 0.0001,
+      "loss": 1.462,
+      "step": 13972
+    },
+    {
+      "epoch": 1.6050772500143586,
+      "grad_norm": 0.6345918774604797,
+      "learning_rate": 0.0001,
+      "loss": 1.5339,
+      "step": 13973
+    },
+    {
+      "epoch": 1.6051921199241859,
+      "grad_norm": 0.5779280066490173,
+      "learning_rate": 0.0001,
+      "loss": 1.3821,
+      "step": 13974
+    },
+    {
+      "epoch": 1.605306989834013,
+      "grad_norm": 0.6681035757064819,
+      "learning_rate": 0.0001,
+      "loss": 1.5948,
+      "step": 13975
+    },
+    {
+      "epoch": 1.60542185974384,
+      "grad_norm": 0.6619514226913452,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 13976
+    },
+    {
+      "epoch": 1.6055367296536671,
+      "grad_norm": 0.5762709379196167,
+      "learning_rate": 0.0001,
+      "loss": 1.4743,
+      "step": 13977
+    },
+    {
+      "epoch": 1.6056515995634943,
+      "grad_norm": 0.597047746181488,
+      "learning_rate": 0.0001,
+      "loss": 1.3797,
+      "step": 13978
+    },
+    {
+      "epoch": 1.6057664694733216,
+      "grad_norm": 0.6428928375244141,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 13979
+    },
+    {
+      "epoch": 1.6058813393831486,
+      "grad_norm": 0.6371673941612244,
+      "learning_rate": 0.0001,
+      "loss": 1.5729,
+      "step": 13980
+    },
+    {
+      "epoch": 1.6059962092929756,
+      "grad_norm": 0.69063401222229,
+      "learning_rate": 0.0001,
+      "loss": 1.3198,
+      "step": 13981
+    },
+    {
+      "epoch": 1.6061110792028028,
+      "grad_norm": 0.6148319244384766,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 13982
+    },
+    {
+      "epoch": 1.60622594911263,
+      "grad_norm": 0.6522954106330872,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 13983
+    },
+    {
+      "epoch": 1.606340819022457,
+      "grad_norm": 0.6170140504837036,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 13984
+    },
+    {
+      "epoch": 1.606455688932284,
+      "grad_norm": 0.5978295803070068,
+      "learning_rate": 0.0001,
+      "loss": 1.4698,
+      "step": 13985
+    },
+    {
+      "epoch": 1.6065705588421113,
+      "grad_norm": 0.6230267882347107,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 13986
+    },
+    {
+      "epoch": 1.6066854287519385,
+      "grad_norm": 0.6035535335540771,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 13987
+    },
+    {
+      "epoch": 1.6068002986617655,
+      "grad_norm": 0.6185717582702637,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 13988
+    },
+    {
+      "epoch": 1.6069151685715926,
+      "grad_norm": 0.5879165530204773,
+      "learning_rate": 0.0001,
+      "loss": 1.1743,
+      "step": 13989
+    },
+    {
+      "epoch": 1.6070300384814198,
+      "grad_norm": 0.6925861239433289,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 13990
+    },
+    {
+      "epoch": 1.607144908391247,
+      "grad_norm": 0.603412926197052,
+      "learning_rate": 0.0001,
+      "loss": 1.3555,
+      "step": 13991
+    },
+    {
+      "epoch": 1.607259778301074,
+      "grad_norm": 0.6258808374404907,
+      "learning_rate": 0.0001,
+      "loss": 1.692,
+      "step": 13992
+    },
+    {
+      "epoch": 1.607374648210901,
+      "grad_norm": 0.6243830919265747,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 13993
+    },
+    {
+      "epoch": 1.6074895181207283,
+      "grad_norm": 0.5801231861114502,
+      "learning_rate": 0.0001,
+      "loss": 1.3985,
+      "step": 13994
+    },
+    {
+      "epoch": 1.6076043880305555,
+      "grad_norm": 0.5816532373428345,
+      "learning_rate": 0.0001,
+      "loss": 1.274,
+      "step": 13995
+    },
+    {
+      "epoch": 1.6077192579403825,
+      "grad_norm": 0.6143868565559387,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 13996
+    },
+    {
+      "epoch": 1.6078341278502095,
+      "grad_norm": 0.5322367548942566,
+      "learning_rate": 0.0001,
+      "loss": 1.4064,
+      "step": 13997
+    },
+    {
+      "epoch": 1.6079489977600367,
+      "grad_norm": 0.6119695901870728,
+      "learning_rate": 0.0001,
+      "loss": 1.3524,
+      "step": 13998
+    },
+    {
+      "epoch": 1.608063867669864,
+      "grad_norm": 0.6040320992469788,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 13999
+    },
+    {
+      "epoch": 1.608178737579691,
+      "grad_norm": 0.5936914682388306,
+      "learning_rate": 0.0001,
+      "loss": 1.3736,
+      "step": 14000
+    },
+    {
+      "epoch": 1.608293607489518,
+      "grad_norm": 0.6493724584579468,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 14001
+    },
+    {
+      "epoch": 1.6084084773993452,
+      "grad_norm": 0.5766366720199585,
+      "learning_rate": 0.0001,
+      "loss": 1.4142,
+      "step": 14002
+    },
+    {
+      "epoch": 1.6085233473091725,
+      "grad_norm": 0.6150953769683838,
+      "learning_rate": 0.0001,
+      "loss": 1.3288,
+      "step": 14003
+    },
+    {
+      "epoch": 1.6086382172189995,
+      "grad_norm": 0.5571459531784058,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 14004
+    },
+    {
+      "epoch": 1.6087530871288265,
+      "grad_norm": 0.6243135333061218,
+      "learning_rate": 0.0001,
+      "loss": 1.3522,
+      "step": 14005
+    },
+    {
+      "epoch": 1.6088679570386537,
+      "grad_norm": 0.5796478390693665,
+      "learning_rate": 0.0001,
+      "loss": 1.3919,
+      "step": 14006
+    },
+    {
+      "epoch": 1.608982826948481,
+      "grad_norm": 0.6436089873313904,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 14007
+    },
+    {
+      "epoch": 1.609097696858308,
+      "grad_norm": 0.6261184811592102,
+      "learning_rate": 0.0001,
+      "loss": 1.5976,
+      "step": 14008
+    },
+    {
+      "epoch": 1.609212566768135,
+      "grad_norm": 0.5776735544204712,
+      "learning_rate": 0.0001,
+      "loss": 1.3081,
+      "step": 14009
+    },
+    {
+      "epoch": 1.6093274366779622,
+      "grad_norm": 0.5873029828071594,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 14010
+    },
+    {
+      "epoch": 1.6094423065877894,
+      "grad_norm": 0.642424464225769,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 14011
+    },
+    {
+      "epoch": 1.6095571764976164,
+      "grad_norm": 0.6408437490463257,
+      "learning_rate": 0.0001,
+      "loss": 1.4495,
+      "step": 14012
+    },
+    {
+      "epoch": 1.6096720464074434,
+      "grad_norm": 0.6098792552947998,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 14013
+    },
+    {
+      "epoch": 1.6097869163172707,
+      "grad_norm": 0.5650305151939392,
+      "learning_rate": 0.0001,
+      "loss": 1.5223,
+      "step": 14014
+    },
+    {
+      "epoch": 1.609901786227098,
+      "grad_norm": 0.6191934943199158,
+      "learning_rate": 0.0001,
+      "loss": 1.3989,
+      "step": 14015
+    },
+    {
+      "epoch": 1.610016656136925,
+      "grad_norm": 0.5842486619949341,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 14016
+    },
+    {
+      "epoch": 1.610131526046752,
+      "grad_norm": 0.6919946670532227,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 14017
+    },
+    {
+      "epoch": 1.6102463959565791,
+      "grad_norm": 0.5900247693061829,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 14018
+    },
+    {
+      "epoch": 1.6103612658664064,
+      "grad_norm": 0.5633094906806946,
+      "learning_rate": 0.0001,
+      "loss": 1.2923,
+      "step": 14019
+    },
+    {
+      "epoch": 1.6104761357762334,
+      "grad_norm": 0.6042163968086243,
+      "learning_rate": 0.0001,
+      "loss": 1.5307,
+      "step": 14020
+    },
+    {
+      "epoch": 1.6105910056860604,
+      "grad_norm": 0.6642841100692749,
+      "learning_rate": 0.0001,
+      "loss": 1.4918,
+      "step": 14021
+    },
+    {
+      "epoch": 1.6107058755958876,
+      "grad_norm": 0.7833337783813477,
+      "learning_rate": 0.0001,
+      "loss": 1.4682,
+      "step": 14022
+    },
+    {
+      "epoch": 1.6108207455057149,
+      "grad_norm": 0.5767188668251038,
+      "learning_rate": 0.0001,
+      "loss": 1.4422,
+      "step": 14023
+    },
+    {
+      "epoch": 1.6109356154155419,
+      "grad_norm": 0.5754833817481995,
+      "learning_rate": 0.0001,
+      "loss": 1.4488,
+      "step": 14024
+    },
+    {
+      "epoch": 1.6110504853253689,
+      "grad_norm": 0.6136239171028137,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 14025
+    },
+    {
+      "epoch": 1.611165355235196,
+      "grad_norm": 0.6007016897201538,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 14026
+    },
+    {
+      "epoch": 1.6112802251450233,
+      "grad_norm": 0.5736799836158752,
+      "learning_rate": 0.0001,
+      "loss": 1.4608,
+      "step": 14027
+    },
+    {
+      "epoch": 1.6113950950548503,
+      "grad_norm": 0.5962418913841248,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 14028
+    },
+    {
+      "epoch": 1.6115099649646774,
+      "grad_norm": 0.6547172665596008,
+      "learning_rate": 0.0001,
+      "loss": 1.4047,
+      "step": 14029
+    },
+    {
+      "epoch": 1.6116248348745046,
+      "grad_norm": 0.6729933023452759,
+      "learning_rate": 0.0001,
+      "loss": 1.57,
+      "step": 14030
+    },
+    {
+      "epoch": 1.6117397047843318,
+      "grad_norm": 0.5763393640518188,
+      "learning_rate": 0.0001,
+      "loss": 1.6013,
+      "step": 14031
+    },
+    {
+      "epoch": 1.6118545746941588,
+      "grad_norm": 0.6308915019035339,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 14032
+    },
+    {
+      "epoch": 1.6119694446039858,
+      "grad_norm": 0.5580734610557556,
+      "learning_rate": 0.0001,
+      "loss": 1.4442,
+      "step": 14033
+    },
+    {
+      "epoch": 1.612084314513813,
+      "grad_norm": 0.5644499063491821,
+      "learning_rate": 0.0001,
+      "loss": 1.3308,
+      "step": 14034
+    },
+    {
+      "epoch": 1.6121991844236403,
+      "grad_norm": 0.6455258131027222,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 14035
+    },
+    {
+      "epoch": 1.6123140543334673,
+      "grad_norm": 0.596584677696228,
+      "learning_rate": 0.0001,
+      "loss": 1.3524,
+      "step": 14036
+    },
+    {
+      "epoch": 1.6124289242432943,
+      "grad_norm": 0.7680908441543579,
+      "learning_rate": 0.0001,
+      "loss": 1.5657,
+      "step": 14037
+    },
+    {
+      "epoch": 1.6125437941531215,
+      "grad_norm": 0.661892294883728,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 14038
+    },
+    {
+      "epoch": 1.6126586640629488,
+      "grad_norm": 0.5909493565559387,
+      "learning_rate": 0.0001,
+      "loss": 1.3543,
+      "step": 14039
+    },
+    {
+      "epoch": 1.6127735339727758,
+      "grad_norm": 0.5718796849250793,
+      "learning_rate": 0.0001,
+      "loss": 1.191,
+      "step": 14040
+    },
+    {
+      "epoch": 1.6128884038826028,
+      "grad_norm": 0.6069730520248413,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 14041
+    },
+    {
+      "epoch": 1.61300327379243,
+      "grad_norm": 0.6135676503181458,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 14042
+    },
+    {
+      "epoch": 1.6131181437022573,
+      "grad_norm": 0.5874907970428467,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 14043
+    },
+    {
+      "epoch": 1.6132330136120843,
+      "grad_norm": 0.6142223477363586,
+      "learning_rate": 0.0001,
+      "loss": 1.3995,
+      "step": 14044
+    },
+    {
+      "epoch": 1.6133478835219113,
+      "grad_norm": 0.6572316884994507,
+      "learning_rate": 0.0001,
+      "loss": 1.6148,
+      "step": 14045
+    },
+    {
+      "epoch": 1.6134627534317385,
+      "grad_norm": 0.6865919828414917,
+      "learning_rate": 0.0001,
+      "loss": 1.4198,
+      "step": 14046
+    },
+    {
+      "epoch": 1.6135776233415657,
+      "grad_norm": 0.57161945104599,
+      "learning_rate": 0.0001,
+      "loss": 1.3409,
+      "step": 14047
+    },
+    {
+      "epoch": 1.6136924932513927,
+      "grad_norm": 0.672649085521698,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 14048
+    },
+    {
+      "epoch": 1.6138073631612198,
+      "grad_norm": 0.5908873677253723,
+      "learning_rate": 0.0001,
+      "loss": 1.3198,
+      "step": 14049
+    },
+    {
+      "epoch": 1.613922233071047,
+      "grad_norm": 0.7088580131530762,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 14050
+    },
+    {
+      "epoch": 1.6140371029808742,
+      "grad_norm": 0.6201817393302917,
+      "learning_rate": 0.0001,
+      "loss": 1.3589,
+      "step": 14051
+    },
+    {
+      "epoch": 1.6141519728907012,
+      "grad_norm": 0.584344208240509,
+      "learning_rate": 0.0001,
+      "loss": 1.4533,
+      "step": 14052
+    },
+    {
+      "epoch": 1.6142668428005285,
+      "grad_norm": 0.5732455849647522,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 14053
+    },
+    {
+      "epoch": 1.6143817127103555,
+      "grad_norm": 0.5996015071868896,
+      "learning_rate": 0.0001,
+      "loss": 1.3556,
+      "step": 14054
+    },
+    {
+      "epoch": 1.6144965826201827,
+      "grad_norm": 0.633338987827301,
+      "learning_rate": 0.0001,
+      "loss": 1.3706,
+      "step": 14055
+    },
+    {
+      "epoch": 1.61461145253001,
+      "grad_norm": 0.6130332350730896,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 14056
+    },
+    {
+      "epoch": 1.614726322439837,
+      "grad_norm": 0.5730156898498535,
+      "learning_rate": 0.0001,
+      "loss": 1.2704,
+      "step": 14057
+    },
+    {
+      "epoch": 1.614841192349664,
+      "grad_norm": 0.5948575139045715,
+      "learning_rate": 0.0001,
+      "loss": 1.5801,
+      "step": 14058
+    },
+    {
+      "epoch": 1.6149560622594912,
+      "grad_norm": 0.6097572445869446,
+      "learning_rate": 0.0001,
+      "loss": 1.3788,
+      "step": 14059
+    },
+    {
+      "epoch": 1.6150709321693184,
+      "grad_norm": 0.5689347386360168,
+      "learning_rate": 0.0001,
+      "loss": 1.3254,
+      "step": 14060
+    },
+    {
+      "epoch": 1.6151858020791454,
+      "grad_norm": 0.6363088488578796,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 14061
+    },
+    {
+      "epoch": 1.6153006719889724,
+      "grad_norm": 0.6014930605888367,
+      "learning_rate": 0.0001,
+      "loss": 1.5373,
+      "step": 14062
+    },
+    {
+      "epoch": 1.6154155418987997,
+      "grad_norm": 0.5899112224578857,
+      "learning_rate": 0.0001,
+      "loss": 1.4386,
+      "step": 14063
+    },
+    {
+      "epoch": 1.6155304118086269,
+      "grad_norm": 0.7439025044441223,
+      "learning_rate": 0.0001,
+      "loss": 1.7865,
+      "step": 14064
+    },
+    {
+      "epoch": 1.615645281718454,
+      "grad_norm": 0.6421499252319336,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 14065
+    },
+    {
+      "epoch": 1.615760151628281,
+      "grad_norm": 0.6733357310295105,
+      "learning_rate": 0.0001,
+      "loss": 1.506,
+      "step": 14066
+    },
+    {
+      "epoch": 1.6158750215381081,
+      "grad_norm": 0.5812866687774658,
+      "learning_rate": 0.0001,
+      "loss": 1.3306,
+      "step": 14067
+    },
+    {
+      "epoch": 1.6159898914479354,
+      "grad_norm": 0.6346286535263062,
+      "learning_rate": 0.0001,
+      "loss": 1.5606,
+      "step": 14068
+    },
+    {
+      "epoch": 1.6161047613577624,
+      "grad_norm": 0.6660323739051819,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 14069
+    },
+    {
+      "epoch": 1.6162196312675894,
+      "grad_norm": 0.6062818169593811,
+      "learning_rate": 0.0001,
+      "loss": 1.4311,
+      "step": 14070
+    },
+    {
+      "epoch": 1.6163345011774166,
+      "grad_norm": 0.6203289031982422,
+      "learning_rate": 0.0001,
+      "loss": 1.5723,
+      "step": 14071
+    },
+    {
+      "epoch": 1.6164493710872438,
+      "grad_norm": 0.6779844760894775,
+      "learning_rate": 0.0001,
+      "loss": 1.3864,
+      "step": 14072
+    },
+    {
+      "epoch": 1.6165642409970709,
+      "grad_norm": 0.6042460203170776,
+      "learning_rate": 0.0001,
+      "loss": 1.5513,
+      "step": 14073
+    },
+    {
+      "epoch": 1.6166791109068979,
+      "grad_norm": 0.6318661570549011,
+      "learning_rate": 0.0001,
+      "loss": 1.3761,
+      "step": 14074
+    },
+    {
+      "epoch": 1.616793980816725,
+      "grad_norm": 0.579138457775116,
+      "learning_rate": 0.0001,
+      "loss": 1.3566,
+      "step": 14075
+    },
+    {
+      "epoch": 1.6169088507265523,
+      "grad_norm": 0.6678400635719299,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 14076
+    },
+    {
+      "epoch": 1.6170237206363793,
+      "grad_norm": 0.6432377099990845,
+      "learning_rate": 0.0001,
+      "loss": 1.3636,
+      "step": 14077
+    },
+    {
+      "epoch": 1.6171385905462063,
+      "grad_norm": 0.6431517004966736,
+      "learning_rate": 0.0001,
+      "loss": 1.4582,
+      "step": 14078
+    },
+    {
+      "epoch": 1.6172534604560336,
+      "grad_norm": 0.6440772414207458,
+      "learning_rate": 0.0001,
+      "loss": 1.3715,
+      "step": 14079
+    },
+    {
+      "epoch": 1.6173683303658608,
+      "grad_norm": 0.5892122387886047,
+      "learning_rate": 0.0001,
+      "loss": 1.4429,
+      "step": 14080
+    },
+    {
+      "epoch": 1.6174832002756878,
+      "grad_norm": 0.5709667801856995,
+      "learning_rate": 0.0001,
+      "loss": 1.4392,
+      "step": 14081
+    },
+    {
+      "epoch": 1.6175980701855148,
+      "grad_norm": 0.5939996242523193,
+      "learning_rate": 0.0001,
+      "loss": 1.4266,
+      "step": 14082
+    },
+    {
+      "epoch": 1.617712940095342,
+      "grad_norm": 0.6955810785293579,
+      "learning_rate": 0.0001,
+      "loss": 1.453,
+      "step": 14083
+    },
+    {
+      "epoch": 1.6178278100051693,
+      "grad_norm": 0.6132066249847412,
+      "learning_rate": 0.0001,
+      "loss": 1.5033,
+      "step": 14084
+    },
+    {
+      "epoch": 1.6179426799149963,
+      "grad_norm": 0.5954875946044922,
+      "learning_rate": 0.0001,
+      "loss": 1.4841,
+      "step": 14085
+    },
+    {
+      "epoch": 1.6180575498248233,
+      "grad_norm": 0.6023247838020325,
+      "learning_rate": 0.0001,
+      "loss": 1.3852,
+      "step": 14086
+    },
+    {
+      "epoch": 1.6181724197346505,
+      "grad_norm": 0.7028385400772095,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 14087
+    },
+    {
+      "epoch": 1.6182872896444778,
+      "grad_norm": 0.6407709121704102,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 14088
+    },
+    {
+      "epoch": 1.6184021595543048,
+      "grad_norm": 0.5889942049980164,
+      "learning_rate": 0.0001,
+      "loss": 1.5281,
+      "step": 14089
+    },
+    {
+      "epoch": 1.6185170294641318,
+      "grad_norm": 0.6771349906921387,
+      "learning_rate": 0.0001,
+      "loss": 1.4593,
+      "step": 14090
+    },
+    {
+      "epoch": 1.618631899373959,
+      "grad_norm": 0.6546837687492371,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 14091
+    },
+    {
+      "epoch": 1.6187467692837862,
+      "grad_norm": 0.5943772196769714,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 14092
+    },
+    {
+      "epoch": 1.6188616391936133,
+      "grad_norm": 0.5858648419380188,
+      "learning_rate": 0.0001,
+      "loss": 1.3868,
+      "step": 14093
+    },
+    {
+      "epoch": 1.6189765091034403,
+      "grad_norm": 0.6995881795883179,
+      "learning_rate": 0.0001,
+      "loss": 1.7393,
+      "step": 14094
+    },
+    {
+      "epoch": 1.6190913790132675,
+      "grad_norm": 0.6692748069763184,
+      "learning_rate": 0.0001,
+      "loss": 1.4476,
+      "step": 14095
+    },
+    {
+      "epoch": 1.6192062489230947,
+      "grad_norm": 0.5856147408485413,
+      "learning_rate": 0.0001,
+      "loss": 1.3743,
+      "step": 14096
+    },
+    {
+      "epoch": 1.6193211188329217,
+      "grad_norm": 0.5996282696723938,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 14097
+    },
+    {
+      "epoch": 1.6194359887427487,
+      "grad_norm": 0.6970905065536499,
+      "learning_rate": 0.0001,
+      "loss": 1.5738,
+      "step": 14098
+    },
+    {
+      "epoch": 1.619550858652576,
+      "grad_norm": 0.6515586972236633,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 14099
+    },
+    {
+      "epoch": 1.6196657285624032,
+      "grad_norm": 0.5993524789810181,
+      "learning_rate": 0.0001,
+      "loss": 1.4535,
+      "step": 14100
+    },
+    {
+      "epoch": 1.6197805984722302,
+      "grad_norm": 0.5940983295440674,
+      "learning_rate": 0.0001,
+      "loss": 1.4715,
+      "step": 14101
+    },
+    {
+      "epoch": 1.6198954683820572,
+      "grad_norm": 0.585989236831665,
+      "learning_rate": 0.0001,
+      "loss": 1.4361,
+      "step": 14102
+    },
+    {
+      "epoch": 1.6200103382918845,
+      "grad_norm": 0.5850023031234741,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 14103
+    },
+    {
+      "epoch": 1.6201252082017117,
+      "grad_norm": 0.5683290362358093,
+      "learning_rate": 0.0001,
+      "loss": 1.1528,
+      "step": 14104
+    },
+    {
+      "epoch": 1.6202400781115387,
+      "grad_norm": 0.5724591016769409,
+      "learning_rate": 0.0001,
+      "loss": 1.4089,
+      "step": 14105
+    },
+    {
+      "epoch": 1.6203549480213657,
+      "grad_norm": 0.7932894825935364,
+      "learning_rate": 0.0001,
+      "loss": 1.6699,
+      "step": 14106
+    },
+    {
+      "epoch": 1.620469817931193,
+      "grad_norm": 0.6048719882965088,
+      "learning_rate": 0.0001,
+      "loss": 1.4477,
+      "step": 14107
+    },
+    {
+      "epoch": 1.6205846878410202,
+      "grad_norm": 0.6853838562965393,
+      "learning_rate": 0.0001,
+      "loss": 1.5403,
+      "step": 14108
+    },
+    {
+      "epoch": 1.6206995577508472,
+      "grad_norm": 0.7041744589805603,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 14109
+    },
+    {
+      "epoch": 1.6208144276606742,
+      "grad_norm": 0.6290716528892517,
+      "learning_rate": 0.0001,
+      "loss": 1.5163,
+      "step": 14110
+    },
+    {
+      "epoch": 1.6209292975705014,
+      "grad_norm": 0.5670915842056274,
+      "learning_rate": 0.0001,
+      "loss": 1.3655,
+      "step": 14111
+    },
+    {
+      "epoch": 1.6210441674803286,
+      "grad_norm": 0.5751041769981384,
+      "learning_rate": 0.0001,
+      "loss": 1.2977,
+      "step": 14112
+    },
+    {
+      "epoch": 1.6211590373901557,
+      "grad_norm": 0.5855926275253296,
+      "learning_rate": 0.0001,
+      "loss": 1.6452,
+      "step": 14113
+    },
+    {
+      "epoch": 1.6212739072999827,
+      "grad_norm": 0.6567171812057495,
+      "learning_rate": 0.0001,
+      "loss": 1.3936,
+      "step": 14114
+    },
+    {
+      "epoch": 1.62138877720981,
+      "grad_norm": 0.5859187841415405,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 14115
+    },
+    {
+      "epoch": 1.6215036471196371,
+      "grad_norm": 0.5759543180465698,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 14116
+    },
+    {
+      "epoch": 1.6216185170294641,
+      "grad_norm": 0.5698956847190857,
+      "learning_rate": 0.0001,
+      "loss": 1.3733,
+      "step": 14117
+    },
+    {
+      "epoch": 1.6217333869392911,
+      "grad_norm": 0.6352313160896301,
+      "learning_rate": 0.0001,
+      "loss": 1.3525,
+      "step": 14118
+    },
+    {
+      "epoch": 1.6218482568491184,
+      "grad_norm": 0.6209557056427002,
+      "learning_rate": 0.0001,
+      "loss": 1.2902,
+      "step": 14119
+    },
+    {
+      "epoch": 1.6219631267589456,
+      "grad_norm": 0.6741890907287598,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 14120
+    },
+    {
+      "epoch": 1.6220779966687726,
+      "grad_norm": 0.5992792248725891,
+      "learning_rate": 0.0001,
+      "loss": 1.3194,
+      "step": 14121
+    },
+    {
+      "epoch": 1.6221928665785996,
+      "grad_norm": 0.6535847187042236,
+      "learning_rate": 0.0001,
+      "loss": 1.285,
+      "step": 14122
+    },
+    {
+      "epoch": 1.6223077364884269,
+      "grad_norm": 0.5931274890899658,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 14123
+    },
+    {
+      "epoch": 1.622422606398254,
+      "grad_norm": 0.6604529023170471,
+      "learning_rate": 0.0001,
+      "loss": 1.4776,
+      "step": 14124
+    },
+    {
+      "epoch": 1.622537476308081,
+      "grad_norm": 0.6257154941558838,
+      "learning_rate": 0.0001,
+      "loss": 1.6617,
+      "step": 14125
+    },
+    {
+      "epoch": 1.622652346217908,
+      "grad_norm": 0.6580725908279419,
+      "learning_rate": 0.0001,
+      "loss": 1.1384,
+      "step": 14126
+    },
+    {
+      "epoch": 1.6227672161277353,
+      "grad_norm": 0.575198233127594,
+      "learning_rate": 0.0001,
+      "loss": 1.3248,
+      "step": 14127
+    },
+    {
+      "epoch": 1.6228820860375626,
+      "grad_norm": 0.591127872467041,
+      "learning_rate": 0.0001,
+      "loss": 1.3635,
+      "step": 14128
+    },
+    {
+      "epoch": 1.6229969559473896,
+      "grad_norm": 0.6647173166275024,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 14129
+    },
+    {
+      "epoch": 1.6231118258572166,
+      "grad_norm": 0.5722182393074036,
+      "learning_rate": 0.0001,
+      "loss": 1.4466,
+      "step": 14130
+    },
+    {
+      "epoch": 1.6232266957670438,
+      "grad_norm": 0.5584431886672974,
+      "learning_rate": 0.0001,
+      "loss": 1.5135,
+      "step": 14131
+    },
+    {
+      "epoch": 1.623341565676871,
+      "grad_norm": 0.6139621138572693,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 14132
+    },
+    {
+      "epoch": 1.623456435586698,
+      "grad_norm": 0.6344627737998962,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 14133
+    },
+    {
+      "epoch": 1.623571305496525,
+      "grad_norm": 0.6031376123428345,
+      "learning_rate": 0.0001,
+      "loss": 1.2777,
+      "step": 14134
+    },
+    {
+      "epoch": 1.6236861754063523,
+      "grad_norm": 0.6405968070030212,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 14135
+    },
+    {
+      "epoch": 1.6238010453161795,
+      "grad_norm": 0.7221595644950867,
+      "learning_rate": 0.0001,
+      "loss": 1.2492,
+      "step": 14136
+    },
+    {
+      "epoch": 1.6239159152260065,
+      "grad_norm": 0.5714577436447144,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 14137
+    },
+    {
+      "epoch": 1.6240307851358335,
+      "grad_norm": 0.607839047908783,
+      "learning_rate": 0.0001,
+      "loss": 1.5385,
+      "step": 14138
+    },
+    {
+      "epoch": 1.6241456550456608,
+      "grad_norm": 0.6695128679275513,
+      "learning_rate": 0.0001,
+      "loss": 1.5084,
+      "step": 14139
+    },
+    {
+      "epoch": 1.624260524955488,
+      "grad_norm": 0.6498796343803406,
+      "learning_rate": 0.0001,
+      "loss": 1.5016,
+      "step": 14140
+    },
+    {
+      "epoch": 1.624375394865315,
+      "grad_norm": 0.582234799861908,
+      "learning_rate": 0.0001,
+      "loss": 1.2691,
+      "step": 14141
+    },
+    {
+      "epoch": 1.624490264775142,
+      "grad_norm": 0.6602511405944824,
+      "learning_rate": 0.0001,
+      "loss": 1.445,
+      "step": 14142
+    },
+    {
+      "epoch": 1.6246051346849693,
+      "grad_norm": 0.6686058640480042,
+      "learning_rate": 0.0001,
+      "loss": 1.4448,
+      "step": 14143
+    },
+    {
+      "epoch": 1.6247200045947965,
+      "grad_norm": 0.65159010887146,
+      "learning_rate": 0.0001,
+      "loss": 1.428,
+      "step": 14144
+    },
+    {
+      "epoch": 1.6248348745046235,
+      "grad_norm": 0.6343734860420227,
+      "learning_rate": 0.0001,
+      "loss": 1.565,
+      "step": 14145
+    },
+    {
+      "epoch": 1.6249497444144505,
+      "grad_norm": 0.5925803184509277,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 14146
+    },
+    {
+      "epoch": 1.6250646143242777,
+      "grad_norm": 0.593427300453186,
+      "learning_rate": 0.0001,
+      "loss": 1.3529,
+      "step": 14147
+    },
+    {
+      "epoch": 1.625179484234105,
+      "grad_norm": 0.6615251302719116,
+      "learning_rate": 0.0001,
+      "loss": 1.466,
+      "step": 14148
+    },
+    {
+      "epoch": 1.625294354143932,
+      "grad_norm": 0.6435301899909973,
+      "learning_rate": 0.0001,
+      "loss": 1.4284,
+      "step": 14149
+    },
+    {
+      "epoch": 1.625409224053759,
+      "grad_norm": 0.5664091110229492,
+      "learning_rate": 0.0001,
+      "loss": 1.3037,
+      "step": 14150
+    },
+    {
+      "epoch": 1.6255240939635862,
+      "grad_norm": 0.5707288980484009,
+      "learning_rate": 0.0001,
+      "loss": 1.4152,
+      "step": 14151
+    },
+    {
+      "epoch": 1.6256389638734134,
+      "grad_norm": 0.6005744934082031,
+      "learning_rate": 0.0001,
+      "loss": 1.4547,
+      "step": 14152
+    },
+    {
+      "epoch": 1.6257538337832405,
+      "grad_norm": 0.6113733649253845,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 14153
+    },
+    {
+      "epoch": 1.6258687036930675,
+      "grad_norm": 0.6435218453407288,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 14154
+    },
+    {
+      "epoch": 1.6259835736028947,
+      "grad_norm": 0.6765085458755493,
+      "learning_rate": 0.0001,
+      "loss": 1.3975,
+      "step": 14155
+    },
+    {
+      "epoch": 1.626098443512722,
+      "grad_norm": 0.6617031693458557,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 14156
+    },
+    {
+      "epoch": 1.626213313422549,
+      "grad_norm": 0.6027497053146362,
+      "learning_rate": 0.0001,
+      "loss": 1.3812,
+      "step": 14157
+    },
+    {
+      "epoch": 1.626328183332376,
+      "grad_norm": 0.6761353611946106,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 14158
+    },
+    {
+      "epoch": 1.6264430532422032,
+      "grad_norm": 0.6267996430397034,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 14159
+    },
+    {
+      "epoch": 1.6265579231520304,
+      "grad_norm": 0.6282928586006165,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 14160
+    },
+    {
+      "epoch": 1.6266727930618574,
+      "grad_norm": 0.5661084651947021,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 14161
+    },
+    {
+      "epoch": 1.6267876629716844,
+      "grad_norm": 0.6607252955436707,
+      "learning_rate": 0.0001,
+      "loss": 1.5683,
+      "step": 14162
+    },
+    {
+      "epoch": 1.6269025328815117,
+      "grad_norm": 0.6275922060012817,
+      "learning_rate": 0.0001,
+      "loss": 1.2246,
+      "step": 14163
+    },
+    {
+      "epoch": 1.6270174027913389,
+      "grad_norm": 0.5838649272918701,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 14164
+    },
+    {
+      "epoch": 1.627132272701166,
+      "grad_norm": 0.627335786819458,
+      "learning_rate": 0.0001,
+      "loss": 1.3975,
+      "step": 14165
+    },
+    {
+      "epoch": 1.627247142610993,
+      "grad_norm": 0.6400899291038513,
+      "learning_rate": 0.0001,
+      "loss": 1.3117,
+      "step": 14166
+    },
+    {
+      "epoch": 1.6273620125208201,
+      "grad_norm": 0.6840837597846985,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 14167
+    },
+    {
+      "epoch": 1.6274768824306474,
+      "grad_norm": 0.6731758713722229,
+      "learning_rate": 0.0001,
+      "loss": 1.3957,
+      "step": 14168
+    },
+    {
+      "epoch": 1.6275917523404744,
+      "grad_norm": 0.6186261177062988,
+      "learning_rate": 0.0001,
+      "loss": 1.3614,
+      "step": 14169
+    },
+    {
+      "epoch": 1.6277066222503014,
+      "grad_norm": 0.6093758344650269,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 14170
+    },
+    {
+      "epoch": 1.6278214921601286,
+      "grad_norm": 0.8518862128257751,
+      "learning_rate": 0.0001,
+      "loss": 1.1226,
+      "step": 14171
+    },
+    {
+      "epoch": 1.6279363620699558,
+      "grad_norm": 0.6340732574462891,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 14172
+    },
+    {
+      "epoch": 1.6280512319797829,
+      "grad_norm": 0.5993664264678955,
+      "learning_rate": 0.0001,
+      "loss": 1.4323,
+      "step": 14173
+    },
+    {
+      "epoch": 1.6281661018896099,
+      "grad_norm": 0.6589725017547607,
+      "learning_rate": 0.0001,
+      "loss": 1.321,
+      "step": 14174
+    },
+    {
+      "epoch": 1.628280971799437,
+      "grad_norm": 0.668163001537323,
+      "learning_rate": 0.0001,
+      "loss": 1.3883,
+      "step": 14175
+    },
+    {
+      "epoch": 1.6283958417092643,
+      "grad_norm": 0.6377967000007629,
+      "learning_rate": 0.0001,
+      "loss": 1.3221,
+      "step": 14176
+    },
+    {
+      "epoch": 1.6285107116190913,
+      "grad_norm": 0.7198302149772644,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 14177
+    },
+    {
+      "epoch": 1.6286255815289183,
+      "grad_norm": 0.6226831078529358,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 14178
+    },
+    {
+      "epoch": 1.6287404514387456,
+      "grad_norm": 0.6308751106262207,
+      "learning_rate": 0.0001,
+      "loss": 1.3146,
+      "step": 14179
+    },
+    {
+      "epoch": 1.6288553213485728,
+      "grad_norm": 0.6031596064567566,
+      "learning_rate": 0.0001,
+      "loss": 1.4562,
+      "step": 14180
+    },
+    {
+      "epoch": 1.6289701912583998,
+      "grad_norm": 0.6148450374603271,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 14181
+    },
+    {
+      "epoch": 1.6290850611682268,
+      "grad_norm": 0.6116688847541809,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 14182
+    },
+    {
+      "epoch": 1.629199931078054,
+      "grad_norm": 0.594800591468811,
+      "learning_rate": 0.0001,
+      "loss": 1.5319,
+      "step": 14183
+    },
+    {
+      "epoch": 1.6293148009878813,
+      "grad_norm": 0.5846831202507019,
+      "learning_rate": 0.0001,
+      "loss": 1.371,
+      "step": 14184
+    },
+    {
+      "epoch": 1.6294296708977083,
+      "grad_norm": 0.5682091116905212,
+      "learning_rate": 0.0001,
+      "loss": 1.2561,
+      "step": 14185
+    },
+    {
+      "epoch": 1.6295445408075353,
+      "grad_norm": 0.6536598801612854,
+      "learning_rate": 0.0001,
+      "loss": 1.5194,
+      "step": 14186
+    },
+    {
+      "epoch": 1.6296594107173625,
+      "grad_norm": 0.6191934943199158,
+      "learning_rate": 0.0001,
+      "loss": 1.5155,
+      "step": 14187
+    },
+    {
+      "epoch": 1.6297742806271898,
+      "grad_norm": 0.5823689699172974,
+      "learning_rate": 0.0001,
+      "loss": 1.4693,
+      "step": 14188
+    },
+    {
+      "epoch": 1.6298891505370168,
+      "grad_norm": 0.6082186698913574,
+      "learning_rate": 0.0001,
+      "loss": 1.4571,
+      "step": 14189
+    },
+    {
+      "epoch": 1.630004020446844,
+      "grad_norm": 0.6357173919677734,
+      "learning_rate": 0.0001,
+      "loss": 1.2842,
+      "step": 14190
+    },
+    {
+      "epoch": 1.630118890356671,
+      "grad_norm": 0.5339798331260681,
+      "learning_rate": 0.0001,
+      "loss": 1.271,
+      "step": 14191
+    },
+    {
+      "epoch": 1.6302337602664982,
+      "grad_norm": 0.6120511889457703,
+      "learning_rate": 0.0001,
+      "loss": 1.3902,
+      "step": 14192
+    },
+    {
+      "epoch": 1.6303486301763255,
+      "grad_norm": 0.6843846440315247,
+      "learning_rate": 0.0001,
+      "loss": 1.4445,
+      "step": 14193
+    },
+    {
+      "epoch": 1.6304635000861525,
+      "grad_norm": 0.5969381928443909,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 14194
+    },
+    {
+      "epoch": 1.6305783699959795,
+      "grad_norm": 0.621590793132782,
+      "learning_rate": 0.0001,
+      "loss": 1.4124,
+      "step": 14195
+    },
+    {
+      "epoch": 1.6306932399058067,
+      "grad_norm": 0.6120715737342834,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 14196
+    },
+    {
+      "epoch": 1.630808109815634,
+      "grad_norm": 0.6380988955497742,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 14197
+    },
+    {
+      "epoch": 1.630922979725461,
+      "grad_norm": 0.6133922338485718,
+      "learning_rate": 0.0001,
+      "loss": 1.4572,
+      "step": 14198
+    },
+    {
+      "epoch": 1.631037849635288,
+      "grad_norm": 0.6428544521331787,
+      "learning_rate": 0.0001,
+      "loss": 1.3983,
+      "step": 14199
+    },
+    {
+      "epoch": 1.6311527195451152,
+      "grad_norm": 0.5923104286193848,
+      "learning_rate": 0.0001,
+      "loss": 1.3451,
+      "step": 14200
+    },
+    {
+      "epoch": 1.6312675894549424,
+      "grad_norm": 0.6585755348205566,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 14201
+    },
+    {
+      "epoch": 1.6313824593647694,
+      "grad_norm": 0.5719717144966125,
+      "learning_rate": 0.0001,
+      "loss": 1.4134,
+      "step": 14202
+    },
+    {
+      "epoch": 1.6314973292745965,
+      "grad_norm": 0.6237934231758118,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 14203
+    },
+    {
+      "epoch": 1.6316121991844237,
+      "grad_norm": 0.6467723250389099,
+      "learning_rate": 0.0001,
+      "loss": 1.3684,
+      "step": 14204
+    },
+    {
+      "epoch": 1.631727069094251,
+      "grad_norm": 0.7436313033103943,
+      "learning_rate": 0.0001,
+      "loss": 1.6965,
+      "step": 14205
+    },
+    {
+      "epoch": 1.631841939004078,
+      "grad_norm": 0.6541860699653625,
+      "learning_rate": 0.0001,
+      "loss": 1.5145,
+      "step": 14206
+    },
+    {
+      "epoch": 1.631956808913905,
+      "grad_norm": 0.6902755498886108,
+      "learning_rate": 0.0001,
+      "loss": 1.7396,
+      "step": 14207
+    },
+    {
+      "epoch": 1.6320716788237322,
+      "grad_norm": 0.5739198923110962,
+      "learning_rate": 0.0001,
+      "loss": 1.4084,
+      "step": 14208
+    },
+    {
+      "epoch": 1.6321865487335594,
+      "grad_norm": 0.6371223330497742,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 14209
+    },
+    {
+      "epoch": 1.6323014186433864,
+      "grad_norm": 0.5753113031387329,
+      "learning_rate": 0.0001,
+      "loss": 1.3629,
+      "step": 14210
+    },
+    {
+      "epoch": 1.6324162885532134,
+      "grad_norm": 0.5805362462997437,
+      "learning_rate": 0.0001,
+      "loss": 1.5456,
+      "step": 14211
+    },
+    {
+      "epoch": 1.6325311584630406,
+      "grad_norm": 0.5889348983764648,
+      "learning_rate": 0.0001,
+      "loss": 1.331,
+      "step": 14212
+    },
+    {
+      "epoch": 1.6326460283728679,
+      "grad_norm": 0.6165045499801636,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 14213
+    },
+    {
+      "epoch": 1.6327608982826949,
+      "grad_norm": 0.6740553975105286,
+      "learning_rate": 0.0001,
+      "loss": 1.4898,
+      "step": 14214
+    },
+    {
+      "epoch": 1.632875768192522,
+      "grad_norm": 0.6100121140480042,
+      "learning_rate": 0.0001,
+      "loss": 1.3378,
+      "step": 14215
+    },
+    {
+      "epoch": 1.6329906381023491,
+      "grad_norm": 0.6665295958518982,
+      "learning_rate": 0.0001,
+      "loss": 1.5055,
+      "step": 14216
+    },
+    {
+      "epoch": 1.6331055080121764,
+      "grad_norm": 0.574766218662262,
+      "learning_rate": 0.0001,
+      "loss": 1.4792,
+      "step": 14217
+    },
+    {
+      "epoch": 1.6332203779220034,
+      "grad_norm": 0.6160637140274048,
+      "learning_rate": 0.0001,
+      "loss": 1.529,
+      "step": 14218
+    },
+    {
+      "epoch": 1.6333352478318304,
+      "grad_norm": 0.6245967745780945,
+      "learning_rate": 0.0001,
+      "loss": 1.5056,
+      "step": 14219
+    },
+    {
+      "epoch": 1.6334501177416576,
+      "grad_norm": 0.6003860235214233,
+      "learning_rate": 0.0001,
+      "loss": 1.4246,
+      "step": 14220
+    },
+    {
+      "epoch": 1.6335649876514848,
+      "grad_norm": 0.5847166776657104,
+      "learning_rate": 0.0001,
+      "loss": 1.3012,
+      "step": 14221
+    },
+    {
+      "epoch": 1.6336798575613118,
+      "grad_norm": 0.567745566368103,
+      "learning_rate": 0.0001,
+      "loss": 1.3486,
+      "step": 14222
+    },
+    {
+      "epoch": 1.6337947274711389,
+      "grad_norm": 0.7059317231178284,
+      "learning_rate": 0.0001,
+      "loss": 1.4865,
+      "step": 14223
+    },
+    {
+      "epoch": 1.633909597380966,
+      "grad_norm": 0.6010088920593262,
+      "learning_rate": 0.0001,
+      "loss": 1.4275,
+      "step": 14224
+    },
+    {
+      "epoch": 1.6340244672907933,
+      "grad_norm": 0.7067935466766357,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 14225
+    },
+    {
+      "epoch": 1.6341393372006203,
+      "grad_norm": 0.6423289775848389,
+      "learning_rate": 0.0001,
+      "loss": 1.4735,
+      "step": 14226
+    },
+    {
+      "epoch": 1.6342542071104473,
+      "grad_norm": 0.6251084804534912,
+      "learning_rate": 0.0001,
+      "loss": 1.3356,
+      "step": 14227
+    },
+    {
+      "epoch": 1.6343690770202746,
+      "grad_norm": 0.6532206535339355,
+      "learning_rate": 0.0001,
+      "loss": 1.3189,
+      "step": 14228
+    },
+    {
+      "epoch": 1.6344839469301018,
+      "grad_norm": 0.5804827213287354,
+      "learning_rate": 0.0001,
+      "loss": 1.527,
+      "step": 14229
+    },
+    {
+      "epoch": 1.6345988168399288,
+      "grad_norm": 0.6210793256759644,
+      "learning_rate": 0.0001,
+      "loss": 1.4906,
+      "step": 14230
+    },
+    {
+      "epoch": 1.6347136867497558,
+      "grad_norm": 0.5873556137084961,
+      "learning_rate": 0.0001,
+      "loss": 1.2527,
+      "step": 14231
+    },
+    {
+      "epoch": 1.634828556659583,
+      "grad_norm": 0.5766886472702026,
+      "learning_rate": 0.0001,
+      "loss": 1.4624,
+      "step": 14232
+    },
+    {
+      "epoch": 1.6349434265694103,
+      "grad_norm": 0.6255853772163391,
+      "learning_rate": 0.0001,
+      "loss": 1.6646,
+      "step": 14233
+    },
+    {
+      "epoch": 1.6350582964792373,
+      "grad_norm": 0.6151155829429626,
+      "learning_rate": 0.0001,
+      "loss": 1.4379,
+      "step": 14234
+    },
+    {
+      "epoch": 1.6351731663890643,
+      "grad_norm": 0.602134644985199,
+      "learning_rate": 0.0001,
+      "loss": 1.4381,
+      "step": 14235
+    },
+    {
+      "epoch": 1.6352880362988915,
+      "grad_norm": 0.5812894105911255,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 14236
+    },
+    {
+      "epoch": 1.6354029062087188,
+      "grad_norm": 0.5972526669502258,
+      "learning_rate": 0.0001,
+      "loss": 1.5256,
+      "step": 14237
+    },
+    {
+      "epoch": 1.6355177761185458,
+      "grad_norm": 0.6522207260131836,
+      "learning_rate": 0.0001,
+      "loss": 1.4439,
+      "step": 14238
+    },
+    {
+      "epoch": 1.6356326460283728,
+      "grad_norm": 0.5848917961120605,
+      "learning_rate": 0.0001,
+      "loss": 1.114,
+      "step": 14239
+    },
+    {
+      "epoch": 1.6357475159382,
+      "grad_norm": 0.6326489448547363,
+      "learning_rate": 0.0001,
+      "loss": 1.3317,
+      "step": 14240
+    },
+    {
+      "epoch": 1.6358623858480272,
+      "grad_norm": 0.6178004741668701,
+      "learning_rate": 0.0001,
+      "loss": 1.3889,
+      "step": 14241
+    },
+    {
+      "epoch": 1.6359772557578542,
+      "grad_norm": 0.6275560259819031,
+      "learning_rate": 0.0001,
+      "loss": 1.3722,
+      "step": 14242
+    },
+    {
+      "epoch": 1.6360921256676813,
+      "grad_norm": 0.5886978507041931,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 14243
+    },
+    {
+      "epoch": 1.6362069955775085,
+      "grad_norm": 0.6700698137283325,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 14244
+    },
+    {
+      "epoch": 1.6363218654873357,
+      "grad_norm": 0.6027423143386841,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 14245
+    },
+    {
+      "epoch": 1.6364367353971627,
+      "grad_norm": 0.6072428226470947,
+      "learning_rate": 0.0001,
+      "loss": 1.4246,
+      "step": 14246
+    },
+    {
+      "epoch": 1.6365516053069897,
+      "grad_norm": 0.6349037289619446,
+      "learning_rate": 0.0001,
+      "loss": 1.0018,
+      "step": 14247
+    },
+    {
+      "epoch": 1.636666475216817,
+      "grad_norm": 0.6661435961723328,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 14248
+    },
+    {
+      "epoch": 1.6367813451266442,
+      "grad_norm": 0.6801328063011169,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 14249
+    },
+    {
+      "epoch": 1.6368962150364712,
+      "grad_norm": 0.6124873161315918,
+      "learning_rate": 0.0001,
+      "loss": 1.3097,
+      "step": 14250
+    },
+    {
+      "epoch": 1.6370110849462982,
+      "grad_norm": 0.6170512437820435,
+      "learning_rate": 0.0001,
+      "loss": 1.3019,
+      "step": 14251
+    },
+    {
+      "epoch": 1.6371259548561254,
+      "grad_norm": 0.6388773918151855,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 14252
+    },
+    {
+      "epoch": 1.6372408247659527,
+      "grad_norm": 0.6021920442581177,
+      "learning_rate": 0.0001,
+      "loss": 1.483,
+      "step": 14253
+    },
+    {
+      "epoch": 1.6373556946757797,
+      "grad_norm": 0.6165171265602112,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 14254
+    },
+    {
+      "epoch": 1.6374705645856067,
+      "grad_norm": 0.5764616131782532,
+      "learning_rate": 0.0001,
+      "loss": 1.2656,
+      "step": 14255
+    },
+    {
+      "epoch": 1.637585434495434,
+      "grad_norm": 0.6353535652160645,
+      "learning_rate": 0.0001,
+      "loss": 1.4185,
+      "step": 14256
+    },
+    {
+      "epoch": 1.6377003044052612,
+      "grad_norm": 0.5799005627632141,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 14257
+    },
+    {
+      "epoch": 1.6378151743150882,
+      "grad_norm": 0.6138670444488525,
+      "learning_rate": 0.0001,
+      "loss": 1.4453,
+      "step": 14258
+    },
+    {
+      "epoch": 1.6379300442249152,
+      "grad_norm": 0.5995337963104248,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 14259
+    },
+    {
+      "epoch": 1.6380449141347424,
+      "grad_norm": 0.5708662271499634,
+      "learning_rate": 0.0001,
+      "loss": 1.3962,
+      "step": 14260
+    },
+    {
+      "epoch": 1.6381597840445696,
+      "grad_norm": 0.6403148174285889,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 14261
+    },
+    {
+      "epoch": 1.6382746539543966,
+      "grad_norm": 0.6167385578155518,
+      "learning_rate": 0.0001,
+      "loss": 1.4572,
+      "step": 14262
+    },
+    {
+      "epoch": 1.6383895238642237,
+      "grad_norm": 0.5929616689682007,
+      "learning_rate": 0.0001,
+      "loss": 1.4688,
+      "step": 14263
+    },
+    {
+      "epoch": 1.6385043937740509,
+      "grad_norm": 0.5963886380195618,
+      "learning_rate": 0.0001,
+      "loss": 1.304,
+      "step": 14264
+    },
+    {
+      "epoch": 1.6386192636838781,
+      "grad_norm": 0.596179723739624,
+      "learning_rate": 0.0001,
+      "loss": 1.3922,
+      "step": 14265
+    },
+    {
+      "epoch": 1.6387341335937051,
+      "grad_norm": 0.6750287413597107,
+      "learning_rate": 0.0001,
+      "loss": 1.4107,
+      "step": 14266
+    },
+    {
+      "epoch": 1.6388490035035321,
+      "grad_norm": 0.6416113376617432,
+      "learning_rate": 0.0001,
+      "loss": 1.4031,
+      "step": 14267
+    },
+    {
+      "epoch": 1.6389638734133594,
+      "grad_norm": 0.547660231590271,
+      "learning_rate": 0.0001,
+      "loss": 1.2633,
+      "step": 14268
+    },
+    {
+      "epoch": 1.6390787433231866,
+      "grad_norm": 0.6302758455276489,
+      "learning_rate": 0.0001,
+      "loss": 1.526,
+      "step": 14269
+    },
+    {
+      "epoch": 1.6391936132330136,
+      "grad_norm": 0.6041958928108215,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 14270
+    },
+    {
+      "epoch": 1.6393084831428406,
+      "grad_norm": 0.5641599893569946,
+      "learning_rate": 0.0001,
+      "loss": 1.2687,
+      "step": 14271
+    },
+    {
+      "epoch": 1.6394233530526678,
+      "grad_norm": 0.6017992496490479,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 14272
+    },
+    {
+      "epoch": 1.639538222962495,
+      "grad_norm": 0.5696545243263245,
+      "learning_rate": 0.0001,
+      "loss": 1.5084,
+      "step": 14273
+    },
+    {
+      "epoch": 1.639653092872322,
+      "grad_norm": 0.6044654250144958,
+      "learning_rate": 0.0001,
+      "loss": 1.4717,
+      "step": 14274
+    },
+    {
+      "epoch": 1.639767962782149,
+      "grad_norm": 0.5639116168022156,
+      "learning_rate": 0.0001,
+      "loss": 1.4058,
+      "step": 14275
+    },
+    {
+      "epoch": 1.6398828326919763,
+      "grad_norm": 0.6422322988510132,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 14276
+    },
+    {
+      "epoch": 1.6399977026018036,
+      "grad_norm": 0.651578426361084,
+      "learning_rate": 0.0001,
+      "loss": 1.3275,
+      "step": 14277
+    },
+    {
+      "epoch": 1.6401125725116306,
+      "grad_norm": 0.6262843012809753,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 14278
+    },
+    {
+      "epoch": 1.6402274424214576,
+      "grad_norm": 0.6277267336845398,
+      "learning_rate": 0.0001,
+      "loss": 1.5122,
+      "step": 14279
+    },
+    {
+      "epoch": 1.6403423123312848,
+      "grad_norm": 0.6524622440338135,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 14280
+    },
+    {
+      "epoch": 1.640457182241112,
+      "grad_norm": 0.6294464468955994,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 14281
+    },
+    {
+      "epoch": 1.640572052150939,
+      "grad_norm": 0.609076201915741,
+      "learning_rate": 0.0001,
+      "loss": 1.4457,
+      "step": 14282
+    },
+    {
+      "epoch": 1.640686922060766,
+      "grad_norm": 0.5797666907310486,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 14283
+    },
+    {
+      "epoch": 1.6408017919705933,
+      "grad_norm": 0.593920111656189,
+      "learning_rate": 0.0001,
+      "loss": 1.3679,
+      "step": 14284
+    },
+    {
+      "epoch": 1.6409166618804205,
+      "grad_norm": 0.6230970025062561,
+      "learning_rate": 0.0001,
+      "loss": 1.4868,
+      "step": 14285
+    },
+    {
+      "epoch": 1.6410315317902475,
+      "grad_norm": 0.6537438631057739,
+      "learning_rate": 0.0001,
+      "loss": 1.5031,
+      "step": 14286
+    },
+    {
+      "epoch": 1.6411464017000745,
+      "grad_norm": 0.5811769962310791,
+      "learning_rate": 0.0001,
+      "loss": 1.3242,
+      "step": 14287
+    },
+    {
+      "epoch": 1.6412612716099018,
+      "grad_norm": 0.6686524748802185,
+      "learning_rate": 0.0001,
+      "loss": 1.6498,
+      "step": 14288
+    },
+    {
+      "epoch": 1.641376141519729,
+      "grad_norm": 0.6294060349464417,
+      "learning_rate": 0.0001,
+      "loss": 1.4349,
+      "step": 14289
+    },
+    {
+      "epoch": 1.641491011429556,
+      "grad_norm": 0.6076821684837341,
+      "learning_rate": 0.0001,
+      "loss": 1.4389,
+      "step": 14290
+    },
+    {
+      "epoch": 1.641605881339383,
+      "grad_norm": 0.6137123107910156,
+      "learning_rate": 0.0001,
+      "loss": 1.4509,
+      "step": 14291
+    },
+    {
+      "epoch": 1.6417207512492102,
+      "grad_norm": 0.5927547812461853,
+      "learning_rate": 0.0001,
+      "loss": 1.3281,
+      "step": 14292
+    },
+    {
+      "epoch": 1.6418356211590375,
+      "grad_norm": 0.6271234750747681,
+      "learning_rate": 0.0001,
+      "loss": 1.4946,
+      "step": 14293
+    },
+    {
+      "epoch": 1.6419504910688645,
+      "grad_norm": 0.6225645542144775,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 14294
+    },
+    {
+      "epoch": 1.6420653609786915,
+      "grad_norm": 0.6010716557502747,
+      "learning_rate": 0.0001,
+      "loss": 1.4282,
+      "step": 14295
+    },
+    {
+      "epoch": 1.6421802308885187,
+      "grad_norm": 0.5875716805458069,
+      "learning_rate": 0.0001,
+      "loss": 1.4042,
+      "step": 14296
+    },
+    {
+      "epoch": 1.642295100798346,
+      "grad_norm": 0.5978166460990906,
+      "learning_rate": 0.0001,
+      "loss": 1.3037,
+      "step": 14297
+    },
+    {
+      "epoch": 1.642409970708173,
+      "grad_norm": 0.6238410472869873,
+      "learning_rate": 0.0001,
+      "loss": 1.5442,
+      "step": 14298
+    },
+    {
+      "epoch": 1.642524840618,
+      "grad_norm": 0.6206856966018677,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 14299
+    },
+    {
+      "epoch": 1.6426397105278272,
+      "grad_norm": 0.6737737655639648,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 14300
+    },
+    {
+      "epoch": 1.6427545804376544,
+      "grad_norm": 0.6297151446342468,
+      "learning_rate": 0.0001,
+      "loss": 1.5897,
+      "step": 14301
+    },
+    {
+      "epoch": 1.6428694503474814,
+      "grad_norm": 0.6496241092681885,
+      "learning_rate": 0.0001,
+      "loss": 1.5686,
+      "step": 14302
+    },
+    {
+      "epoch": 1.6429843202573085,
+      "grad_norm": 0.6007269620895386,
+      "learning_rate": 0.0001,
+      "loss": 1.4899,
+      "step": 14303
+    },
+    {
+      "epoch": 1.6430991901671357,
+      "grad_norm": 0.7573440670967102,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 14304
+    },
+    {
+      "epoch": 1.643214060076963,
+      "grad_norm": 0.6531268954277039,
+      "learning_rate": 0.0001,
+      "loss": 1.44,
+      "step": 14305
+    },
+    {
+      "epoch": 1.64332892998679,
+      "grad_norm": 0.6082087159156799,
+      "learning_rate": 0.0001,
+      "loss": 1.5434,
+      "step": 14306
+    },
+    {
+      "epoch": 1.643443799896617,
+      "grad_norm": 0.641038179397583,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 14307
+    },
+    {
+      "epoch": 1.6435586698064442,
+      "grad_norm": 0.6176828145980835,
+      "learning_rate": 0.0001,
+      "loss": 1.3881,
+      "step": 14308
+    },
+    {
+      "epoch": 1.6436735397162714,
+      "grad_norm": 0.6139614582061768,
+      "learning_rate": 0.0001,
+      "loss": 1.4972,
+      "step": 14309
+    },
+    {
+      "epoch": 1.6437884096260984,
+      "grad_norm": 0.6146692037582397,
+      "learning_rate": 0.0001,
+      "loss": 1.3981,
+      "step": 14310
+    },
+    {
+      "epoch": 1.6439032795359254,
+      "grad_norm": 0.6403799057006836,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 14311
+    },
+    {
+      "epoch": 1.6440181494457526,
+      "grad_norm": 0.6941229104995728,
+      "learning_rate": 0.0001,
+      "loss": 1.3688,
+      "step": 14312
+    },
+    {
+      "epoch": 1.6441330193555799,
+      "grad_norm": 0.6253212094306946,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 14313
+    },
+    {
+      "epoch": 1.6442478892654069,
+      "grad_norm": 0.6304717063903809,
+      "learning_rate": 0.0001,
+      "loss": 1.3255,
+      "step": 14314
+    },
+    {
+      "epoch": 1.644362759175234,
+      "grad_norm": 0.6108677387237549,
+      "learning_rate": 0.0001,
+      "loss": 1.549,
+      "step": 14315
+    },
+    {
+      "epoch": 1.6444776290850611,
+      "grad_norm": 0.7766926288604736,
+      "learning_rate": 0.0001,
+      "loss": 1.6879,
+      "step": 14316
+    },
+    {
+      "epoch": 1.6445924989948884,
+      "grad_norm": 0.6032610535621643,
+      "learning_rate": 0.0001,
+      "loss": 1.3837,
+      "step": 14317
+    },
+    {
+      "epoch": 1.6447073689047154,
+      "grad_norm": 0.6643081307411194,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 14318
+    },
+    {
+      "epoch": 1.6448222388145424,
+      "grad_norm": 0.6397324800491333,
+      "learning_rate": 0.0001,
+      "loss": 1.3905,
+      "step": 14319
+    },
+    {
+      "epoch": 1.6449371087243696,
+      "grad_norm": 0.5864359736442566,
+      "learning_rate": 0.0001,
+      "loss": 1.2971,
+      "step": 14320
+    },
+    {
+      "epoch": 1.6450519786341968,
+      "grad_norm": 0.6553377509117126,
+      "learning_rate": 0.0001,
+      "loss": 1.3454,
+      "step": 14321
+    },
+    {
+      "epoch": 1.6451668485440238,
+      "grad_norm": 0.7063130736351013,
+      "learning_rate": 0.0001,
+      "loss": 1.2324,
+      "step": 14322
+    },
+    {
+      "epoch": 1.6452817184538508,
+      "grad_norm": 0.6475198864936829,
+      "learning_rate": 0.0001,
+      "loss": 1.4901,
+      "step": 14323
+    },
+    {
+      "epoch": 1.645396588363678,
+      "grad_norm": 0.6414164304733276,
+      "learning_rate": 0.0001,
+      "loss": 1.4304,
+      "step": 14324
+    },
+    {
+      "epoch": 1.6455114582735053,
+      "grad_norm": 0.6583581566810608,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 14325
+    },
+    {
+      "epoch": 1.6456263281833323,
+      "grad_norm": 0.6059905290603638,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 14326
+    },
+    {
+      "epoch": 1.6457411980931596,
+      "grad_norm": 0.6345920562744141,
+      "learning_rate": 0.0001,
+      "loss": 1.639,
+      "step": 14327
+    },
+    {
+      "epoch": 1.6458560680029866,
+      "grad_norm": 0.620621919631958,
+      "learning_rate": 0.0001,
+      "loss": 1.5369,
+      "step": 14328
+    },
+    {
+      "epoch": 1.6459709379128138,
+      "grad_norm": 0.6474280953407288,
+      "learning_rate": 0.0001,
+      "loss": 1.4068,
+      "step": 14329
+    },
+    {
+      "epoch": 1.646085807822641,
+      "grad_norm": 0.5968835949897766,
+      "learning_rate": 0.0001,
+      "loss": 1.4271,
+      "step": 14330
+    },
+    {
+      "epoch": 1.646200677732468,
+      "grad_norm": 0.6671695113182068,
+      "learning_rate": 0.0001,
+      "loss": 1.1918,
+      "step": 14331
+    },
+    {
+      "epoch": 1.646315547642295,
+      "grad_norm": 0.622138500213623,
+      "learning_rate": 0.0001,
+      "loss": 1.2681,
+      "step": 14332
+    },
+    {
+      "epoch": 1.6464304175521223,
+      "grad_norm": 0.587945818901062,
+      "learning_rate": 0.0001,
+      "loss": 1.391,
+      "step": 14333
+    },
+    {
+      "epoch": 1.6465452874619495,
+      "grad_norm": 0.6441559791564941,
+      "learning_rate": 0.0001,
+      "loss": 1.5257,
+      "step": 14334
+    },
+    {
+      "epoch": 1.6466601573717765,
+      "grad_norm": 0.6618092656135559,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 14335
+    },
+    {
+      "epoch": 1.6467750272816035,
+      "grad_norm": 0.5723019242286682,
+      "learning_rate": 0.0001,
+      "loss": 1.5012,
+      "step": 14336
+    },
+    {
+      "epoch": 1.6468898971914308,
+      "grad_norm": 0.6029953956604004,
+      "learning_rate": 0.0001,
+      "loss": 1.3561,
+      "step": 14337
+    },
+    {
+      "epoch": 1.647004767101258,
+      "grad_norm": 0.646026074886322,
+      "learning_rate": 0.0001,
+      "loss": 1.5387,
+      "step": 14338
+    },
+    {
+      "epoch": 1.647119637011085,
+      "grad_norm": 0.6190069913864136,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 14339
+    },
+    {
+      "epoch": 1.647234506920912,
+      "grad_norm": 0.6896335482597351,
+      "learning_rate": 0.0001,
+      "loss": 1.6861,
+      "step": 14340
+    },
+    {
+      "epoch": 1.6473493768307392,
+      "grad_norm": 0.611846387386322,
+      "learning_rate": 0.0001,
+      "loss": 1.4243,
+      "step": 14341
+    },
+    {
+      "epoch": 1.6474642467405665,
+      "grad_norm": 0.5718992948532104,
+      "learning_rate": 0.0001,
+      "loss": 1.2889,
+      "step": 14342
+    },
+    {
+      "epoch": 1.6475791166503935,
+      "grad_norm": 0.5699307918548584,
+      "learning_rate": 0.0001,
+      "loss": 1.4659,
+      "step": 14343
+    },
+    {
+      "epoch": 1.6476939865602205,
+      "grad_norm": 0.5619608163833618,
+      "learning_rate": 0.0001,
+      "loss": 1.2261,
+      "step": 14344
+    },
+    {
+      "epoch": 1.6478088564700477,
+      "grad_norm": 0.6159090399742126,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 14345
+    },
+    {
+      "epoch": 1.647923726379875,
+      "grad_norm": 0.5962672233581543,
+      "learning_rate": 0.0001,
+      "loss": 1.1608,
+      "step": 14346
+    },
+    {
+      "epoch": 1.648038596289702,
+      "grad_norm": 0.6491231918334961,
+      "learning_rate": 0.0001,
+      "loss": 1.4453,
+      "step": 14347
+    },
+    {
+      "epoch": 1.648153466199529,
+      "grad_norm": 0.6284507513046265,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 14348
+    },
+    {
+      "epoch": 1.6482683361093562,
+      "grad_norm": 0.5969189405441284,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 14349
+    },
+    {
+      "epoch": 1.6483832060191834,
+      "grad_norm": 0.5988194346427917,
+      "learning_rate": 0.0001,
+      "loss": 1.352,
+      "step": 14350
+    },
+    {
+      "epoch": 1.6484980759290104,
+      "grad_norm": 0.6822172999382019,
+      "learning_rate": 0.0001,
+      "loss": 1.5764,
+      "step": 14351
+    },
+    {
+      "epoch": 1.6486129458388374,
+      "grad_norm": 0.6363539695739746,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 14352
+    },
+    {
+      "epoch": 1.6487278157486647,
+      "grad_norm": 0.65143221616745,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 14353
+    },
+    {
+      "epoch": 1.648842685658492,
+      "grad_norm": 0.6166749000549316,
+      "learning_rate": 0.0001,
+      "loss": 1.4744,
+      "step": 14354
+    },
+    {
+      "epoch": 1.648957555568319,
+      "grad_norm": 0.6183647513389587,
+      "learning_rate": 0.0001,
+      "loss": 1.4902,
+      "step": 14355
+    },
+    {
+      "epoch": 1.649072425478146,
+      "grad_norm": 0.5944738984107971,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 14356
+    },
+    {
+      "epoch": 1.6491872953879732,
+      "grad_norm": 0.6394641995429993,
+      "learning_rate": 0.0001,
+      "loss": 1.3743,
+      "step": 14357
+    },
+    {
+      "epoch": 1.6493021652978004,
+      "grad_norm": 0.6612902283668518,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 14358
+    },
+    {
+      "epoch": 1.6494170352076274,
+      "grad_norm": 0.6034130454063416,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 14359
+    },
+    {
+      "epoch": 1.6495319051174544,
+      "grad_norm": 0.5944687128067017,
+      "learning_rate": 0.0001,
+      "loss": 1.4008,
+      "step": 14360
+    },
+    {
+      "epoch": 1.6496467750272816,
+      "grad_norm": 0.6049661040306091,
+      "learning_rate": 0.0001,
+      "loss": 1.5983,
+      "step": 14361
+    },
+    {
+      "epoch": 1.6497616449371089,
+      "grad_norm": 0.5754156112670898,
+      "learning_rate": 0.0001,
+      "loss": 1.4454,
+      "step": 14362
+    },
+    {
+      "epoch": 1.6498765148469359,
+      "grad_norm": 0.6179141998291016,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 14363
+    },
+    {
+      "epoch": 1.6499913847567629,
+      "grad_norm": 0.5943996906280518,
+      "learning_rate": 0.0001,
+      "loss": 1.4576,
+      "step": 14364
+    },
+    {
+      "epoch": 1.65010625466659,
+      "grad_norm": 0.6303023099899292,
+      "learning_rate": 0.0001,
+      "loss": 1.4891,
+      "step": 14365
+    },
+    {
+      "epoch": 1.6502211245764173,
+      "grad_norm": 0.5632081627845764,
+      "learning_rate": 0.0001,
+      "loss": 1.3568,
+      "step": 14366
+    },
+    {
+      "epoch": 1.6503359944862444,
+      "grad_norm": 0.574306070804596,
+      "learning_rate": 0.0001,
+      "loss": 1.2801,
+      "step": 14367
+    },
+    {
+      "epoch": 1.6504508643960714,
+      "grad_norm": 0.5834609270095825,
+      "learning_rate": 0.0001,
+      "loss": 1.3699,
+      "step": 14368
+    },
+    {
+      "epoch": 1.6505657343058986,
+      "grad_norm": 0.5941557288169861,
+      "learning_rate": 0.0001,
+      "loss": 1.398,
+      "step": 14369
+    },
+    {
+      "epoch": 1.6506806042157258,
+      "grad_norm": 0.6643593907356262,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 14370
+    },
+    {
+      "epoch": 1.6507954741255528,
+      "grad_norm": 0.5751389861106873,
+      "learning_rate": 0.0001,
+      "loss": 1.4632,
+      "step": 14371
+    },
+    {
+      "epoch": 1.6509103440353798,
+      "grad_norm": 0.6634930968284607,
+      "learning_rate": 0.0001,
+      "loss": 1.6016,
+      "step": 14372
+    },
+    {
+      "epoch": 1.651025213945207,
+      "grad_norm": 0.6502936482429504,
+      "learning_rate": 0.0001,
+      "loss": 1.4685,
+      "step": 14373
+    },
+    {
+      "epoch": 1.6511400838550343,
+      "grad_norm": 0.6059218049049377,
+      "learning_rate": 0.0001,
+      "loss": 1.4038,
+      "step": 14374
+    },
+    {
+      "epoch": 1.6512549537648613,
+      "grad_norm": 0.6395809650421143,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 14375
+    },
+    {
+      "epoch": 1.6513698236746883,
+      "grad_norm": 0.614824116230011,
+      "learning_rate": 0.0001,
+      "loss": 1.3675,
+      "step": 14376
+    },
+    {
+      "epoch": 1.6514846935845156,
+      "grad_norm": 0.6158643364906311,
+      "learning_rate": 0.0001,
+      "loss": 1.6117,
+      "step": 14377
+    },
+    {
+      "epoch": 1.6515995634943428,
+      "grad_norm": 0.5627233982086182,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 14378
+    },
+    {
+      "epoch": 1.6517144334041698,
+      "grad_norm": 0.7553406953811646,
+      "learning_rate": 0.0001,
+      "loss": 1.4885,
+      "step": 14379
+    },
+    {
+      "epoch": 1.6518293033139968,
+      "grad_norm": 0.5953923463821411,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 14380
+    },
+    {
+      "epoch": 1.651944173223824,
+      "grad_norm": 0.6409009695053101,
+      "learning_rate": 0.0001,
+      "loss": 1.4745,
+      "step": 14381
+    },
+    {
+      "epoch": 1.6520590431336513,
+      "grad_norm": 0.5870394110679626,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 14382
+    },
+    {
+      "epoch": 1.6521739130434783,
+      "grad_norm": 0.6005561947822571,
+      "learning_rate": 0.0001,
+      "loss": 1.5428,
+      "step": 14383
+    },
+    {
+      "epoch": 1.6522887829533053,
+      "grad_norm": 0.651766836643219,
+      "learning_rate": 0.0001,
+      "loss": 1.4715,
+      "step": 14384
+    },
+    {
+      "epoch": 1.6524036528631325,
+      "grad_norm": 0.5785371661186218,
+      "learning_rate": 0.0001,
+      "loss": 1.5496,
+      "step": 14385
+    },
+    {
+      "epoch": 1.6525185227729597,
+      "grad_norm": 0.6617969274520874,
+      "learning_rate": 0.0001,
+      "loss": 1.4035,
+      "step": 14386
+    },
+    {
+      "epoch": 1.6526333926827868,
+      "grad_norm": 0.6037518382072449,
+      "learning_rate": 0.0001,
+      "loss": 1.5064,
+      "step": 14387
+    },
+    {
+      "epoch": 1.6527482625926138,
+      "grad_norm": 0.6930570006370544,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 14388
+    },
+    {
+      "epoch": 1.652863132502441,
+      "grad_norm": 0.6605214476585388,
+      "learning_rate": 0.0001,
+      "loss": 1.4112,
+      "step": 14389
+    },
+    {
+      "epoch": 1.6529780024122682,
+      "grad_norm": 0.6189714074134827,
+      "learning_rate": 0.0001,
+      "loss": 1.3354,
+      "step": 14390
+    },
+    {
+      "epoch": 1.6530928723220952,
+      "grad_norm": 0.6060426235198975,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 14391
+    },
+    {
+      "epoch": 1.6532077422319222,
+      "grad_norm": 0.5811141729354858,
+      "learning_rate": 0.0001,
+      "loss": 1.2768,
+      "step": 14392
+    },
+    {
+      "epoch": 1.6533226121417495,
+      "grad_norm": 0.608869731426239,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 14393
+    },
+    {
+      "epoch": 1.6534374820515767,
+      "grad_norm": 0.6065149307250977,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 14394
+    },
+    {
+      "epoch": 1.6535523519614037,
+      "grad_norm": 0.7323920726776123,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 14395
+    },
+    {
+      "epoch": 1.6536672218712307,
+      "grad_norm": 0.604194700717926,
+      "learning_rate": 0.0001,
+      "loss": 1.4543,
+      "step": 14396
+    },
+    {
+      "epoch": 1.653782091781058,
+      "grad_norm": 0.6321179866790771,
+      "learning_rate": 0.0001,
+      "loss": 1.4389,
+      "step": 14397
+    },
+    {
+      "epoch": 1.6538969616908852,
+      "grad_norm": 0.6846322417259216,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 14398
+    },
+    {
+      "epoch": 1.6540118316007122,
+      "grad_norm": 0.6531241536140442,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 14399
+    },
+    {
+      "epoch": 1.6541267015105392,
+      "grad_norm": 0.5736318826675415,
+      "learning_rate": 0.0001,
+      "loss": 1.359,
+      "step": 14400
+    },
+    {
+      "epoch": 1.6542415714203664,
+      "grad_norm": 0.5621281266212463,
+      "learning_rate": 0.0001,
+      "loss": 1.2091,
+      "step": 14401
+    },
+    {
+      "epoch": 1.6543564413301937,
+      "grad_norm": 0.6186012029647827,
+      "learning_rate": 0.0001,
+      "loss": 1.3328,
+      "step": 14402
+    },
+    {
+      "epoch": 1.6544713112400207,
+      "grad_norm": 0.6069703698158264,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 14403
+    },
+    {
+      "epoch": 1.6545861811498477,
+      "grad_norm": 0.6218054890632629,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 14404
+    },
+    {
+      "epoch": 1.654701051059675,
+      "grad_norm": 0.6225132346153259,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 14405
+    },
+    {
+      "epoch": 1.6548159209695021,
+      "grad_norm": 0.6091734170913696,
+      "learning_rate": 0.0001,
+      "loss": 1.3794,
+      "step": 14406
+    },
+    {
+      "epoch": 1.6549307908793291,
+      "grad_norm": 0.569509744644165,
+      "learning_rate": 0.0001,
+      "loss": 1.2694,
+      "step": 14407
+    },
+    {
+      "epoch": 1.6550456607891562,
+      "grad_norm": 0.690952718257904,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 14408
+    },
+    {
+      "epoch": 1.6551605306989834,
+      "grad_norm": 0.6232033371925354,
+      "learning_rate": 0.0001,
+      "loss": 1.3905,
+      "step": 14409
+    },
+    {
+      "epoch": 1.6552754006088106,
+      "grad_norm": 0.6081985235214233,
+      "learning_rate": 0.0001,
+      "loss": 1.2826,
+      "step": 14410
+    },
+    {
+      "epoch": 1.6553902705186376,
+      "grad_norm": 0.6122736930847168,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 14411
+    },
+    {
+      "epoch": 1.6555051404284646,
+      "grad_norm": 0.5784592628479004,
+      "learning_rate": 0.0001,
+      "loss": 1.6078,
+      "step": 14412
+    },
+    {
+      "epoch": 1.6556200103382919,
+      "grad_norm": 0.5675357580184937,
+      "learning_rate": 0.0001,
+      "loss": 1.4174,
+      "step": 14413
+    },
+    {
+      "epoch": 1.655734880248119,
+      "grad_norm": 0.5335760116577148,
+      "learning_rate": 0.0001,
+      "loss": 1.1964,
+      "step": 14414
+    },
+    {
+      "epoch": 1.655849750157946,
+      "grad_norm": 0.604422390460968,
+      "learning_rate": 0.0001,
+      "loss": 1.4652,
+      "step": 14415
+    },
+    {
+      "epoch": 1.6559646200677731,
+      "grad_norm": 0.6323688626289368,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 14416
+    },
+    {
+      "epoch": 1.6560794899776003,
+      "grad_norm": 0.5734161734580994,
+      "learning_rate": 0.0001,
+      "loss": 1.3093,
+      "step": 14417
+    },
+    {
+      "epoch": 1.6561943598874276,
+      "grad_norm": 0.6848605275154114,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 14418
+    },
+    {
+      "epoch": 1.6563092297972546,
+      "grad_norm": 0.5923926830291748,
+      "learning_rate": 0.0001,
+      "loss": 1.3412,
+      "step": 14419
+    },
+    {
+      "epoch": 1.6564240997070816,
+      "grad_norm": 0.7171801328659058,
+      "learning_rate": 0.0001,
+      "loss": 1.4567,
+      "step": 14420
+    },
+    {
+      "epoch": 1.6565389696169088,
+      "grad_norm": 0.6611701846122742,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 14421
+    },
+    {
+      "epoch": 1.656653839526736,
+      "grad_norm": 0.6520163416862488,
+      "learning_rate": 0.0001,
+      "loss": 1.495,
+      "step": 14422
+    },
+    {
+      "epoch": 1.656768709436563,
+      "grad_norm": 0.6139830350875854,
+      "learning_rate": 0.0001,
+      "loss": 1.6852,
+      "step": 14423
+    },
+    {
+      "epoch": 1.65688357934639,
+      "grad_norm": 0.5797949433326721,
+      "learning_rate": 0.0001,
+      "loss": 1.3619,
+      "step": 14424
+    },
+    {
+      "epoch": 1.6569984492562173,
+      "grad_norm": 0.5873029828071594,
+      "learning_rate": 0.0001,
+      "loss": 1.3233,
+      "step": 14425
+    },
+    {
+      "epoch": 1.6571133191660445,
+      "grad_norm": 0.6242226362228394,
+      "learning_rate": 0.0001,
+      "loss": 1.2352,
+      "step": 14426
+    },
+    {
+      "epoch": 1.6572281890758715,
+      "grad_norm": 0.589346170425415,
+      "learning_rate": 0.0001,
+      "loss": 1.327,
+      "step": 14427
+    },
+    {
+      "epoch": 1.6573430589856986,
+      "grad_norm": 0.6410361528396606,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 14428
+    },
+    {
+      "epoch": 1.6574579288955258,
+      "grad_norm": 0.5922463536262512,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 14429
+    },
+    {
+      "epoch": 1.657572798805353,
+      "grad_norm": 0.5775658488273621,
+      "learning_rate": 0.0001,
+      "loss": 1.2298,
+      "step": 14430
+    },
+    {
+      "epoch": 1.65768766871518,
+      "grad_norm": 0.7526030540466309,
+      "learning_rate": 0.0001,
+      "loss": 1.5982,
+      "step": 14431
+    },
+    {
+      "epoch": 1.657802538625007,
+      "grad_norm": 0.5999011993408203,
+      "learning_rate": 0.0001,
+      "loss": 1.3619,
+      "step": 14432
+    },
+    {
+      "epoch": 1.6579174085348343,
+      "grad_norm": 0.6628460884094238,
+      "learning_rate": 0.0001,
+      "loss": 1.6252,
+      "step": 14433
+    },
+    {
+      "epoch": 1.6580322784446615,
+      "grad_norm": 0.6481192708015442,
+      "learning_rate": 0.0001,
+      "loss": 1.4441,
+      "step": 14434
+    },
+    {
+      "epoch": 1.6581471483544885,
+      "grad_norm": 0.6300088167190552,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 14435
+    },
+    {
+      "epoch": 1.6582620182643155,
+      "grad_norm": 0.5989202260971069,
+      "learning_rate": 0.0001,
+      "loss": 1.2056,
+      "step": 14436
+    },
+    {
+      "epoch": 1.6583768881741427,
+      "grad_norm": 0.7720966339111328,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 14437
+    },
+    {
+      "epoch": 1.65849175808397,
+      "grad_norm": 0.6756711602210999,
+      "learning_rate": 0.0001,
+      "loss": 1.4533,
+      "step": 14438
+    },
+    {
+      "epoch": 1.658606627993797,
+      "grad_norm": 0.6091902852058411,
+      "learning_rate": 0.0001,
+      "loss": 1.4482,
+      "step": 14439
+    },
+    {
+      "epoch": 1.658721497903624,
+      "grad_norm": 0.6059051156044006,
+      "learning_rate": 0.0001,
+      "loss": 1.4326,
+      "step": 14440
+    },
+    {
+      "epoch": 1.6588363678134512,
+      "grad_norm": 0.6013321876525879,
+      "learning_rate": 0.0001,
+      "loss": 1.352,
+      "step": 14441
+    },
+    {
+      "epoch": 1.6589512377232785,
+      "grad_norm": 0.5931786298751831,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 14442
+    },
+    {
+      "epoch": 1.6590661076331055,
+      "grad_norm": 0.5912080407142639,
+      "learning_rate": 0.0001,
+      "loss": 1.4353,
+      "step": 14443
+    },
+    {
+      "epoch": 1.6591809775429325,
+      "grad_norm": 0.614547073841095,
+      "learning_rate": 0.0001,
+      "loss": 1.4778,
+      "step": 14444
+    },
+    {
+      "epoch": 1.6592958474527597,
+      "grad_norm": 0.6366251707077026,
+      "learning_rate": 0.0001,
+      "loss": 1.4466,
+      "step": 14445
+    },
+    {
+      "epoch": 1.659410717362587,
+      "grad_norm": 0.5798001289367676,
+      "learning_rate": 0.0001,
+      "loss": 1.3829,
+      "step": 14446
+    },
+    {
+      "epoch": 1.659525587272414,
+      "grad_norm": 0.5841000080108643,
+      "learning_rate": 0.0001,
+      "loss": 1.5427,
+      "step": 14447
+    },
+    {
+      "epoch": 1.659640457182241,
+      "grad_norm": 0.6051506400108337,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 14448
+    },
+    {
+      "epoch": 1.6597553270920682,
+      "grad_norm": 0.5879446864128113,
+      "learning_rate": 0.0001,
+      "loss": 1.4215,
+      "step": 14449
+    },
+    {
+      "epoch": 1.6598701970018954,
+      "grad_norm": 0.661629855632782,
+      "learning_rate": 0.0001,
+      "loss": 1.4418,
+      "step": 14450
+    },
+    {
+      "epoch": 1.6599850669117224,
+      "grad_norm": 0.6741747856140137,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 14451
+    },
+    {
+      "epoch": 1.6600999368215494,
+      "grad_norm": 0.5925803184509277,
+      "learning_rate": 0.0001,
+      "loss": 1.3766,
+      "step": 14452
+    },
+    {
+      "epoch": 1.6602148067313767,
+      "grad_norm": 0.6182143688201904,
+      "learning_rate": 0.0001,
+      "loss": 1.4005,
+      "step": 14453
+    },
+    {
+      "epoch": 1.660329676641204,
+      "grad_norm": 0.6143870949745178,
+      "learning_rate": 0.0001,
+      "loss": 1.4139,
+      "step": 14454
+    },
+    {
+      "epoch": 1.660444546551031,
+      "grad_norm": 0.6046828031539917,
+      "learning_rate": 0.0001,
+      "loss": 1.3055,
+      "step": 14455
+    },
+    {
+      "epoch": 1.660559416460858,
+      "grad_norm": 0.767876148223877,
+      "learning_rate": 0.0001,
+      "loss": 1.5525,
+      "step": 14456
+    },
+    {
+      "epoch": 1.6606742863706851,
+      "grad_norm": 0.6341820359230042,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 14457
+    },
+    {
+      "epoch": 1.6607891562805124,
+      "grad_norm": 0.6907703876495361,
+      "learning_rate": 0.0001,
+      "loss": 1.3277,
+      "step": 14458
+    },
+    {
+      "epoch": 1.6609040261903394,
+      "grad_norm": 0.6609397530555725,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 14459
+    },
+    {
+      "epoch": 1.6610188961001664,
+      "grad_norm": 0.6220353245735168,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 14460
+    },
+    {
+      "epoch": 1.6611337660099936,
+      "grad_norm": 0.5976139903068542,
+      "learning_rate": 0.0001,
+      "loss": 1.4258,
+      "step": 14461
+    },
+    {
+      "epoch": 1.6612486359198209,
+      "grad_norm": 0.654104471206665,
+      "learning_rate": 0.0001,
+      "loss": 1.53,
+      "step": 14462
+    },
+    {
+      "epoch": 1.6613635058296479,
+      "grad_norm": 0.6497766971588135,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 14463
+    },
+    {
+      "epoch": 1.661478375739475,
+      "grad_norm": 0.5906299948692322,
+      "learning_rate": 0.0001,
+      "loss": 1.5512,
+      "step": 14464
+    },
+    {
+      "epoch": 1.661593245649302,
+      "grad_norm": 0.6207028031349182,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 14465
+    },
+    {
+      "epoch": 1.6617081155591293,
+      "grad_norm": 0.6227574348449707,
+      "learning_rate": 0.0001,
+      "loss": 1.4097,
+      "step": 14466
+    },
+    {
+      "epoch": 1.6618229854689566,
+      "grad_norm": 0.7923548817634583,
+      "learning_rate": 0.0001,
+      "loss": 1.0432,
+      "step": 14467
+    },
+    {
+      "epoch": 1.6619378553787836,
+      "grad_norm": 0.6193511486053467,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 14468
+    },
+    {
+      "epoch": 1.6620527252886106,
+      "grad_norm": 0.6481901407241821,
+      "learning_rate": 0.0001,
+      "loss": 1.2574,
+      "step": 14469
+    },
+    {
+      "epoch": 1.6621675951984378,
+      "grad_norm": 0.7560790181159973,
+      "learning_rate": 0.0001,
+      "loss": 1.7806,
+      "step": 14470
+    },
+    {
+      "epoch": 1.662282465108265,
+      "grad_norm": 0.6491854786872864,
+      "learning_rate": 0.0001,
+      "loss": 1.6269,
+      "step": 14471
+    },
+    {
+      "epoch": 1.662397335018092,
+      "grad_norm": 0.6301493048667908,
+      "learning_rate": 0.0001,
+      "loss": 1.4367,
+      "step": 14472
+    },
+    {
+      "epoch": 1.662512204927919,
+      "grad_norm": 0.6142545342445374,
+      "learning_rate": 0.0001,
+      "loss": 1.5011,
+      "step": 14473
+    },
+    {
+      "epoch": 1.6626270748377463,
+      "grad_norm": 0.6091681718826294,
+      "learning_rate": 0.0001,
+      "loss": 1.4234,
+      "step": 14474
+    },
+    {
+      "epoch": 1.6627419447475735,
+      "grad_norm": 0.5943846106529236,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 14475
+    },
+    {
+      "epoch": 1.6628568146574005,
+      "grad_norm": 0.7121033668518066,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 14476
+    },
+    {
+      "epoch": 1.6629716845672275,
+      "grad_norm": 0.575641930103302,
+      "learning_rate": 0.0001,
+      "loss": 1.283,
+      "step": 14477
+    },
+    {
+      "epoch": 1.6630865544770548,
+      "grad_norm": 0.62906414270401,
+      "learning_rate": 0.0001,
+      "loss": 1.6762,
+      "step": 14478
+    },
+    {
+      "epoch": 1.663201424386882,
+      "grad_norm": 0.5758494734764099,
+      "learning_rate": 0.0001,
+      "loss": 1.3509,
+      "step": 14479
+    },
+    {
+      "epoch": 1.663316294296709,
+      "grad_norm": 0.581095278263092,
+      "learning_rate": 0.0001,
+      "loss": 1.376,
+      "step": 14480
+    },
+    {
+      "epoch": 1.663431164206536,
+      "grad_norm": 0.6003084182739258,
+      "learning_rate": 0.0001,
+      "loss": 1.4864,
+      "step": 14481
+    },
+    {
+      "epoch": 1.6635460341163633,
+      "grad_norm": 0.6477168202400208,
+      "learning_rate": 0.0001,
+      "loss": 1.3539,
+      "step": 14482
+    },
+    {
+      "epoch": 1.6636609040261905,
+      "grad_norm": 0.6288648843765259,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 14483
+    },
+    {
+      "epoch": 1.6637757739360175,
+      "grad_norm": 0.6509358882904053,
+      "learning_rate": 0.0001,
+      "loss": 1.3245,
+      "step": 14484
+    },
+    {
+      "epoch": 1.6638906438458445,
+      "grad_norm": 0.6689729690551758,
+      "learning_rate": 0.0001,
+      "loss": 1.4742,
+      "step": 14485
+    },
+    {
+      "epoch": 1.6640055137556717,
+      "grad_norm": 0.6662051677703857,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 14486
+    },
+    {
+      "epoch": 1.664120383665499,
+      "grad_norm": 0.6444410681724548,
+      "learning_rate": 0.0001,
+      "loss": 1.5203,
+      "step": 14487
+    },
+    {
+      "epoch": 1.664235253575326,
+      "grad_norm": 0.5951128602027893,
+      "learning_rate": 0.0001,
+      "loss": 1.4918,
+      "step": 14488
+    },
+    {
+      "epoch": 1.664350123485153,
+      "grad_norm": 0.6148644089698792,
+      "learning_rate": 0.0001,
+      "loss": 1.2399,
+      "step": 14489
+    },
+    {
+      "epoch": 1.6644649933949802,
+      "grad_norm": 0.6238715052604675,
+      "learning_rate": 0.0001,
+      "loss": 1.6374,
+      "step": 14490
+    },
+    {
+      "epoch": 1.6645798633048074,
+      "grad_norm": 0.6118984818458557,
+      "learning_rate": 0.0001,
+      "loss": 1.5081,
+      "step": 14491
+    },
+    {
+      "epoch": 1.6646947332146345,
+      "grad_norm": 0.6418251991271973,
+      "learning_rate": 0.0001,
+      "loss": 1.4167,
+      "step": 14492
+    },
+    {
+      "epoch": 1.6648096031244615,
+      "grad_norm": 0.6296436786651611,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 14493
+    },
+    {
+      "epoch": 1.6649244730342887,
+      "grad_norm": 0.6078484654426575,
+      "learning_rate": 0.0001,
+      "loss": 1.4571,
+      "step": 14494
+    },
+    {
+      "epoch": 1.665039342944116,
+      "grad_norm": 0.5940951108932495,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 14495
+    },
+    {
+      "epoch": 1.665154212853943,
+      "grad_norm": 0.6265494227409363,
+      "learning_rate": 0.0001,
+      "loss": 1.4058,
+      "step": 14496
+    },
+    {
+      "epoch": 1.66526908276377,
+      "grad_norm": 0.551169753074646,
+      "learning_rate": 0.0001,
+      "loss": 1.4444,
+      "step": 14497
+    },
+    {
+      "epoch": 1.6653839526735972,
+      "grad_norm": 0.6252605319023132,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 14498
+    },
+    {
+      "epoch": 1.6654988225834244,
+      "grad_norm": 0.6075454950332642,
+      "learning_rate": 0.0001,
+      "loss": 1.4124,
+      "step": 14499
+    },
+    {
+      "epoch": 1.6656136924932514,
+      "grad_norm": 0.6732686161994934,
+      "learning_rate": 0.0001,
+      "loss": 1.6495,
+      "step": 14500
+    },
+    {
+      "epoch": 1.6657285624030784,
+      "grad_norm": 0.6578999161720276,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 14501
+    },
+    {
+      "epoch": 1.6658434323129057,
+      "grad_norm": 0.6686480045318604,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 14502
+    },
+    {
+      "epoch": 1.665958302222733,
+      "grad_norm": 0.5849835276603699,
+      "learning_rate": 0.0001,
+      "loss": 1.3953,
+      "step": 14503
+    },
+    {
+      "epoch": 1.66607317213256,
+      "grad_norm": 0.6391957402229309,
+      "learning_rate": 0.0001,
+      "loss": 1.5159,
+      "step": 14504
+    },
+    {
+      "epoch": 1.666188042042387,
+      "grad_norm": 0.6156760454177856,
+      "learning_rate": 0.0001,
+      "loss": 1.2013,
+      "step": 14505
+    },
+    {
+      "epoch": 1.6663029119522141,
+      "grad_norm": 0.627495527267456,
+      "learning_rate": 0.0001,
+      "loss": 1.7254,
+      "step": 14506
+    },
+    {
+      "epoch": 1.6664177818620414,
+      "grad_norm": 0.6318630576133728,
+      "learning_rate": 0.0001,
+      "loss": 1.495,
+      "step": 14507
+    },
+    {
+      "epoch": 1.6665326517718684,
+      "grad_norm": 0.6174932718276978,
+      "learning_rate": 0.0001,
+      "loss": 1.508,
+      "step": 14508
+    },
+    {
+      "epoch": 1.6666475216816954,
+      "grad_norm": 0.6004678606987,
+      "learning_rate": 0.0001,
+      "loss": 1.3112,
+      "step": 14509
+    },
+    {
+      "epoch": 1.6667623915915226,
+      "grad_norm": 0.5961777567863464,
+      "learning_rate": 0.0001,
+      "loss": 1.3316,
+      "step": 14510
+    },
+    {
+      "epoch": 1.6668772615013498,
+      "grad_norm": 0.6679588556289673,
+      "learning_rate": 0.0001,
+      "loss": 1.4856,
+      "step": 14511
+    },
+    {
+      "epoch": 1.6669921314111769,
+      "grad_norm": 0.6514487862586975,
+      "learning_rate": 0.0001,
+      "loss": 1.5705,
+      "step": 14512
+    },
+    {
+      "epoch": 1.6671070013210039,
+      "grad_norm": 0.6601291298866272,
+      "learning_rate": 0.0001,
+      "loss": 1.4138,
+      "step": 14513
+    },
+    {
+      "epoch": 1.667221871230831,
+      "grad_norm": 0.6563011407852173,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 14514
+    },
+    {
+      "epoch": 1.6673367411406583,
+      "grad_norm": 0.6405079960823059,
+      "learning_rate": 0.0001,
+      "loss": 1.484,
+      "step": 14515
+    },
+    {
+      "epoch": 1.6674516110504853,
+      "grad_norm": 0.6541486978530884,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 14516
+    },
+    {
+      "epoch": 1.6675664809603123,
+      "grad_norm": 0.5867092609405518,
+      "learning_rate": 0.0001,
+      "loss": 1.4034,
+      "step": 14517
+    },
+    {
+      "epoch": 1.6676813508701396,
+      "grad_norm": 0.5766440033912659,
+      "learning_rate": 0.0001,
+      "loss": 1.4268,
+      "step": 14518
+    },
+    {
+      "epoch": 1.6677962207799668,
+      "grad_norm": 0.5653627514839172,
+      "learning_rate": 0.0001,
+      "loss": 1.3912,
+      "step": 14519
+    },
+    {
+      "epoch": 1.6679110906897938,
+      "grad_norm": 0.6476209759712219,
+      "learning_rate": 0.0001,
+      "loss": 1.5981,
+      "step": 14520
+    },
+    {
+      "epoch": 1.6680259605996208,
+      "grad_norm": 0.5878192186355591,
+      "learning_rate": 0.0001,
+      "loss": 1.3328,
+      "step": 14521
+    },
+    {
+      "epoch": 1.668140830509448,
+      "grad_norm": 0.5736682415008545,
+      "learning_rate": 0.0001,
+      "loss": 1.3202,
+      "step": 14522
+    },
+    {
+      "epoch": 1.6682557004192753,
+      "grad_norm": 0.6797529458999634,
+      "learning_rate": 0.0001,
+      "loss": 1.7396,
+      "step": 14523
+    },
+    {
+      "epoch": 1.6683705703291023,
+      "grad_norm": 0.6123679876327515,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 14524
+    },
+    {
+      "epoch": 1.6684854402389293,
+      "grad_norm": 0.567111074924469,
+      "learning_rate": 0.0001,
+      "loss": 1.4072,
+      "step": 14525
+    },
+    {
+      "epoch": 1.6686003101487565,
+      "grad_norm": 0.5874302983283997,
+      "learning_rate": 0.0001,
+      "loss": 1.3295,
+      "step": 14526
+    },
+    {
+      "epoch": 1.6687151800585838,
+      "grad_norm": 0.6208910346031189,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 14527
+    },
+    {
+      "epoch": 1.6688300499684108,
+      "grad_norm": 0.5815864205360413,
+      "learning_rate": 0.0001,
+      "loss": 1.2403,
+      "step": 14528
+    },
+    {
+      "epoch": 1.6689449198782378,
+      "grad_norm": 0.6493234634399414,
+      "learning_rate": 0.0001,
+      "loss": 1.4954,
+      "step": 14529
+    },
+    {
+      "epoch": 1.669059789788065,
+      "grad_norm": 0.556931734085083,
+      "learning_rate": 0.0001,
+      "loss": 1.3694,
+      "step": 14530
+    },
+    {
+      "epoch": 1.6691746596978922,
+      "grad_norm": 0.6887297034263611,
+      "learning_rate": 0.0001,
+      "loss": 1.6801,
+      "step": 14531
+    },
+    {
+      "epoch": 1.6692895296077193,
+      "grad_norm": 0.688178300857544,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 14532
+    },
+    {
+      "epoch": 1.6694043995175463,
+      "grad_norm": 0.6757493019104004,
+      "learning_rate": 0.0001,
+      "loss": 1.5948,
+      "step": 14533
+    },
+    {
+      "epoch": 1.6695192694273735,
+      "grad_norm": 0.5804457068443298,
+      "learning_rate": 0.0001,
+      "loss": 1.3788,
+      "step": 14534
+    },
+    {
+      "epoch": 1.6696341393372007,
+      "grad_norm": 0.5876379013061523,
+      "learning_rate": 0.0001,
+      "loss": 1.1737,
+      "step": 14535
+    },
+    {
+      "epoch": 1.6697490092470277,
+      "grad_norm": 0.6323837637901306,
+      "learning_rate": 0.0001,
+      "loss": 1.3878,
+      "step": 14536
+    },
+    {
+      "epoch": 1.6698638791568547,
+      "grad_norm": 0.6448050141334534,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 14537
+    },
+    {
+      "epoch": 1.669978749066682,
+      "grad_norm": 0.6202848553657532,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 14538
+    },
+    {
+      "epoch": 1.6700936189765092,
+      "grad_norm": 0.6006600856781006,
+      "learning_rate": 0.0001,
+      "loss": 1.5798,
+      "step": 14539
+    },
+    {
+      "epoch": 1.6702084888863362,
+      "grad_norm": 0.6253841519355774,
+      "learning_rate": 0.0001,
+      "loss": 1.361,
+      "step": 14540
+    },
+    {
+      "epoch": 1.6703233587961632,
+      "grad_norm": 0.6643701791763306,
+      "learning_rate": 0.0001,
+      "loss": 1.3992,
+      "step": 14541
+    },
+    {
+      "epoch": 1.6704382287059905,
+      "grad_norm": 0.6138646602630615,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 14542
+    },
+    {
+      "epoch": 1.6705530986158177,
+      "grad_norm": 0.614280641078949,
+      "learning_rate": 0.0001,
+      "loss": 1.4676,
+      "step": 14543
+    },
+    {
+      "epoch": 1.6706679685256447,
+      "grad_norm": 0.5600206851959229,
+      "learning_rate": 0.0001,
+      "loss": 1.3489,
+      "step": 14544
+    },
+    {
+      "epoch": 1.6707828384354717,
+      "grad_norm": 0.6660794615745544,
+      "learning_rate": 0.0001,
+      "loss": 1.6009,
+      "step": 14545
+    },
+    {
+      "epoch": 1.670897708345299,
+      "grad_norm": 0.6051227450370789,
+      "learning_rate": 0.0001,
+      "loss": 1.4396,
+      "step": 14546
+    },
+    {
+      "epoch": 1.6710125782551262,
+      "grad_norm": 0.5952056646347046,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 14547
+    },
+    {
+      "epoch": 1.6711274481649532,
+      "grad_norm": 0.8179489374160767,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 14548
+    },
+    {
+      "epoch": 1.6712423180747802,
+      "grad_norm": 0.5879932045936584,
+      "learning_rate": 0.0001,
+      "loss": 1.2387,
+      "step": 14549
+    },
+    {
+      "epoch": 1.6713571879846074,
+      "grad_norm": 0.6535231471061707,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 14550
+    },
+    {
+      "epoch": 1.6714720578944346,
+      "grad_norm": 0.6634051203727722,
+      "learning_rate": 0.0001,
+      "loss": 1.5402,
+      "step": 14551
+    },
+    {
+      "epoch": 1.6715869278042617,
+      "grad_norm": 0.6484226584434509,
+      "learning_rate": 0.0001,
+      "loss": 1.4585,
+      "step": 14552
+    },
+    {
+      "epoch": 1.6717017977140887,
+      "grad_norm": 0.6908429861068726,
+      "learning_rate": 0.0001,
+      "loss": 1.4969,
+      "step": 14553
+    },
+    {
+      "epoch": 1.671816667623916,
+      "grad_norm": 0.6413170695304871,
+      "learning_rate": 0.0001,
+      "loss": 1.3641,
+      "step": 14554
+    },
+    {
+      "epoch": 1.6719315375337431,
+      "grad_norm": 0.6127830147743225,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 14555
+    },
+    {
+      "epoch": 1.6720464074435701,
+      "grad_norm": 0.6318410038948059,
+      "learning_rate": 0.0001,
+      "loss": 1.5818,
+      "step": 14556
+    },
+    {
+      "epoch": 1.6721612773533971,
+      "grad_norm": 0.6313050389289856,
+      "learning_rate": 0.0001,
+      "loss": 1.4046,
+      "step": 14557
+    },
+    {
+      "epoch": 1.6722761472632244,
+      "grad_norm": 0.6314618587493896,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 14558
+    },
+    {
+      "epoch": 1.6723910171730516,
+      "grad_norm": 0.6210103034973145,
+      "learning_rate": 0.0001,
+      "loss": 1.2459,
+      "step": 14559
+    },
+    {
+      "epoch": 1.6725058870828786,
+      "grad_norm": 0.6324970126152039,
+      "learning_rate": 0.0001,
+      "loss": 1.3796,
+      "step": 14560
+    },
+    {
+      "epoch": 1.6726207569927056,
+      "grad_norm": 0.679050087928772,
+      "learning_rate": 0.0001,
+      "loss": 1.4594,
+      "step": 14561
+    },
+    {
+      "epoch": 1.6727356269025329,
+      "grad_norm": 0.5676151514053345,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 14562
+    },
+    {
+      "epoch": 1.67285049681236,
+      "grad_norm": 0.5721042156219482,
+      "learning_rate": 0.0001,
+      "loss": 1.3224,
+      "step": 14563
+    },
+    {
+      "epoch": 1.672965366722187,
+      "grad_norm": 0.5606781840324402,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 14564
+    },
+    {
+      "epoch": 1.673080236632014,
+      "grad_norm": 0.602632462978363,
+      "learning_rate": 0.0001,
+      "loss": 1.316,
+      "step": 14565
+    },
+    {
+      "epoch": 1.6731951065418413,
+      "grad_norm": 0.6165114641189575,
+      "learning_rate": 0.0001,
+      "loss": 1.485,
+      "step": 14566
+    },
+    {
+      "epoch": 1.6733099764516686,
+      "grad_norm": 0.6127558350563049,
+      "learning_rate": 0.0001,
+      "loss": 1.5381,
+      "step": 14567
+    },
+    {
+      "epoch": 1.6734248463614956,
+      "grad_norm": 0.6075685024261475,
+      "learning_rate": 0.0001,
+      "loss": 1.3019,
+      "step": 14568
+    },
+    {
+      "epoch": 1.6735397162713226,
+      "grad_norm": 0.6863022446632385,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 14569
+    },
+    {
+      "epoch": 1.6736545861811498,
+      "grad_norm": 0.6382628679275513,
+      "learning_rate": 0.0001,
+      "loss": 1.5063,
+      "step": 14570
+    },
+    {
+      "epoch": 1.673769456090977,
+      "grad_norm": 0.6291472911834717,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 14571
+    },
+    {
+      "epoch": 1.673884326000804,
+      "grad_norm": 0.6682292222976685,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 14572
+    },
+    {
+      "epoch": 1.673999195910631,
+      "grad_norm": 0.6253922581672668,
+      "learning_rate": 0.0001,
+      "loss": 1.4154,
+      "step": 14573
+    },
+    {
+      "epoch": 1.6741140658204583,
+      "grad_norm": 0.6436946988105774,
+      "learning_rate": 0.0001,
+      "loss": 1.246,
+      "step": 14574
+    },
+    {
+      "epoch": 1.6742289357302855,
+      "grad_norm": 0.6284177303314209,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 14575
+    },
+    {
+      "epoch": 1.6743438056401125,
+      "grad_norm": 0.6123202443122864,
+      "learning_rate": 0.0001,
+      "loss": 1.4549,
+      "step": 14576
+    },
+    {
+      "epoch": 1.6744586755499395,
+      "grad_norm": 0.635627269744873,
+      "learning_rate": 0.0001,
+      "loss": 1.4781,
+      "step": 14577
+    },
+    {
+      "epoch": 1.6745735454597668,
+      "grad_norm": 0.637518048286438,
+      "learning_rate": 0.0001,
+      "loss": 1.5056,
+      "step": 14578
+    },
+    {
+      "epoch": 1.674688415369594,
+      "grad_norm": 0.6325652003288269,
+      "learning_rate": 0.0001,
+      "loss": 1.3018,
+      "step": 14579
+    },
+    {
+      "epoch": 1.674803285279421,
+      "grad_norm": 0.6435227394104004,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 14580
+    },
+    {
+      "epoch": 1.674918155189248,
+      "grad_norm": 0.6192457675933838,
+      "learning_rate": 0.0001,
+      "loss": 1.3374,
+      "step": 14581
+    },
+    {
+      "epoch": 1.6750330250990753,
+      "grad_norm": 0.5760471820831299,
+      "learning_rate": 0.0001,
+      "loss": 1.2543,
+      "step": 14582
+    },
+    {
+      "epoch": 1.6751478950089025,
+      "grad_norm": 0.6090351343154907,
+      "learning_rate": 0.0001,
+      "loss": 1.245,
+      "step": 14583
+    },
+    {
+      "epoch": 1.6752627649187295,
+      "grad_norm": 0.5881224274635315,
+      "learning_rate": 0.0001,
+      "loss": 1.4009,
+      "step": 14584
+    },
+    {
+      "epoch": 1.6753776348285565,
+      "grad_norm": 0.7061769366264343,
+      "learning_rate": 0.0001,
+      "loss": 1.6304,
+      "step": 14585
+    },
+    {
+      "epoch": 1.6754925047383837,
+      "grad_norm": 0.6568925380706787,
+      "learning_rate": 0.0001,
+      "loss": 1.4963,
+      "step": 14586
+    },
+    {
+      "epoch": 1.675607374648211,
+      "grad_norm": 0.6131194233894348,
+      "learning_rate": 0.0001,
+      "loss": 1.3247,
+      "step": 14587
+    },
+    {
+      "epoch": 1.675722244558038,
+      "grad_norm": 0.6096778512001038,
+      "learning_rate": 0.0001,
+      "loss": 1.1763,
+      "step": 14588
+    },
+    {
+      "epoch": 1.675837114467865,
+      "grad_norm": 0.7438702583312988,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 14589
+    },
+    {
+      "epoch": 1.6759519843776922,
+      "grad_norm": 0.6072378754615784,
+      "learning_rate": 0.0001,
+      "loss": 1.4269,
+      "step": 14590
+    },
+    {
+      "epoch": 1.6760668542875194,
+      "grad_norm": 0.6651720404624939,
+      "learning_rate": 0.0001,
+      "loss": 1.3208,
+      "step": 14591
+    },
+    {
+      "epoch": 1.6761817241973465,
+      "grad_norm": 0.7027106881141663,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 14592
+    },
+    {
+      "epoch": 1.6762965941071735,
+      "grad_norm": 0.5823193788528442,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 14593
+    },
+    {
+      "epoch": 1.6764114640170007,
+      "grad_norm": 0.5353499054908752,
+      "learning_rate": 0.0001,
+      "loss": 1.0777,
+      "step": 14594
+    },
+    {
+      "epoch": 1.676526333926828,
+      "grad_norm": 0.7175213098526001,
+      "learning_rate": 0.0001,
+      "loss": 1.42,
+      "step": 14595
+    },
+    {
+      "epoch": 1.676641203836655,
+      "grad_norm": 0.5985221862792969,
+      "learning_rate": 0.0001,
+      "loss": 1.4292,
+      "step": 14596
+    },
+    {
+      "epoch": 1.676756073746482,
+      "grad_norm": 0.6454430818557739,
+      "learning_rate": 0.0001,
+      "loss": 1.4714,
+      "step": 14597
+    },
+    {
+      "epoch": 1.6768709436563092,
+      "grad_norm": 0.6255630850791931,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 14598
+    },
+    {
+      "epoch": 1.6769858135661364,
+      "grad_norm": 0.6272461414337158,
+      "learning_rate": 0.0001,
+      "loss": 1.528,
+      "step": 14599
+    },
+    {
+      "epoch": 1.6771006834759634,
+      "grad_norm": 0.6686280369758606,
+      "learning_rate": 0.0001,
+      "loss": 1.711,
+      "step": 14600
+    },
+    {
+      "epoch": 1.6772155533857906,
+      "grad_norm": 0.5825721025466919,
+      "learning_rate": 0.0001,
+      "loss": 1.3257,
+      "step": 14601
+    },
+    {
+      "epoch": 1.6773304232956177,
+      "grad_norm": 0.6497721672058105,
+      "learning_rate": 0.0001,
+      "loss": 1.6273,
+      "step": 14602
+    },
+    {
+      "epoch": 1.6774452932054449,
+      "grad_norm": 0.6572535634040833,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 14603
+    },
+    {
+      "epoch": 1.6775601631152721,
+      "grad_norm": 0.590975284576416,
+      "learning_rate": 0.0001,
+      "loss": 1.476,
+      "step": 14604
+    },
+    {
+      "epoch": 1.6776750330250991,
+      "grad_norm": 0.6207257509231567,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 14605
+    },
+    {
+      "epoch": 1.6777899029349261,
+      "grad_norm": 0.6565500497817993,
+      "learning_rate": 0.0001,
+      "loss": 1.6584,
+      "step": 14606
+    },
+    {
+      "epoch": 1.6779047728447534,
+      "grad_norm": 0.6749123334884644,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 14607
+    },
+    {
+      "epoch": 1.6780196427545806,
+      "grad_norm": 0.6280671954154968,
+      "learning_rate": 0.0001,
+      "loss": 1.5966,
+      "step": 14608
+    },
+    {
+      "epoch": 1.6781345126644076,
+      "grad_norm": 0.6080400347709656,
+      "learning_rate": 0.0001,
+      "loss": 1.4813,
+      "step": 14609
+    },
+    {
+      "epoch": 1.6782493825742346,
+      "grad_norm": 0.6583701968193054,
+      "learning_rate": 0.0001,
+      "loss": 1.5606,
+      "step": 14610
+    },
+    {
+      "epoch": 1.6783642524840618,
+      "grad_norm": 0.7133932113647461,
+      "learning_rate": 0.0001,
+      "loss": 1.4113,
+      "step": 14611
+    },
+    {
+      "epoch": 1.678479122393889,
+      "grad_norm": 0.647867739200592,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 14612
+    },
+    {
+      "epoch": 1.678593992303716,
+      "grad_norm": 0.5785742402076721,
+      "learning_rate": 0.0001,
+      "loss": 1.2819,
+      "step": 14613
+    },
+    {
+      "epoch": 1.678708862213543,
+      "grad_norm": 0.6787799596786499,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 14614
+    },
+    {
+      "epoch": 1.6788237321233703,
+      "grad_norm": 0.581734299659729,
+      "learning_rate": 0.0001,
+      "loss": 1.2793,
+      "step": 14615
+    },
+    {
+      "epoch": 1.6789386020331976,
+      "grad_norm": 0.5504398345947266,
+      "learning_rate": 0.0001,
+      "loss": 1.3114,
+      "step": 14616
+    },
+    {
+      "epoch": 1.6790534719430246,
+      "grad_norm": 0.5877566337585449,
+      "learning_rate": 0.0001,
+      "loss": 1.3758,
+      "step": 14617
+    },
+    {
+      "epoch": 1.6791683418528516,
+      "grad_norm": 0.7053650617599487,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 14618
+    },
+    {
+      "epoch": 1.6792832117626788,
+      "grad_norm": 0.644305408000946,
+      "learning_rate": 0.0001,
+      "loss": 1.233,
+      "step": 14619
+    },
+    {
+      "epoch": 1.679398081672506,
+      "grad_norm": 0.6391613483428955,
+      "learning_rate": 0.0001,
+      "loss": 1.3591,
+      "step": 14620
+    },
+    {
+      "epoch": 1.679512951582333,
+      "grad_norm": 0.598298192024231,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 14621
+    },
+    {
+      "epoch": 1.67962782149216,
+      "grad_norm": 0.60166996717453,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 14622
+    },
+    {
+      "epoch": 1.6797426914019873,
+      "grad_norm": 0.6578497290611267,
+      "learning_rate": 0.0001,
+      "loss": 1.3284,
+      "step": 14623
+    },
+    {
+      "epoch": 1.6798575613118145,
+      "grad_norm": 0.6441363096237183,
+      "learning_rate": 0.0001,
+      "loss": 1.3718,
+      "step": 14624
+    },
+    {
+      "epoch": 1.6799724312216415,
+      "grad_norm": 0.6314264535903931,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 14625
+    },
+    {
+      "epoch": 1.6800873011314685,
+      "grad_norm": 0.6196305155754089,
+      "learning_rate": 0.0001,
+      "loss": 1.503,
+      "step": 14626
+    },
+    {
+      "epoch": 1.6802021710412958,
+      "grad_norm": 0.6449001431465149,
+      "learning_rate": 0.0001,
+      "loss": 1.4048,
+      "step": 14627
+    },
+    {
+      "epoch": 1.680317040951123,
+      "grad_norm": 0.6065531969070435,
+      "learning_rate": 0.0001,
+      "loss": 1.396,
+      "step": 14628
+    },
+    {
+      "epoch": 1.68043191086095,
+      "grad_norm": 0.6479570269584656,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 14629
+    },
+    {
+      "epoch": 1.680546780770777,
+      "grad_norm": 0.6452540159225464,
+      "learning_rate": 0.0001,
+      "loss": 1.4435,
+      "step": 14630
+    },
+    {
+      "epoch": 1.6806616506806042,
+      "grad_norm": 0.6001893877983093,
+      "learning_rate": 0.0001,
+      "loss": 1.3286,
+      "step": 14631
+    },
+    {
+      "epoch": 1.6807765205904315,
+      "grad_norm": 0.6244105696678162,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 14632
+    },
+    {
+      "epoch": 1.6808913905002585,
+      "grad_norm": 0.6581660509109497,
+      "learning_rate": 0.0001,
+      "loss": 1.4778,
+      "step": 14633
+    },
+    {
+      "epoch": 1.6810062604100855,
+      "grad_norm": 0.7391045093536377,
+      "learning_rate": 0.0001,
+      "loss": 1.5467,
+      "step": 14634
+    },
+    {
+      "epoch": 1.6811211303199127,
+      "grad_norm": 0.6155508160591125,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 14635
+    },
+    {
+      "epoch": 1.68123600022974,
+      "grad_norm": 0.673112154006958,
+      "learning_rate": 0.0001,
+      "loss": 1.6676,
+      "step": 14636
+    },
+    {
+      "epoch": 1.681350870139567,
+      "grad_norm": 0.6411565542221069,
+      "learning_rate": 0.0001,
+      "loss": 1.4505,
+      "step": 14637
+    },
+    {
+      "epoch": 1.681465740049394,
+      "grad_norm": 0.7229892015457153,
+      "learning_rate": 0.0001,
+      "loss": 1.4545,
+      "step": 14638
+    },
+    {
+      "epoch": 1.6815806099592212,
+      "grad_norm": 0.5898306369781494,
+      "learning_rate": 0.0001,
+      "loss": 1.4702,
+      "step": 14639
+    },
+    {
+      "epoch": 1.6816954798690484,
+      "grad_norm": 0.6058882474899292,
+      "learning_rate": 0.0001,
+      "loss": 1.5154,
+      "step": 14640
+    },
+    {
+      "epoch": 1.6818103497788754,
+      "grad_norm": 0.7117508053779602,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 14641
+    },
+    {
+      "epoch": 1.6819252196887025,
+      "grad_norm": 0.7228211164474487,
+      "learning_rate": 0.0001,
+      "loss": 1.3576,
+      "step": 14642
+    },
+    {
+      "epoch": 1.6820400895985297,
+      "grad_norm": 0.5465759634971619,
+      "learning_rate": 0.0001,
+      "loss": 1.3254,
+      "step": 14643
+    },
+    {
+      "epoch": 1.682154959508357,
+      "grad_norm": 0.7400708198547363,
+      "learning_rate": 0.0001,
+      "loss": 1.7752,
+      "step": 14644
+    },
+    {
+      "epoch": 1.682269829418184,
+      "grad_norm": 0.5981911420822144,
+      "learning_rate": 0.0001,
+      "loss": 1.4287,
+      "step": 14645
+    },
+    {
+      "epoch": 1.682384699328011,
+      "grad_norm": 0.6448339223861694,
+      "learning_rate": 0.0001,
+      "loss": 1.6,
+      "step": 14646
+    },
+    {
+      "epoch": 1.6824995692378382,
+      "grad_norm": 0.560468316078186,
+      "learning_rate": 0.0001,
+      "loss": 1.453,
+      "step": 14647
+    },
+    {
+      "epoch": 1.6826144391476654,
+      "grad_norm": 0.6334208250045776,
+      "learning_rate": 0.0001,
+      "loss": 1.3525,
+      "step": 14648
+    },
+    {
+      "epoch": 1.6827293090574924,
+      "grad_norm": 0.6362658739089966,
+      "learning_rate": 0.0001,
+      "loss": 1.5981,
+      "step": 14649
+    },
+    {
+      "epoch": 1.6828441789673194,
+      "grad_norm": 0.644481360912323,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 14650
+    },
+    {
+      "epoch": 1.6829590488771466,
+      "grad_norm": 0.6140202879905701,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 14651
+    },
+    {
+      "epoch": 1.6830739187869739,
+      "grad_norm": 0.6383975744247437,
+      "learning_rate": 0.0001,
+      "loss": 1.4495,
+      "step": 14652
+    },
+    {
+      "epoch": 1.6831887886968009,
+      "grad_norm": 0.6611577868461609,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 14653
+    },
+    {
+      "epoch": 1.683303658606628,
+      "grad_norm": 0.6007152199745178,
+      "learning_rate": 0.0001,
+      "loss": 1.3869,
+      "step": 14654
+    },
+    {
+      "epoch": 1.6834185285164551,
+      "grad_norm": 0.6275546550750732,
+      "learning_rate": 0.0001,
+      "loss": 1.368,
+      "step": 14655
+    },
+    {
+      "epoch": 1.6835333984262824,
+      "grad_norm": 0.649321973323822,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 14656
+    },
+    {
+      "epoch": 1.6836482683361094,
+      "grad_norm": 0.5704925060272217,
+      "learning_rate": 0.0001,
+      "loss": 1.403,
+      "step": 14657
+    },
+    {
+      "epoch": 1.6837631382459364,
+      "grad_norm": 0.6267727017402649,
+      "learning_rate": 0.0001,
+      "loss": 1.4628,
+      "step": 14658
+    },
+    {
+      "epoch": 1.6838780081557636,
+      "grad_norm": 0.6294718980789185,
+      "learning_rate": 0.0001,
+      "loss": 1.3524,
+      "step": 14659
+    },
+    {
+      "epoch": 1.6839928780655908,
+      "grad_norm": 0.583004891872406,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 14660
+    },
+    {
+      "epoch": 1.6841077479754178,
+      "grad_norm": 0.616648256778717,
+      "learning_rate": 0.0001,
+      "loss": 1.4662,
+      "step": 14661
+    },
+    {
+      "epoch": 1.6842226178852449,
+      "grad_norm": 0.6571115255355835,
+      "learning_rate": 0.0001,
+      "loss": 1.4202,
+      "step": 14662
+    },
+    {
+      "epoch": 1.684337487795072,
+      "grad_norm": 0.7342918515205383,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 14663
+    },
+    {
+      "epoch": 1.6844523577048993,
+      "grad_norm": 0.6430667042732239,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 14664
+    },
+    {
+      "epoch": 1.6845672276147263,
+      "grad_norm": 0.6896945238113403,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 14665
+    },
+    {
+      "epoch": 1.6846820975245533,
+      "grad_norm": 0.6307993531227112,
+      "learning_rate": 0.0001,
+      "loss": 1.2423,
+      "step": 14666
+    },
+    {
+      "epoch": 1.6847969674343806,
+      "grad_norm": 0.6446820497512817,
+      "learning_rate": 0.0001,
+      "loss": 1.259,
+      "step": 14667
+    },
+    {
+      "epoch": 1.6849118373442078,
+      "grad_norm": 0.6242921948432922,
+      "learning_rate": 0.0001,
+      "loss": 1.4609,
+      "step": 14668
+    },
+    {
+      "epoch": 1.6850267072540348,
+      "grad_norm": 0.6997743248939514,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 14669
+    },
+    {
+      "epoch": 1.6851415771638618,
+      "grad_norm": 0.6224357485771179,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 14670
+    },
+    {
+      "epoch": 1.685256447073689,
+      "grad_norm": 0.5937080383300781,
+      "learning_rate": 0.0001,
+      "loss": 1.4058,
+      "step": 14671
+    },
+    {
+      "epoch": 1.6853713169835163,
+      "grad_norm": 0.6098648309707642,
+      "learning_rate": 0.0001,
+      "loss": 1.4876,
+      "step": 14672
+    },
+    {
+      "epoch": 1.6854861868933433,
+      "grad_norm": 0.6156720519065857,
+      "learning_rate": 0.0001,
+      "loss": 1.4901,
+      "step": 14673
+    },
+    {
+      "epoch": 1.6856010568031703,
+      "grad_norm": 0.6202700734138489,
+      "learning_rate": 0.0001,
+      "loss": 1.4744,
+      "step": 14674
+    },
+    {
+      "epoch": 1.6857159267129975,
+      "grad_norm": 0.6744952201843262,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 14675
+    },
+    {
+      "epoch": 1.6858307966228248,
+      "grad_norm": 0.6229870319366455,
+      "learning_rate": 0.0001,
+      "loss": 1.4296,
+      "step": 14676
+    },
+    {
+      "epoch": 1.6859456665326518,
+      "grad_norm": 0.6743441224098206,
+      "learning_rate": 0.0001,
+      "loss": 1.4914,
+      "step": 14677
+    },
+    {
+      "epoch": 1.6860605364424788,
+      "grad_norm": 0.6453379988670349,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 14678
+    },
+    {
+      "epoch": 1.686175406352306,
+      "grad_norm": 0.6405113935470581,
+      "learning_rate": 0.0001,
+      "loss": 1.4714,
+      "step": 14679
+    },
+    {
+      "epoch": 1.6862902762621332,
+      "grad_norm": 0.610621988773346,
+      "learning_rate": 0.0001,
+      "loss": 1.5101,
+      "step": 14680
+    },
+    {
+      "epoch": 1.6864051461719602,
+      "grad_norm": 0.5977540016174316,
+      "learning_rate": 0.0001,
+      "loss": 1.4575,
+      "step": 14681
+    },
+    {
+      "epoch": 1.6865200160817873,
+      "grad_norm": 0.6138947606086731,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 14682
+    },
+    {
+      "epoch": 1.6866348859916145,
+      "grad_norm": 0.6411577463150024,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 14683
+    },
+    {
+      "epoch": 1.6867497559014417,
+      "grad_norm": 0.6003040671348572,
+      "learning_rate": 0.0001,
+      "loss": 1.3768,
+      "step": 14684
+    },
+    {
+      "epoch": 1.6868646258112687,
+      "grad_norm": 0.5844460129737854,
+      "learning_rate": 0.0001,
+      "loss": 1.351,
+      "step": 14685
+    },
+    {
+      "epoch": 1.6869794957210957,
+      "grad_norm": 0.7202532887458801,
+      "learning_rate": 0.0001,
+      "loss": 1.6241,
+      "step": 14686
+    },
+    {
+      "epoch": 1.687094365630923,
+      "grad_norm": 0.6601413488388062,
+      "learning_rate": 0.0001,
+      "loss": 1.793,
+      "step": 14687
+    },
+    {
+      "epoch": 1.6872092355407502,
+      "grad_norm": 0.6533530354499817,
+      "learning_rate": 0.0001,
+      "loss": 1.4675,
+      "step": 14688
+    },
+    {
+      "epoch": 1.6873241054505772,
+      "grad_norm": 0.5973615646362305,
+      "learning_rate": 0.0001,
+      "loss": 1.4479,
+      "step": 14689
+    },
+    {
+      "epoch": 1.6874389753604042,
+      "grad_norm": 0.6223400831222534,
+      "learning_rate": 0.0001,
+      "loss": 1.5005,
+      "step": 14690
+    },
+    {
+      "epoch": 1.6875538452702314,
+      "grad_norm": 0.7212362885475159,
+      "learning_rate": 0.0001,
+      "loss": 1.7079,
+      "step": 14691
+    },
+    {
+      "epoch": 1.6876687151800587,
+      "grad_norm": 0.6673148274421692,
+      "learning_rate": 0.0001,
+      "loss": 1.428,
+      "step": 14692
+    },
+    {
+      "epoch": 1.6877835850898857,
+      "grad_norm": 0.6568313241004944,
+      "learning_rate": 0.0001,
+      "loss": 1.3289,
+      "step": 14693
+    },
+    {
+      "epoch": 1.6878984549997127,
+      "grad_norm": 0.6275414228439331,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 14694
+    },
+    {
+      "epoch": 1.68801332490954,
+      "grad_norm": 0.6030680537223816,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 14695
+    },
+    {
+      "epoch": 1.6881281948193672,
+      "grad_norm": 0.6563968658447266,
+      "learning_rate": 0.0001,
+      "loss": 1.3393,
+      "step": 14696
+    },
+    {
+      "epoch": 1.6882430647291942,
+      "grad_norm": 0.683423638343811,
+      "learning_rate": 0.0001,
+      "loss": 1.5572,
+      "step": 14697
+    },
+    {
+      "epoch": 1.6883579346390212,
+      "grad_norm": 0.6408438086509705,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 14698
+    },
+    {
+      "epoch": 1.6884728045488484,
+      "grad_norm": 0.6201786398887634,
+      "learning_rate": 0.0001,
+      "loss": 1.3984,
+      "step": 14699
+    },
+    {
+      "epoch": 1.6885876744586756,
+      "grad_norm": 0.6087800860404968,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 14700
+    },
+    {
+      "epoch": 1.6887025443685026,
+      "grad_norm": 0.6137244701385498,
+      "learning_rate": 0.0001,
+      "loss": 1.3528,
+      "step": 14701
+    },
+    {
+      "epoch": 1.6888174142783297,
+      "grad_norm": 0.5812850594520569,
+      "learning_rate": 0.0001,
+      "loss": 1.3308,
+      "step": 14702
+    },
+    {
+      "epoch": 1.6889322841881569,
+      "grad_norm": 0.6570180058479309,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 14703
+    },
+    {
+      "epoch": 1.6890471540979841,
+      "grad_norm": 0.6571214199066162,
+      "learning_rate": 0.0001,
+      "loss": 1.3125,
+      "step": 14704
+    },
+    {
+      "epoch": 1.6891620240078111,
+      "grad_norm": 0.6408611536026001,
+      "learning_rate": 0.0001,
+      "loss": 1.3928,
+      "step": 14705
+    },
+    {
+      "epoch": 1.6892768939176381,
+      "grad_norm": 0.6647818088531494,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 14706
+    },
+    {
+      "epoch": 1.6893917638274654,
+      "grad_norm": 0.6333121061325073,
+      "learning_rate": 0.0001,
+      "loss": 1.4169,
+      "step": 14707
+    },
+    {
+      "epoch": 1.6895066337372926,
+      "grad_norm": 0.6847052574157715,
+      "learning_rate": 0.0001,
+      "loss": 1.383,
+      "step": 14708
+    },
+    {
+      "epoch": 1.6896215036471196,
+      "grad_norm": 0.6509214639663696,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 14709
+    },
+    {
+      "epoch": 1.6897363735569466,
+      "grad_norm": 0.6630228757858276,
+      "learning_rate": 0.0001,
+      "loss": 1.4381,
+      "step": 14710
+    },
+    {
+      "epoch": 1.6898512434667738,
+      "grad_norm": 0.5848700404167175,
+      "learning_rate": 0.0001,
+      "loss": 1.3055,
+      "step": 14711
+    },
+    {
+      "epoch": 1.689966113376601,
+      "grad_norm": 0.6231377720832825,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 14712
+    },
+    {
+      "epoch": 1.690080983286428,
+      "grad_norm": 0.5822433233261108,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 14713
+    },
+    {
+      "epoch": 1.690195853196255,
+      "grad_norm": 0.5917012095451355,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 14714
+    },
+    {
+      "epoch": 1.6903107231060823,
+      "grad_norm": 0.5557729601860046,
+      "learning_rate": 0.0001,
+      "loss": 1.4386,
+      "step": 14715
+    },
+    {
+      "epoch": 1.6904255930159096,
+      "grad_norm": 0.6072548627853394,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 14716
+    },
+    {
+      "epoch": 1.6905404629257366,
+      "grad_norm": 0.6061117053031921,
+      "learning_rate": 0.0001,
+      "loss": 1.2462,
+      "step": 14717
+    },
+    {
+      "epoch": 1.6906553328355636,
+      "grad_norm": 0.6177441477775574,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 14718
+    },
+    {
+      "epoch": 1.6907702027453908,
+      "grad_norm": 0.6135041117668152,
+      "learning_rate": 0.0001,
+      "loss": 1.4055,
+      "step": 14719
+    },
+    {
+      "epoch": 1.690885072655218,
+      "grad_norm": 0.5727534890174866,
+      "learning_rate": 0.0001,
+      "loss": 1.2838,
+      "step": 14720
+    },
+    {
+      "epoch": 1.690999942565045,
+      "grad_norm": 0.6418645977973938,
+      "learning_rate": 0.0001,
+      "loss": 1.3512,
+      "step": 14721
+    },
+    {
+      "epoch": 1.691114812474872,
+      "grad_norm": 0.6252636909484863,
+      "learning_rate": 0.0001,
+      "loss": 1.5915,
+      "step": 14722
+    },
+    {
+      "epoch": 1.6912296823846993,
+      "grad_norm": 0.6545863151550293,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 14723
+    },
+    {
+      "epoch": 1.6913445522945265,
+      "grad_norm": 0.6704011559486389,
+      "learning_rate": 0.0001,
+      "loss": 1.4523,
+      "step": 14724
+    },
+    {
+      "epoch": 1.6914594222043535,
+      "grad_norm": 0.6669784188270569,
+      "learning_rate": 0.0001,
+      "loss": 1.6168,
+      "step": 14725
+    },
+    {
+      "epoch": 1.6915742921141805,
+      "grad_norm": 0.5946208238601685,
+      "learning_rate": 0.0001,
+      "loss": 1.5255,
+      "step": 14726
+    },
+    {
+      "epoch": 1.6916891620240078,
+      "grad_norm": 0.6300133466720581,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 14727
+    },
+    {
+      "epoch": 1.691804031933835,
+      "grad_norm": 0.6053428053855896,
+      "learning_rate": 0.0001,
+      "loss": 1.6368,
+      "step": 14728
+    },
+    {
+      "epoch": 1.691918901843662,
+      "grad_norm": 0.5880074501037598,
+      "learning_rate": 0.0001,
+      "loss": 1.3086,
+      "step": 14729
+    },
+    {
+      "epoch": 1.692033771753489,
+      "grad_norm": 0.6411834955215454,
+      "learning_rate": 0.0001,
+      "loss": 1.3448,
+      "step": 14730
+    },
+    {
+      "epoch": 1.6921486416633162,
+      "grad_norm": 0.6445175409317017,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 14731
+    },
+    {
+      "epoch": 1.6922635115731435,
+      "grad_norm": 0.5757904052734375,
+      "learning_rate": 0.0001,
+      "loss": 1.3863,
+      "step": 14732
+    },
+    {
+      "epoch": 1.6923783814829705,
+      "grad_norm": 0.6143217086791992,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 14733
+    },
+    {
+      "epoch": 1.6924932513927975,
+      "grad_norm": 0.6285145282745361,
+      "learning_rate": 0.0001,
+      "loss": 1.4525,
+      "step": 14734
+    },
+    {
+      "epoch": 1.6926081213026247,
+      "grad_norm": 0.6272004246711731,
+      "learning_rate": 0.0001,
+      "loss": 1.3839,
+      "step": 14735
+    },
+    {
+      "epoch": 1.692722991212452,
+      "grad_norm": 0.6545289754867554,
+      "learning_rate": 0.0001,
+      "loss": 1.4322,
+      "step": 14736
+    },
+    {
+      "epoch": 1.692837861122279,
+      "grad_norm": 0.6652606725692749,
+      "learning_rate": 0.0001,
+      "loss": 1.6324,
+      "step": 14737
+    },
+    {
+      "epoch": 1.6929527310321062,
+      "grad_norm": 0.6375353932380676,
+      "learning_rate": 0.0001,
+      "loss": 1.4693,
+      "step": 14738
+    },
+    {
+      "epoch": 1.6930676009419332,
+      "grad_norm": 0.6303648948669434,
+      "learning_rate": 0.0001,
+      "loss": 1.4073,
+      "step": 14739
+    },
+    {
+      "epoch": 1.6931824708517604,
+      "grad_norm": 0.6370104551315308,
+      "learning_rate": 0.0001,
+      "loss": 1.2702,
+      "step": 14740
+    },
+    {
+      "epoch": 1.6932973407615877,
+      "grad_norm": 0.723127007484436,
+      "learning_rate": 0.0001,
+      "loss": 1.4016,
+      "step": 14741
+    },
+    {
+      "epoch": 1.6934122106714147,
+      "grad_norm": 0.628434419631958,
+      "learning_rate": 0.0001,
+      "loss": 1.3892,
+      "step": 14742
+    },
+    {
+      "epoch": 1.6935270805812417,
+      "grad_norm": 0.5962461233139038,
+      "learning_rate": 0.0001,
+      "loss": 1.4348,
+      "step": 14743
+    },
+    {
+      "epoch": 1.693641950491069,
+      "grad_norm": 0.6001591682434082,
+      "learning_rate": 0.0001,
+      "loss": 1.4057,
+      "step": 14744
+    },
+    {
+      "epoch": 1.6937568204008961,
+      "grad_norm": 0.6299112439155579,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 14745
+    },
+    {
+      "epoch": 1.6938716903107232,
+      "grad_norm": 0.6881070137023926,
+      "learning_rate": 0.0001,
+      "loss": 1.666,
+      "step": 14746
+    },
+    {
+      "epoch": 1.6939865602205502,
+      "grad_norm": 0.5913798809051514,
+      "learning_rate": 0.0001,
+      "loss": 1.3889,
+      "step": 14747
+    },
+    {
+      "epoch": 1.6941014301303774,
+      "grad_norm": 0.627502977848053,
+      "learning_rate": 0.0001,
+      "loss": 1.5266,
+      "step": 14748
+    },
+    {
+      "epoch": 1.6942163000402046,
+      "grad_norm": 0.6564575433731079,
+      "learning_rate": 0.0001,
+      "loss": 1.4907,
+      "step": 14749
+    },
+    {
+      "epoch": 1.6943311699500316,
+      "grad_norm": 0.5897772908210754,
+      "learning_rate": 0.0001,
+      "loss": 1.3894,
+      "step": 14750
+    },
+    {
+      "epoch": 1.6944460398598586,
+      "grad_norm": 0.5913833975791931,
+      "learning_rate": 0.0001,
+      "loss": 1.2416,
+      "step": 14751
+    },
+    {
+      "epoch": 1.6945609097696859,
+      "grad_norm": 0.658932626247406,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 14752
+    },
+    {
+      "epoch": 1.694675779679513,
+      "grad_norm": 0.6177611351013184,
+      "learning_rate": 0.0001,
+      "loss": 1.4164,
+      "step": 14753
+    },
+    {
+      "epoch": 1.6947906495893401,
+      "grad_norm": 0.6406357884407043,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 14754
+    },
+    {
+      "epoch": 1.6949055194991671,
+      "grad_norm": 0.6174083948135376,
+      "learning_rate": 0.0001,
+      "loss": 1.4722,
+      "step": 14755
+    },
+    {
+      "epoch": 1.6950203894089944,
+      "grad_norm": 0.653312087059021,
+      "learning_rate": 0.0001,
+      "loss": 1.4612,
+      "step": 14756
+    },
+    {
+      "epoch": 1.6951352593188216,
+      "grad_norm": 0.6030119061470032,
+      "learning_rate": 0.0001,
+      "loss": 1.3403,
+      "step": 14757
+    },
+    {
+      "epoch": 1.6952501292286486,
+      "grad_norm": 0.6403154730796814,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 14758
+    },
+    {
+      "epoch": 1.6953649991384756,
+      "grad_norm": 0.6424520015716553,
+      "learning_rate": 0.0001,
+      "loss": 1.5404,
+      "step": 14759
+    },
+    {
+      "epoch": 1.6954798690483028,
+      "grad_norm": 0.6321309804916382,
+      "learning_rate": 0.0001,
+      "loss": 1.3525,
+      "step": 14760
+    },
+    {
+      "epoch": 1.69559473895813,
+      "grad_norm": 0.5908375382423401,
+      "learning_rate": 0.0001,
+      "loss": 1.4181,
+      "step": 14761
+    },
+    {
+      "epoch": 1.695709608867957,
+      "grad_norm": 0.6151296496391296,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 14762
+    },
+    {
+      "epoch": 1.695824478777784,
+      "grad_norm": 0.5596523880958557,
+      "learning_rate": 0.0001,
+      "loss": 1.2521,
+      "step": 14763
+    },
+    {
+      "epoch": 1.6959393486876113,
+      "grad_norm": 0.6613996028900146,
+      "learning_rate": 0.0001,
+      "loss": 1.5143,
+      "step": 14764
+    },
+    {
+      "epoch": 1.6960542185974385,
+      "grad_norm": 0.6622706055641174,
+      "learning_rate": 0.0001,
+      "loss": 1.6614,
+      "step": 14765
+    },
+    {
+      "epoch": 1.6961690885072656,
+      "grad_norm": 0.5737489461898804,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 14766
+    },
+    {
+      "epoch": 1.6962839584170926,
+      "grad_norm": 0.6686919927597046,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 14767
+    },
+    {
+      "epoch": 1.6963988283269198,
+      "grad_norm": 0.6299922466278076,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 14768
+    },
+    {
+      "epoch": 1.696513698236747,
+      "grad_norm": 0.7300532460212708,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 14769
+    },
+    {
+      "epoch": 1.696628568146574,
+      "grad_norm": 0.5972233414649963,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 14770
+    },
+    {
+      "epoch": 1.696743438056401,
+      "grad_norm": 0.6257323026657104,
+      "learning_rate": 0.0001,
+      "loss": 1.3211,
+      "step": 14771
+    },
+    {
+      "epoch": 1.6968583079662283,
+      "grad_norm": 0.6275603175163269,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 14772
+    },
+    {
+      "epoch": 1.6969731778760555,
+      "grad_norm": 0.5872368216514587,
+      "learning_rate": 0.0001,
+      "loss": 1.3845,
+      "step": 14773
+    },
+    {
+      "epoch": 1.6970880477858825,
+      "grad_norm": 0.5436793565750122,
+      "learning_rate": 0.0001,
+      "loss": 1.1879,
+      "step": 14774
+    },
+    {
+      "epoch": 1.6972029176957095,
+      "grad_norm": 0.5925673246383667,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 14775
+    },
+    {
+      "epoch": 1.6973177876055368,
+      "grad_norm": 0.622356653213501,
+      "learning_rate": 0.0001,
+      "loss": 1.2646,
+      "step": 14776
+    },
+    {
+      "epoch": 1.697432657515364,
+      "grad_norm": 0.644891083240509,
+      "learning_rate": 0.0001,
+      "loss": 1.4162,
+      "step": 14777
+    },
+    {
+      "epoch": 1.697547527425191,
+      "grad_norm": 0.6052226424217224,
+      "learning_rate": 0.0001,
+      "loss": 1.3639,
+      "step": 14778
+    },
+    {
+      "epoch": 1.697662397335018,
+      "grad_norm": 0.6266080737113953,
+      "learning_rate": 0.0001,
+      "loss": 1.3751,
+      "step": 14779
+    },
+    {
+      "epoch": 1.6977772672448452,
+      "grad_norm": 0.6198167204856873,
+      "learning_rate": 0.0001,
+      "loss": 1.5003,
+      "step": 14780
+    },
+    {
+      "epoch": 1.6978921371546725,
+      "grad_norm": 0.6346505284309387,
+      "learning_rate": 0.0001,
+      "loss": 1.4329,
+      "step": 14781
+    },
+    {
+      "epoch": 1.6980070070644995,
+      "grad_norm": 0.6443406939506531,
+      "learning_rate": 0.0001,
+      "loss": 1.3899,
+      "step": 14782
+    },
+    {
+      "epoch": 1.6981218769743265,
+      "grad_norm": 0.6747581958770752,
+      "learning_rate": 0.0001,
+      "loss": 1.4359,
+      "step": 14783
+    },
+    {
+      "epoch": 1.6982367468841537,
+      "grad_norm": 0.6033055186271667,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 14784
+    },
+    {
+      "epoch": 1.698351616793981,
+      "grad_norm": 0.6114761829376221,
+      "learning_rate": 0.0001,
+      "loss": 1.4088,
+      "step": 14785
+    },
+    {
+      "epoch": 1.698466486703808,
+      "grad_norm": 0.5994529724121094,
+      "learning_rate": 0.0001,
+      "loss": 1.4139,
+      "step": 14786
+    },
+    {
+      "epoch": 1.698581356613635,
+      "grad_norm": 0.6032854318618774,
+      "learning_rate": 0.0001,
+      "loss": 1.3684,
+      "step": 14787
+    },
+    {
+      "epoch": 1.6986962265234622,
+      "grad_norm": 0.579774796962738,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 14788
+    },
+    {
+      "epoch": 1.6988110964332894,
+      "grad_norm": 0.600327730178833,
+      "learning_rate": 0.0001,
+      "loss": 1.3107,
+      "step": 14789
+    },
+    {
+      "epoch": 1.6989259663431164,
+      "grad_norm": 0.6498262882232666,
+      "learning_rate": 0.0001,
+      "loss": 1.5339,
+      "step": 14790
+    },
+    {
+      "epoch": 1.6990408362529434,
+      "grad_norm": 0.6453105211257935,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 14791
+    },
+    {
+      "epoch": 1.6991557061627707,
+      "grad_norm": 0.6776260137557983,
+      "learning_rate": 0.0001,
+      "loss": 1.4309,
+      "step": 14792
+    },
+    {
+      "epoch": 1.699270576072598,
+      "grad_norm": 0.5969988107681274,
+      "learning_rate": 0.0001,
+      "loss": 1.4604,
+      "step": 14793
+    },
+    {
+      "epoch": 1.699385445982425,
+      "grad_norm": 0.585665762424469,
+      "learning_rate": 0.0001,
+      "loss": 1.3594,
+      "step": 14794
+    },
+    {
+      "epoch": 1.699500315892252,
+      "grad_norm": 0.6432570815086365,
+      "learning_rate": 0.0001,
+      "loss": 1.5177,
+      "step": 14795
+    },
+    {
+      "epoch": 1.6996151858020792,
+      "grad_norm": 0.6451871395111084,
+      "learning_rate": 0.0001,
+      "loss": 1.3841,
+      "step": 14796
+    },
+    {
+      "epoch": 1.6997300557119064,
+      "grad_norm": 0.6202559471130371,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 14797
+    },
+    {
+      "epoch": 1.6998449256217334,
+      "grad_norm": 0.6412597894668579,
+      "learning_rate": 0.0001,
+      "loss": 1.4781,
+      "step": 14798
+    },
+    {
+      "epoch": 1.6999597955315604,
+      "grad_norm": 0.6816568374633789,
+      "learning_rate": 0.0001,
+      "loss": 1.4192,
+      "step": 14799
+    },
+    {
+      "epoch": 1.7000746654413876,
+      "grad_norm": 0.6473395824432373,
+      "learning_rate": 0.0001,
+      "loss": 1.5152,
+      "step": 14800
+    },
+    {
+      "epoch": 1.7001895353512149,
+      "grad_norm": 0.6455778479576111,
+      "learning_rate": 0.0001,
+      "loss": 1.2928,
+      "step": 14801
+    },
+    {
+      "epoch": 1.7003044052610419,
+      "grad_norm": 0.5928786396980286,
+      "learning_rate": 0.0001,
+      "loss": 1.2802,
+      "step": 14802
+    },
+    {
+      "epoch": 1.7004192751708689,
+      "grad_norm": 0.6279940605163574,
+      "learning_rate": 0.0001,
+      "loss": 1.2901,
+      "step": 14803
+    },
+    {
+      "epoch": 1.7005341450806961,
+      "grad_norm": 0.6435730457305908,
+      "learning_rate": 0.0001,
+      "loss": 1.4583,
+      "step": 14804
+    },
+    {
+      "epoch": 1.7006490149905233,
+      "grad_norm": 0.6767624616622925,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 14805
+    },
+    {
+      "epoch": 1.7007638849003504,
+      "grad_norm": 0.6514124870300293,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 14806
+    },
+    {
+      "epoch": 1.7008787548101774,
+      "grad_norm": 0.6798242926597595,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 14807
+    },
+    {
+      "epoch": 1.7009936247200046,
+      "grad_norm": 0.6122627258300781,
+      "learning_rate": 0.0001,
+      "loss": 1.6312,
+      "step": 14808
+    },
+    {
+      "epoch": 1.7011084946298318,
+      "grad_norm": 0.6929191946983337,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 14809
+    },
+    {
+      "epoch": 1.7012233645396588,
+      "grad_norm": 0.6241300702095032,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 14810
+    },
+    {
+      "epoch": 1.7013382344494858,
+      "grad_norm": 0.7182061076164246,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 14811
+    },
+    {
+      "epoch": 1.701453104359313,
+      "grad_norm": 0.6124372482299805,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 14812
+    },
+    {
+      "epoch": 1.7015679742691403,
+      "grad_norm": 0.766761064529419,
+      "learning_rate": 0.0001,
+      "loss": 1.4963,
+      "step": 14813
+    },
+    {
+      "epoch": 1.7016828441789673,
+      "grad_norm": 0.664625346660614,
+      "learning_rate": 0.0001,
+      "loss": 1.2332,
+      "step": 14814
+    },
+    {
+      "epoch": 1.7017977140887943,
+      "grad_norm": 0.641326367855072,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 14815
+    },
+    {
+      "epoch": 1.7019125839986216,
+      "grad_norm": 0.6896613240242004,
+      "learning_rate": 0.0001,
+      "loss": 1.4501,
+      "step": 14816
+    },
+    {
+      "epoch": 1.7020274539084488,
+      "grad_norm": 0.6627568602561951,
+      "learning_rate": 0.0001,
+      "loss": 1.7495,
+      "step": 14817
+    },
+    {
+      "epoch": 1.7021423238182758,
+      "grad_norm": 0.6433359384536743,
+      "learning_rate": 0.0001,
+      "loss": 1.5985,
+      "step": 14818
+    },
+    {
+      "epoch": 1.7022571937281028,
+      "grad_norm": 0.5926421880722046,
+      "learning_rate": 0.0001,
+      "loss": 1.419,
+      "step": 14819
+    },
+    {
+      "epoch": 1.70237206363793,
+      "grad_norm": 0.6341785192489624,
+      "learning_rate": 0.0001,
+      "loss": 1.3109,
+      "step": 14820
+    },
+    {
+      "epoch": 1.7024869335477573,
+      "grad_norm": 0.5665348768234253,
+      "learning_rate": 0.0001,
+      "loss": 1.2676,
+      "step": 14821
+    },
+    {
+      "epoch": 1.7026018034575843,
+      "grad_norm": 0.6543471217155457,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 14822
+    },
+    {
+      "epoch": 1.7027166733674113,
+      "grad_norm": 0.6324287056922913,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 14823
+    },
+    {
+      "epoch": 1.7028315432772385,
+      "grad_norm": 0.6068698167800903,
+      "learning_rate": 0.0001,
+      "loss": 1.3328,
+      "step": 14824
+    },
+    {
+      "epoch": 1.7029464131870657,
+      "grad_norm": 0.6536720991134644,
+      "learning_rate": 0.0001,
+      "loss": 1.422,
+      "step": 14825
+    },
+    {
+      "epoch": 1.7030612830968928,
+      "grad_norm": 0.681016206741333,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 14826
+    },
+    {
+      "epoch": 1.7031761530067198,
+      "grad_norm": 0.6520375609397888,
+      "learning_rate": 0.0001,
+      "loss": 1.565,
+      "step": 14827
+    },
+    {
+      "epoch": 1.703291022916547,
+      "grad_norm": 0.6231850385665894,
+      "learning_rate": 0.0001,
+      "loss": 1.2909,
+      "step": 14828
+    },
+    {
+      "epoch": 1.7034058928263742,
+      "grad_norm": 0.6289814114570618,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 14829
+    },
+    {
+      "epoch": 1.7035207627362012,
+      "grad_norm": 0.687447726726532,
+      "learning_rate": 0.0001,
+      "loss": 1.2311,
+      "step": 14830
+    },
+    {
+      "epoch": 1.7036356326460282,
+      "grad_norm": 0.6084256172180176,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 14831
+    },
+    {
+      "epoch": 1.7037505025558555,
+      "grad_norm": 0.6083263754844666,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 14832
+    },
+    {
+      "epoch": 1.7038653724656827,
+      "grad_norm": 0.612534761428833,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 14833
+    },
+    {
+      "epoch": 1.7039802423755097,
+      "grad_norm": 0.6613317728042603,
+      "learning_rate": 0.0001,
+      "loss": 1.6517,
+      "step": 14834
+    },
+    {
+      "epoch": 1.7040951122853367,
+      "grad_norm": 0.6363469958305359,
+      "learning_rate": 0.0001,
+      "loss": 1.5,
+      "step": 14835
+    },
+    {
+      "epoch": 1.704209982195164,
+      "grad_norm": 0.6956435441970825,
+      "learning_rate": 0.0001,
+      "loss": 1.6757,
+      "step": 14836
+    },
+    {
+      "epoch": 1.7043248521049912,
+      "grad_norm": 0.649994969367981,
+      "learning_rate": 0.0001,
+      "loss": 1.6309,
+      "step": 14837
+    },
+    {
+      "epoch": 1.7044397220148182,
+      "grad_norm": 0.6042583584785461,
+      "learning_rate": 0.0001,
+      "loss": 1.3096,
+      "step": 14838
+    },
+    {
+      "epoch": 1.7045545919246452,
+      "grad_norm": 0.6033551692962646,
+      "learning_rate": 0.0001,
+      "loss": 1.4756,
+      "step": 14839
+    },
+    {
+      "epoch": 1.7046694618344724,
+      "grad_norm": 0.6003922820091248,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 14840
+    },
+    {
+      "epoch": 1.7047843317442997,
+      "grad_norm": 0.6671146154403687,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 14841
+    },
+    {
+      "epoch": 1.7048992016541267,
+      "grad_norm": 0.5957947373390198,
+      "learning_rate": 0.0001,
+      "loss": 1.3582,
+      "step": 14842
+    },
+    {
+      "epoch": 1.7050140715639537,
+      "grad_norm": 0.6179354786872864,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 14843
+    },
+    {
+      "epoch": 1.705128941473781,
+      "grad_norm": 0.6118759512901306,
+      "learning_rate": 0.0001,
+      "loss": 1.3776,
+      "step": 14844
+    },
+    {
+      "epoch": 1.7052438113836081,
+      "grad_norm": 0.5971879959106445,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 14845
+    },
+    {
+      "epoch": 1.7053586812934352,
+      "grad_norm": 0.5953097343444824,
+      "learning_rate": 0.0001,
+      "loss": 1.4166,
+      "step": 14846
+    },
+    {
+      "epoch": 1.7054735512032622,
+      "grad_norm": 0.6157338619232178,
+      "learning_rate": 0.0001,
+      "loss": 1.4949,
+      "step": 14847
+    },
+    {
+      "epoch": 1.7055884211130894,
+      "grad_norm": 0.6230495572090149,
+      "learning_rate": 0.0001,
+      "loss": 1.4628,
+      "step": 14848
+    },
+    {
+      "epoch": 1.7057032910229166,
+      "grad_norm": 0.6791070699691772,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 14849
+    },
+    {
+      "epoch": 1.7058181609327436,
+      "grad_norm": 0.5627061128616333,
+      "learning_rate": 0.0001,
+      "loss": 1.2639,
+      "step": 14850
+    },
+    {
+      "epoch": 1.7059330308425706,
+      "grad_norm": 0.6061202883720398,
+      "learning_rate": 0.0001,
+      "loss": 1.34,
+      "step": 14851
+    },
+    {
+      "epoch": 1.7060479007523979,
+      "grad_norm": 0.6992841362953186,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 14852
+    },
+    {
+      "epoch": 1.706162770662225,
+      "grad_norm": 0.6365556120872498,
+      "learning_rate": 0.0001,
+      "loss": 1.632,
+      "step": 14853
+    },
+    {
+      "epoch": 1.7062776405720521,
+      "grad_norm": 0.6153625845909119,
+      "learning_rate": 0.0001,
+      "loss": 1.2521,
+      "step": 14854
+    },
+    {
+      "epoch": 1.7063925104818791,
+      "grad_norm": 0.6573269963264465,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 14855
+    },
+    {
+      "epoch": 1.7065073803917064,
+      "grad_norm": 0.6061968207359314,
+      "learning_rate": 0.0001,
+      "loss": 1.4127,
+      "step": 14856
+    },
+    {
+      "epoch": 1.7066222503015336,
+      "grad_norm": 0.6206647753715515,
+      "learning_rate": 0.0001,
+      "loss": 1.3738,
+      "step": 14857
+    },
+    {
+      "epoch": 1.7067371202113606,
+      "grad_norm": 0.5990536212921143,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 14858
+    },
+    {
+      "epoch": 1.7068519901211876,
+      "grad_norm": 0.6412786245346069,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 14859
+    },
+    {
+      "epoch": 1.7069668600310148,
+      "grad_norm": 0.6077423691749573,
+      "learning_rate": 0.0001,
+      "loss": 1.4176,
+      "step": 14860
+    },
+    {
+      "epoch": 1.707081729940842,
+      "grad_norm": 0.6050223708152771,
+      "learning_rate": 0.0001,
+      "loss": 1.3676,
+      "step": 14861
+    },
+    {
+      "epoch": 1.707196599850669,
+      "grad_norm": 0.6094151735305786,
+      "learning_rate": 0.0001,
+      "loss": 1.4076,
+      "step": 14862
+    },
+    {
+      "epoch": 1.707311469760496,
+      "grad_norm": 0.6265443563461304,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 14863
+    },
+    {
+      "epoch": 1.7074263396703233,
+      "grad_norm": 0.6976631283760071,
+      "learning_rate": 0.0001,
+      "loss": 1.3129,
+      "step": 14864
+    },
+    {
+      "epoch": 1.7075412095801505,
+      "grad_norm": 0.6556214094161987,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 14865
+    },
+    {
+      "epoch": 1.7076560794899776,
+      "grad_norm": 0.5818542838096619,
+      "learning_rate": 0.0001,
+      "loss": 1.3679,
+      "step": 14866
+    },
+    {
+      "epoch": 1.7077709493998046,
+      "grad_norm": 0.5856722593307495,
+      "learning_rate": 0.0001,
+      "loss": 1.3271,
+      "step": 14867
+    },
+    {
+      "epoch": 1.7078858193096318,
+      "grad_norm": 0.5733777284622192,
+      "learning_rate": 0.0001,
+      "loss": 1.3968,
+      "step": 14868
+    },
+    {
+      "epoch": 1.708000689219459,
+      "grad_norm": 0.6631439924240112,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 14869
+    },
+    {
+      "epoch": 1.708115559129286,
+      "grad_norm": 0.6170898079872131,
+      "learning_rate": 0.0001,
+      "loss": 1.4636,
+      "step": 14870
+    },
+    {
+      "epoch": 1.708230429039113,
+      "grad_norm": 0.700430154800415,
+      "learning_rate": 0.0001,
+      "loss": 1.4296,
+      "step": 14871
+    },
+    {
+      "epoch": 1.7083452989489403,
+      "grad_norm": 0.6631823182106018,
+      "learning_rate": 0.0001,
+      "loss": 1.4876,
+      "step": 14872
+    },
+    {
+      "epoch": 1.7084601688587675,
+      "grad_norm": 0.6962304711341858,
+      "learning_rate": 0.0001,
+      "loss": 1.4368,
+      "step": 14873
+    },
+    {
+      "epoch": 1.7085750387685945,
+      "grad_norm": 0.5741162896156311,
+      "learning_rate": 0.0001,
+      "loss": 1.3373,
+      "step": 14874
+    },
+    {
+      "epoch": 1.7086899086784217,
+      "grad_norm": 0.6080014109611511,
+      "learning_rate": 0.0001,
+      "loss": 1.3203,
+      "step": 14875
+    },
+    {
+      "epoch": 1.7088047785882488,
+      "grad_norm": 0.6642729043960571,
+      "learning_rate": 0.0001,
+      "loss": 1.5342,
+      "step": 14876
+    },
+    {
+      "epoch": 1.708919648498076,
+      "grad_norm": 0.659485936164856,
+      "learning_rate": 0.0001,
+      "loss": 1.4116,
+      "step": 14877
+    },
+    {
+      "epoch": 1.7090345184079032,
+      "grad_norm": 0.6787165999412537,
+      "learning_rate": 0.0001,
+      "loss": 1.4067,
+      "step": 14878
+    },
+    {
+      "epoch": 1.7091493883177302,
+      "grad_norm": 0.6123861074447632,
+      "learning_rate": 0.0001,
+      "loss": 1.4821,
+      "step": 14879
+    },
+    {
+      "epoch": 1.7092642582275572,
+      "grad_norm": 0.608695387840271,
+      "learning_rate": 0.0001,
+      "loss": 1.4714,
+      "step": 14880
+    },
+    {
+      "epoch": 1.7093791281373845,
+      "grad_norm": 0.5870606899261475,
+      "learning_rate": 0.0001,
+      "loss": 1.4441,
+      "step": 14881
+    },
+    {
+      "epoch": 1.7094939980472117,
+      "grad_norm": 0.6159781813621521,
+      "learning_rate": 0.0001,
+      "loss": 1.2932,
+      "step": 14882
+    },
+    {
+      "epoch": 1.7096088679570387,
+      "grad_norm": 0.6811216473579407,
+      "learning_rate": 0.0001,
+      "loss": 1.4543,
+      "step": 14883
+    },
+    {
+      "epoch": 1.7097237378668657,
+      "grad_norm": 0.601915180683136,
+      "learning_rate": 0.0001,
+      "loss": 1.2939,
+      "step": 14884
+    },
+    {
+      "epoch": 1.709838607776693,
+      "grad_norm": 0.5753575563430786,
+      "learning_rate": 0.0001,
+      "loss": 1.4496,
+      "step": 14885
+    },
+    {
+      "epoch": 1.7099534776865202,
+      "grad_norm": 0.6104464530944824,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 14886
+    },
+    {
+      "epoch": 1.7100683475963472,
+      "grad_norm": 0.5839319229125977,
+      "learning_rate": 0.0001,
+      "loss": 1.2978,
+      "step": 14887
+    },
+    {
+      "epoch": 1.7101832175061742,
+      "grad_norm": 0.6496050953865051,
+      "learning_rate": 0.0001,
+      "loss": 1.6834,
+      "step": 14888
+    },
+    {
+      "epoch": 1.7102980874160014,
+      "grad_norm": 0.6293419599533081,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 14889
+    },
+    {
+      "epoch": 1.7104129573258287,
+      "grad_norm": 0.6812129616737366,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 14890
+    },
+    {
+      "epoch": 1.7105278272356557,
+      "grad_norm": 0.614952802658081,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 14891
+    },
+    {
+      "epoch": 1.7106426971454827,
+      "grad_norm": 0.6444807648658752,
+      "learning_rate": 0.0001,
+      "loss": 1.3534,
+      "step": 14892
+    },
+    {
+      "epoch": 1.71075756705531,
+      "grad_norm": 0.7069858312606812,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 14893
+    },
+    {
+      "epoch": 1.7108724369651371,
+      "grad_norm": 0.6311841607093811,
+      "learning_rate": 0.0001,
+      "loss": 1.4556,
+      "step": 14894
+    },
+    {
+      "epoch": 1.7109873068749641,
+      "grad_norm": 0.645004153251648,
+      "learning_rate": 0.0001,
+      "loss": 1.4576,
+      "step": 14895
+    },
+    {
+      "epoch": 1.7111021767847912,
+      "grad_norm": 0.613746166229248,
+      "learning_rate": 0.0001,
+      "loss": 1.351,
+      "step": 14896
+    },
+    {
+      "epoch": 1.7112170466946184,
+      "grad_norm": 0.653880774974823,
+      "learning_rate": 0.0001,
+      "loss": 1.4966,
+      "step": 14897
+    },
+    {
+      "epoch": 1.7113319166044456,
+      "grad_norm": 0.6353497505187988,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 14898
+    },
+    {
+      "epoch": 1.7114467865142726,
+      "grad_norm": 0.6185076236724854,
+      "learning_rate": 0.0001,
+      "loss": 1.3926,
+      "step": 14899
+    },
+    {
+      "epoch": 1.7115616564240996,
+      "grad_norm": 0.6185590028762817,
+      "learning_rate": 0.0001,
+      "loss": 1.4641,
+      "step": 14900
+    },
+    {
+      "epoch": 1.7116765263339269,
+      "grad_norm": 0.6211035847663879,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 14901
+    },
+    {
+      "epoch": 1.711791396243754,
+      "grad_norm": 0.5859823226928711,
+      "learning_rate": 0.0001,
+      "loss": 1.3688,
+      "step": 14902
+    },
+    {
+      "epoch": 1.711906266153581,
+      "grad_norm": 0.5817895531654358,
+      "learning_rate": 0.0001,
+      "loss": 1.3958,
+      "step": 14903
+    },
+    {
+      "epoch": 1.7120211360634081,
+      "grad_norm": 0.6351556181907654,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 14904
+    },
+    {
+      "epoch": 1.7121360059732353,
+      "grad_norm": 0.6283468008041382,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 14905
+    },
+    {
+      "epoch": 1.7122508758830626,
+      "grad_norm": 0.6462883353233337,
+      "learning_rate": 0.0001,
+      "loss": 1.5196,
+      "step": 14906
+    },
+    {
+      "epoch": 1.7123657457928896,
+      "grad_norm": 0.6173494458198547,
+      "learning_rate": 0.0001,
+      "loss": 1.4087,
+      "step": 14907
+    },
+    {
+      "epoch": 1.7124806157027166,
+      "grad_norm": 0.5699242949485779,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 14908
+    },
+    {
+      "epoch": 1.7125954856125438,
+      "grad_norm": 0.6764625310897827,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 14909
+    },
+    {
+      "epoch": 1.712710355522371,
+      "grad_norm": 0.6505712866783142,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 14910
+    },
+    {
+      "epoch": 1.712825225432198,
+      "grad_norm": 0.6406113505363464,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 14911
+    },
+    {
+      "epoch": 1.712940095342025,
+      "grad_norm": 0.621518075466156,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 14912
+    },
+    {
+      "epoch": 1.7130549652518523,
+      "grad_norm": 0.6753479838371277,
+      "learning_rate": 0.0001,
+      "loss": 1.7465,
+      "step": 14913
+    },
+    {
+      "epoch": 1.7131698351616795,
+      "grad_norm": 0.7874284982681274,
+      "learning_rate": 0.0001,
+      "loss": 1.5676,
+      "step": 14914
+    },
+    {
+      "epoch": 1.7132847050715065,
+      "grad_norm": 0.6016520857810974,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 14915
+    },
+    {
+      "epoch": 1.7133995749813336,
+      "grad_norm": 0.6585147976875305,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 14916
+    },
+    {
+      "epoch": 1.7135144448911608,
+      "grad_norm": 0.5877991318702698,
+      "learning_rate": 0.0001,
+      "loss": 1.4722,
+      "step": 14917
+    },
+    {
+      "epoch": 1.713629314800988,
+      "grad_norm": 0.5988264083862305,
+      "learning_rate": 0.0001,
+      "loss": 1.6484,
+      "step": 14918
+    },
+    {
+      "epoch": 1.713744184710815,
+      "grad_norm": 0.7021918296813965,
+      "learning_rate": 0.0001,
+      "loss": 1.5297,
+      "step": 14919
+    },
+    {
+      "epoch": 1.713859054620642,
+      "grad_norm": 0.6114031076431274,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 14920
+    },
+    {
+      "epoch": 1.7139739245304693,
+      "grad_norm": 0.646767258644104,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 14921
+    },
+    {
+      "epoch": 1.7140887944402965,
+      "grad_norm": 0.598789393901825,
+      "learning_rate": 0.0001,
+      "loss": 1.3379,
+      "step": 14922
+    },
+    {
+      "epoch": 1.7142036643501235,
+      "grad_norm": 0.6237581372261047,
+      "learning_rate": 0.0001,
+      "loss": 1.4198,
+      "step": 14923
+    },
+    {
+      "epoch": 1.7143185342599505,
+      "grad_norm": 0.613361120223999,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 14924
+    },
+    {
+      "epoch": 1.7144334041697777,
+      "grad_norm": 0.654420018196106,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 14925
+    },
+    {
+      "epoch": 1.714548274079605,
+      "grad_norm": 0.6176064014434814,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 14926
+    },
+    {
+      "epoch": 1.714663143989432,
+      "grad_norm": 0.6290966272354126,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 14927
+    },
+    {
+      "epoch": 1.714778013899259,
+      "grad_norm": 0.6438091397285461,
+      "learning_rate": 0.0001,
+      "loss": 1.3984,
+      "step": 14928
+    },
+    {
+      "epoch": 1.7148928838090862,
+      "grad_norm": 0.6490173935890198,
+      "learning_rate": 0.0001,
+      "loss": 1.5162,
+      "step": 14929
+    },
+    {
+      "epoch": 1.7150077537189135,
+      "grad_norm": 0.5619537830352783,
+      "learning_rate": 0.0001,
+      "loss": 1.4052,
+      "step": 14930
+    },
+    {
+      "epoch": 1.7151226236287405,
+      "grad_norm": 0.5740304589271545,
+      "learning_rate": 0.0001,
+      "loss": 1.3887,
+      "step": 14931
+    },
+    {
+      "epoch": 1.7152374935385675,
+      "grad_norm": 0.5972532629966736,
+      "learning_rate": 0.0001,
+      "loss": 1.488,
+      "step": 14932
+    },
+    {
+      "epoch": 1.7153523634483947,
+      "grad_norm": 0.6191779375076294,
+      "learning_rate": 0.0001,
+      "loss": 1.3594,
+      "step": 14933
+    },
+    {
+      "epoch": 1.715467233358222,
+      "grad_norm": 0.6333027482032776,
+      "learning_rate": 0.0001,
+      "loss": 1.3897,
+      "step": 14934
+    },
+    {
+      "epoch": 1.715582103268049,
+      "grad_norm": 0.5880884528160095,
+      "learning_rate": 0.0001,
+      "loss": 1.243,
+      "step": 14935
+    },
+    {
+      "epoch": 1.715696973177876,
+      "grad_norm": 0.7410986423492432,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 14936
+    },
+    {
+      "epoch": 1.7158118430877032,
+      "grad_norm": 0.7331770062446594,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 14937
+    },
+    {
+      "epoch": 1.7159267129975304,
+      "grad_norm": 0.6188743114471436,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 14938
+    },
+    {
+      "epoch": 1.7160415829073574,
+      "grad_norm": 0.6449677348136902,
+      "learning_rate": 0.0001,
+      "loss": 1.8204,
+      "step": 14939
+    },
+    {
+      "epoch": 1.7161564528171844,
+      "grad_norm": 0.6289303302764893,
+      "learning_rate": 0.0001,
+      "loss": 1.5189,
+      "step": 14940
+    },
+    {
+      "epoch": 1.7162713227270117,
+      "grad_norm": 0.5749282836914062,
+      "learning_rate": 0.0001,
+      "loss": 1.3974,
+      "step": 14941
+    },
+    {
+      "epoch": 1.716386192636839,
+      "grad_norm": 0.6385175585746765,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 14942
+    },
+    {
+      "epoch": 1.716501062546666,
+      "grad_norm": 0.6189382672309875,
+      "learning_rate": 0.0001,
+      "loss": 1.5049,
+      "step": 14943
+    },
+    {
+      "epoch": 1.716615932456493,
+      "grad_norm": 0.7417494058609009,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 14944
+    },
+    {
+      "epoch": 1.7167308023663201,
+      "grad_norm": 0.5887095332145691,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 14945
+    },
+    {
+      "epoch": 1.7168456722761474,
+      "grad_norm": 0.591689944267273,
+      "learning_rate": 0.0001,
+      "loss": 1.3988,
+      "step": 14946
+    },
+    {
+      "epoch": 1.7169605421859744,
+      "grad_norm": 0.7112393975257874,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 14947
+    },
+    {
+      "epoch": 1.7170754120958014,
+      "grad_norm": 0.6112138032913208,
+      "learning_rate": 0.0001,
+      "loss": 1.3103,
+      "step": 14948
+    },
+    {
+      "epoch": 1.7171902820056286,
+      "grad_norm": 0.6022821664810181,
+      "learning_rate": 0.0001,
+      "loss": 1.2665,
+      "step": 14949
+    },
+    {
+      "epoch": 1.7173051519154559,
+      "grad_norm": 0.6003796458244324,
+      "learning_rate": 0.0001,
+      "loss": 1.3495,
+      "step": 14950
+    },
+    {
+      "epoch": 1.7174200218252829,
+      "grad_norm": 0.7552711963653564,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 14951
+    },
+    {
+      "epoch": 1.7175348917351099,
+      "grad_norm": 0.6487937569618225,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 14952
+    },
+    {
+      "epoch": 1.717649761644937,
+      "grad_norm": 0.5925029516220093,
+      "learning_rate": 0.0001,
+      "loss": 1.3924,
+      "step": 14953
+    },
+    {
+      "epoch": 1.7177646315547643,
+      "grad_norm": 0.6977973580360413,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 14954
+    },
+    {
+      "epoch": 1.7178795014645913,
+      "grad_norm": 0.6343498826026917,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 14955
+    },
+    {
+      "epoch": 1.7179943713744183,
+      "grad_norm": 0.608635663986206,
+      "learning_rate": 0.0001,
+      "loss": 1.3089,
+      "step": 14956
+    },
+    {
+      "epoch": 1.7181092412842456,
+      "grad_norm": 0.6102705001831055,
+      "learning_rate": 0.0001,
+      "loss": 1.4239,
+      "step": 14957
+    },
+    {
+      "epoch": 1.7182241111940728,
+      "grad_norm": 0.6915038824081421,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 14958
+    },
+    {
+      "epoch": 1.7183389811038998,
+      "grad_norm": 0.6009780764579773,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 14959
+    },
+    {
+      "epoch": 1.7184538510137268,
+      "grad_norm": 0.5998890399932861,
+      "learning_rate": 0.0001,
+      "loss": 1.4216,
+      "step": 14960
+    },
+    {
+      "epoch": 1.718568720923554,
+      "grad_norm": 0.6564615964889526,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 14961
+    },
+    {
+      "epoch": 1.7186835908333813,
+      "grad_norm": 0.5995262861251831,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 14962
+    },
+    {
+      "epoch": 1.7187984607432083,
+      "grad_norm": 0.5520354509353638,
+      "learning_rate": 0.0001,
+      "loss": 1.3103,
+      "step": 14963
+    },
+    {
+      "epoch": 1.7189133306530353,
+      "grad_norm": 0.5722663402557373,
+      "learning_rate": 0.0001,
+      "loss": 1.2938,
+      "step": 14964
+    },
+    {
+      "epoch": 1.7190282005628625,
+      "grad_norm": 0.6511684060096741,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 14965
+    },
+    {
+      "epoch": 1.7191430704726898,
+      "grad_norm": 0.6239314675331116,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 14966
+    },
+    {
+      "epoch": 1.7192579403825168,
+      "grad_norm": 0.6765071153640747,
+      "learning_rate": 0.0001,
+      "loss": 1.7592,
+      "step": 14967
+    },
+    {
+      "epoch": 1.7193728102923438,
+      "grad_norm": 0.6002155542373657,
+      "learning_rate": 0.0001,
+      "loss": 1.4651,
+      "step": 14968
+    },
+    {
+      "epoch": 1.719487680202171,
+      "grad_norm": 0.6126995086669922,
+      "learning_rate": 0.0001,
+      "loss": 1.3875,
+      "step": 14969
+    },
+    {
+      "epoch": 1.7196025501119983,
+      "grad_norm": 0.6200347542762756,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 14970
+    },
+    {
+      "epoch": 1.7197174200218253,
+      "grad_norm": 0.6229084730148315,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 14971
+    },
+    {
+      "epoch": 1.7198322899316523,
+      "grad_norm": 0.5846944451332092,
+      "learning_rate": 0.0001,
+      "loss": 1.2907,
+      "step": 14972
+    },
+    {
+      "epoch": 1.7199471598414795,
+      "grad_norm": 0.6202731132507324,
+      "learning_rate": 0.0001,
+      "loss": 1.4005,
+      "step": 14973
+    },
+    {
+      "epoch": 1.7200620297513067,
+      "grad_norm": 0.5932221412658691,
+      "learning_rate": 0.0001,
+      "loss": 1.3362,
+      "step": 14974
+    },
+    {
+      "epoch": 1.7201768996611337,
+      "grad_norm": 0.5857172012329102,
+      "learning_rate": 0.0001,
+      "loss": 1.3536,
+      "step": 14975
+    },
+    {
+      "epoch": 1.7202917695709607,
+      "grad_norm": 0.6796727776527405,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 14976
+    },
+    {
+      "epoch": 1.720406639480788,
+      "grad_norm": 0.6333981156349182,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 14977
+    },
+    {
+      "epoch": 1.7205215093906152,
+      "grad_norm": 0.6612197160720825,
+      "learning_rate": 0.0001,
+      "loss": 1.3956,
+      "step": 14978
+    },
+    {
+      "epoch": 1.7206363793004422,
+      "grad_norm": 0.6356850862503052,
+      "learning_rate": 0.0001,
+      "loss": 1.4098,
+      "step": 14979
+    },
+    {
+      "epoch": 1.7207512492102692,
+      "grad_norm": 0.6078984141349792,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 14980
+    },
+    {
+      "epoch": 1.7208661191200965,
+      "grad_norm": 0.6189718842506409,
+      "learning_rate": 0.0001,
+      "loss": 1.4904,
+      "step": 14981
+    },
+    {
+      "epoch": 1.7209809890299237,
+      "grad_norm": 0.5823473334312439,
+      "learning_rate": 0.0001,
+      "loss": 1.3974,
+      "step": 14982
+    },
+    {
+      "epoch": 1.7210958589397507,
+      "grad_norm": 0.6768457889556885,
+      "learning_rate": 0.0001,
+      "loss": 1.5662,
+      "step": 14983
+    },
+    {
+      "epoch": 1.7212107288495777,
+      "grad_norm": 0.6085696220397949,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 14984
+    },
+    {
+      "epoch": 1.721325598759405,
+      "grad_norm": 0.6197189092636108,
+      "learning_rate": 0.0001,
+      "loss": 1.4594,
+      "step": 14985
+    },
+    {
+      "epoch": 1.7214404686692322,
+      "grad_norm": 0.6606781482696533,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 14986
+    },
+    {
+      "epoch": 1.7215553385790592,
+      "grad_norm": 0.5829959511756897,
+      "learning_rate": 0.0001,
+      "loss": 1.2158,
+      "step": 14987
+    },
+    {
+      "epoch": 1.7216702084888862,
+      "grad_norm": 0.6431682705879211,
+      "learning_rate": 0.0001,
+      "loss": 1.3648,
+      "step": 14988
+    },
+    {
+      "epoch": 1.7217850783987134,
+      "grad_norm": 0.6825764775276184,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 14989
+    },
+    {
+      "epoch": 1.7218999483085407,
+      "grad_norm": 0.5983292460441589,
+      "learning_rate": 0.0001,
+      "loss": 1.3756,
+      "step": 14990
+    },
+    {
+      "epoch": 1.7220148182183677,
+      "grad_norm": 0.6294005513191223,
+      "learning_rate": 0.0001,
+      "loss": 1.2658,
+      "step": 14991
+    },
+    {
+      "epoch": 1.7221296881281947,
+      "grad_norm": 0.6360632181167603,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 14992
+    },
+    {
+      "epoch": 1.722244558038022,
+      "grad_norm": 0.5845391154289246,
+      "learning_rate": 0.0001,
+      "loss": 1.2789,
+      "step": 14993
+    },
+    {
+      "epoch": 1.7223594279478491,
+      "grad_norm": 0.7193369269371033,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 14994
+    },
+    {
+      "epoch": 1.7224742978576761,
+      "grad_norm": 0.583917498588562,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 14995
+    },
+    {
+      "epoch": 1.7225891677675031,
+      "grad_norm": 0.6654466986656189,
+      "learning_rate": 0.0001,
+      "loss": 1.5591,
+      "step": 14996
+    },
+    {
+      "epoch": 1.7227040376773304,
+      "grad_norm": 0.5956998467445374,
+      "learning_rate": 0.0001,
+      "loss": 1.2864,
+      "step": 14997
+    },
+    {
+      "epoch": 1.7228189075871576,
+      "grad_norm": 0.6005557179450989,
+      "learning_rate": 0.0001,
+      "loss": 1.2126,
+      "step": 14998
+    },
+    {
+      "epoch": 1.7229337774969846,
+      "grad_norm": 0.6229847073554993,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 14999
+    },
+    {
+      "epoch": 1.7230486474068116,
+      "grad_norm": 0.5950932502746582,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 15000
+    },
+    {
+      "epoch": 1.7231635173166389,
+      "grad_norm": 0.627495288848877,
+      "learning_rate": 0.0001,
+      "loss": 1.4347,
+      "step": 15001
+    },
+    {
+      "epoch": 1.723278387226466,
+      "grad_norm": 0.5470762252807617,
+      "learning_rate": 0.0001,
+      "loss": 1.1846,
+      "step": 15002
+    },
+    {
+      "epoch": 1.723393257136293,
+      "grad_norm": 0.5681390166282654,
+      "learning_rate": 0.0001,
+      "loss": 1.3358,
+      "step": 15003
+    },
+    {
+      "epoch": 1.72350812704612,
+      "grad_norm": 0.5831446647644043,
+      "learning_rate": 0.0001,
+      "loss": 1.4402,
+      "step": 15004
+    },
+    {
+      "epoch": 1.7236229969559473,
+      "grad_norm": 0.6146191358566284,
+      "learning_rate": 0.0001,
+      "loss": 1.4376,
+      "step": 15005
+    },
+    {
+      "epoch": 1.7237378668657746,
+      "grad_norm": 0.6659471988677979,
+      "learning_rate": 0.0001,
+      "loss": 1.3841,
+      "step": 15006
+    },
+    {
+      "epoch": 1.7238527367756016,
+      "grad_norm": 0.6301655173301697,
+      "learning_rate": 0.0001,
+      "loss": 1.4554,
+      "step": 15007
+    },
+    {
+      "epoch": 1.7239676066854286,
+      "grad_norm": 0.6276340484619141,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 15008
+    },
+    {
+      "epoch": 1.7240824765952558,
+      "grad_norm": 0.5837733745574951,
+      "learning_rate": 0.0001,
+      "loss": 1.4584,
+      "step": 15009
+    },
+    {
+      "epoch": 1.724197346505083,
+      "grad_norm": 0.5841119885444641,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 15010
+    },
+    {
+      "epoch": 1.72431221641491,
+      "grad_norm": 0.6044458746910095,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 15011
+    },
+    {
+      "epoch": 1.7244270863247373,
+      "grad_norm": 0.6079664826393127,
+      "learning_rate": 0.0001,
+      "loss": 1.4079,
+      "step": 15012
+    },
+    {
+      "epoch": 1.7245419562345643,
+      "grad_norm": 0.8847506642341614,
+      "learning_rate": 0.0001,
+      "loss": 1.6107,
+      "step": 15013
+    },
+    {
+      "epoch": 1.7246568261443915,
+      "grad_norm": 0.6034465432167053,
+      "learning_rate": 0.0001,
+      "loss": 1.2711,
+      "step": 15014
+    },
+    {
+      "epoch": 1.7247716960542188,
+      "grad_norm": 0.8095591068267822,
+      "learning_rate": 0.0001,
+      "loss": 1.4583,
+      "step": 15015
+    },
+    {
+      "epoch": 1.7248865659640458,
+      "grad_norm": 0.5719621777534485,
+      "learning_rate": 0.0001,
+      "loss": 1.427,
+      "step": 15016
+    },
+    {
+      "epoch": 1.7250014358738728,
+      "grad_norm": 0.6097038388252258,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 15017
+    },
+    {
+      "epoch": 1.7251163057837,
+      "grad_norm": 0.64508455991745,
+      "learning_rate": 0.0001,
+      "loss": 1.5549,
+      "step": 15018
+    },
+    {
+      "epoch": 1.7252311756935272,
+      "grad_norm": 0.5853515863418579,
+      "learning_rate": 0.0001,
+      "loss": 1.3652,
+      "step": 15019
+    },
+    {
+      "epoch": 1.7253460456033543,
+      "grad_norm": 0.6391353011131287,
+      "learning_rate": 0.0001,
+      "loss": 1.4132,
+      "step": 15020
+    },
+    {
+      "epoch": 1.7254609155131813,
+      "grad_norm": 0.6568958163261414,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 15021
+    },
+    {
+      "epoch": 1.7255757854230085,
+      "grad_norm": 0.6337956190109253,
+      "learning_rate": 0.0001,
+      "loss": 1.5081,
+      "step": 15022
+    },
+    {
+      "epoch": 1.7256906553328357,
+      "grad_norm": 0.682131290435791,
+      "learning_rate": 0.0001,
+      "loss": 1.4523,
+      "step": 15023
+    },
+    {
+      "epoch": 1.7258055252426627,
+      "grad_norm": 0.6506146192550659,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 15024
+    },
+    {
+      "epoch": 1.7259203951524897,
+      "grad_norm": 0.6392216086387634,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 15025
+    },
+    {
+      "epoch": 1.726035265062317,
+      "grad_norm": 0.6349377632141113,
+      "learning_rate": 0.0001,
+      "loss": 1.3924,
+      "step": 15026
+    },
+    {
+      "epoch": 1.7261501349721442,
+      "grad_norm": 0.588093638420105,
+      "learning_rate": 0.0001,
+      "loss": 1.5297,
+      "step": 15027
+    },
+    {
+      "epoch": 1.7262650048819712,
+      "grad_norm": 0.6122344732284546,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 15028
+    },
+    {
+      "epoch": 1.7263798747917982,
+      "grad_norm": 0.6537800431251526,
+      "learning_rate": 0.0001,
+      "loss": 1.5847,
+      "step": 15029
+    },
+    {
+      "epoch": 1.7264947447016255,
+      "grad_norm": 0.6109773516654968,
+      "learning_rate": 0.0001,
+      "loss": 1.4354,
+      "step": 15030
+    },
+    {
+      "epoch": 1.7266096146114527,
+      "grad_norm": 0.576797604560852,
+      "learning_rate": 0.0001,
+      "loss": 1.3143,
+      "step": 15031
+    },
+    {
+      "epoch": 1.7267244845212797,
+      "grad_norm": 0.6462949514389038,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 15032
+    },
+    {
+      "epoch": 1.7268393544311067,
+      "grad_norm": 0.6329190731048584,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 15033
+    },
+    {
+      "epoch": 1.726954224340934,
+      "grad_norm": 0.5958088040351868,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 15034
+    },
+    {
+      "epoch": 1.7270690942507612,
+      "grad_norm": 0.630052387714386,
+      "learning_rate": 0.0001,
+      "loss": 1.3768,
+      "step": 15035
+    },
+    {
+      "epoch": 1.7271839641605882,
+      "grad_norm": 0.667178750038147,
+      "learning_rate": 0.0001,
+      "loss": 1.3865,
+      "step": 15036
+    },
+    {
+      "epoch": 1.7272988340704152,
+      "grad_norm": 0.7113081216812134,
+      "learning_rate": 0.0001,
+      "loss": 1.7048,
+      "step": 15037
+    },
+    {
+      "epoch": 1.7274137039802424,
+      "grad_norm": 0.6631393432617188,
+      "learning_rate": 0.0001,
+      "loss": 1.4899,
+      "step": 15038
+    },
+    {
+      "epoch": 1.7275285738900696,
+      "grad_norm": 0.6248582005500793,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 15039
+    },
+    {
+      "epoch": 1.7276434437998966,
+      "grad_norm": 0.6079177260398865,
+      "learning_rate": 0.0001,
+      "loss": 1.3492,
+      "step": 15040
+    },
+    {
+      "epoch": 1.7277583137097237,
+      "grad_norm": 0.5928921699523926,
+      "learning_rate": 0.0001,
+      "loss": 1.4487,
+      "step": 15041
+    },
+    {
+      "epoch": 1.727873183619551,
+      "grad_norm": 0.6063787341117859,
+      "learning_rate": 0.0001,
+      "loss": 1.225,
+      "step": 15042
+    },
+    {
+      "epoch": 1.7279880535293781,
+      "grad_norm": 0.6443949937820435,
+      "learning_rate": 0.0001,
+      "loss": 1.5223,
+      "step": 15043
+    },
+    {
+      "epoch": 1.7281029234392051,
+      "grad_norm": 0.6002311110496521,
+      "learning_rate": 0.0001,
+      "loss": 1.474,
+      "step": 15044
+    },
+    {
+      "epoch": 1.7282177933490321,
+      "grad_norm": 0.6543359756469727,
+      "learning_rate": 0.0001,
+      "loss": 1.4626,
+      "step": 15045
+    },
+    {
+      "epoch": 1.7283326632588594,
+      "grad_norm": 0.6799443364143372,
+      "learning_rate": 0.0001,
+      "loss": 1.5299,
+      "step": 15046
+    },
+    {
+      "epoch": 1.7284475331686866,
+      "grad_norm": 0.6205649971961975,
+      "learning_rate": 0.0001,
+      "loss": 1.2928,
+      "step": 15047
+    },
+    {
+      "epoch": 1.7285624030785136,
+      "grad_norm": 0.589119553565979,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 15048
+    },
+    {
+      "epoch": 1.7286772729883406,
+      "grad_norm": 0.6019067168235779,
+      "learning_rate": 0.0001,
+      "loss": 1.3567,
+      "step": 15049
+    },
+    {
+      "epoch": 1.7287921428981678,
+      "grad_norm": 0.5993046760559082,
+      "learning_rate": 0.0001,
+      "loss": 1.3804,
+      "step": 15050
+    },
+    {
+      "epoch": 1.728907012807995,
+      "grad_norm": 0.6053866147994995,
+      "learning_rate": 0.0001,
+      "loss": 1.4004,
+      "step": 15051
+    },
+    {
+      "epoch": 1.729021882717822,
+      "grad_norm": 0.5996253490447998,
+      "learning_rate": 0.0001,
+      "loss": 1.1198,
+      "step": 15052
+    },
+    {
+      "epoch": 1.729136752627649,
+      "grad_norm": 0.6605576872825623,
+      "learning_rate": 0.0001,
+      "loss": 1.4068,
+      "step": 15053
+    },
+    {
+      "epoch": 1.7292516225374763,
+      "grad_norm": 0.5937269926071167,
+      "learning_rate": 0.0001,
+      "loss": 1.4028,
+      "step": 15054
+    },
+    {
+      "epoch": 1.7293664924473036,
+      "grad_norm": 0.6884070038795471,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 15055
+    },
+    {
+      "epoch": 1.7294813623571306,
+      "grad_norm": 0.6444214582443237,
+      "learning_rate": 0.0001,
+      "loss": 1.4598,
+      "step": 15056
+    },
+    {
+      "epoch": 1.7295962322669576,
+      "grad_norm": 0.6065829992294312,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 15057
+    },
+    {
+      "epoch": 1.7297111021767848,
+      "grad_norm": 0.6256981492042542,
+      "learning_rate": 0.0001,
+      "loss": 1.2783,
+      "step": 15058
+    },
+    {
+      "epoch": 1.729825972086612,
+      "grad_norm": 0.641852617263794,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 15059
+    },
+    {
+      "epoch": 1.729940841996439,
+      "grad_norm": 0.5563915967941284,
+      "learning_rate": 0.0001,
+      "loss": 1.4621,
+      "step": 15060
+    },
+    {
+      "epoch": 1.730055711906266,
+      "grad_norm": 0.6508216261863708,
+      "learning_rate": 0.0001,
+      "loss": 1.3978,
+      "step": 15061
+    },
+    {
+      "epoch": 1.7301705818160933,
+      "grad_norm": 0.61888188123703,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 15062
+    },
+    {
+      "epoch": 1.7302854517259205,
+      "grad_norm": 0.6529124975204468,
+      "learning_rate": 0.0001,
+      "loss": 1.3718,
+      "step": 15063
+    },
+    {
+      "epoch": 1.7304003216357475,
+      "grad_norm": 0.5647428631782532,
+      "learning_rate": 0.0001,
+      "loss": 1.4557,
+      "step": 15064
+    },
+    {
+      "epoch": 1.7305151915455745,
+      "grad_norm": 0.710871696472168,
+      "learning_rate": 0.0001,
+      "loss": 1.6701,
+      "step": 15065
+    },
+    {
+      "epoch": 1.7306300614554018,
+      "grad_norm": 0.6646068692207336,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 15066
+    },
+    {
+      "epoch": 1.730744931365229,
+      "grad_norm": 0.6500128507614136,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 15067
+    },
+    {
+      "epoch": 1.730859801275056,
+      "grad_norm": 0.6322208642959595,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 15068
+    },
+    {
+      "epoch": 1.730974671184883,
+      "grad_norm": 0.5904911756515503,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 15069
+    },
+    {
+      "epoch": 1.7310895410947102,
+      "grad_norm": 0.639840841293335,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 15070
+    },
+    {
+      "epoch": 1.7312044110045375,
+      "grad_norm": 0.6180384159088135,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 15071
+    },
+    {
+      "epoch": 1.7313192809143645,
+      "grad_norm": 0.5702844858169556,
+      "learning_rate": 0.0001,
+      "loss": 1.3738,
+      "step": 15072
+    },
+    {
+      "epoch": 1.7314341508241915,
+      "grad_norm": 0.6036345362663269,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 15073
+    },
+    {
+      "epoch": 1.7315490207340187,
+      "grad_norm": 0.6159979104995728,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 15074
+    },
+    {
+      "epoch": 1.731663890643846,
+      "grad_norm": 0.5626546740531921,
+      "learning_rate": 0.0001,
+      "loss": 1.422,
+      "step": 15075
+    },
+    {
+      "epoch": 1.731778760553673,
+      "grad_norm": 0.6199283599853516,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 15076
+    },
+    {
+      "epoch": 1.7318936304635,
+      "grad_norm": 0.6209295392036438,
+      "learning_rate": 0.0001,
+      "loss": 1.5764,
+      "step": 15077
+    },
+    {
+      "epoch": 1.7320085003733272,
+      "grad_norm": 0.574235200881958,
+      "learning_rate": 0.0001,
+      "loss": 1.4249,
+      "step": 15078
+    },
+    {
+      "epoch": 1.7321233702831544,
+      "grad_norm": 0.6106351017951965,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 15079
+    },
+    {
+      "epoch": 1.7322382401929814,
+      "grad_norm": 0.6253929138183594,
+      "learning_rate": 0.0001,
+      "loss": 1.3034,
+      "step": 15080
+    },
+    {
+      "epoch": 1.7323531101028085,
+      "grad_norm": 0.6872732639312744,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 15081
+    },
+    {
+      "epoch": 1.7324679800126357,
+      "grad_norm": 0.6001285314559937,
+      "learning_rate": 0.0001,
+      "loss": 1.3798,
+      "step": 15082
+    },
+    {
+      "epoch": 1.732582849922463,
+      "grad_norm": 0.5744822025299072,
+      "learning_rate": 0.0001,
+      "loss": 1.5976,
+      "step": 15083
+    },
+    {
+      "epoch": 1.73269771983229,
+      "grad_norm": 0.5871395468711853,
+      "learning_rate": 0.0001,
+      "loss": 1.4265,
+      "step": 15084
+    },
+    {
+      "epoch": 1.732812589742117,
+      "grad_norm": 0.586462140083313,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 15085
+    },
+    {
+      "epoch": 1.7329274596519442,
+      "grad_norm": 0.6341403722763062,
+      "learning_rate": 0.0001,
+      "loss": 1.4599,
+      "step": 15086
+    },
+    {
+      "epoch": 1.7330423295617714,
+      "grad_norm": 0.5986180901527405,
+      "learning_rate": 0.0001,
+      "loss": 1.3682,
+      "step": 15087
+    },
+    {
+      "epoch": 1.7331571994715984,
+      "grad_norm": 0.6208639144897461,
+      "learning_rate": 0.0001,
+      "loss": 1.5485,
+      "step": 15088
+    },
+    {
+      "epoch": 1.7332720693814254,
+      "grad_norm": 0.684834361076355,
+      "learning_rate": 0.0001,
+      "loss": 1.5264,
+      "step": 15089
+    },
+    {
+      "epoch": 1.7333869392912526,
+      "grad_norm": 0.692686915397644,
+      "learning_rate": 0.0001,
+      "loss": 1.441,
+      "step": 15090
+    },
+    {
+      "epoch": 1.7335018092010799,
+      "grad_norm": 0.621494710445404,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 15091
+    },
+    {
+      "epoch": 1.7336166791109069,
+      "grad_norm": 0.6340914368629456,
+      "learning_rate": 0.0001,
+      "loss": 1.5213,
+      "step": 15092
+    },
+    {
+      "epoch": 1.733731549020734,
+      "grad_norm": 0.6137570142745972,
+      "learning_rate": 0.0001,
+      "loss": 1.3492,
+      "step": 15093
+    },
+    {
+      "epoch": 1.7338464189305611,
+      "grad_norm": 0.6438025832176208,
+      "learning_rate": 0.0001,
+      "loss": 1.4728,
+      "step": 15094
+    },
+    {
+      "epoch": 1.7339612888403884,
+      "grad_norm": 0.5989691615104675,
+      "learning_rate": 0.0001,
+      "loss": 1.4918,
+      "step": 15095
+    },
+    {
+      "epoch": 1.7340761587502154,
+      "grad_norm": 0.6587399840354919,
+      "learning_rate": 0.0001,
+      "loss": 1.2788,
+      "step": 15096
+    },
+    {
+      "epoch": 1.7341910286600424,
+      "grad_norm": 0.6119810342788696,
+      "learning_rate": 0.0001,
+      "loss": 1.3324,
+      "step": 15097
+    },
+    {
+      "epoch": 1.7343058985698696,
+      "grad_norm": 0.6475637555122375,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 15098
+    },
+    {
+      "epoch": 1.7344207684796968,
+      "grad_norm": 0.5902281403541565,
+      "learning_rate": 0.0001,
+      "loss": 1.4495,
+      "step": 15099
+    },
+    {
+      "epoch": 1.7345356383895238,
+      "grad_norm": 0.5807440876960754,
+      "learning_rate": 0.0001,
+      "loss": 1.3451,
+      "step": 15100
+    },
+    {
+      "epoch": 1.7346505082993509,
+      "grad_norm": 0.6495022773742676,
+      "learning_rate": 0.0001,
+      "loss": 1.5186,
+      "step": 15101
+    },
+    {
+      "epoch": 1.734765378209178,
+      "grad_norm": 0.65216064453125,
+      "learning_rate": 0.0001,
+      "loss": 1.4218,
+      "step": 15102
+    },
+    {
+      "epoch": 1.7348802481190053,
+      "grad_norm": 0.6560716032981873,
+      "learning_rate": 0.0001,
+      "loss": 1.4779,
+      "step": 15103
+    },
+    {
+      "epoch": 1.7349951180288323,
+      "grad_norm": 0.5937643051147461,
+      "learning_rate": 0.0001,
+      "loss": 1.3291,
+      "step": 15104
+    },
+    {
+      "epoch": 1.7351099879386593,
+      "grad_norm": 0.6294172406196594,
+      "learning_rate": 0.0001,
+      "loss": 1.2777,
+      "step": 15105
+    },
+    {
+      "epoch": 1.7352248578484866,
+      "grad_norm": 0.6666632890701294,
+      "learning_rate": 0.0001,
+      "loss": 1.3235,
+      "step": 15106
+    },
+    {
+      "epoch": 1.7353397277583138,
+      "grad_norm": 0.6566066741943359,
+      "learning_rate": 0.0001,
+      "loss": 1.3513,
+      "step": 15107
+    },
+    {
+      "epoch": 1.7354545976681408,
+      "grad_norm": 0.6714562773704529,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 15108
+    },
+    {
+      "epoch": 1.7355694675779678,
+      "grad_norm": 0.5992361307144165,
+      "learning_rate": 0.0001,
+      "loss": 1.3927,
+      "step": 15109
+    },
+    {
+      "epoch": 1.735684337487795,
+      "grad_norm": 0.5907565951347351,
+      "learning_rate": 0.0001,
+      "loss": 1.5398,
+      "step": 15110
+    },
+    {
+      "epoch": 1.7357992073976223,
+      "grad_norm": 0.6381534934043884,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 15111
+    },
+    {
+      "epoch": 1.7359140773074493,
+      "grad_norm": 0.5985099077224731,
+      "learning_rate": 0.0001,
+      "loss": 1.4673,
+      "step": 15112
+    },
+    {
+      "epoch": 1.7360289472172763,
+      "grad_norm": 0.566951334476471,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 15113
+    },
+    {
+      "epoch": 1.7361438171271035,
+      "grad_norm": 0.6164551973342896,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 15114
+    },
+    {
+      "epoch": 1.7362586870369308,
+      "grad_norm": 0.5941165089607239,
+      "learning_rate": 0.0001,
+      "loss": 1.3014,
+      "step": 15115
+    },
+    {
+      "epoch": 1.7363735569467578,
+      "grad_norm": 0.5940799117088318,
+      "learning_rate": 0.0001,
+      "loss": 1.6121,
+      "step": 15116
+    },
+    {
+      "epoch": 1.7364884268565848,
+      "grad_norm": 0.629859209060669,
+      "learning_rate": 0.0001,
+      "loss": 1.6983,
+      "step": 15117
+    },
+    {
+      "epoch": 1.736603296766412,
+      "grad_norm": 0.6390464305877686,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 15118
+    },
+    {
+      "epoch": 1.7367181666762392,
+      "grad_norm": 0.6173864006996155,
+      "learning_rate": 0.0001,
+      "loss": 1.3426,
+      "step": 15119
+    },
+    {
+      "epoch": 1.7368330365860662,
+      "grad_norm": 0.5565081238746643,
+      "learning_rate": 0.0001,
+      "loss": 1.4207,
+      "step": 15120
+    },
+    {
+      "epoch": 1.7369479064958933,
+      "grad_norm": 0.6064382791519165,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 15121
+    },
+    {
+      "epoch": 1.7370627764057205,
+      "grad_norm": 0.6307348608970642,
+      "learning_rate": 0.0001,
+      "loss": 1.3269,
+      "step": 15122
+    },
+    {
+      "epoch": 1.7371776463155477,
+      "grad_norm": 0.63545823097229,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 15123
+    },
+    {
+      "epoch": 1.7372925162253747,
+      "grad_norm": 0.6427724957466125,
+      "learning_rate": 0.0001,
+      "loss": 1.4303,
+      "step": 15124
+    },
+    {
+      "epoch": 1.7374073861352017,
+      "grad_norm": 0.5886901021003723,
+      "learning_rate": 0.0001,
+      "loss": 1.2971,
+      "step": 15125
+    },
+    {
+      "epoch": 1.737522256045029,
+      "grad_norm": 0.6295827031135559,
+      "learning_rate": 0.0001,
+      "loss": 1.2244,
+      "step": 15126
+    },
+    {
+      "epoch": 1.7376371259548562,
+      "grad_norm": 0.5705087780952454,
+      "learning_rate": 0.0001,
+      "loss": 1.2789,
+      "step": 15127
+    },
+    {
+      "epoch": 1.7377519958646832,
+      "grad_norm": 0.6834404468536377,
+      "learning_rate": 0.0001,
+      "loss": 1.3771,
+      "step": 15128
+    },
+    {
+      "epoch": 1.7378668657745102,
+      "grad_norm": 0.6291454434394836,
+      "learning_rate": 0.0001,
+      "loss": 1.3769,
+      "step": 15129
+    },
+    {
+      "epoch": 1.7379817356843374,
+      "grad_norm": 0.7055854201316833,
+      "learning_rate": 0.0001,
+      "loss": 1.609,
+      "step": 15130
+    },
+    {
+      "epoch": 1.7380966055941647,
+      "grad_norm": 0.6519612669944763,
+      "learning_rate": 0.0001,
+      "loss": 1.3599,
+      "step": 15131
+    },
+    {
+      "epoch": 1.7382114755039917,
+      "grad_norm": 0.7939478754997253,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 15132
+    },
+    {
+      "epoch": 1.7383263454138187,
+      "grad_norm": 0.5996159315109253,
+      "learning_rate": 0.0001,
+      "loss": 1.4736,
+      "step": 15133
+    },
+    {
+      "epoch": 1.738441215323646,
+      "grad_norm": 0.6544761061668396,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 15134
+    },
+    {
+      "epoch": 1.7385560852334732,
+      "grad_norm": 0.5604496002197266,
+      "learning_rate": 0.0001,
+      "loss": 1.2495,
+      "step": 15135
+    },
+    {
+      "epoch": 1.7386709551433002,
+      "grad_norm": 0.6021955609321594,
+      "learning_rate": 0.0001,
+      "loss": 1.4273,
+      "step": 15136
+    },
+    {
+      "epoch": 1.7387858250531272,
+      "grad_norm": 0.6894492506980896,
+      "learning_rate": 0.0001,
+      "loss": 1.4778,
+      "step": 15137
+    },
+    {
+      "epoch": 1.7389006949629544,
+      "grad_norm": 0.5934672355651855,
+      "learning_rate": 0.0001,
+      "loss": 1.4288,
+      "step": 15138
+    },
+    {
+      "epoch": 1.7390155648727816,
+      "grad_norm": 0.6965713500976562,
+      "learning_rate": 0.0001,
+      "loss": 1.5802,
+      "step": 15139
+    },
+    {
+      "epoch": 1.7391304347826086,
+      "grad_norm": 0.6328989863395691,
+      "learning_rate": 0.0001,
+      "loss": 1.3815,
+      "step": 15140
+    },
+    {
+      "epoch": 1.7392453046924357,
+      "grad_norm": 0.6331512928009033,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 15141
+    },
+    {
+      "epoch": 1.7393601746022629,
+      "grad_norm": 0.7313364148139954,
+      "learning_rate": 0.0001,
+      "loss": 1.6402,
+      "step": 15142
+    },
+    {
+      "epoch": 1.7394750445120901,
+      "grad_norm": 0.6274045705795288,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 15143
+    },
+    {
+      "epoch": 1.7395899144219171,
+      "grad_norm": 0.6259190440177917,
+      "learning_rate": 0.0001,
+      "loss": 1.4399,
+      "step": 15144
+    },
+    {
+      "epoch": 1.7397047843317441,
+      "grad_norm": 0.6094817519187927,
+      "learning_rate": 0.0001,
+      "loss": 1.5018,
+      "step": 15145
+    },
+    {
+      "epoch": 1.7398196542415714,
+      "grad_norm": 0.6577267646789551,
+      "learning_rate": 0.0001,
+      "loss": 1.3699,
+      "step": 15146
+    },
+    {
+      "epoch": 1.7399345241513986,
+      "grad_norm": 0.6378180980682373,
+      "learning_rate": 0.0001,
+      "loss": 1.469,
+      "step": 15147
+    },
+    {
+      "epoch": 1.7400493940612256,
+      "grad_norm": 0.6114896535873413,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 15148
+    },
+    {
+      "epoch": 1.7401642639710528,
+      "grad_norm": 0.6004807353019714,
+      "learning_rate": 0.0001,
+      "loss": 1.4736,
+      "step": 15149
+    },
+    {
+      "epoch": 1.7402791338808798,
+      "grad_norm": 0.5777592658996582,
+      "learning_rate": 0.0001,
+      "loss": 1.463,
+      "step": 15150
+    },
+    {
+      "epoch": 1.740394003790707,
+      "grad_norm": 0.667898416519165,
+      "learning_rate": 0.0001,
+      "loss": 1.4367,
+      "step": 15151
+    },
+    {
+      "epoch": 1.7405088737005343,
+      "grad_norm": 0.6375964879989624,
+      "learning_rate": 0.0001,
+      "loss": 1.4712,
+      "step": 15152
+    },
+    {
+      "epoch": 1.7406237436103613,
+      "grad_norm": 0.6091217398643494,
+      "learning_rate": 0.0001,
+      "loss": 1.4241,
+      "step": 15153
+    },
+    {
+      "epoch": 1.7407386135201883,
+      "grad_norm": 0.6524693965911865,
+      "learning_rate": 0.0001,
+      "loss": 1.3071,
+      "step": 15154
+    },
+    {
+      "epoch": 1.7408534834300156,
+      "grad_norm": 0.6394875645637512,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 15155
+    },
+    {
+      "epoch": 1.7409683533398428,
+      "grad_norm": 0.6616272926330566,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 15156
+    },
+    {
+      "epoch": 1.7410832232496698,
+      "grad_norm": 0.6061643958091736,
+      "learning_rate": 0.0001,
+      "loss": 1.4134,
+      "step": 15157
+    },
+    {
+      "epoch": 1.7411980931594968,
+      "grad_norm": 0.5694348216056824,
+      "learning_rate": 0.0001,
+      "loss": 1.44,
+      "step": 15158
+    },
+    {
+      "epoch": 1.741312963069324,
+      "grad_norm": 0.651995062828064,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 15159
+    },
+    {
+      "epoch": 1.7414278329791513,
+      "grad_norm": 0.6059070229530334,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 15160
+    },
+    {
+      "epoch": 1.7415427028889783,
+      "grad_norm": 0.58353590965271,
+      "learning_rate": 0.0001,
+      "loss": 1.4756,
+      "step": 15161
+    },
+    {
+      "epoch": 1.7416575727988053,
+      "grad_norm": 0.5728119611740112,
+      "learning_rate": 0.0001,
+      "loss": 1.3305,
+      "step": 15162
+    },
+    {
+      "epoch": 1.7417724427086325,
+      "grad_norm": 0.5594245791435242,
+      "learning_rate": 0.0001,
+      "loss": 1.295,
+      "step": 15163
+    },
+    {
+      "epoch": 1.7418873126184597,
+      "grad_norm": 0.6906068325042725,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 15164
+    },
+    {
+      "epoch": 1.7420021825282868,
+      "grad_norm": 0.6946594715118408,
+      "learning_rate": 0.0001,
+      "loss": 1.5164,
+      "step": 15165
+    },
+    {
+      "epoch": 1.7421170524381138,
+      "grad_norm": 0.6139326095581055,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 15166
+    },
+    {
+      "epoch": 1.742231922347941,
+      "grad_norm": 0.6352965235710144,
+      "learning_rate": 0.0001,
+      "loss": 1.4031,
+      "step": 15167
+    },
+    {
+      "epoch": 1.7423467922577682,
+      "grad_norm": 0.6800305843353271,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 15168
+    },
+    {
+      "epoch": 1.7424616621675952,
+      "grad_norm": 0.5998824834823608,
+      "learning_rate": 0.0001,
+      "loss": 1.3361,
+      "step": 15169
+    },
+    {
+      "epoch": 1.7425765320774222,
+      "grad_norm": 0.6264121532440186,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 15170
+    },
+    {
+      "epoch": 1.7426914019872495,
+      "grad_norm": 0.6085196733474731,
+      "learning_rate": 0.0001,
+      "loss": 1.3164,
+      "step": 15171
+    },
+    {
+      "epoch": 1.7428062718970767,
+      "grad_norm": 0.6375176310539246,
+      "learning_rate": 0.0001,
+      "loss": 1.4239,
+      "step": 15172
+    },
+    {
+      "epoch": 1.7429211418069037,
+      "grad_norm": 0.6847028732299805,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 15173
+    },
+    {
+      "epoch": 1.7430360117167307,
+      "grad_norm": 0.8053979873657227,
+      "learning_rate": 0.0001,
+      "loss": 1.5727,
+      "step": 15174
+    },
+    {
+      "epoch": 1.743150881626558,
+      "grad_norm": 0.6131575107574463,
+      "learning_rate": 0.0001,
+      "loss": 1.3573,
+      "step": 15175
+    },
+    {
+      "epoch": 1.7432657515363852,
+      "grad_norm": 0.570138692855835,
+      "learning_rate": 0.0001,
+      "loss": 1.2953,
+      "step": 15176
+    },
+    {
+      "epoch": 1.7433806214462122,
+      "grad_norm": 0.6050834059715271,
+      "learning_rate": 0.0001,
+      "loss": 1.2771,
+      "step": 15177
+    },
+    {
+      "epoch": 1.7434954913560392,
+      "grad_norm": 0.6091510653495789,
+      "learning_rate": 0.0001,
+      "loss": 1.4857,
+      "step": 15178
+    },
+    {
+      "epoch": 1.7436103612658664,
+      "grad_norm": 0.7901420593261719,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 15179
+    },
+    {
+      "epoch": 1.7437252311756937,
+      "grad_norm": 0.675471842288971,
+      "learning_rate": 0.0001,
+      "loss": 1.3616,
+      "step": 15180
+    },
+    {
+      "epoch": 1.7438401010855207,
+      "grad_norm": 0.6008798480033875,
+      "learning_rate": 0.0001,
+      "loss": 1.4139,
+      "step": 15181
+    },
+    {
+      "epoch": 1.7439549709953477,
+      "grad_norm": 0.6464613080024719,
+      "learning_rate": 0.0001,
+      "loss": 1.3417,
+      "step": 15182
+    },
+    {
+      "epoch": 1.744069840905175,
+      "grad_norm": 0.6285551190376282,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 15183
+    },
+    {
+      "epoch": 1.7441847108150021,
+      "grad_norm": 0.6213048100471497,
+      "learning_rate": 0.0001,
+      "loss": 1.3166,
+      "step": 15184
+    },
+    {
+      "epoch": 1.7442995807248292,
+      "grad_norm": 0.6194262504577637,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 15185
+    },
+    {
+      "epoch": 1.7444144506346562,
+      "grad_norm": 0.6530826687812805,
+      "learning_rate": 0.0001,
+      "loss": 1.563,
+      "step": 15186
+    },
+    {
+      "epoch": 1.7445293205444834,
+      "grad_norm": 0.6780787110328674,
+      "learning_rate": 0.0001,
+      "loss": 1.2973,
+      "step": 15187
+    },
+    {
+      "epoch": 1.7446441904543106,
+      "grad_norm": 0.7618767023086548,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 15188
+    },
+    {
+      "epoch": 1.7447590603641376,
+      "grad_norm": 0.6271486282348633,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 15189
+    },
+    {
+      "epoch": 1.7448739302739646,
+      "grad_norm": 0.5610907077789307,
+      "learning_rate": 0.0001,
+      "loss": 1.2431,
+      "step": 15190
+    },
+    {
+      "epoch": 1.7449888001837919,
+      "grad_norm": 0.6444764733314514,
+      "learning_rate": 0.0001,
+      "loss": 1.3007,
+      "step": 15191
+    },
+    {
+      "epoch": 1.745103670093619,
+      "grad_norm": 0.653785765171051,
+      "learning_rate": 0.0001,
+      "loss": 1.7039,
+      "step": 15192
+    },
+    {
+      "epoch": 1.7452185400034461,
+      "grad_norm": 0.6614614129066467,
+      "learning_rate": 0.0001,
+      "loss": 1.5011,
+      "step": 15193
+    },
+    {
+      "epoch": 1.7453334099132731,
+      "grad_norm": 0.6090264320373535,
+      "learning_rate": 0.0001,
+      "loss": 1.4303,
+      "step": 15194
+    },
+    {
+      "epoch": 1.7454482798231004,
+      "grad_norm": 0.5970935821533203,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 15195
+    },
+    {
+      "epoch": 1.7455631497329276,
+      "grad_norm": 0.6474558711051941,
+      "learning_rate": 0.0001,
+      "loss": 1.5577,
+      "step": 15196
+    },
+    {
+      "epoch": 1.7456780196427546,
+      "grad_norm": 0.6696446537971497,
+      "learning_rate": 0.0001,
+      "loss": 1.3981,
+      "step": 15197
+    },
+    {
+      "epoch": 1.7457928895525816,
+      "grad_norm": 0.6368044018745422,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 15198
+    },
+    {
+      "epoch": 1.7459077594624088,
+      "grad_norm": 0.5889632105827332,
+      "learning_rate": 0.0001,
+      "loss": 1.3663,
+      "step": 15199
+    },
+    {
+      "epoch": 1.746022629372236,
+      "grad_norm": 0.6324240565299988,
+      "learning_rate": 0.0001,
+      "loss": 1.4025,
+      "step": 15200
+    },
+    {
+      "epoch": 1.746137499282063,
+      "grad_norm": 0.664838969707489,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 15201
+    },
+    {
+      "epoch": 1.74625236919189,
+      "grad_norm": 0.609592616558075,
+      "learning_rate": 0.0001,
+      "loss": 1.4405,
+      "step": 15202
+    },
+    {
+      "epoch": 1.7463672391017173,
+      "grad_norm": 0.64817214012146,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 15203
+    },
+    {
+      "epoch": 1.7464821090115445,
+      "grad_norm": 0.5915722846984863,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 15204
+    },
+    {
+      "epoch": 1.7465969789213716,
+      "grad_norm": 0.6245658993721008,
+      "learning_rate": 0.0001,
+      "loss": 1.4266,
+      "step": 15205
+    },
+    {
+      "epoch": 1.7467118488311986,
+      "grad_norm": 0.5552368760108948,
+      "learning_rate": 0.0001,
+      "loss": 1.3063,
+      "step": 15206
+    },
+    {
+      "epoch": 1.7468267187410258,
+      "grad_norm": 0.6671764254570007,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 15207
+    },
+    {
+      "epoch": 1.746941588650853,
+      "grad_norm": 0.6359453201293945,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 15208
+    },
+    {
+      "epoch": 1.74705645856068,
+      "grad_norm": 0.6473332643508911,
+      "learning_rate": 0.0001,
+      "loss": 1.4903,
+      "step": 15209
+    },
+    {
+      "epoch": 1.747171328470507,
+      "grad_norm": 0.609463095664978,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 15210
+    },
+    {
+      "epoch": 1.7472861983803343,
+      "grad_norm": 0.7541261911392212,
+      "learning_rate": 0.0001,
+      "loss": 1.4196,
+      "step": 15211
+    },
+    {
+      "epoch": 1.7474010682901615,
+      "grad_norm": 0.6469170451164246,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 15212
+    },
+    {
+      "epoch": 1.7475159381999885,
+      "grad_norm": 0.6460253000259399,
+      "learning_rate": 0.0001,
+      "loss": 1.4009,
+      "step": 15213
+    },
+    {
+      "epoch": 1.7476308081098155,
+      "grad_norm": 0.6805127859115601,
+      "learning_rate": 0.0001,
+      "loss": 1.4109,
+      "step": 15214
+    },
+    {
+      "epoch": 1.7477456780196428,
+      "grad_norm": 0.6314157843589783,
+      "learning_rate": 0.0001,
+      "loss": 1.3916,
+      "step": 15215
+    },
+    {
+      "epoch": 1.74786054792947,
+      "grad_norm": 0.5894502997398376,
+      "learning_rate": 0.0001,
+      "loss": 1.2586,
+      "step": 15216
+    },
+    {
+      "epoch": 1.747975417839297,
+      "grad_norm": 0.6559074521064758,
+      "learning_rate": 0.0001,
+      "loss": 1.6868,
+      "step": 15217
+    },
+    {
+      "epoch": 1.748090287749124,
+      "grad_norm": 0.624843955039978,
+      "learning_rate": 0.0001,
+      "loss": 1.3526,
+      "step": 15218
+    },
+    {
+      "epoch": 1.7482051576589512,
+      "grad_norm": 0.6877935528755188,
+      "learning_rate": 0.0001,
+      "loss": 1.4833,
+      "step": 15219
+    },
+    {
+      "epoch": 1.7483200275687785,
+      "grad_norm": 0.6488064527511597,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 15220
+    },
+    {
+      "epoch": 1.7484348974786055,
+      "grad_norm": 0.7320221066474915,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 15221
+    },
+    {
+      "epoch": 1.7485497673884325,
+      "grad_norm": 0.6499512195587158,
+      "learning_rate": 0.0001,
+      "loss": 1.4216,
+      "step": 15222
+    },
+    {
+      "epoch": 1.7486646372982597,
+      "grad_norm": 0.6216921806335449,
+      "learning_rate": 0.0001,
+      "loss": 1.4012,
+      "step": 15223
+    },
+    {
+      "epoch": 1.748779507208087,
+      "grad_norm": 0.6483284831047058,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 15224
+    },
+    {
+      "epoch": 1.748894377117914,
+      "grad_norm": 0.6150051951408386,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 15225
+    },
+    {
+      "epoch": 1.749009247027741,
+      "grad_norm": 0.5937765836715698,
+      "learning_rate": 0.0001,
+      "loss": 1.4394,
+      "step": 15226
+    },
+    {
+      "epoch": 1.7491241169375682,
+      "grad_norm": 0.5825061202049255,
+      "learning_rate": 0.0001,
+      "loss": 1.3277,
+      "step": 15227
+    },
+    {
+      "epoch": 1.7492389868473954,
+      "grad_norm": 0.7533363103866577,
+      "learning_rate": 0.0001,
+      "loss": 1.7251,
+      "step": 15228
+    },
+    {
+      "epoch": 1.7493538567572224,
+      "grad_norm": 0.6395822167396545,
+      "learning_rate": 0.0001,
+      "loss": 1.444,
+      "step": 15229
+    },
+    {
+      "epoch": 1.7494687266670494,
+      "grad_norm": 0.5838786959648132,
+      "learning_rate": 0.0001,
+      "loss": 1.5254,
+      "step": 15230
+    },
+    {
+      "epoch": 1.7495835965768767,
+      "grad_norm": 0.6399153470993042,
+      "learning_rate": 0.0001,
+      "loss": 1.2255,
+      "step": 15231
+    },
+    {
+      "epoch": 1.749698466486704,
+      "grad_norm": 0.667523205280304,
+      "learning_rate": 0.0001,
+      "loss": 1.4569,
+      "step": 15232
+    },
+    {
+      "epoch": 1.749813336396531,
+      "grad_norm": 0.6431367993354797,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 15233
+    },
+    {
+      "epoch": 1.749928206306358,
+      "grad_norm": 0.6183831095695496,
+      "learning_rate": 0.0001,
+      "loss": 1.3978,
+      "step": 15234
+    },
+    {
+      "epoch": 1.7500430762161852,
+      "grad_norm": 0.655053436756134,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 15235
+    },
+    {
+      "epoch": 1.7501579461260124,
+      "grad_norm": 0.7972126603126526,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 15236
+    },
+    {
+      "epoch": 1.7502728160358394,
+      "grad_norm": 0.6024256944656372,
+      "learning_rate": 0.0001,
+      "loss": 1.4008,
+      "step": 15237
+    },
+    {
+      "epoch": 1.7503876859456664,
+      "grad_norm": 0.5974478125572205,
+      "learning_rate": 0.0001,
+      "loss": 1.453,
+      "step": 15238
+    },
+    {
+      "epoch": 1.7505025558554936,
+      "grad_norm": 0.7210254669189453,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 15239
+    },
+    {
+      "epoch": 1.7506174257653209,
+      "grad_norm": 0.6723774075508118,
+      "learning_rate": 0.0001,
+      "loss": 1.4211,
+      "step": 15240
+    },
+    {
+      "epoch": 1.7507322956751479,
+      "grad_norm": 0.6472156047821045,
+      "learning_rate": 0.0001,
+      "loss": 1.5018,
+      "step": 15241
+    },
+    {
+      "epoch": 1.7508471655849749,
+      "grad_norm": 0.5895910859107971,
+      "learning_rate": 0.0001,
+      "loss": 1.2851,
+      "step": 15242
+    },
+    {
+      "epoch": 1.7509620354948021,
+      "grad_norm": 0.6794853210449219,
+      "learning_rate": 0.0001,
+      "loss": 1.3592,
+      "step": 15243
+    },
+    {
+      "epoch": 1.7510769054046293,
+      "grad_norm": 0.6932422518730164,
+      "learning_rate": 0.0001,
+      "loss": 1.6878,
+      "step": 15244
+    },
+    {
+      "epoch": 1.7511917753144564,
+      "grad_norm": 0.6949918866157532,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 15245
+    },
+    {
+      "epoch": 1.7513066452242834,
+      "grad_norm": 0.6456189751625061,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 15246
+    },
+    {
+      "epoch": 1.7514215151341106,
+      "grad_norm": 0.6524791121482849,
+      "learning_rate": 0.0001,
+      "loss": 1.5143,
+      "step": 15247
+    },
+    {
+      "epoch": 1.7515363850439378,
+      "grad_norm": 0.5790715217590332,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 15248
+    },
+    {
+      "epoch": 1.7516512549537648,
+      "grad_norm": 0.5909721255302429,
+      "learning_rate": 0.0001,
+      "loss": 1.3346,
+      "step": 15249
+    },
+    {
+      "epoch": 1.7517661248635918,
+      "grad_norm": 0.5983031988143921,
+      "learning_rate": 0.0001,
+      "loss": 1.3218,
+      "step": 15250
+    },
+    {
+      "epoch": 1.751880994773419,
+      "grad_norm": 0.5888960361480713,
+      "learning_rate": 0.0001,
+      "loss": 1.4113,
+      "step": 15251
+    },
+    {
+      "epoch": 1.7519958646832463,
+      "grad_norm": 0.6211053729057312,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 15252
+    },
+    {
+      "epoch": 1.7521107345930733,
+      "grad_norm": 0.6472274661064148,
+      "learning_rate": 0.0001,
+      "loss": 1.6397,
+      "step": 15253
+    },
+    {
+      "epoch": 1.7522256045029003,
+      "grad_norm": 0.5877479314804077,
+      "learning_rate": 0.0001,
+      "loss": 1.3772,
+      "step": 15254
+    },
+    {
+      "epoch": 1.7523404744127276,
+      "grad_norm": 0.6007200479507446,
+      "learning_rate": 0.0001,
+      "loss": 1.3258,
+      "step": 15255
+    },
+    {
+      "epoch": 1.7524553443225548,
+      "grad_norm": 0.6134718656539917,
+      "learning_rate": 0.0001,
+      "loss": 1.3475,
+      "step": 15256
+    },
+    {
+      "epoch": 1.7525702142323818,
+      "grad_norm": 0.6493116021156311,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 15257
+    },
+    {
+      "epoch": 1.7526850841422088,
+      "grad_norm": 0.6316721439361572,
+      "learning_rate": 0.0001,
+      "loss": 1.5339,
+      "step": 15258
+    },
+    {
+      "epoch": 1.752799954052036,
+      "grad_norm": 0.63750159740448,
+      "learning_rate": 0.0001,
+      "loss": 1.4438,
+      "step": 15259
+    },
+    {
+      "epoch": 1.7529148239618633,
+      "grad_norm": 0.6934375762939453,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 15260
+    },
+    {
+      "epoch": 1.7530296938716903,
+      "grad_norm": 0.6297123432159424,
+      "learning_rate": 0.0001,
+      "loss": 1.4865,
+      "step": 15261
+    },
+    {
+      "epoch": 1.7531445637815173,
+      "grad_norm": 0.582813560962677,
+      "learning_rate": 0.0001,
+      "loss": 1.1494,
+      "step": 15262
+    },
+    {
+      "epoch": 1.7532594336913445,
+      "grad_norm": 0.6479847431182861,
+      "learning_rate": 0.0001,
+      "loss": 1.3812,
+      "step": 15263
+    },
+    {
+      "epoch": 1.7533743036011717,
+      "grad_norm": 0.6570015549659729,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 15264
+    },
+    {
+      "epoch": 1.7534891735109988,
+      "grad_norm": 0.6873630285263062,
+      "learning_rate": 0.0001,
+      "loss": 1.2798,
+      "step": 15265
+    },
+    {
+      "epoch": 1.7536040434208258,
+      "grad_norm": 0.6358373165130615,
+      "learning_rate": 0.0001,
+      "loss": 1.4256,
+      "step": 15266
+    },
+    {
+      "epoch": 1.753718913330653,
+      "grad_norm": 0.5748741626739502,
+      "learning_rate": 0.0001,
+      "loss": 1.3127,
+      "step": 15267
+    },
+    {
+      "epoch": 1.7538337832404802,
+      "grad_norm": 0.6858731508255005,
+      "learning_rate": 0.0001,
+      "loss": 1.4542,
+      "step": 15268
+    },
+    {
+      "epoch": 1.7539486531503072,
+      "grad_norm": 0.5939129590988159,
+      "learning_rate": 0.0001,
+      "loss": 1.266,
+      "step": 15269
+    },
+    {
+      "epoch": 1.7540635230601342,
+      "grad_norm": 0.65522301197052,
+      "learning_rate": 0.0001,
+      "loss": 1.3433,
+      "step": 15270
+    },
+    {
+      "epoch": 1.7541783929699615,
+      "grad_norm": 0.6629993915557861,
+      "learning_rate": 0.0001,
+      "loss": 1.4215,
+      "step": 15271
+    },
+    {
+      "epoch": 1.7542932628797887,
+      "grad_norm": 0.6635041236877441,
+      "learning_rate": 0.0001,
+      "loss": 1.5512,
+      "step": 15272
+    },
+    {
+      "epoch": 1.7544081327896157,
+      "grad_norm": 0.6313446164131165,
+      "learning_rate": 0.0001,
+      "loss": 1.6941,
+      "step": 15273
+    },
+    {
+      "epoch": 1.7545230026994427,
+      "grad_norm": 0.6612875461578369,
+      "learning_rate": 0.0001,
+      "loss": 1.3453,
+      "step": 15274
+    },
+    {
+      "epoch": 1.75463787260927,
+      "grad_norm": 0.6233786940574646,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 15275
+    },
+    {
+      "epoch": 1.7547527425190972,
+      "grad_norm": 0.688138484954834,
+      "learning_rate": 0.0001,
+      "loss": 1.4463,
+      "step": 15276
+    },
+    {
+      "epoch": 1.7548676124289242,
+      "grad_norm": 0.5722640156745911,
+      "learning_rate": 0.0001,
+      "loss": 1.173,
+      "step": 15277
+    },
+    {
+      "epoch": 1.7549824823387512,
+      "grad_norm": 0.6150445342063904,
+      "learning_rate": 0.0001,
+      "loss": 1.4503,
+      "step": 15278
+    },
+    {
+      "epoch": 1.7550973522485784,
+      "grad_norm": 0.6374548673629761,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 15279
+    },
+    {
+      "epoch": 1.7552122221584057,
+      "grad_norm": 0.5925512313842773,
+      "learning_rate": 0.0001,
+      "loss": 1.3682,
+      "step": 15280
+    },
+    {
+      "epoch": 1.7553270920682327,
+      "grad_norm": 0.6822425723075867,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 15281
+    },
+    {
+      "epoch": 1.7554419619780597,
+      "grad_norm": 0.6651631593704224,
+      "learning_rate": 0.0001,
+      "loss": 1.7616,
+      "step": 15282
+    },
+    {
+      "epoch": 1.755556831887887,
+      "grad_norm": 0.6492800712585449,
+      "learning_rate": 0.0001,
+      "loss": 1.4807,
+      "step": 15283
+    },
+    {
+      "epoch": 1.7556717017977141,
+      "grad_norm": 0.5847377777099609,
+      "learning_rate": 0.0001,
+      "loss": 1.4534,
+      "step": 15284
+    },
+    {
+      "epoch": 1.7557865717075412,
+      "grad_norm": 0.6187418699264526,
+      "learning_rate": 0.0001,
+      "loss": 1.3685,
+      "step": 15285
+    },
+    {
+      "epoch": 1.7559014416173682,
+      "grad_norm": 0.5958352088928223,
+      "learning_rate": 0.0001,
+      "loss": 1.3842,
+      "step": 15286
+    },
+    {
+      "epoch": 1.7560163115271954,
+      "grad_norm": 0.6693896055221558,
+      "learning_rate": 0.0001,
+      "loss": 1.4296,
+      "step": 15287
+    },
+    {
+      "epoch": 1.7561311814370226,
+      "grad_norm": 0.6221320033073425,
+      "learning_rate": 0.0001,
+      "loss": 1.3786,
+      "step": 15288
+    },
+    {
+      "epoch": 1.7562460513468499,
+      "grad_norm": 0.6341426968574524,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 15289
+    },
+    {
+      "epoch": 1.7563609212566769,
+      "grad_norm": 0.5881104469299316,
+      "learning_rate": 0.0001,
+      "loss": 1.4077,
+      "step": 15290
+    },
+    {
+      "epoch": 1.7564757911665039,
+      "grad_norm": 0.6449071764945984,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 15291
+    },
+    {
+      "epoch": 1.756590661076331,
+      "grad_norm": 0.6435044407844543,
+      "learning_rate": 0.0001,
+      "loss": 1.486,
+      "step": 15292
+    },
+    {
+      "epoch": 1.7567055309861583,
+      "grad_norm": 0.5955443978309631,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 15293
+    },
+    {
+      "epoch": 1.7568204008959853,
+      "grad_norm": 0.6793741583824158,
+      "learning_rate": 0.0001,
+      "loss": 1.5893,
+      "step": 15294
+    },
+    {
+      "epoch": 1.7569352708058124,
+      "grad_norm": 0.6085398197174072,
+      "learning_rate": 0.0001,
+      "loss": 1.4329,
+      "step": 15295
+    },
+    {
+      "epoch": 1.7570501407156396,
+      "grad_norm": 0.6038925051689148,
+      "learning_rate": 0.0001,
+      "loss": 1.3851,
+      "step": 15296
+    },
+    {
+      "epoch": 1.7571650106254668,
+      "grad_norm": 0.6389901041984558,
+      "learning_rate": 0.0001,
+      "loss": 1.4429,
+      "step": 15297
+    },
+    {
+      "epoch": 1.7572798805352938,
+      "grad_norm": 0.5799313187599182,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 15298
+    },
+    {
+      "epoch": 1.7573947504451208,
+      "grad_norm": 0.6270938515663147,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 15299
+    },
+    {
+      "epoch": 1.757509620354948,
+      "grad_norm": 0.5713139176368713,
+      "learning_rate": 0.0001,
+      "loss": 1.4485,
+      "step": 15300
+    },
+    {
+      "epoch": 1.7576244902647753,
+      "grad_norm": 0.5795127153396606,
+      "learning_rate": 0.0001,
+      "loss": 1.2822,
+      "step": 15301
+    },
+    {
+      "epoch": 1.7577393601746023,
+      "grad_norm": 0.5726675391197205,
+      "learning_rate": 0.0001,
+      "loss": 1.0326,
+      "step": 15302
+    },
+    {
+      "epoch": 1.7578542300844293,
+      "grad_norm": 0.6082470417022705,
+      "learning_rate": 0.0001,
+      "loss": 1.4728,
+      "step": 15303
+    },
+    {
+      "epoch": 1.7579690999942565,
+      "grad_norm": 0.6625812649726868,
+      "learning_rate": 0.0001,
+      "loss": 1.4903,
+      "step": 15304
+    },
+    {
+      "epoch": 1.7580839699040838,
+      "grad_norm": 0.6669724583625793,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 15305
+    },
+    {
+      "epoch": 1.7581988398139108,
+      "grad_norm": 0.6124973893165588,
+      "learning_rate": 0.0001,
+      "loss": 1.5021,
+      "step": 15306
+    },
+    {
+      "epoch": 1.7583137097237378,
+      "grad_norm": 0.6731737852096558,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 15307
+    },
+    {
+      "epoch": 1.758428579633565,
+      "grad_norm": 0.597639799118042,
+      "learning_rate": 0.0001,
+      "loss": 1.2609,
+      "step": 15308
+    },
+    {
+      "epoch": 1.7585434495433923,
+      "grad_norm": 0.6299818158149719,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 15309
+    },
+    {
+      "epoch": 1.7586583194532193,
+      "grad_norm": 0.578934371471405,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 15310
+    },
+    {
+      "epoch": 1.7587731893630463,
+      "grad_norm": 0.6151703596115112,
+      "learning_rate": 0.0001,
+      "loss": 1.3717,
+      "step": 15311
+    },
+    {
+      "epoch": 1.7588880592728735,
+      "grad_norm": 0.6442793011665344,
+      "learning_rate": 0.0001,
+      "loss": 1.4778,
+      "step": 15312
+    },
+    {
+      "epoch": 1.7590029291827007,
+      "grad_norm": 0.6389909982681274,
+      "learning_rate": 0.0001,
+      "loss": 1.4665,
+      "step": 15313
+    },
+    {
+      "epoch": 1.7591177990925277,
+      "grad_norm": 0.5759516954421997,
+      "learning_rate": 0.0001,
+      "loss": 1.354,
+      "step": 15314
+    },
+    {
+      "epoch": 1.7592326690023548,
+      "grad_norm": 0.6577784419059753,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 15315
+    },
+    {
+      "epoch": 1.759347538912182,
+      "grad_norm": 0.6161843538284302,
+      "learning_rate": 0.0001,
+      "loss": 1.3574,
+      "step": 15316
+    },
+    {
+      "epoch": 1.7594624088220092,
+      "grad_norm": 0.6002083420753479,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 15317
+    },
+    {
+      "epoch": 1.7595772787318362,
+      "grad_norm": 0.6231681704521179,
+      "learning_rate": 0.0001,
+      "loss": 1.2959,
+      "step": 15318
+    },
+    {
+      "epoch": 1.7596921486416632,
+      "grad_norm": 0.6722519397735596,
+      "learning_rate": 0.0001,
+      "loss": 1.3182,
+      "step": 15319
+    },
+    {
+      "epoch": 1.7598070185514905,
+      "grad_norm": 0.6488606929779053,
+      "learning_rate": 0.0001,
+      "loss": 1.249,
+      "step": 15320
+    },
+    {
+      "epoch": 1.7599218884613177,
+      "grad_norm": 0.6526792049407959,
+      "learning_rate": 0.0001,
+      "loss": 1.389,
+      "step": 15321
+    },
+    {
+      "epoch": 1.7600367583711447,
+      "grad_norm": 0.8293418884277344,
+      "learning_rate": 0.0001,
+      "loss": 1.5196,
+      "step": 15322
+    },
+    {
+      "epoch": 1.7601516282809717,
+      "grad_norm": 0.6284173130989075,
+      "learning_rate": 0.0001,
+      "loss": 1.4126,
+      "step": 15323
+    },
+    {
+      "epoch": 1.760266498190799,
+      "grad_norm": 0.6085017919540405,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 15324
+    },
+    {
+      "epoch": 1.7603813681006262,
+      "grad_norm": 0.6490882039070129,
+      "learning_rate": 0.0001,
+      "loss": 1.5043,
+      "step": 15325
+    },
+    {
+      "epoch": 1.7604962380104532,
+      "grad_norm": 0.6944969296455383,
+      "learning_rate": 0.0001,
+      "loss": 1.4436,
+      "step": 15326
+    },
+    {
+      "epoch": 1.7606111079202802,
+      "grad_norm": 0.5754250288009644,
+      "learning_rate": 0.0001,
+      "loss": 1.4301,
+      "step": 15327
+    },
+    {
+      "epoch": 1.7607259778301074,
+      "grad_norm": 0.5691632628440857,
+      "learning_rate": 0.0001,
+      "loss": 1.2594,
+      "step": 15328
+    },
+    {
+      "epoch": 1.7608408477399347,
+      "grad_norm": 0.6530894041061401,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 15329
+    },
+    {
+      "epoch": 1.7609557176497617,
+      "grad_norm": 0.5945919156074524,
+      "learning_rate": 0.0001,
+      "loss": 1.5005,
+      "step": 15330
+    },
+    {
+      "epoch": 1.7610705875595887,
+      "grad_norm": 0.6224496364593506,
+      "learning_rate": 0.0001,
+      "loss": 1.371,
+      "step": 15331
+    },
+    {
+      "epoch": 1.761185457469416,
+      "grad_norm": 0.5837387442588806,
+      "learning_rate": 0.0001,
+      "loss": 1.2299,
+      "step": 15332
+    },
+    {
+      "epoch": 1.7613003273792431,
+      "grad_norm": 0.6480647325515747,
+      "learning_rate": 0.0001,
+      "loss": 1.4484,
+      "step": 15333
+    },
+    {
+      "epoch": 1.7614151972890701,
+      "grad_norm": 0.655102014541626,
+      "learning_rate": 0.0001,
+      "loss": 1.6464,
+      "step": 15334
+    },
+    {
+      "epoch": 1.7615300671988972,
+      "grad_norm": 0.624885618686676,
+      "learning_rate": 0.0001,
+      "loss": 1.504,
+      "step": 15335
+    },
+    {
+      "epoch": 1.7616449371087244,
+      "grad_norm": 0.7076594233512878,
+      "learning_rate": 0.0001,
+      "loss": 1.2598,
+      "step": 15336
+    },
+    {
+      "epoch": 1.7617598070185516,
+      "grad_norm": 0.6876631379127502,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 15337
+    },
+    {
+      "epoch": 1.7618746769283786,
+      "grad_norm": 0.642560601234436,
+      "learning_rate": 0.0001,
+      "loss": 1.474,
+      "step": 15338
+    },
+    {
+      "epoch": 1.7619895468382056,
+      "grad_norm": 0.6945380568504333,
+      "learning_rate": 0.0001,
+      "loss": 1.4465,
+      "step": 15339
+    },
+    {
+      "epoch": 1.7621044167480329,
+      "grad_norm": 0.6342966556549072,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 15340
+    },
+    {
+      "epoch": 1.76221928665786,
+      "grad_norm": 0.6334778666496277,
+      "learning_rate": 0.0001,
+      "loss": 1.4535,
+      "step": 15341
+    },
+    {
+      "epoch": 1.762334156567687,
+      "grad_norm": 0.5924774408340454,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 15342
+    },
+    {
+      "epoch": 1.7624490264775141,
+      "grad_norm": 0.6480969190597534,
+      "learning_rate": 0.0001,
+      "loss": 1.3745,
+      "step": 15343
+    },
+    {
+      "epoch": 1.7625638963873413,
+      "grad_norm": 0.635136067867279,
+      "learning_rate": 0.0001,
+      "loss": 1.4212,
+      "step": 15344
+    },
+    {
+      "epoch": 1.7626787662971686,
+      "grad_norm": 0.6227051615715027,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 15345
+    },
+    {
+      "epoch": 1.7627936362069956,
+      "grad_norm": 0.5813191533088684,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 15346
+    },
+    {
+      "epoch": 1.7629085061168226,
+      "grad_norm": 0.7114011645317078,
+      "learning_rate": 0.0001,
+      "loss": 1.4692,
+      "step": 15347
+    },
+    {
+      "epoch": 1.7630233760266498,
+      "grad_norm": 0.6176375150680542,
+      "learning_rate": 0.0001,
+      "loss": 1.4317,
+      "step": 15348
+    },
+    {
+      "epoch": 1.763138245936477,
+      "grad_norm": 0.7010447382926941,
+      "learning_rate": 0.0001,
+      "loss": 1.6766,
+      "step": 15349
+    },
+    {
+      "epoch": 1.763253115846304,
+      "grad_norm": 0.62742680311203,
+      "learning_rate": 0.0001,
+      "loss": 1.3897,
+      "step": 15350
+    },
+    {
+      "epoch": 1.763367985756131,
+      "grad_norm": 0.6509758830070496,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 15351
+    },
+    {
+      "epoch": 1.7634828556659583,
+      "grad_norm": 0.6291651725769043,
+      "learning_rate": 0.0001,
+      "loss": 1.1953,
+      "step": 15352
+    },
+    {
+      "epoch": 1.7635977255757855,
+      "grad_norm": 0.6242035031318665,
+      "learning_rate": 0.0001,
+      "loss": 1.4124,
+      "step": 15353
+    },
+    {
+      "epoch": 1.7637125954856125,
+      "grad_norm": 0.5866750478744507,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 15354
+    },
+    {
+      "epoch": 1.7638274653954396,
+      "grad_norm": 0.6242676973342896,
+      "learning_rate": 0.0001,
+      "loss": 1.4799,
+      "step": 15355
+    },
+    {
+      "epoch": 1.7639423353052668,
+      "grad_norm": 0.6475937366485596,
+      "learning_rate": 0.0001,
+      "loss": 1.3933,
+      "step": 15356
+    },
+    {
+      "epoch": 1.764057205215094,
+      "grad_norm": 0.6821995377540588,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 15357
+    },
+    {
+      "epoch": 1.764172075124921,
+      "grad_norm": 0.6403782963752747,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 15358
+    },
+    {
+      "epoch": 1.764286945034748,
+      "grad_norm": 0.5568740963935852,
+      "learning_rate": 0.0001,
+      "loss": 1.3501,
+      "step": 15359
+    },
+    {
+      "epoch": 1.7644018149445753,
+      "grad_norm": 0.7482900619506836,
+      "learning_rate": 0.0001,
+      "loss": 1.7237,
+      "step": 15360
+    },
+    {
+      "epoch": 1.7645166848544025,
+      "grad_norm": 0.7028294801712036,
+      "learning_rate": 0.0001,
+      "loss": 1.4546,
+      "step": 15361
+    },
+    {
+      "epoch": 1.7646315547642295,
+      "grad_norm": 0.6432153582572937,
+      "learning_rate": 0.0001,
+      "loss": 1.4991,
+      "step": 15362
+    },
+    {
+      "epoch": 1.7647464246740565,
+      "grad_norm": 0.5955196619033813,
+      "learning_rate": 0.0001,
+      "loss": 1.4428,
+      "step": 15363
+    },
+    {
+      "epoch": 1.7648612945838837,
+      "grad_norm": 0.6378939747810364,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 15364
+    },
+    {
+      "epoch": 1.764976164493711,
+      "grad_norm": 0.6353983283042908,
+      "learning_rate": 0.0001,
+      "loss": 1.4364,
+      "step": 15365
+    },
+    {
+      "epoch": 1.765091034403538,
+      "grad_norm": 0.6289456486701965,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 15366
+    },
+    {
+      "epoch": 1.765205904313365,
+      "grad_norm": 0.6229878664016724,
+      "learning_rate": 0.0001,
+      "loss": 1.2323,
+      "step": 15367
+    },
+    {
+      "epoch": 1.7653207742231922,
+      "grad_norm": 0.6140499114990234,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 15368
+    },
+    {
+      "epoch": 1.7654356441330195,
+      "grad_norm": 0.6184989213943481,
+      "learning_rate": 0.0001,
+      "loss": 1.3525,
+      "step": 15369
+    },
+    {
+      "epoch": 1.7655505140428465,
+      "grad_norm": 0.6055330634117126,
+      "learning_rate": 0.0001,
+      "loss": 1.2379,
+      "step": 15370
+    },
+    {
+      "epoch": 1.7656653839526735,
+      "grad_norm": 0.6226028203964233,
+      "learning_rate": 0.0001,
+      "loss": 1.511,
+      "step": 15371
+    },
+    {
+      "epoch": 1.7657802538625007,
+      "grad_norm": 0.6356282830238342,
+      "learning_rate": 0.0001,
+      "loss": 1.281,
+      "step": 15372
+    },
+    {
+      "epoch": 1.765895123772328,
+      "grad_norm": 0.5963003635406494,
+      "learning_rate": 0.0001,
+      "loss": 1.3596,
+      "step": 15373
+    },
+    {
+      "epoch": 1.766009993682155,
+      "grad_norm": 0.598003625869751,
+      "learning_rate": 0.0001,
+      "loss": 1.4494,
+      "step": 15374
+    },
+    {
+      "epoch": 1.766124863591982,
+      "grad_norm": 0.6972695589065552,
+      "learning_rate": 0.0001,
+      "loss": 1.2956,
+      "step": 15375
+    },
+    {
+      "epoch": 1.7662397335018092,
+      "grad_norm": 0.6252386569976807,
+      "learning_rate": 0.0001,
+      "loss": 1.4696,
+      "step": 15376
+    },
+    {
+      "epoch": 1.7663546034116364,
+      "grad_norm": 0.7368474006652832,
+      "learning_rate": 0.0001,
+      "loss": 1.6705,
+      "step": 15377
+    },
+    {
+      "epoch": 1.7664694733214634,
+      "grad_norm": 0.7342138886451721,
+      "learning_rate": 0.0001,
+      "loss": 1.4238,
+      "step": 15378
+    },
+    {
+      "epoch": 1.7665843432312904,
+      "grad_norm": 0.5836483836174011,
+      "learning_rate": 0.0001,
+      "loss": 1.3567,
+      "step": 15379
+    },
+    {
+      "epoch": 1.7666992131411177,
+      "grad_norm": 0.6157330274581909,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 15380
+    },
+    {
+      "epoch": 1.766814083050945,
+      "grad_norm": 0.630933940410614,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 15381
+    },
+    {
+      "epoch": 1.766928952960772,
+      "grad_norm": 0.6310692429542542,
+      "learning_rate": 0.0001,
+      "loss": 1.4669,
+      "step": 15382
+    },
+    {
+      "epoch": 1.767043822870599,
+      "grad_norm": 0.6385482549667358,
+      "learning_rate": 0.0001,
+      "loss": 1.2872,
+      "step": 15383
+    },
+    {
+      "epoch": 1.7671586927804261,
+      "grad_norm": 0.6223406195640564,
+      "learning_rate": 0.0001,
+      "loss": 1.2195,
+      "step": 15384
+    },
+    {
+      "epoch": 1.7672735626902534,
+      "grad_norm": 0.5795210003852844,
+      "learning_rate": 0.0001,
+      "loss": 1.3071,
+      "step": 15385
+    },
+    {
+      "epoch": 1.7673884326000804,
+      "grad_norm": 0.6244056224822998,
+      "learning_rate": 0.0001,
+      "loss": 1.3275,
+      "step": 15386
+    },
+    {
+      "epoch": 1.7675033025099074,
+      "grad_norm": 0.7372899055480957,
+      "learning_rate": 0.0001,
+      "loss": 1.6583,
+      "step": 15387
+    },
+    {
+      "epoch": 1.7676181724197346,
+      "grad_norm": 0.6304470300674438,
+      "learning_rate": 0.0001,
+      "loss": 1.3243,
+      "step": 15388
+    },
+    {
+      "epoch": 1.7677330423295619,
+      "grad_norm": 0.6503053307533264,
+      "learning_rate": 0.0001,
+      "loss": 1.3891,
+      "step": 15389
+    },
+    {
+      "epoch": 1.7678479122393889,
+      "grad_norm": 0.6643422245979309,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 15390
+    },
+    {
+      "epoch": 1.7679627821492159,
+      "grad_norm": 0.6259305477142334,
+      "learning_rate": 0.0001,
+      "loss": 1.3976,
+      "step": 15391
+    },
+    {
+      "epoch": 1.768077652059043,
+      "grad_norm": 0.6734678745269775,
+      "learning_rate": 0.0001,
+      "loss": 1.4114,
+      "step": 15392
+    },
+    {
+      "epoch": 1.7681925219688703,
+      "grad_norm": 0.6242077946662903,
+      "learning_rate": 0.0001,
+      "loss": 1.4955,
+      "step": 15393
+    },
+    {
+      "epoch": 1.7683073918786973,
+      "grad_norm": 0.6714116930961609,
+      "learning_rate": 0.0001,
+      "loss": 1.3507,
+      "step": 15394
+    },
+    {
+      "epoch": 1.7684222617885244,
+      "grad_norm": 0.5949355363845825,
+      "learning_rate": 0.0001,
+      "loss": 1.4515,
+      "step": 15395
+    },
+    {
+      "epoch": 1.7685371316983516,
+      "grad_norm": 0.5709308981895447,
+      "learning_rate": 0.0001,
+      "loss": 1.3621,
+      "step": 15396
+    },
+    {
+      "epoch": 1.7686520016081788,
+      "grad_norm": 0.6260430812835693,
+      "learning_rate": 0.0001,
+      "loss": 1.4699,
+      "step": 15397
+    },
+    {
+      "epoch": 1.7687668715180058,
+      "grad_norm": 0.6341519355773926,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 15398
+    },
+    {
+      "epoch": 1.7688817414278328,
+      "grad_norm": 0.6372659802436829,
+      "learning_rate": 0.0001,
+      "loss": 1.4443,
+      "step": 15399
+    },
+    {
+      "epoch": 1.76899661133766,
+      "grad_norm": 0.5935454964637756,
+      "learning_rate": 0.0001,
+      "loss": 1.2865,
+      "step": 15400
+    },
+    {
+      "epoch": 1.7691114812474873,
+      "grad_norm": 0.6392391324043274,
+      "learning_rate": 0.0001,
+      "loss": 1.5012,
+      "step": 15401
+    },
+    {
+      "epoch": 1.7692263511573143,
+      "grad_norm": 0.6767097115516663,
+      "learning_rate": 0.0001,
+      "loss": 1.5747,
+      "step": 15402
+    },
+    {
+      "epoch": 1.7693412210671413,
+      "grad_norm": 0.575113832950592,
+      "learning_rate": 0.0001,
+      "loss": 1.4338,
+      "step": 15403
+    },
+    {
+      "epoch": 1.7694560909769685,
+      "grad_norm": 0.6148155927658081,
+      "learning_rate": 0.0001,
+      "loss": 1.4446,
+      "step": 15404
+    },
+    {
+      "epoch": 1.7695709608867958,
+      "grad_norm": 0.6530346274375916,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 15405
+    },
+    {
+      "epoch": 1.7696858307966228,
+      "grad_norm": 0.624439001083374,
+      "learning_rate": 0.0001,
+      "loss": 1.4565,
+      "step": 15406
+    },
+    {
+      "epoch": 1.7698007007064498,
+      "grad_norm": 0.633822500705719,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 15407
+    },
+    {
+      "epoch": 1.769915570616277,
+      "grad_norm": 0.6398473381996155,
+      "learning_rate": 0.0001,
+      "loss": 1.3496,
+      "step": 15408
+    },
+    {
+      "epoch": 1.7700304405261043,
+      "grad_norm": 0.6082612872123718,
+      "learning_rate": 0.0001,
+      "loss": 1.2266,
+      "step": 15409
+    },
+    {
+      "epoch": 1.7701453104359313,
+      "grad_norm": 0.6723073124885559,
+      "learning_rate": 0.0001,
+      "loss": 1.3755,
+      "step": 15410
+    },
+    {
+      "epoch": 1.7702601803457583,
+      "grad_norm": 0.6552845239639282,
+      "learning_rate": 0.0001,
+      "loss": 1.5456,
+      "step": 15411
+    },
+    {
+      "epoch": 1.7703750502555855,
+      "grad_norm": 0.682786226272583,
+      "learning_rate": 0.0001,
+      "loss": 1.4209,
+      "step": 15412
+    },
+    {
+      "epoch": 1.7704899201654127,
+      "grad_norm": 0.6715489625930786,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 15413
+    },
+    {
+      "epoch": 1.7706047900752397,
+      "grad_norm": 0.6266937851905823,
+      "learning_rate": 0.0001,
+      "loss": 1.4389,
+      "step": 15414
+    },
+    {
+      "epoch": 1.7707196599850668,
+      "grad_norm": 0.6482228636741638,
+      "learning_rate": 0.0001,
+      "loss": 1.4746,
+      "step": 15415
+    },
+    {
+      "epoch": 1.770834529894894,
+      "grad_norm": 0.6691463589668274,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 15416
+    },
+    {
+      "epoch": 1.7709493998047212,
+      "grad_norm": 0.6793537735939026,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 15417
+    },
+    {
+      "epoch": 1.7710642697145482,
+      "grad_norm": 0.604053258895874,
+      "learning_rate": 0.0001,
+      "loss": 1.4415,
+      "step": 15418
+    },
+    {
+      "epoch": 1.7711791396243752,
+      "grad_norm": 0.7533746361732483,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 15419
+    },
+    {
+      "epoch": 1.7712940095342025,
+      "grad_norm": 0.6014366745948792,
+      "learning_rate": 0.0001,
+      "loss": 1.37,
+      "step": 15420
+    },
+    {
+      "epoch": 1.7714088794440297,
+      "grad_norm": 0.6073466539382935,
+      "learning_rate": 0.0001,
+      "loss": 1.4946,
+      "step": 15421
+    },
+    {
+      "epoch": 1.7715237493538567,
+      "grad_norm": 0.6218917965888977,
+      "learning_rate": 0.0001,
+      "loss": 1.5332,
+      "step": 15422
+    },
+    {
+      "epoch": 1.7716386192636837,
+      "grad_norm": 0.6328446865081787,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 15423
+    },
+    {
+      "epoch": 1.771753489173511,
+      "grad_norm": 0.5783063173294067,
+      "learning_rate": 0.0001,
+      "loss": 1.2782,
+      "step": 15424
+    },
+    {
+      "epoch": 1.7718683590833382,
+      "grad_norm": 0.5905358791351318,
+      "learning_rate": 0.0001,
+      "loss": 1.528,
+      "step": 15425
+    },
+    {
+      "epoch": 1.7719832289931654,
+      "grad_norm": 0.6171038150787354,
+      "learning_rate": 0.0001,
+      "loss": 1.4173,
+      "step": 15426
+    },
+    {
+      "epoch": 1.7720980989029924,
+      "grad_norm": 0.5971978902816772,
+      "learning_rate": 0.0001,
+      "loss": 1.401,
+      "step": 15427
+    },
+    {
+      "epoch": 1.7722129688128194,
+      "grad_norm": 0.6217857003211975,
+      "learning_rate": 0.0001,
+      "loss": 1.4225,
+      "step": 15428
+    },
+    {
+      "epoch": 1.7723278387226467,
+      "grad_norm": 0.6341108679771423,
+      "learning_rate": 0.0001,
+      "loss": 1.3223,
+      "step": 15429
+    },
+    {
+      "epoch": 1.7724427086324739,
+      "grad_norm": 0.5544351935386658,
+      "learning_rate": 0.0001,
+      "loss": 1.4596,
+      "step": 15430
+    },
+    {
+      "epoch": 1.772557578542301,
+      "grad_norm": 0.6070437431335449,
+      "learning_rate": 0.0001,
+      "loss": 1.4921,
+      "step": 15431
+    },
+    {
+      "epoch": 1.772672448452128,
+      "grad_norm": 0.679899275302887,
+      "learning_rate": 0.0001,
+      "loss": 1.5184,
+      "step": 15432
+    },
+    {
+      "epoch": 1.7727873183619551,
+      "grad_norm": 0.6571292877197266,
+      "learning_rate": 0.0001,
+      "loss": 1.3402,
+      "step": 15433
+    },
+    {
+      "epoch": 1.7729021882717824,
+      "grad_norm": 0.6752861738204956,
+      "learning_rate": 0.0001,
+      "loss": 1.4057,
+      "step": 15434
+    },
+    {
+      "epoch": 1.7730170581816094,
+      "grad_norm": 0.6188561916351318,
+      "learning_rate": 0.0001,
+      "loss": 1.4021,
+      "step": 15435
+    },
+    {
+      "epoch": 1.7731319280914364,
+      "grad_norm": 0.645605206489563,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 15436
+    },
+    {
+      "epoch": 1.7732467980012636,
+      "grad_norm": 0.6331257820129395,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 15437
+    },
+    {
+      "epoch": 1.7733616679110908,
+      "grad_norm": 0.6232604384422302,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 15438
+    },
+    {
+      "epoch": 1.7734765378209179,
+      "grad_norm": 0.6299999952316284,
+      "learning_rate": 0.0001,
+      "loss": 1.4565,
+      "step": 15439
+    },
+    {
+      "epoch": 1.7735914077307449,
+      "grad_norm": 0.6201632618904114,
+      "learning_rate": 0.0001,
+      "loss": 1.4274,
+      "step": 15440
+    },
+    {
+      "epoch": 1.773706277640572,
+      "grad_norm": 0.6291703581809998,
+      "learning_rate": 0.0001,
+      "loss": 1.5339,
+      "step": 15441
+    },
+    {
+      "epoch": 1.7738211475503993,
+      "grad_norm": 0.6148052215576172,
+      "learning_rate": 0.0001,
+      "loss": 1.4622,
+      "step": 15442
+    },
+    {
+      "epoch": 1.7739360174602263,
+      "grad_norm": 0.6915506720542908,
+      "learning_rate": 0.0001,
+      "loss": 1.4758,
+      "step": 15443
+    },
+    {
+      "epoch": 1.7740508873700533,
+      "grad_norm": 0.6690899133682251,
+      "learning_rate": 0.0001,
+      "loss": 1.3532,
+      "step": 15444
+    },
+    {
+      "epoch": 1.7741657572798806,
+      "grad_norm": 0.6253287196159363,
+      "learning_rate": 0.0001,
+      "loss": 1.5029,
+      "step": 15445
+    },
+    {
+      "epoch": 1.7742806271897078,
+      "grad_norm": 0.66769939661026,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 15446
+    },
+    {
+      "epoch": 1.7743954970995348,
+      "grad_norm": 0.6366974115371704,
+      "learning_rate": 0.0001,
+      "loss": 1.4671,
+      "step": 15447
+    },
+    {
+      "epoch": 1.7745103670093618,
+      "grad_norm": 0.6763840913772583,
+      "learning_rate": 0.0001,
+      "loss": 1.575,
+      "step": 15448
+    },
+    {
+      "epoch": 1.774625236919189,
+      "grad_norm": 0.6336223483085632,
+      "learning_rate": 0.0001,
+      "loss": 1.5122,
+      "step": 15449
+    },
+    {
+      "epoch": 1.7747401068290163,
+      "grad_norm": 0.6298586130142212,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 15450
+    },
+    {
+      "epoch": 1.7748549767388433,
+      "grad_norm": 0.6808599829673767,
+      "learning_rate": 0.0001,
+      "loss": 1.4164,
+      "step": 15451
+    },
+    {
+      "epoch": 1.7749698466486703,
+      "grad_norm": 0.638803243637085,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 15452
+    },
+    {
+      "epoch": 1.7750847165584975,
+      "grad_norm": 0.6190632581710815,
+      "learning_rate": 0.0001,
+      "loss": 1.3554,
+      "step": 15453
+    },
+    {
+      "epoch": 1.7751995864683248,
+      "grad_norm": 0.6603758335113525,
+      "learning_rate": 0.0001,
+      "loss": 1.5177,
+      "step": 15454
+    },
+    {
+      "epoch": 1.7753144563781518,
+      "grad_norm": 0.6279389262199402,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 15455
+    },
+    {
+      "epoch": 1.7754293262879788,
+      "grad_norm": 0.6289408206939697,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 15456
+    },
+    {
+      "epoch": 1.775544196197806,
+      "grad_norm": 0.5870251059532166,
+      "learning_rate": 0.0001,
+      "loss": 1.3102,
+      "step": 15457
+    },
+    {
+      "epoch": 1.7756590661076332,
+      "grad_norm": 0.6192272305488586,
+      "learning_rate": 0.0001,
+      "loss": 1.3527,
+      "step": 15458
+    },
+    {
+      "epoch": 1.7757739360174603,
+      "grad_norm": 0.626162588596344,
+      "learning_rate": 0.0001,
+      "loss": 1.4285,
+      "step": 15459
+    },
+    {
+      "epoch": 1.7758888059272873,
+      "grad_norm": 0.6371445059776306,
+      "learning_rate": 0.0001,
+      "loss": 1.4435,
+      "step": 15460
+    },
+    {
+      "epoch": 1.7760036758371145,
+      "grad_norm": 0.7196854948997498,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 15461
+    },
+    {
+      "epoch": 1.7761185457469417,
+      "grad_norm": 0.6295327544212341,
+      "learning_rate": 0.0001,
+      "loss": 1.4168,
+      "step": 15462
+    },
+    {
+      "epoch": 1.7762334156567687,
+      "grad_norm": 0.6012369394302368,
+      "learning_rate": 0.0001,
+      "loss": 1.1423,
+      "step": 15463
+    },
+    {
+      "epoch": 1.7763482855665957,
+      "grad_norm": 0.6109943985939026,
+      "learning_rate": 0.0001,
+      "loss": 1.7749,
+      "step": 15464
+    },
+    {
+      "epoch": 1.776463155476423,
+      "grad_norm": 0.6386968493461609,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 15465
+    },
+    {
+      "epoch": 1.7765780253862502,
+      "grad_norm": 0.6312034130096436,
+      "learning_rate": 0.0001,
+      "loss": 1.3813,
+      "step": 15466
+    },
+    {
+      "epoch": 1.7766928952960772,
+      "grad_norm": 0.6818085312843323,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 15467
+    },
+    {
+      "epoch": 1.7768077652059042,
+      "grad_norm": 0.616159975528717,
+      "learning_rate": 0.0001,
+      "loss": 1.3302,
+      "step": 15468
+    },
+    {
+      "epoch": 1.7769226351157315,
+      "grad_norm": 0.6063972115516663,
+      "learning_rate": 0.0001,
+      "loss": 1.6571,
+      "step": 15469
+    },
+    {
+      "epoch": 1.7770375050255587,
+      "grad_norm": 0.6804771423339844,
+      "learning_rate": 0.0001,
+      "loss": 1.6351,
+      "step": 15470
+    },
+    {
+      "epoch": 1.7771523749353857,
+      "grad_norm": 0.6419492959976196,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 15471
+    },
+    {
+      "epoch": 1.7772672448452127,
+      "grad_norm": 0.6010777354240417,
+      "learning_rate": 0.0001,
+      "loss": 1.419,
+      "step": 15472
+    },
+    {
+      "epoch": 1.77738211475504,
+      "grad_norm": 0.6770652532577515,
+      "learning_rate": 0.0001,
+      "loss": 1.523,
+      "step": 15473
+    },
+    {
+      "epoch": 1.7774969846648672,
+      "grad_norm": 0.7224645614624023,
+      "learning_rate": 0.0001,
+      "loss": 1.4757,
+      "step": 15474
+    },
+    {
+      "epoch": 1.7776118545746942,
+      "grad_norm": 0.6119104623794556,
+      "learning_rate": 0.0001,
+      "loss": 1.3286,
+      "step": 15475
+    },
+    {
+      "epoch": 1.7777267244845212,
+      "grad_norm": 0.6447218060493469,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 15476
+    },
+    {
+      "epoch": 1.7778415943943484,
+      "grad_norm": 0.6411008238792419,
+      "learning_rate": 0.0001,
+      "loss": 1.5644,
+      "step": 15477
+    },
+    {
+      "epoch": 1.7779564643041756,
+      "grad_norm": 0.6330268383026123,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 15478
+    },
+    {
+      "epoch": 1.7780713342140027,
+      "grad_norm": 0.5914243459701538,
+      "learning_rate": 0.0001,
+      "loss": 1.5313,
+      "step": 15479
+    },
+    {
+      "epoch": 1.7781862041238297,
+      "grad_norm": 0.6504284739494324,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 15480
+    },
+    {
+      "epoch": 1.778301074033657,
+      "grad_norm": 0.6057482361793518,
+      "learning_rate": 0.0001,
+      "loss": 1.3601,
+      "step": 15481
+    },
+    {
+      "epoch": 1.7784159439434841,
+      "grad_norm": 0.6217951774597168,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 15482
+    },
+    {
+      "epoch": 1.7785308138533111,
+      "grad_norm": 0.6592935919761658,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 15483
+    },
+    {
+      "epoch": 1.7786456837631381,
+      "grad_norm": 0.6222397089004517,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 15484
+    },
+    {
+      "epoch": 1.7787605536729654,
+      "grad_norm": 0.6175488233566284,
+      "learning_rate": 0.0001,
+      "loss": 1.4577,
+      "step": 15485
+    },
+    {
+      "epoch": 1.7788754235827926,
+      "grad_norm": 0.6524683833122253,
+      "learning_rate": 0.0001,
+      "loss": 1.45,
+      "step": 15486
+    },
+    {
+      "epoch": 1.7789902934926196,
+      "grad_norm": 0.6026191711425781,
+      "learning_rate": 0.0001,
+      "loss": 1.3723,
+      "step": 15487
+    },
+    {
+      "epoch": 1.7791051634024466,
+      "grad_norm": 0.6724210977554321,
+      "learning_rate": 0.0001,
+      "loss": 1.3315,
+      "step": 15488
+    },
+    {
+      "epoch": 1.7792200333122739,
+      "grad_norm": 0.738828718662262,
+      "learning_rate": 0.0001,
+      "loss": 1.6497,
+      "step": 15489
+    },
+    {
+      "epoch": 1.779334903222101,
+      "grad_norm": 0.7495474219322205,
+      "learning_rate": 0.0001,
+      "loss": 1.5393,
+      "step": 15490
+    },
+    {
+      "epoch": 1.779449773131928,
+      "grad_norm": 0.5927098989486694,
+      "learning_rate": 0.0001,
+      "loss": 1.4034,
+      "step": 15491
+    },
+    {
+      "epoch": 1.779564643041755,
+      "grad_norm": 0.6002185940742493,
+      "learning_rate": 0.0001,
+      "loss": 1.4098,
+      "step": 15492
+    },
+    {
+      "epoch": 1.7796795129515823,
+      "grad_norm": 0.6237373948097229,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 15493
+    },
+    {
+      "epoch": 1.7797943828614096,
+      "grad_norm": 0.606011152267456,
+      "learning_rate": 0.0001,
+      "loss": 1.4736,
+      "step": 15494
+    },
+    {
+      "epoch": 1.7799092527712366,
+      "grad_norm": 0.6501420736312866,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 15495
+    },
+    {
+      "epoch": 1.7800241226810636,
+      "grad_norm": 0.6354816555976868,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 15496
+    },
+    {
+      "epoch": 1.7801389925908908,
+      "grad_norm": 0.5931168794631958,
+      "learning_rate": 0.0001,
+      "loss": 1.4077,
+      "step": 15497
+    },
+    {
+      "epoch": 1.780253862500718,
+      "grad_norm": 0.626844048500061,
+      "learning_rate": 0.0001,
+      "loss": 1.5079,
+      "step": 15498
+    },
+    {
+      "epoch": 1.780368732410545,
+      "grad_norm": 0.6105011105537415,
+      "learning_rate": 0.0001,
+      "loss": 1.2362,
+      "step": 15499
+    },
+    {
+      "epoch": 1.780483602320372,
+      "grad_norm": 0.6420881152153015,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 15500
+    },
+    {
+      "epoch": 1.7805984722301993,
+      "grad_norm": 0.6629682779312134,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 15501
+    },
+    {
+      "epoch": 1.7807133421400265,
+      "grad_norm": 0.5871655941009521,
+      "learning_rate": 0.0001,
+      "loss": 1.3112,
+      "step": 15502
+    },
+    {
+      "epoch": 1.7808282120498535,
+      "grad_norm": 0.6966809034347534,
+      "learning_rate": 0.0001,
+      "loss": 1.5694,
+      "step": 15503
+    },
+    {
+      "epoch": 1.7809430819596805,
+      "grad_norm": 0.6312387585639954,
+      "learning_rate": 0.0001,
+      "loss": 1.5059,
+      "step": 15504
+    },
+    {
+      "epoch": 1.7810579518695078,
+      "grad_norm": 0.6314889192581177,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 15505
+    },
+    {
+      "epoch": 1.781172821779335,
+      "grad_norm": 0.657773494720459,
+      "learning_rate": 0.0001,
+      "loss": 1.2146,
+      "step": 15506
+    },
+    {
+      "epoch": 1.781287691689162,
+      "grad_norm": 0.6220430731773376,
+      "learning_rate": 0.0001,
+      "loss": 1.3241,
+      "step": 15507
+    },
+    {
+      "epoch": 1.781402561598989,
+      "grad_norm": 0.7031171321868896,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 15508
+    },
+    {
+      "epoch": 1.7815174315088163,
+      "grad_norm": 0.6239733099937439,
+      "learning_rate": 0.0001,
+      "loss": 1.481,
+      "step": 15509
+    },
+    {
+      "epoch": 1.7816323014186435,
+      "grad_norm": 0.5512365698814392,
+      "learning_rate": 0.0001,
+      "loss": 1.1257,
+      "step": 15510
+    },
+    {
+      "epoch": 1.7817471713284705,
+      "grad_norm": 0.6748344302177429,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 15511
+    },
+    {
+      "epoch": 1.7818620412382975,
+      "grad_norm": 0.6460611820220947,
+      "learning_rate": 0.0001,
+      "loss": 1.3985,
+      "step": 15512
+    },
+    {
+      "epoch": 1.7819769111481247,
+      "grad_norm": 0.6497007608413696,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 15513
+    },
+    {
+      "epoch": 1.782091781057952,
+      "grad_norm": 0.6263741850852966,
+      "learning_rate": 0.0001,
+      "loss": 1.4769,
+      "step": 15514
+    },
+    {
+      "epoch": 1.782206650967779,
+      "grad_norm": 0.6193414330482483,
+      "learning_rate": 0.0001,
+      "loss": 1.4357,
+      "step": 15515
+    },
+    {
+      "epoch": 1.782321520877606,
+      "grad_norm": 0.6849316358566284,
+      "learning_rate": 0.0001,
+      "loss": 1.3156,
+      "step": 15516
+    },
+    {
+      "epoch": 1.7824363907874332,
+      "grad_norm": 0.5991883873939514,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 15517
+    },
+    {
+      "epoch": 1.7825512606972604,
+      "grad_norm": 0.608816921710968,
+      "learning_rate": 0.0001,
+      "loss": 1.4711,
+      "step": 15518
+    },
+    {
+      "epoch": 1.7826661306070875,
+      "grad_norm": 0.6820626258850098,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 15519
+    },
+    {
+      "epoch": 1.7827810005169145,
+      "grad_norm": 0.6117185354232788,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 15520
+    },
+    {
+      "epoch": 1.7828958704267417,
+      "grad_norm": 0.5957820415496826,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 15521
+    },
+    {
+      "epoch": 1.783010740336569,
+      "grad_norm": 0.6155253648757935,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 15522
+    },
+    {
+      "epoch": 1.783125610246396,
+      "grad_norm": 0.5900723934173584,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 15523
+    },
+    {
+      "epoch": 1.783240480156223,
+      "grad_norm": 0.6127877831459045,
+      "learning_rate": 0.0001,
+      "loss": 1.485,
+      "step": 15524
+    },
+    {
+      "epoch": 1.7833553500660502,
+      "grad_norm": 0.6130275130271912,
+      "learning_rate": 0.0001,
+      "loss": 1.2872,
+      "step": 15525
+    },
+    {
+      "epoch": 1.7834702199758774,
+      "grad_norm": 0.6365898251533508,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 15526
+    },
+    {
+      "epoch": 1.7835850898857044,
+      "grad_norm": 0.6927840113639832,
+      "learning_rate": 0.0001,
+      "loss": 1.4844,
+      "step": 15527
+    },
+    {
+      "epoch": 1.7836999597955314,
+      "grad_norm": 0.6556539535522461,
+      "learning_rate": 0.0001,
+      "loss": 1.489,
+      "step": 15528
+    },
+    {
+      "epoch": 1.7838148297053587,
+      "grad_norm": 0.6017742156982422,
+      "learning_rate": 0.0001,
+      "loss": 1.3716,
+      "step": 15529
+    },
+    {
+      "epoch": 1.7839296996151859,
+      "grad_norm": 0.6629507541656494,
+      "learning_rate": 0.0001,
+      "loss": 1.5471,
+      "step": 15530
+    },
+    {
+      "epoch": 1.784044569525013,
+      "grad_norm": 0.6213143467903137,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 15531
+    },
+    {
+      "epoch": 1.78415943943484,
+      "grad_norm": 0.6131148338317871,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 15532
+    },
+    {
+      "epoch": 1.7842743093446671,
+      "grad_norm": 0.6498699188232422,
+      "learning_rate": 0.0001,
+      "loss": 1.5769,
+      "step": 15533
+    },
+    {
+      "epoch": 1.7843891792544944,
+      "grad_norm": 0.6236017346382141,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 15534
+    },
+    {
+      "epoch": 1.7845040491643214,
+      "grad_norm": 0.6245088577270508,
+      "learning_rate": 0.0001,
+      "loss": 1.4866,
+      "step": 15535
+    },
+    {
+      "epoch": 1.7846189190741484,
+      "grad_norm": 0.6300262212753296,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 15536
+    },
+    {
+      "epoch": 1.7847337889839756,
+      "grad_norm": 0.6133743524551392,
+      "learning_rate": 0.0001,
+      "loss": 1.339,
+      "step": 15537
+    },
+    {
+      "epoch": 1.7848486588938028,
+      "grad_norm": 0.5978740453720093,
+      "learning_rate": 0.0001,
+      "loss": 1.38,
+      "step": 15538
+    },
+    {
+      "epoch": 1.7849635288036299,
+      "grad_norm": 0.583224892616272,
+      "learning_rate": 0.0001,
+      "loss": 1.4004,
+      "step": 15539
+    },
+    {
+      "epoch": 1.7850783987134569,
+      "grad_norm": 0.6395849585533142,
+      "learning_rate": 0.0001,
+      "loss": 1.2943,
+      "step": 15540
+    },
+    {
+      "epoch": 1.785193268623284,
+      "grad_norm": 0.5944806933403015,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 15541
+    },
+    {
+      "epoch": 1.7853081385331113,
+      "grad_norm": 0.5802217125892639,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 15542
+    },
+    {
+      "epoch": 1.7854230084429383,
+      "grad_norm": 0.6466507911682129,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 15543
+    },
+    {
+      "epoch": 1.7855378783527653,
+      "grad_norm": 0.6032712459564209,
+      "learning_rate": 0.0001,
+      "loss": 1.3153,
+      "step": 15544
+    },
+    {
+      "epoch": 1.7856527482625926,
+      "grad_norm": 0.6779415011405945,
+      "learning_rate": 0.0001,
+      "loss": 1.3325,
+      "step": 15545
+    },
+    {
+      "epoch": 1.7857676181724198,
+      "grad_norm": 0.6788262128829956,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 15546
+    },
+    {
+      "epoch": 1.7858824880822468,
+      "grad_norm": 0.6456349492073059,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 15547
+    },
+    {
+      "epoch": 1.7859973579920738,
+      "grad_norm": 0.6493527889251709,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 15548
+    },
+    {
+      "epoch": 1.786112227901901,
+      "grad_norm": 0.5926679968833923,
+      "learning_rate": 0.0001,
+      "loss": 1.2383,
+      "step": 15549
+    },
+    {
+      "epoch": 1.7862270978117283,
+      "grad_norm": 0.6278696656227112,
+      "learning_rate": 0.0001,
+      "loss": 1.4126,
+      "step": 15550
+    },
+    {
+      "epoch": 1.7863419677215553,
+      "grad_norm": 0.6778453588485718,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 15551
+    },
+    {
+      "epoch": 1.7864568376313823,
+      "grad_norm": 0.6517847180366516,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 15552
+    },
+    {
+      "epoch": 1.7865717075412095,
+      "grad_norm": 0.6085006594657898,
+      "learning_rate": 0.0001,
+      "loss": 1.3612,
+      "step": 15553
+    },
+    {
+      "epoch": 1.7866865774510368,
+      "grad_norm": 0.6282052397727966,
+      "learning_rate": 0.0001,
+      "loss": 1.5781,
+      "step": 15554
+    },
+    {
+      "epoch": 1.7868014473608638,
+      "grad_norm": 0.6484823226928711,
+      "learning_rate": 0.0001,
+      "loss": 1.6014,
+      "step": 15555
+    },
+    {
+      "epoch": 1.7869163172706908,
+      "grad_norm": 0.6144720315933228,
+      "learning_rate": 0.0001,
+      "loss": 1.4621,
+      "step": 15556
+    },
+    {
+      "epoch": 1.787031187180518,
+      "grad_norm": 0.6251539587974548,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 15557
+    },
+    {
+      "epoch": 1.7871460570903452,
+      "grad_norm": 0.7098304033279419,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 15558
+    },
+    {
+      "epoch": 1.7872609270001723,
+      "grad_norm": 0.6878634095191956,
+      "learning_rate": 0.0001,
+      "loss": 1.5715,
+      "step": 15559
+    },
+    {
+      "epoch": 1.7873757969099993,
+      "grad_norm": 0.6232366561889648,
+      "learning_rate": 0.0001,
+      "loss": 1.4746,
+      "step": 15560
+    },
+    {
+      "epoch": 1.7874906668198265,
+      "grad_norm": 0.6680793166160583,
+      "learning_rate": 0.0001,
+      "loss": 1.4533,
+      "step": 15561
+    },
+    {
+      "epoch": 1.7876055367296537,
+      "grad_norm": 0.5885732173919678,
+      "learning_rate": 0.0001,
+      "loss": 1.3305,
+      "step": 15562
+    },
+    {
+      "epoch": 1.787720406639481,
+      "grad_norm": 0.7211229801177979,
+      "learning_rate": 0.0001,
+      "loss": 1.6991,
+      "step": 15563
+    },
+    {
+      "epoch": 1.787835276549308,
+      "grad_norm": 0.6472933888435364,
+      "learning_rate": 0.0001,
+      "loss": 1.5143,
+      "step": 15564
+    },
+    {
+      "epoch": 1.787950146459135,
+      "grad_norm": 0.6679675579071045,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 15565
+    },
+    {
+      "epoch": 1.7880650163689622,
+      "grad_norm": 0.6244038939476013,
+      "learning_rate": 0.0001,
+      "loss": 1.4162,
+      "step": 15566
+    },
+    {
+      "epoch": 1.7881798862787894,
+      "grad_norm": 0.6286173462867737,
+      "learning_rate": 0.0001,
+      "loss": 1.2787,
+      "step": 15567
+    },
+    {
+      "epoch": 1.7882947561886164,
+      "grad_norm": 0.6652150750160217,
+      "learning_rate": 0.0001,
+      "loss": 1.3333,
+      "step": 15568
+    },
+    {
+      "epoch": 1.7884096260984435,
+      "grad_norm": 0.5869207382202148,
+      "learning_rate": 0.0001,
+      "loss": 1.3693,
+      "step": 15569
+    },
+    {
+      "epoch": 1.7885244960082707,
+      "grad_norm": 0.6084496974945068,
+      "learning_rate": 0.0001,
+      "loss": 1.419,
+      "step": 15570
+    },
+    {
+      "epoch": 1.788639365918098,
+      "grad_norm": 0.6759893894195557,
+      "learning_rate": 0.0001,
+      "loss": 1.2883,
+      "step": 15571
+    },
+    {
+      "epoch": 1.788754235827925,
+      "grad_norm": 0.5952495336532593,
+      "learning_rate": 0.0001,
+      "loss": 1.4366,
+      "step": 15572
+    },
+    {
+      "epoch": 1.788869105737752,
+      "grad_norm": 0.5997838377952576,
+      "learning_rate": 0.0001,
+      "loss": 1.4331,
+      "step": 15573
+    },
+    {
+      "epoch": 1.7889839756475792,
+      "grad_norm": 0.6798581480979919,
+      "learning_rate": 0.0001,
+      "loss": 1.6547,
+      "step": 15574
+    },
+    {
+      "epoch": 1.7890988455574064,
+      "grad_norm": 0.6956671476364136,
+      "learning_rate": 0.0001,
+      "loss": 1.7073,
+      "step": 15575
+    },
+    {
+      "epoch": 1.7892137154672334,
+      "grad_norm": 0.642201840877533,
+      "learning_rate": 0.0001,
+      "loss": 1.3273,
+      "step": 15576
+    },
+    {
+      "epoch": 1.7893285853770604,
+      "grad_norm": 0.6423424482345581,
+      "learning_rate": 0.0001,
+      "loss": 1.4838,
+      "step": 15577
+    },
+    {
+      "epoch": 1.7894434552868876,
+      "grad_norm": 0.6451569199562073,
+      "learning_rate": 0.0001,
+      "loss": 1.4567,
+      "step": 15578
+    },
+    {
+      "epoch": 1.7895583251967149,
+      "grad_norm": 0.623618483543396,
+      "learning_rate": 0.0001,
+      "loss": 1.3361,
+      "step": 15579
+    },
+    {
+      "epoch": 1.7896731951065419,
+      "grad_norm": 0.6370881795883179,
+      "learning_rate": 0.0001,
+      "loss": 1.4506,
+      "step": 15580
+    },
+    {
+      "epoch": 1.789788065016369,
+      "grad_norm": 0.6714301705360413,
+      "learning_rate": 0.0001,
+      "loss": 1.48,
+      "step": 15581
+    },
+    {
+      "epoch": 1.7899029349261961,
+      "grad_norm": 0.7274426817893982,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 15582
+    },
+    {
+      "epoch": 1.7900178048360234,
+      "grad_norm": 0.7094860672950745,
+      "learning_rate": 0.0001,
+      "loss": 1.5049,
+      "step": 15583
+    },
+    {
+      "epoch": 1.7901326747458504,
+      "grad_norm": 0.6510686278343201,
+      "learning_rate": 0.0001,
+      "loss": 1.1147,
+      "step": 15584
+    },
+    {
+      "epoch": 1.7902475446556774,
+      "grad_norm": 0.6724952459335327,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 15585
+    },
+    {
+      "epoch": 1.7903624145655046,
+      "grad_norm": 0.6730976700782776,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 15586
+    },
+    {
+      "epoch": 1.7904772844753318,
+      "grad_norm": 0.5927944779396057,
+      "learning_rate": 0.0001,
+      "loss": 1.269,
+      "step": 15587
+    },
+    {
+      "epoch": 1.7905921543851588,
+      "grad_norm": 0.6580035090446472,
+      "learning_rate": 0.0001,
+      "loss": 1.4699,
+      "step": 15588
+    },
+    {
+      "epoch": 1.7907070242949858,
+      "grad_norm": 0.5773756504058838,
+      "learning_rate": 0.0001,
+      "loss": 1.4074,
+      "step": 15589
+    },
+    {
+      "epoch": 1.790821894204813,
+      "grad_norm": 0.660759449005127,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 15590
+    },
+    {
+      "epoch": 1.7909367641146403,
+      "grad_norm": 0.7201364040374756,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 15591
+    },
+    {
+      "epoch": 1.7910516340244673,
+      "grad_norm": 0.6264855861663818,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 15592
+    },
+    {
+      "epoch": 1.7911665039342943,
+      "grad_norm": 0.6170802712440491,
+      "learning_rate": 0.0001,
+      "loss": 1.3783,
+      "step": 15593
+    },
+    {
+      "epoch": 1.7912813738441216,
+      "grad_norm": 0.6500568389892578,
+      "learning_rate": 0.0001,
+      "loss": 1.449,
+      "step": 15594
+    },
+    {
+      "epoch": 1.7913962437539488,
+      "grad_norm": 0.6140516996383667,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 15595
+    },
+    {
+      "epoch": 1.7915111136637758,
+      "grad_norm": 0.5915245413780212,
+      "learning_rate": 0.0001,
+      "loss": 1.3078,
+      "step": 15596
+    },
+    {
+      "epoch": 1.7916259835736028,
+      "grad_norm": 0.5879004001617432,
+      "learning_rate": 0.0001,
+      "loss": 1.3216,
+      "step": 15597
+    },
+    {
+      "epoch": 1.79174085348343,
+      "grad_norm": 0.6277405619621277,
+      "learning_rate": 0.0001,
+      "loss": 1.4248,
+      "step": 15598
+    },
+    {
+      "epoch": 1.7918557233932573,
+      "grad_norm": 0.5921669006347656,
+      "learning_rate": 0.0001,
+      "loss": 1.3723,
+      "step": 15599
+    },
+    {
+      "epoch": 1.7919705933030843,
+      "grad_norm": 0.5975223183631897,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 15600
+    },
+    {
+      "epoch": 1.7920854632129113,
+      "grad_norm": 0.6387137770652771,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 15601
+    },
+    {
+      "epoch": 1.7922003331227385,
+      "grad_norm": 0.5730849504470825,
+      "learning_rate": 0.0001,
+      "loss": 1.3881,
+      "step": 15602
+    },
+    {
+      "epoch": 1.7923152030325658,
+      "grad_norm": 0.5885117650032043,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 15603
+    },
+    {
+      "epoch": 1.7924300729423928,
+      "grad_norm": 0.6096714735031128,
+      "learning_rate": 0.0001,
+      "loss": 1.4063,
+      "step": 15604
+    },
+    {
+      "epoch": 1.7925449428522198,
+      "grad_norm": 0.6324706077575684,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 15605
+    },
+    {
+      "epoch": 1.792659812762047,
+      "grad_norm": 0.6507951617240906,
+      "learning_rate": 0.0001,
+      "loss": 1.4797,
+      "step": 15606
+    },
+    {
+      "epoch": 1.7927746826718742,
+      "grad_norm": 0.6201066970825195,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 15607
+    },
+    {
+      "epoch": 1.7928895525817012,
+      "grad_norm": 0.6061670780181885,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 15608
+    },
+    {
+      "epoch": 1.7930044224915282,
+      "grad_norm": 0.6483253240585327,
+      "learning_rate": 0.0001,
+      "loss": 1.3321,
+      "step": 15609
+    },
+    {
+      "epoch": 1.7931192924013555,
+      "grad_norm": 0.608324408531189,
+      "learning_rate": 0.0001,
+      "loss": 1.5719,
+      "step": 15610
+    },
+    {
+      "epoch": 1.7932341623111827,
+      "grad_norm": 0.5885353684425354,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 15611
+    },
+    {
+      "epoch": 1.7933490322210097,
+      "grad_norm": 0.612770676612854,
+      "learning_rate": 0.0001,
+      "loss": 1.4332,
+      "step": 15612
+    },
+    {
+      "epoch": 1.7934639021308367,
+      "grad_norm": 0.656052827835083,
+      "learning_rate": 0.0001,
+      "loss": 1.3348,
+      "step": 15613
+    },
+    {
+      "epoch": 1.793578772040664,
+      "grad_norm": 0.6733161211013794,
+      "learning_rate": 0.0001,
+      "loss": 1.5747,
+      "step": 15614
+    },
+    {
+      "epoch": 1.7936936419504912,
+      "grad_norm": 0.6594151258468628,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 15615
+    },
+    {
+      "epoch": 1.7938085118603182,
+      "grad_norm": 0.6126337051391602,
+      "learning_rate": 0.0001,
+      "loss": 1.3746,
+      "step": 15616
+    },
+    {
+      "epoch": 1.7939233817701452,
+      "grad_norm": 0.613390326499939,
+      "learning_rate": 0.0001,
+      "loss": 1.4174,
+      "step": 15617
+    },
+    {
+      "epoch": 1.7940382516799724,
+      "grad_norm": 0.6970869898796082,
+      "learning_rate": 0.0001,
+      "loss": 1.4908,
+      "step": 15618
+    },
+    {
+      "epoch": 1.7941531215897997,
+      "grad_norm": 0.5497614145278931,
+      "learning_rate": 0.0001,
+      "loss": 1.3542,
+      "step": 15619
+    },
+    {
+      "epoch": 1.7942679914996267,
+      "grad_norm": 0.5766857862472534,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 15620
+    },
+    {
+      "epoch": 1.7943828614094537,
+      "grad_norm": 0.6197019219398499,
+      "learning_rate": 0.0001,
+      "loss": 1.2267,
+      "step": 15621
+    },
+    {
+      "epoch": 1.794497731319281,
+      "grad_norm": 0.6978227496147156,
+      "learning_rate": 0.0001,
+      "loss": 1.3925,
+      "step": 15622
+    },
+    {
+      "epoch": 1.7946126012291082,
+      "grad_norm": 0.6291922330856323,
+      "learning_rate": 0.0001,
+      "loss": 1.3681,
+      "step": 15623
+    },
+    {
+      "epoch": 1.7947274711389352,
+      "grad_norm": 0.6435182690620422,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 15624
+    },
+    {
+      "epoch": 1.7948423410487622,
+      "grad_norm": 0.6218845248222351,
+      "learning_rate": 0.0001,
+      "loss": 1.4538,
+      "step": 15625
+    },
+    {
+      "epoch": 1.7949572109585894,
+      "grad_norm": 0.5924429297447205,
+      "learning_rate": 0.0001,
+      "loss": 1.4747,
+      "step": 15626
+    },
+    {
+      "epoch": 1.7950720808684166,
+      "grad_norm": 0.632346510887146,
+      "learning_rate": 0.0001,
+      "loss": 1.5101,
+      "step": 15627
+    },
+    {
+      "epoch": 1.7951869507782436,
+      "grad_norm": 0.5771130323410034,
+      "learning_rate": 0.0001,
+      "loss": 1.4203,
+      "step": 15628
+    },
+    {
+      "epoch": 1.7953018206880706,
+      "grad_norm": 0.5830177068710327,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 15629
+    },
+    {
+      "epoch": 1.7954166905978979,
+      "grad_norm": 0.685839831829071,
+      "learning_rate": 0.0001,
+      "loss": 1.3029,
+      "step": 15630
+    },
+    {
+      "epoch": 1.795531560507725,
+      "grad_norm": 0.5754895806312561,
+      "learning_rate": 0.0001,
+      "loss": 1.4049,
+      "step": 15631
+    },
+    {
+      "epoch": 1.7956464304175521,
+      "grad_norm": 0.6010079979896545,
+      "learning_rate": 0.0001,
+      "loss": 1.3411,
+      "step": 15632
+    },
+    {
+      "epoch": 1.7957613003273791,
+      "grad_norm": 0.6301531791687012,
+      "learning_rate": 0.0001,
+      "loss": 1.3216,
+      "step": 15633
+    },
+    {
+      "epoch": 1.7958761702372064,
+      "grad_norm": 0.5766407251358032,
+      "learning_rate": 0.0001,
+      "loss": 1.3345,
+      "step": 15634
+    },
+    {
+      "epoch": 1.7959910401470336,
+      "grad_norm": 0.6998764872550964,
+      "learning_rate": 0.0001,
+      "loss": 1.7205,
+      "step": 15635
+    },
+    {
+      "epoch": 1.7961059100568606,
+      "grad_norm": 0.6410335898399353,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 15636
+    },
+    {
+      "epoch": 1.7962207799666876,
+      "grad_norm": 0.5890425443649292,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 15637
+    },
+    {
+      "epoch": 1.7963356498765148,
+      "grad_norm": 0.717514157295227,
+      "learning_rate": 0.0001,
+      "loss": 1.4209,
+      "step": 15638
+    },
+    {
+      "epoch": 1.796450519786342,
+      "grad_norm": 0.6638051271438599,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 15639
+    },
+    {
+      "epoch": 1.796565389696169,
+      "grad_norm": 0.6315420269966125,
+      "learning_rate": 0.0001,
+      "loss": 1.3122,
+      "step": 15640
+    },
+    {
+      "epoch": 1.796680259605996,
+      "grad_norm": 0.6300930380821228,
+      "learning_rate": 0.0001,
+      "loss": 1.4313,
+      "step": 15641
+    },
+    {
+      "epoch": 1.7967951295158233,
+      "grad_norm": 0.6064184904098511,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 15642
+    },
+    {
+      "epoch": 1.7969099994256506,
+      "grad_norm": 0.5937897562980652,
+      "learning_rate": 0.0001,
+      "loss": 1.44,
+      "step": 15643
+    },
+    {
+      "epoch": 1.7970248693354776,
+      "grad_norm": 0.5919926166534424,
+      "learning_rate": 0.0001,
+      "loss": 1.4487,
+      "step": 15644
+    },
+    {
+      "epoch": 1.7971397392453046,
+      "grad_norm": 0.6598802208900452,
+      "learning_rate": 0.0001,
+      "loss": 1.3712,
+      "step": 15645
+    },
+    {
+      "epoch": 1.7972546091551318,
+      "grad_norm": 0.6512289047241211,
+      "learning_rate": 0.0001,
+      "loss": 1.2151,
+      "step": 15646
+    },
+    {
+      "epoch": 1.797369479064959,
+      "grad_norm": 0.6600479483604431,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 15647
+    },
+    {
+      "epoch": 1.797484348974786,
+      "grad_norm": 0.6048285365104675,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 15648
+    },
+    {
+      "epoch": 1.797599218884613,
+      "grad_norm": 0.6369686126708984,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 15649
+    },
+    {
+      "epoch": 1.7977140887944403,
+      "grad_norm": 0.6395129561424255,
+      "learning_rate": 0.0001,
+      "loss": 1.4658,
+      "step": 15650
+    },
+    {
+      "epoch": 1.7978289587042675,
+      "grad_norm": 0.6294239163398743,
+      "learning_rate": 0.0001,
+      "loss": 1.3939,
+      "step": 15651
+    },
+    {
+      "epoch": 1.7979438286140945,
+      "grad_norm": 0.6349182724952698,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 15652
+    },
+    {
+      "epoch": 1.7980586985239215,
+      "grad_norm": 0.7371272444725037,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 15653
+    },
+    {
+      "epoch": 1.7981735684337488,
+      "grad_norm": 0.724577784538269,
+      "learning_rate": 0.0001,
+      "loss": 1.3706,
+      "step": 15654
+    },
+    {
+      "epoch": 1.798288438343576,
+      "grad_norm": 0.6599369645118713,
+      "learning_rate": 0.0001,
+      "loss": 1.4384,
+      "step": 15655
+    },
+    {
+      "epoch": 1.798403308253403,
+      "grad_norm": 0.6410521864891052,
+      "learning_rate": 0.0001,
+      "loss": 1.4447,
+      "step": 15656
+    },
+    {
+      "epoch": 1.79851817816323,
+      "grad_norm": 0.6453815698623657,
+      "learning_rate": 0.0001,
+      "loss": 1.2947,
+      "step": 15657
+    },
+    {
+      "epoch": 1.7986330480730572,
+      "grad_norm": 0.5987499356269836,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 15658
+    },
+    {
+      "epoch": 1.7987479179828845,
+      "grad_norm": 0.6878007054328918,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 15659
+    },
+    {
+      "epoch": 1.7988627878927115,
+      "grad_norm": 0.6186679005622864,
+      "learning_rate": 0.0001,
+      "loss": 1.4337,
+      "step": 15660
+    },
+    {
+      "epoch": 1.7989776578025385,
+      "grad_norm": 0.6373911499977112,
+      "learning_rate": 0.0001,
+      "loss": 1.3869,
+      "step": 15661
+    },
+    {
+      "epoch": 1.7990925277123657,
+      "grad_norm": 0.626069962978363,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 15662
+    },
+    {
+      "epoch": 1.799207397622193,
+      "grad_norm": 0.6917441487312317,
+      "learning_rate": 0.0001,
+      "loss": 1.6685,
+      "step": 15663
+    },
+    {
+      "epoch": 1.79932226753202,
+      "grad_norm": 0.6298771500587463,
+      "learning_rate": 0.0001,
+      "loss": 1.4287,
+      "step": 15664
+    },
+    {
+      "epoch": 1.799437137441847,
+      "grad_norm": 0.6612205505371094,
+      "learning_rate": 0.0001,
+      "loss": 1.4862,
+      "step": 15665
+    },
+    {
+      "epoch": 1.7995520073516742,
+      "grad_norm": 0.6308835744857788,
+      "learning_rate": 0.0001,
+      "loss": 1.3969,
+      "step": 15666
+    },
+    {
+      "epoch": 1.7996668772615014,
+      "grad_norm": 0.6707907319068909,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 15667
+    },
+    {
+      "epoch": 1.7997817471713284,
+      "grad_norm": 0.5699948072433472,
+      "learning_rate": 0.0001,
+      "loss": 1.2569,
+      "step": 15668
+    },
+    {
+      "epoch": 1.7998966170811554,
+      "grad_norm": 0.6121656894683838,
+      "learning_rate": 0.0001,
+      "loss": 1.5105,
+      "step": 15669
+    },
+    {
+      "epoch": 1.8000114869909827,
+      "grad_norm": 0.640649676322937,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 15670
+    },
+    {
+      "epoch": 1.80012635690081,
+      "grad_norm": 0.595537543296814,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 15671
+    },
+    {
+      "epoch": 1.800241226810637,
+      "grad_norm": 0.6392516493797302,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 15672
+    },
+    {
+      "epoch": 1.800356096720464,
+      "grad_norm": 0.6374161243438721,
+      "learning_rate": 0.0001,
+      "loss": 1.3964,
+      "step": 15673
+    },
+    {
+      "epoch": 1.8004709666302912,
+      "grad_norm": 0.6486839652061462,
+      "learning_rate": 0.0001,
+      "loss": 1.6107,
+      "step": 15674
+    },
+    {
+      "epoch": 1.8005858365401184,
+      "grad_norm": 0.7054831981658936,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 15675
+    },
+    {
+      "epoch": 1.8007007064499454,
+      "grad_norm": 0.6817604899406433,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 15676
+    },
+    {
+      "epoch": 1.8008155763597724,
+      "grad_norm": 0.6403077840805054,
+      "learning_rate": 0.0001,
+      "loss": 1.4614,
+      "step": 15677
+    },
+    {
+      "epoch": 1.8009304462695996,
+      "grad_norm": 0.6834919452667236,
+      "learning_rate": 0.0001,
+      "loss": 1.3062,
+      "step": 15678
+    },
+    {
+      "epoch": 1.8010453161794269,
+      "grad_norm": 0.6138955354690552,
+      "learning_rate": 0.0001,
+      "loss": 1.235,
+      "step": 15679
+    },
+    {
+      "epoch": 1.8011601860892539,
+      "grad_norm": 0.6135299205780029,
+      "learning_rate": 0.0001,
+      "loss": 1.3112,
+      "step": 15680
+    },
+    {
+      "epoch": 1.8012750559990809,
+      "grad_norm": 0.6524873971939087,
+      "learning_rate": 0.0001,
+      "loss": 1.3524,
+      "step": 15681
+    },
+    {
+      "epoch": 1.8013899259089081,
+      "grad_norm": 0.635270357131958,
+      "learning_rate": 0.0001,
+      "loss": 1.1356,
+      "step": 15682
+    },
+    {
+      "epoch": 1.8015047958187353,
+      "grad_norm": 0.6525683403015137,
+      "learning_rate": 0.0001,
+      "loss": 1.3779,
+      "step": 15683
+    },
+    {
+      "epoch": 1.8016196657285624,
+      "grad_norm": 0.6439868807792664,
+      "learning_rate": 0.0001,
+      "loss": 1.4131,
+      "step": 15684
+    },
+    {
+      "epoch": 1.8017345356383894,
+      "grad_norm": 0.6394991278648376,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 15685
+    },
+    {
+      "epoch": 1.8018494055482166,
+      "grad_norm": 0.6692410707473755,
+      "learning_rate": 0.0001,
+      "loss": 1.477,
+      "step": 15686
+    },
+    {
+      "epoch": 1.8019642754580438,
+      "grad_norm": 0.5906667709350586,
+      "learning_rate": 0.0001,
+      "loss": 1.2574,
+      "step": 15687
+    },
+    {
+      "epoch": 1.8020791453678708,
+      "grad_norm": 0.6098186373710632,
+      "learning_rate": 0.0001,
+      "loss": 1.3739,
+      "step": 15688
+    },
+    {
+      "epoch": 1.8021940152776978,
+      "grad_norm": 0.636909544467926,
+      "learning_rate": 0.0001,
+      "loss": 1.4122,
+      "step": 15689
+    },
+    {
+      "epoch": 1.802308885187525,
+      "grad_norm": 0.606322169303894,
+      "learning_rate": 0.0001,
+      "loss": 1.4376,
+      "step": 15690
+    },
+    {
+      "epoch": 1.8024237550973523,
+      "grad_norm": 0.6577963829040527,
+      "learning_rate": 0.0001,
+      "loss": 1.3516,
+      "step": 15691
+    },
+    {
+      "epoch": 1.8025386250071793,
+      "grad_norm": 0.6309029459953308,
+      "learning_rate": 0.0001,
+      "loss": 1.516,
+      "step": 15692
+    },
+    {
+      "epoch": 1.8026534949170063,
+      "grad_norm": 0.585844099521637,
+      "learning_rate": 0.0001,
+      "loss": 1.4676,
+      "step": 15693
+    },
+    {
+      "epoch": 1.8027683648268336,
+      "grad_norm": 0.6345848441123962,
+      "learning_rate": 0.0001,
+      "loss": 1.3576,
+      "step": 15694
+    },
+    {
+      "epoch": 1.8028832347366608,
+      "grad_norm": 0.6085816621780396,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 15695
+    },
+    {
+      "epoch": 1.8029981046464878,
+      "grad_norm": 0.5922706127166748,
+      "learning_rate": 0.0001,
+      "loss": 1.4294,
+      "step": 15696
+    },
+    {
+      "epoch": 1.8031129745563148,
+      "grad_norm": 0.7549859285354614,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 15697
+    },
+    {
+      "epoch": 1.803227844466142,
+      "grad_norm": 0.5932414531707764,
+      "learning_rate": 0.0001,
+      "loss": 1.3871,
+      "step": 15698
+    },
+    {
+      "epoch": 1.8033427143759693,
+      "grad_norm": 0.6508196592330933,
+      "learning_rate": 0.0001,
+      "loss": 1.4249,
+      "step": 15699
+    },
+    {
+      "epoch": 1.8034575842857965,
+      "grad_norm": 0.6208270788192749,
+      "learning_rate": 0.0001,
+      "loss": 1.5313,
+      "step": 15700
+    },
+    {
+      "epoch": 1.8035724541956235,
+      "grad_norm": 0.5751149654388428,
+      "learning_rate": 0.0001,
+      "loss": 1.333,
+      "step": 15701
+    },
+    {
+      "epoch": 1.8036873241054505,
+      "grad_norm": 0.6106820702552795,
+      "learning_rate": 0.0001,
+      "loss": 1.2523,
+      "step": 15702
+    },
+    {
+      "epoch": 1.8038021940152777,
+      "grad_norm": 0.6373929977416992,
+      "learning_rate": 0.0001,
+      "loss": 1.4284,
+      "step": 15703
+    },
+    {
+      "epoch": 1.803917063925105,
+      "grad_norm": 0.6269072890281677,
+      "learning_rate": 0.0001,
+      "loss": 1.4818,
+      "step": 15704
+    },
+    {
+      "epoch": 1.804031933834932,
+      "grad_norm": 0.6745073199272156,
+      "learning_rate": 0.0001,
+      "loss": 1.3644,
+      "step": 15705
+    },
+    {
+      "epoch": 1.804146803744759,
+      "grad_norm": 0.7023970484733582,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 15706
+    },
+    {
+      "epoch": 1.8042616736545862,
+      "grad_norm": 0.6291683912277222,
+      "learning_rate": 0.0001,
+      "loss": 1.3183,
+      "step": 15707
+    },
+    {
+      "epoch": 1.8043765435644135,
+      "grad_norm": 0.6405826807022095,
+      "learning_rate": 0.0001,
+      "loss": 1.5059,
+      "step": 15708
+    },
+    {
+      "epoch": 1.8044914134742405,
+      "grad_norm": 0.6145171523094177,
+      "learning_rate": 0.0001,
+      "loss": 1.4438,
+      "step": 15709
+    },
+    {
+      "epoch": 1.8046062833840675,
+      "grad_norm": 0.6371951103210449,
+      "learning_rate": 0.0001,
+      "loss": 1.562,
+      "step": 15710
+    },
+    {
+      "epoch": 1.8047211532938947,
+      "grad_norm": 0.6146504878997803,
+      "learning_rate": 0.0001,
+      "loss": 1.4305,
+      "step": 15711
+    },
+    {
+      "epoch": 1.804836023203722,
+      "grad_norm": 0.6177845597267151,
+      "learning_rate": 0.0001,
+      "loss": 1.2665,
+      "step": 15712
+    },
+    {
+      "epoch": 1.804950893113549,
+      "grad_norm": 0.6438102722167969,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 15713
+    },
+    {
+      "epoch": 1.805065763023376,
+      "grad_norm": 0.6426578164100647,
+      "learning_rate": 0.0001,
+      "loss": 1.4181,
+      "step": 15714
+    },
+    {
+      "epoch": 1.8051806329332032,
+      "grad_norm": 0.7150061130523682,
+      "learning_rate": 0.0001,
+      "loss": 1.4365,
+      "step": 15715
+    },
+    {
+      "epoch": 1.8052955028430304,
+      "grad_norm": 0.6319580674171448,
+      "learning_rate": 0.0001,
+      "loss": 1.4184,
+      "step": 15716
+    },
+    {
+      "epoch": 1.8054103727528574,
+      "grad_norm": 0.6289169788360596,
+      "learning_rate": 0.0001,
+      "loss": 1.4949,
+      "step": 15717
+    },
+    {
+      "epoch": 1.8055252426626844,
+      "grad_norm": 0.6072092056274414,
+      "learning_rate": 0.0001,
+      "loss": 1.5016,
+      "step": 15718
+    },
+    {
+      "epoch": 1.8056401125725117,
+      "grad_norm": 0.71548992395401,
+      "learning_rate": 0.0001,
+      "loss": 1.7456,
+      "step": 15719
+    },
+    {
+      "epoch": 1.805754982482339,
+      "grad_norm": 0.622413694858551,
+      "learning_rate": 0.0001,
+      "loss": 1.4269,
+      "step": 15720
+    },
+    {
+      "epoch": 1.805869852392166,
+      "grad_norm": 0.6449538469314575,
+      "learning_rate": 0.0001,
+      "loss": 1.2798,
+      "step": 15721
+    },
+    {
+      "epoch": 1.805984722301993,
+      "grad_norm": 0.6331883668899536,
+      "learning_rate": 0.0001,
+      "loss": 1.4222,
+      "step": 15722
+    },
+    {
+      "epoch": 1.8060995922118201,
+      "grad_norm": 0.57768315076828,
+      "learning_rate": 0.0001,
+      "loss": 1.2396,
+      "step": 15723
+    },
+    {
+      "epoch": 1.8062144621216474,
+      "grad_norm": 0.6328497529029846,
+      "learning_rate": 0.0001,
+      "loss": 1.3806,
+      "step": 15724
+    },
+    {
+      "epoch": 1.8063293320314744,
+      "grad_norm": 0.69939124584198,
+      "learning_rate": 0.0001,
+      "loss": 1.3731,
+      "step": 15725
+    },
+    {
+      "epoch": 1.8064442019413014,
+      "grad_norm": 0.6895657181739807,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 15726
+    },
+    {
+      "epoch": 1.8065590718511286,
+      "grad_norm": 0.6120685935020447,
+      "learning_rate": 0.0001,
+      "loss": 1.1978,
+      "step": 15727
+    },
+    {
+      "epoch": 1.8066739417609559,
+      "grad_norm": 0.6782362461090088,
+      "learning_rate": 0.0001,
+      "loss": 1.2976,
+      "step": 15728
+    },
+    {
+      "epoch": 1.8067888116707829,
+      "grad_norm": 0.6087343096733093,
+      "learning_rate": 0.0001,
+      "loss": 1.3276,
+      "step": 15729
+    },
+    {
+      "epoch": 1.8069036815806099,
+      "grad_norm": 0.6072776913642883,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 15730
+    },
+    {
+      "epoch": 1.807018551490437,
+      "grad_norm": 0.6058279275894165,
+      "learning_rate": 0.0001,
+      "loss": 1.3737,
+      "step": 15731
+    },
+    {
+      "epoch": 1.8071334214002643,
+      "grad_norm": 0.63099604845047,
+      "learning_rate": 0.0001,
+      "loss": 1.413,
+      "step": 15732
+    },
+    {
+      "epoch": 1.8072482913100913,
+      "grad_norm": 0.6059746146202087,
+      "learning_rate": 0.0001,
+      "loss": 1.4017,
+      "step": 15733
+    },
+    {
+      "epoch": 1.8073631612199184,
+      "grad_norm": 0.6642100214958191,
+      "learning_rate": 0.0001,
+      "loss": 1.5973,
+      "step": 15734
+    },
+    {
+      "epoch": 1.8074780311297456,
+      "grad_norm": 0.6303525567054749,
+      "learning_rate": 0.0001,
+      "loss": 1.5459,
+      "step": 15735
+    },
+    {
+      "epoch": 1.8075929010395728,
+      "grad_norm": 0.605383038520813,
+      "learning_rate": 0.0001,
+      "loss": 1.3552,
+      "step": 15736
+    },
+    {
+      "epoch": 1.8077077709493998,
+      "grad_norm": 0.6636791229248047,
+      "learning_rate": 0.0001,
+      "loss": 1.3234,
+      "step": 15737
+    },
+    {
+      "epoch": 1.8078226408592268,
+      "grad_norm": 0.6717275381088257,
+      "learning_rate": 0.0001,
+      "loss": 1.5471,
+      "step": 15738
+    },
+    {
+      "epoch": 1.807937510769054,
+      "grad_norm": 0.6487286686897278,
+      "learning_rate": 0.0001,
+      "loss": 1.5063,
+      "step": 15739
+    },
+    {
+      "epoch": 1.8080523806788813,
+      "grad_norm": 0.6421818137168884,
+      "learning_rate": 0.0001,
+      "loss": 1.4977,
+      "step": 15740
+    },
+    {
+      "epoch": 1.8081672505887083,
+      "grad_norm": 0.6361395716667175,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 15741
+    },
+    {
+      "epoch": 1.8082821204985353,
+      "grad_norm": 0.6091173887252808,
+      "learning_rate": 0.0001,
+      "loss": 1.5454,
+      "step": 15742
+    },
+    {
+      "epoch": 1.8083969904083625,
+      "grad_norm": 0.6464824080467224,
+      "learning_rate": 0.0001,
+      "loss": 1.4512,
+      "step": 15743
+    },
+    {
+      "epoch": 1.8085118603181898,
+      "grad_norm": 0.6446046829223633,
+      "learning_rate": 0.0001,
+      "loss": 1.451,
+      "step": 15744
+    },
+    {
+      "epoch": 1.8086267302280168,
+      "grad_norm": 0.7475030422210693,
+      "learning_rate": 0.0001,
+      "loss": 1.7174,
+      "step": 15745
+    },
+    {
+      "epoch": 1.8087416001378438,
+      "grad_norm": 0.6352036595344543,
+      "learning_rate": 0.0001,
+      "loss": 1.3802,
+      "step": 15746
+    },
+    {
+      "epoch": 1.808856470047671,
+      "grad_norm": 0.65043044090271,
+      "learning_rate": 0.0001,
+      "loss": 1.4297,
+      "step": 15747
+    },
+    {
+      "epoch": 1.8089713399574983,
+      "grad_norm": 0.6558632254600525,
+      "learning_rate": 0.0001,
+      "loss": 1.4044,
+      "step": 15748
+    },
+    {
+      "epoch": 1.8090862098673253,
+      "grad_norm": 0.6217056512832642,
+      "learning_rate": 0.0001,
+      "loss": 1.2797,
+      "step": 15749
+    },
+    {
+      "epoch": 1.8092010797771523,
+      "grad_norm": 0.6086947917938232,
+      "learning_rate": 0.0001,
+      "loss": 1.3942,
+      "step": 15750
+    },
+    {
+      "epoch": 1.8093159496869795,
+      "grad_norm": 0.5924387574195862,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 15751
+    },
+    {
+      "epoch": 1.8094308195968067,
+      "grad_norm": 0.6517341732978821,
+      "learning_rate": 0.0001,
+      "loss": 1.4954,
+      "step": 15752
+    },
+    {
+      "epoch": 1.8095456895066337,
+      "grad_norm": 0.6108118891716003,
+      "learning_rate": 0.0001,
+      "loss": 1.3201,
+      "step": 15753
+    },
+    {
+      "epoch": 1.8096605594164608,
+      "grad_norm": 0.6316987872123718,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 15754
+    },
+    {
+      "epoch": 1.809775429326288,
+      "grad_norm": 0.6387416124343872,
+      "learning_rate": 0.0001,
+      "loss": 1.5644,
+      "step": 15755
+    },
+    {
+      "epoch": 1.8098902992361152,
+      "grad_norm": 0.604150116443634,
+      "learning_rate": 0.0001,
+      "loss": 1.185,
+      "step": 15756
+    },
+    {
+      "epoch": 1.8100051691459422,
+      "grad_norm": 0.6756969690322876,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 15757
+    },
+    {
+      "epoch": 1.8101200390557692,
+      "grad_norm": 0.6240161061286926,
+      "learning_rate": 0.0001,
+      "loss": 1.3972,
+      "step": 15758
+    },
+    {
+      "epoch": 1.8102349089655965,
+      "grad_norm": 0.6883102655410767,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 15759
+    },
+    {
+      "epoch": 1.8103497788754237,
+      "grad_norm": 0.5605351328849792,
+      "learning_rate": 0.0001,
+      "loss": 1.3037,
+      "step": 15760
+    },
+    {
+      "epoch": 1.8104646487852507,
+      "grad_norm": 0.6002980470657349,
+      "learning_rate": 0.0001,
+      "loss": 1.2442,
+      "step": 15761
+    },
+    {
+      "epoch": 1.8105795186950777,
+      "grad_norm": 0.6384274959564209,
+      "learning_rate": 0.0001,
+      "loss": 1.4306,
+      "step": 15762
+    },
+    {
+      "epoch": 1.810694388604905,
+      "grad_norm": 0.6188052296638489,
+      "learning_rate": 0.0001,
+      "loss": 1.4172,
+      "step": 15763
+    },
+    {
+      "epoch": 1.8108092585147322,
+      "grad_norm": 0.6061704158782959,
+      "learning_rate": 0.0001,
+      "loss": 1.2811,
+      "step": 15764
+    },
+    {
+      "epoch": 1.8109241284245592,
+      "grad_norm": 0.6506741642951965,
+      "learning_rate": 0.0001,
+      "loss": 1.4402,
+      "step": 15765
+    },
+    {
+      "epoch": 1.8110389983343862,
+      "grad_norm": 0.6865687966346741,
+      "learning_rate": 0.0001,
+      "loss": 1.4626,
+      "step": 15766
+    },
+    {
+      "epoch": 1.8111538682442134,
+      "grad_norm": 0.6647955179214478,
+      "learning_rate": 0.0001,
+      "loss": 1.3912,
+      "step": 15767
+    },
+    {
+      "epoch": 1.8112687381540407,
+      "grad_norm": 0.6481913924217224,
+      "learning_rate": 0.0001,
+      "loss": 1.4325,
+      "step": 15768
+    },
+    {
+      "epoch": 1.8113836080638677,
+      "grad_norm": 0.7807742953300476,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 15769
+    },
+    {
+      "epoch": 1.8114984779736947,
+      "grad_norm": 0.663368284702301,
+      "learning_rate": 0.0001,
+      "loss": 1.2071,
+      "step": 15770
+    },
+    {
+      "epoch": 1.811613347883522,
+      "grad_norm": 0.6617883443832397,
+      "learning_rate": 0.0001,
+      "loss": 1.462,
+      "step": 15771
+    },
+    {
+      "epoch": 1.8117282177933491,
+      "grad_norm": 0.6506766676902771,
+      "learning_rate": 0.0001,
+      "loss": 1.4079,
+      "step": 15772
+    },
+    {
+      "epoch": 1.8118430877031761,
+      "grad_norm": 0.655314564704895,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 15773
+    },
+    {
+      "epoch": 1.8119579576130032,
+      "grad_norm": 0.5995591878890991,
+      "learning_rate": 0.0001,
+      "loss": 1.4349,
+      "step": 15774
+    },
+    {
+      "epoch": 1.8120728275228304,
+      "grad_norm": 0.7130681276321411,
+      "learning_rate": 0.0001,
+      "loss": 1.28,
+      "step": 15775
+    },
+    {
+      "epoch": 1.8121876974326576,
+      "grad_norm": 0.6497383713722229,
+      "learning_rate": 0.0001,
+      "loss": 1.3618,
+      "step": 15776
+    },
+    {
+      "epoch": 1.8123025673424846,
+      "grad_norm": 0.6399600505828857,
+      "learning_rate": 0.0001,
+      "loss": 1.5071,
+      "step": 15777
+    },
+    {
+      "epoch": 1.8124174372523116,
+      "grad_norm": 0.6517605185508728,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 15778
+    },
+    {
+      "epoch": 1.8125323071621389,
+      "grad_norm": 0.7210836410522461,
+      "learning_rate": 0.0001,
+      "loss": 1.3193,
+      "step": 15779
+    },
+    {
+      "epoch": 1.812647177071966,
+      "grad_norm": 0.6529654264450073,
+      "learning_rate": 0.0001,
+      "loss": 1.3761,
+      "step": 15780
+    },
+    {
+      "epoch": 1.812762046981793,
+      "grad_norm": 0.5776856541633606,
+      "learning_rate": 0.0001,
+      "loss": 1.2645,
+      "step": 15781
+    },
+    {
+      "epoch": 1.8128769168916201,
+      "grad_norm": 0.6165531277656555,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 15782
+    },
+    {
+      "epoch": 1.8129917868014473,
+      "grad_norm": 0.6355653405189514,
+      "learning_rate": 0.0001,
+      "loss": 1.4083,
+      "step": 15783
+    },
+    {
+      "epoch": 1.8131066567112746,
+      "grad_norm": 0.6272133588790894,
+      "learning_rate": 0.0001,
+      "loss": 1.5475,
+      "step": 15784
+    },
+    {
+      "epoch": 1.8132215266211016,
+      "grad_norm": 0.6231955885887146,
+      "learning_rate": 0.0001,
+      "loss": 1.1145,
+      "step": 15785
+    },
+    {
+      "epoch": 1.8133363965309286,
+      "grad_norm": 0.5937713980674744,
+      "learning_rate": 0.0001,
+      "loss": 1.3303,
+      "step": 15786
+    },
+    {
+      "epoch": 1.8134512664407558,
+      "grad_norm": 0.6835618615150452,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 15787
+    },
+    {
+      "epoch": 1.813566136350583,
+      "grad_norm": 0.700341522693634,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 15788
+    },
+    {
+      "epoch": 1.81368100626041,
+      "grad_norm": 0.6277741193771362,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 15789
+    },
+    {
+      "epoch": 1.813795876170237,
+      "grad_norm": 0.6309356689453125,
+      "learning_rate": 0.0001,
+      "loss": 1.5235,
+      "step": 15790
+    },
+    {
+      "epoch": 1.8139107460800643,
+      "grad_norm": 0.5971214175224304,
+      "learning_rate": 0.0001,
+      "loss": 1.3632,
+      "step": 15791
+    },
+    {
+      "epoch": 1.8140256159898915,
+      "grad_norm": 0.7022644281387329,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 15792
+    },
+    {
+      "epoch": 1.8141404858997185,
+      "grad_norm": 0.5725907683372498,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 15793
+    },
+    {
+      "epoch": 1.8142553558095456,
+      "grad_norm": 0.6333192586898804,
+      "learning_rate": 0.0001,
+      "loss": 1.3691,
+      "step": 15794
+    },
+    {
+      "epoch": 1.8143702257193728,
+      "grad_norm": 0.6314082145690918,
+      "learning_rate": 0.0001,
+      "loss": 1.5763,
+      "step": 15795
+    },
+    {
+      "epoch": 1.8144850956292,
+      "grad_norm": 0.6020960211753845,
+      "learning_rate": 0.0001,
+      "loss": 1.3315,
+      "step": 15796
+    },
+    {
+      "epoch": 1.814599965539027,
+      "grad_norm": 0.636551022529602,
+      "learning_rate": 0.0001,
+      "loss": 1.2842,
+      "step": 15797
+    },
+    {
+      "epoch": 1.814714835448854,
+      "grad_norm": 0.6058814525604248,
+      "learning_rate": 0.0001,
+      "loss": 1.4964,
+      "step": 15798
+    },
+    {
+      "epoch": 1.8148297053586813,
+      "grad_norm": 0.581272304058075,
+      "learning_rate": 0.0001,
+      "loss": 1.4039,
+      "step": 15799
+    },
+    {
+      "epoch": 1.8149445752685085,
+      "grad_norm": 0.6171589493751526,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 15800
+    },
+    {
+      "epoch": 1.8150594451783355,
+      "grad_norm": 0.6367219090461731,
+      "learning_rate": 0.0001,
+      "loss": 1.3735,
+      "step": 15801
+    },
+    {
+      "epoch": 1.8151743150881625,
+      "grad_norm": 0.6684643030166626,
+      "learning_rate": 0.0001,
+      "loss": 1.7442,
+      "step": 15802
+    },
+    {
+      "epoch": 1.8152891849979897,
+      "grad_norm": 0.5805231332778931,
+      "learning_rate": 0.0001,
+      "loss": 1.4201,
+      "step": 15803
+    },
+    {
+      "epoch": 1.815404054907817,
+      "grad_norm": 0.6159259676933289,
+      "learning_rate": 0.0001,
+      "loss": 1.2066,
+      "step": 15804
+    },
+    {
+      "epoch": 1.815518924817644,
+      "grad_norm": 0.5856978297233582,
+      "learning_rate": 0.0001,
+      "loss": 1.3786,
+      "step": 15805
+    },
+    {
+      "epoch": 1.815633794727471,
+      "grad_norm": 0.6367132663726807,
+      "learning_rate": 0.0001,
+      "loss": 1.3513,
+      "step": 15806
+    },
+    {
+      "epoch": 1.8157486646372982,
+      "grad_norm": 0.6610946655273438,
+      "learning_rate": 0.0001,
+      "loss": 1.4352,
+      "step": 15807
+    },
+    {
+      "epoch": 1.8158635345471255,
+      "grad_norm": 0.6742769479751587,
+      "learning_rate": 0.0001,
+      "loss": 1.4922,
+      "step": 15808
+    },
+    {
+      "epoch": 1.8159784044569525,
+      "grad_norm": 0.6435980200767517,
+      "learning_rate": 0.0001,
+      "loss": 1.5316,
+      "step": 15809
+    },
+    {
+      "epoch": 1.8160932743667795,
+      "grad_norm": 0.6399965882301331,
+      "learning_rate": 0.0001,
+      "loss": 1.4247,
+      "step": 15810
+    },
+    {
+      "epoch": 1.8162081442766067,
+      "grad_norm": 0.6221297979354858,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 15811
+    },
+    {
+      "epoch": 1.816323014186434,
+      "grad_norm": 0.6029366254806519,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 15812
+    },
+    {
+      "epoch": 1.816437884096261,
+      "grad_norm": 0.628516435623169,
+      "learning_rate": 0.0001,
+      "loss": 1.1754,
+      "step": 15813
+    },
+    {
+      "epoch": 1.816552754006088,
+      "grad_norm": 0.7524257898330688,
+      "learning_rate": 0.0001,
+      "loss": 1.6957,
+      "step": 15814
+    },
+    {
+      "epoch": 1.8166676239159152,
+      "grad_norm": 0.6547483205795288,
+      "learning_rate": 0.0001,
+      "loss": 1.3974,
+      "step": 15815
+    },
+    {
+      "epoch": 1.8167824938257424,
+      "grad_norm": 0.7047604918479919,
+      "learning_rate": 0.0001,
+      "loss": 1.3708,
+      "step": 15816
+    },
+    {
+      "epoch": 1.8168973637355694,
+      "grad_norm": 0.6523260474205017,
+      "learning_rate": 0.0001,
+      "loss": 1.4342,
+      "step": 15817
+    },
+    {
+      "epoch": 1.8170122336453964,
+      "grad_norm": 0.6819579005241394,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 15818
+    },
+    {
+      "epoch": 1.8171271035552237,
+      "grad_norm": 0.6569975018501282,
+      "learning_rate": 0.0001,
+      "loss": 1.3422,
+      "step": 15819
+    },
+    {
+      "epoch": 1.817241973465051,
+      "grad_norm": 0.6896679401397705,
+      "learning_rate": 0.0001,
+      "loss": 1.4424,
+      "step": 15820
+    },
+    {
+      "epoch": 1.817356843374878,
+      "grad_norm": 0.6512749195098877,
+      "learning_rate": 0.0001,
+      "loss": 1.3904,
+      "step": 15821
+    },
+    {
+      "epoch": 1.817471713284705,
+      "grad_norm": 0.6123222708702087,
+      "learning_rate": 0.0001,
+      "loss": 1.3856,
+      "step": 15822
+    },
+    {
+      "epoch": 1.8175865831945321,
+      "grad_norm": 0.5971189141273499,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 15823
+    },
+    {
+      "epoch": 1.8177014531043594,
+      "grad_norm": 0.5992557406425476,
+      "learning_rate": 0.0001,
+      "loss": 1.4017,
+      "step": 15824
+    },
+    {
+      "epoch": 1.8178163230141864,
+      "grad_norm": 0.6497430205345154,
+      "learning_rate": 0.0001,
+      "loss": 1.4175,
+      "step": 15825
+    },
+    {
+      "epoch": 1.8179311929240134,
+      "grad_norm": 0.6796419620513916,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 15826
+    },
+    {
+      "epoch": 1.8180460628338406,
+      "grad_norm": 0.6003844738006592,
+      "learning_rate": 0.0001,
+      "loss": 1.3818,
+      "step": 15827
+    },
+    {
+      "epoch": 1.8181609327436679,
+      "grad_norm": 0.7161847949028015,
+      "learning_rate": 0.0001,
+      "loss": 1.4173,
+      "step": 15828
+    },
+    {
+      "epoch": 1.8182758026534949,
+      "grad_norm": 0.6299294829368591,
+      "learning_rate": 0.0001,
+      "loss": 1.4459,
+      "step": 15829
+    },
+    {
+      "epoch": 1.8183906725633219,
+      "grad_norm": 0.7790576815605164,
+      "learning_rate": 0.0001,
+      "loss": 1.7595,
+      "step": 15830
+    },
+    {
+      "epoch": 1.818505542473149,
+      "grad_norm": 0.5734109282493591,
+      "learning_rate": 0.0001,
+      "loss": 1.3666,
+      "step": 15831
+    },
+    {
+      "epoch": 1.8186204123829763,
+      "grad_norm": 0.6383972764015198,
+      "learning_rate": 0.0001,
+      "loss": 1.6129,
+      "step": 15832
+    },
+    {
+      "epoch": 1.8187352822928033,
+      "grad_norm": 0.6514372825622559,
+      "learning_rate": 0.0001,
+      "loss": 1.4179,
+      "step": 15833
+    },
+    {
+      "epoch": 1.8188501522026304,
+      "grad_norm": 0.6539475321769714,
+      "learning_rate": 0.0001,
+      "loss": 1.3834,
+      "step": 15834
+    },
+    {
+      "epoch": 1.8189650221124576,
+      "grad_norm": 0.6508055925369263,
+      "learning_rate": 0.0001,
+      "loss": 1.1476,
+      "step": 15835
+    },
+    {
+      "epoch": 1.8190798920222848,
+      "grad_norm": 0.6467891335487366,
+      "learning_rate": 0.0001,
+      "loss": 1.4836,
+      "step": 15836
+    },
+    {
+      "epoch": 1.819194761932112,
+      "grad_norm": 0.6416442394256592,
+      "learning_rate": 0.0001,
+      "loss": 1.4443,
+      "step": 15837
+    },
+    {
+      "epoch": 1.819309631841939,
+      "grad_norm": 0.6353035569190979,
+      "learning_rate": 0.0001,
+      "loss": 1.3113,
+      "step": 15838
+    },
+    {
+      "epoch": 1.819424501751766,
+      "grad_norm": 0.6308345198631287,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 15839
+    },
+    {
+      "epoch": 1.8195393716615933,
+      "grad_norm": 0.66878741979599,
+      "learning_rate": 0.0001,
+      "loss": 1.5101,
+      "step": 15840
+    },
+    {
+      "epoch": 1.8196542415714205,
+      "grad_norm": 0.6357544660568237,
+      "learning_rate": 0.0001,
+      "loss": 1.4058,
+      "step": 15841
+    },
+    {
+      "epoch": 1.8197691114812475,
+      "grad_norm": 0.6938048601150513,
+      "learning_rate": 0.0001,
+      "loss": 1.3966,
+      "step": 15842
+    },
+    {
+      "epoch": 1.8198839813910745,
+      "grad_norm": 0.6165516376495361,
+      "learning_rate": 0.0001,
+      "loss": 1.392,
+      "step": 15843
+    },
+    {
+      "epoch": 1.8199988513009018,
+      "grad_norm": 0.6373364329338074,
+      "learning_rate": 0.0001,
+      "loss": 1.2863,
+      "step": 15844
+    },
+    {
+      "epoch": 1.820113721210729,
+      "grad_norm": 0.6185095310211182,
+      "learning_rate": 0.0001,
+      "loss": 1.3833,
+      "step": 15845
+    },
+    {
+      "epoch": 1.820228591120556,
+      "grad_norm": 0.6635664105415344,
+      "learning_rate": 0.0001,
+      "loss": 1.3862,
+      "step": 15846
+    },
+    {
+      "epoch": 1.820343461030383,
+      "grad_norm": 0.6329711675643921,
+      "learning_rate": 0.0001,
+      "loss": 1.3956,
+      "step": 15847
+    },
+    {
+      "epoch": 1.8204583309402103,
+      "grad_norm": 0.6014331579208374,
+      "learning_rate": 0.0001,
+      "loss": 1.41,
+      "step": 15848
+    },
+    {
+      "epoch": 1.8205732008500375,
+      "grad_norm": 0.6105222702026367,
+      "learning_rate": 0.0001,
+      "loss": 1.234,
+      "step": 15849
+    },
+    {
+      "epoch": 1.8206880707598645,
+      "grad_norm": 0.648215651512146,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 15850
+    },
+    {
+      "epoch": 1.8208029406696915,
+      "grad_norm": 0.5845022201538086,
+      "learning_rate": 0.0001,
+      "loss": 1.141,
+      "step": 15851
+    },
+    {
+      "epoch": 1.8209178105795187,
+      "grad_norm": 0.6215102076530457,
+      "learning_rate": 0.0001,
+      "loss": 1.472,
+      "step": 15852
+    },
+    {
+      "epoch": 1.821032680489346,
+      "grad_norm": 0.6202508807182312,
+      "learning_rate": 0.0001,
+      "loss": 1.3187,
+      "step": 15853
+    },
+    {
+      "epoch": 1.821147550399173,
+      "grad_norm": 0.638676106929779,
+      "learning_rate": 0.0001,
+      "loss": 1.3897,
+      "step": 15854
+    },
+    {
+      "epoch": 1.821262420309,
+      "grad_norm": 0.7090446352958679,
+      "learning_rate": 0.0001,
+      "loss": 1.4151,
+      "step": 15855
+    },
+    {
+      "epoch": 1.8213772902188272,
+      "grad_norm": 0.650833249092102,
+      "learning_rate": 0.0001,
+      "loss": 1.4007,
+      "step": 15856
+    },
+    {
+      "epoch": 1.8214921601286544,
+      "grad_norm": 0.7622561454772949,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 15857
+    },
+    {
+      "epoch": 1.8216070300384815,
+      "grad_norm": 0.6946950554847717,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 15858
+    },
+    {
+      "epoch": 1.8217218999483085,
+      "grad_norm": 0.6929587721824646,
+      "learning_rate": 0.0001,
+      "loss": 1.3965,
+      "step": 15859
+    },
+    {
+      "epoch": 1.8218367698581357,
+      "grad_norm": 0.6364197731018066,
+      "learning_rate": 0.0001,
+      "loss": 1.4878,
+      "step": 15860
+    },
+    {
+      "epoch": 1.821951639767963,
+      "grad_norm": 0.7382113933563232,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 15861
+    },
+    {
+      "epoch": 1.82206650967779,
+      "grad_norm": 0.6655716896057129,
+      "learning_rate": 0.0001,
+      "loss": 1.4976,
+      "step": 15862
+    },
+    {
+      "epoch": 1.822181379587617,
+      "grad_norm": 0.597642719745636,
+      "learning_rate": 0.0001,
+      "loss": 1.3425,
+      "step": 15863
+    },
+    {
+      "epoch": 1.8222962494974442,
+      "grad_norm": 0.6217376589775085,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 15864
+    },
+    {
+      "epoch": 1.8224111194072714,
+      "grad_norm": 0.6502073407173157,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 15865
+    },
+    {
+      "epoch": 1.8225259893170984,
+      "grad_norm": 0.6081299185752869,
+      "learning_rate": 0.0001,
+      "loss": 1.3242,
+      "step": 15866
+    },
+    {
+      "epoch": 1.8226408592269254,
+      "grad_norm": 0.6899523735046387,
+      "learning_rate": 0.0001,
+      "loss": 1.093,
+      "step": 15867
+    },
+    {
+      "epoch": 1.8227557291367527,
+      "grad_norm": 0.6164363622665405,
+      "learning_rate": 0.0001,
+      "loss": 1.5016,
+      "step": 15868
+    },
+    {
+      "epoch": 1.8228705990465799,
+      "grad_norm": 0.6042718887329102,
+      "learning_rate": 0.0001,
+      "loss": 1.4358,
+      "step": 15869
+    },
+    {
+      "epoch": 1.822985468956407,
+      "grad_norm": 0.6475842595100403,
+      "learning_rate": 0.0001,
+      "loss": 1.595,
+      "step": 15870
+    },
+    {
+      "epoch": 1.823100338866234,
+      "grad_norm": 0.6268951892852783,
+      "learning_rate": 0.0001,
+      "loss": 1.4257,
+      "step": 15871
+    },
+    {
+      "epoch": 1.8232152087760611,
+      "grad_norm": 0.6121448874473572,
+      "learning_rate": 0.0001,
+      "loss": 1.4889,
+      "step": 15872
+    },
+    {
+      "epoch": 1.8233300786858884,
+      "grad_norm": 0.6150333881378174,
+      "learning_rate": 0.0001,
+      "loss": 1.3803,
+      "step": 15873
+    },
+    {
+      "epoch": 1.8234449485957154,
+      "grad_norm": 0.6406527757644653,
+      "learning_rate": 0.0001,
+      "loss": 1.3256,
+      "step": 15874
+    },
+    {
+      "epoch": 1.8235598185055424,
+      "grad_norm": 0.6126410961151123,
+      "learning_rate": 0.0001,
+      "loss": 1.3716,
+      "step": 15875
+    },
+    {
+      "epoch": 1.8236746884153696,
+      "grad_norm": 0.6979601383209229,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 15876
+    },
+    {
+      "epoch": 1.8237895583251968,
+      "grad_norm": 0.6526153087615967,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 15877
+    },
+    {
+      "epoch": 1.8239044282350239,
+      "grad_norm": 0.6208426356315613,
+      "learning_rate": 0.0001,
+      "loss": 1.3515,
+      "step": 15878
+    },
+    {
+      "epoch": 1.8240192981448509,
+      "grad_norm": 0.5797458291053772,
+      "learning_rate": 0.0001,
+      "loss": 1.393,
+      "step": 15879
+    },
+    {
+      "epoch": 1.824134168054678,
+      "grad_norm": 0.5991585850715637,
+      "learning_rate": 0.0001,
+      "loss": 1.4375,
+      "step": 15880
+    },
+    {
+      "epoch": 1.8242490379645053,
+      "grad_norm": 0.6601519584655762,
+      "learning_rate": 0.0001,
+      "loss": 1.5897,
+      "step": 15881
+    },
+    {
+      "epoch": 1.8243639078743323,
+      "grad_norm": 0.6615204811096191,
+      "learning_rate": 0.0001,
+      "loss": 1.5966,
+      "step": 15882
+    },
+    {
+      "epoch": 1.8244787777841593,
+      "grad_norm": 0.6218351125717163,
+      "learning_rate": 0.0001,
+      "loss": 1.478,
+      "step": 15883
+    },
+    {
+      "epoch": 1.8245936476939866,
+      "grad_norm": 0.6256325244903564,
+      "learning_rate": 0.0001,
+      "loss": 1.6471,
+      "step": 15884
+    },
+    {
+      "epoch": 1.8247085176038138,
+      "grad_norm": 0.6107257008552551,
+      "learning_rate": 0.0001,
+      "loss": 1.4236,
+      "step": 15885
+    },
+    {
+      "epoch": 1.8248233875136408,
+      "grad_norm": 0.7533160448074341,
+      "learning_rate": 0.0001,
+      "loss": 1.425,
+      "step": 15886
+    },
+    {
+      "epoch": 1.8249382574234678,
+      "grad_norm": 0.6379408836364746,
+      "learning_rate": 0.0001,
+      "loss": 1.363,
+      "step": 15887
+    },
+    {
+      "epoch": 1.825053127333295,
+      "grad_norm": 0.598540723323822,
+      "learning_rate": 0.0001,
+      "loss": 1.4424,
+      "step": 15888
+    },
+    {
+      "epoch": 1.8251679972431223,
+      "grad_norm": 0.6070877313613892,
+      "learning_rate": 0.0001,
+      "loss": 1.3913,
+      "step": 15889
+    },
+    {
+      "epoch": 1.8252828671529493,
+      "grad_norm": 0.6327671408653259,
+      "learning_rate": 0.0001,
+      "loss": 1.4302,
+      "step": 15890
+    },
+    {
+      "epoch": 1.8253977370627763,
+      "grad_norm": 0.6539307236671448,
+      "learning_rate": 0.0001,
+      "loss": 1.4466,
+      "step": 15891
+    },
+    {
+      "epoch": 1.8255126069726035,
+      "grad_norm": 0.6575409173965454,
+      "learning_rate": 0.0001,
+      "loss": 1.2643,
+      "step": 15892
+    },
+    {
+      "epoch": 1.8256274768824308,
+      "grad_norm": 0.6441643834114075,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 15893
+    },
+    {
+      "epoch": 1.8257423467922578,
+      "grad_norm": 0.6457522511482239,
+      "learning_rate": 0.0001,
+      "loss": 1.3737,
+      "step": 15894
+    },
+    {
+      "epoch": 1.8258572167020848,
+      "grad_norm": 0.6049162149429321,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 15895
+    },
+    {
+      "epoch": 1.825972086611912,
+      "grad_norm": 0.6564713716506958,
+      "learning_rate": 0.0001,
+      "loss": 1.3928,
+      "step": 15896
+    },
+    {
+      "epoch": 1.8260869565217392,
+      "grad_norm": 0.6829112768173218,
+      "learning_rate": 0.0001,
+      "loss": 1.5156,
+      "step": 15897
+    },
+    {
+      "epoch": 1.8262018264315663,
+      "grad_norm": 0.6410605311393738,
+      "learning_rate": 0.0001,
+      "loss": 1.3843,
+      "step": 15898
+    },
+    {
+      "epoch": 1.8263166963413933,
+      "grad_norm": 0.6501928567886353,
+      "learning_rate": 0.0001,
+      "loss": 1.4678,
+      "step": 15899
+    },
+    {
+      "epoch": 1.8264315662512205,
+      "grad_norm": 0.6255066990852356,
+      "learning_rate": 0.0001,
+      "loss": 1.4146,
+      "step": 15900
+    },
+    {
+      "epoch": 1.8265464361610477,
+      "grad_norm": 0.6358596086502075,
+      "learning_rate": 0.0001,
+      "loss": 1.462,
+      "step": 15901
+    },
+    {
+      "epoch": 1.8266613060708747,
+      "grad_norm": 0.6268904805183411,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 15902
+    },
+    {
+      "epoch": 1.8267761759807017,
+      "grad_norm": 0.5953584909439087,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 15903
+    },
+    {
+      "epoch": 1.826891045890529,
+      "grad_norm": 0.5962085127830505,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 15904
+    },
+    {
+      "epoch": 1.8270059158003562,
+      "grad_norm": 0.6590402722358704,
+      "learning_rate": 0.0001,
+      "loss": 1.4404,
+      "step": 15905
+    },
+    {
+      "epoch": 1.8271207857101832,
+      "grad_norm": 0.5863921642303467,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 15906
+    },
+    {
+      "epoch": 1.8272356556200102,
+      "grad_norm": 0.5984483361244202,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 15907
+    },
+    {
+      "epoch": 1.8273505255298375,
+      "grad_norm": 0.6065948009490967,
+      "learning_rate": 0.0001,
+      "loss": 1.2615,
+      "step": 15908
+    },
+    {
+      "epoch": 1.8274653954396647,
+      "grad_norm": 0.5679669976234436,
+      "learning_rate": 0.0001,
+      "loss": 1.4875,
+      "step": 15909
+    },
+    {
+      "epoch": 1.8275802653494917,
+      "grad_norm": 0.5973514318466187,
+      "learning_rate": 0.0001,
+      "loss": 1.3,
+      "step": 15910
+    },
+    {
+      "epoch": 1.8276951352593187,
+      "grad_norm": 0.644055962562561,
+      "learning_rate": 0.0001,
+      "loss": 1.4505,
+      "step": 15911
+    },
+    {
+      "epoch": 1.827810005169146,
+      "grad_norm": 0.6656803488731384,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 15912
+    },
+    {
+      "epoch": 1.8279248750789732,
+      "grad_norm": 0.6298542022705078,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 15913
+    },
+    {
+      "epoch": 1.8280397449888002,
+      "grad_norm": 0.649831235408783,
+      "learning_rate": 0.0001,
+      "loss": 1.4071,
+      "step": 15914
+    },
+    {
+      "epoch": 1.8281546148986272,
+      "grad_norm": 0.675029456615448,
+      "learning_rate": 0.0001,
+      "loss": 1.5452,
+      "step": 15915
+    },
+    {
+      "epoch": 1.8282694848084544,
+      "grad_norm": 0.5778508186340332,
+      "learning_rate": 0.0001,
+      "loss": 1.365,
+      "step": 15916
+    },
+    {
+      "epoch": 1.8283843547182816,
+      "grad_norm": 0.6754083633422852,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 15917
+    },
+    {
+      "epoch": 1.8284992246281087,
+      "grad_norm": 0.6041424870491028,
+      "learning_rate": 0.0001,
+      "loss": 1.4145,
+      "step": 15918
+    },
+    {
+      "epoch": 1.8286140945379357,
+      "grad_norm": 0.6053421497344971,
+      "learning_rate": 0.0001,
+      "loss": 1.3666,
+      "step": 15919
+    },
+    {
+      "epoch": 1.828728964447763,
+      "grad_norm": 0.5872570276260376,
+      "learning_rate": 0.0001,
+      "loss": 1.4642,
+      "step": 15920
+    },
+    {
+      "epoch": 1.8288438343575901,
+      "grad_norm": 0.663306474685669,
+      "learning_rate": 0.0001,
+      "loss": 1.5113,
+      "step": 15921
+    },
+    {
+      "epoch": 1.8289587042674171,
+      "grad_norm": 0.6472460627555847,
+      "learning_rate": 0.0001,
+      "loss": 1.4313,
+      "step": 15922
+    },
+    {
+      "epoch": 1.8290735741772441,
+      "grad_norm": 0.6650756001472473,
+      "learning_rate": 0.0001,
+      "loss": 1.4514,
+      "step": 15923
+    },
+    {
+      "epoch": 1.8291884440870714,
+      "grad_norm": 0.5926016569137573,
+      "learning_rate": 0.0001,
+      "loss": 1.4217,
+      "step": 15924
+    },
+    {
+      "epoch": 1.8293033139968986,
+      "grad_norm": 0.7057727575302124,
+      "learning_rate": 0.0001,
+      "loss": 1.4764,
+      "step": 15925
+    },
+    {
+      "epoch": 1.8294181839067256,
+      "grad_norm": 0.6436575055122375,
+      "learning_rate": 0.0001,
+      "loss": 1.4334,
+      "step": 15926
+    },
+    {
+      "epoch": 1.8295330538165526,
+      "grad_norm": 0.6020593643188477,
+      "learning_rate": 0.0001,
+      "loss": 1.4711,
+      "step": 15927
+    },
+    {
+      "epoch": 1.8296479237263799,
+      "grad_norm": 0.5867114663124084,
+      "learning_rate": 0.0001,
+      "loss": 1.3176,
+      "step": 15928
+    },
+    {
+      "epoch": 1.829762793636207,
+      "grad_norm": 0.6493946313858032,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 15929
+    },
+    {
+      "epoch": 1.829877663546034,
+      "grad_norm": 0.6781548261642456,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 15930
+    },
+    {
+      "epoch": 1.829992533455861,
+      "grad_norm": 0.5895503759384155,
+      "learning_rate": 0.0001,
+      "loss": 1.2183,
+      "step": 15931
+    },
+    {
+      "epoch": 1.8301074033656883,
+      "grad_norm": 0.7099341750144958,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 15932
+    },
+    {
+      "epoch": 1.8302222732755156,
+      "grad_norm": 0.7078137397766113,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 15933
+    },
+    {
+      "epoch": 1.8303371431853426,
+      "grad_norm": 0.6459853649139404,
+      "learning_rate": 0.0001,
+      "loss": 1.5702,
+      "step": 15934
+    },
+    {
+      "epoch": 1.8304520130951696,
+      "grad_norm": 0.6130335330963135,
+      "learning_rate": 0.0001,
+      "loss": 1.3144,
+      "step": 15935
+    },
+    {
+      "epoch": 1.8305668830049968,
+      "grad_norm": 0.6197957396507263,
+      "learning_rate": 0.0001,
+      "loss": 1.4387,
+      "step": 15936
+    },
+    {
+      "epoch": 1.830681752914824,
+      "grad_norm": 0.667946457862854,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 15937
+    },
+    {
+      "epoch": 1.830796622824651,
+      "grad_norm": 0.5950310230255127,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 15938
+    },
+    {
+      "epoch": 1.830911492734478,
+      "grad_norm": 0.632938802242279,
+      "learning_rate": 0.0001,
+      "loss": 1.4152,
+      "step": 15939
+    },
+    {
+      "epoch": 1.8310263626443053,
+      "grad_norm": 0.5838363766670227,
+      "learning_rate": 0.0001,
+      "loss": 1.4048,
+      "step": 15940
+    },
+    {
+      "epoch": 1.8311412325541325,
+      "grad_norm": 0.7285007834434509,
+      "learning_rate": 0.0001,
+      "loss": 1.5162,
+      "step": 15941
+    },
+    {
+      "epoch": 1.8312561024639595,
+      "grad_norm": 0.6085467338562012,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 15942
+    },
+    {
+      "epoch": 1.8313709723737865,
+      "grad_norm": 0.6019995808601379,
+      "learning_rate": 0.0001,
+      "loss": 1.438,
+      "step": 15943
+    },
+    {
+      "epoch": 1.8314858422836138,
+      "grad_norm": 0.6792658567428589,
+      "learning_rate": 0.0001,
+      "loss": 1.6684,
+      "step": 15944
+    },
+    {
+      "epoch": 1.831600712193441,
+      "grad_norm": 0.602642834186554,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 15945
+    },
+    {
+      "epoch": 1.831715582103268,
+      "grad_norm": 0.6296089887619019,
+      "learning_rate": 0.0001,
+      "loss": 1.4179,
+      "step": 15946
+    },
+    {
+      "epoch": 1.831830452013095,
+      "grad_norm": 0.5773423314094543,
+      "learning_rate": 0.0001,
+      "loss": 1.5549,
+      "step": 15947
+    },
+    {
+      "epoch": 1.8319453219229223,
+      "grad_norm": 0.6150087714195251,
+      "learning_rate": 0.0001,
+      "loss": 1.4553,
+      "step": 15948
+    },
+    {
+      "epoch": 1.8320601918327495,
+      "grad_norm": 0.6208438277244568,
+      "learning_rate": 0.0001,
+      "loss": 1.4832,
+      "step": 15949
+    },
+    {
+      "epoch": 1.8321750617425765,
+      "grad_norm": 0.6598103046417236,
+      "learning_rate": 0.0001,
+      "loss": 1.3849,
+      "step": 15950
+    },
+    {
+      "epoch": 1.8322899316524035,
+      "grad_norm": 0.5817321538925171,
+      "learning_rate": 0.0001,
+      "loss": 1.2799,
+      "step": 15951
+    },
+    {
+      "epoch": 1.8324048015622307,
+      "grad_norm": 0.6446582078933716,
+      "learning_rate": 0.0001,
+      "loss": 1.4087,
+      "step": 15952
+    },
+    {
+      "epoch": 1.832519671472058,
+      "grad_norm": 0.5805476903915405,
+      "learning_rate": 0.0001,
+      "loss": 1.2953,
+      "step": 15953
+    },
+    {
+      "epoch": 1.832634541381885,
+      "grad_norm": 0.6229280233383179,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 15954
+    },
+    {
+      "epoch": 1.832749411291712,
+      "grad_norm": 0.6367110013961792,
+      "learning_rate": 0.0001,
+      "loss": 1.4145,
+      "step": 15955
+    },
+    {
+      "epoch": 1.8328642812015392,
+      "grad_norm": 0.6071845889091492,
+      "learning_rate": 0.0001,
+      "loss": 1.2801,
+      "step": 15956
+    },
+    {
+      "epoch": 1.8329791511113664,
+      "grad_norm": 0.6400860548019409,
+      "learning_rate": 0.0001,
+      "loss": 1.4126,
+      "step": 15957
+    },
+    {
+      "epoch": 1.8330940210211935,
+      "grad_norm": 0.6502026915550232,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 15958
+    },
+    {
+      "epoch": 1.8332088909310205,
+      "grad_norm": 0.5755997896194458,
+      "learning_rate": 0.0001,
+      "loss": 1.3486,
+      "step": 15959
+    },
+    {
+      "epoch": 1.8333237608408477,
+      "grad_norm": 0.6484699845314026,
+      "learning_rate": 0.0001,
+      "loss": 1.5004,
+      "step": 15960
+    },
+    {
+      "epoch": 1.833438630750675,
+      "grad_norm": 0.6342812180519104,
+      "learning_rate": 0.0001,
+      "loss": 1.3965,
+      "step": 15961
+    },
+    {
+      "epoch": 1.833553500660502,
+      "grad_norm": 0.676463782787323,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 15962
+    },
+    {
+      "epoch": 1.833668370570329,
+      "grad_norm": 0.6405801773071289,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 15963
+    },
+    {
+      "epoch": 1.8337832404801562,
+      "grad_norm": 0.614596426486969,
+      "learning_rate": 0.0001,
+      "loss": 1.3663,
+      "step": 15964
+    },
+    {
+      "epoch": 1.8338981103899834,
+      "grad_norm": 0.6203700304031372,
+      "learning_rate": 0.0001,
+      "loss": 1.4567,
+      "step": 15965
+    },
+    {
+      "epoch": 1.8340129802998104,
+      "grad_norm": 0.6081563234329224,
+      "learning_rate": 0.0001,
+      "loss": 1.5254,
+      "step": 15966
+    },
+    {
+      "epoch": 1.8341278502096374,
+      "grad_norm": 0.6504876017570496,
+      "learning_rate": 0.0001,
+      "loss": 1.4266,
+      "step": 15967
+    },
+    {
+      "epoch": 1.8342427201194647,
+      "grad_norm": 0.6089107990264893,
+      "learning_rate": 0.0001,
+      "loss": 1.4368,
+      "step": 15968
+    },
+    {
+      "epoch": 1.8343575900292919,
+      "grad_norm": 0.696449875831604,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 15969
+    },
+    {
+      "epoch": 1.834472459939119,
+      "grad_norm": 0.6738762259483337,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 15970
+    },
+    {
+      "epoch": 1.834587329848946,
+      "grad_norm": 0.6419105529785156,
+      "learning_rate": 0.0001,
+      "loss": 1.465,
+      "step": 15971
+    },
+    {
+      "epoch": 1.8347021997587731,
+      "grad_norm": 0.6349779963493347,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 15972
+    },
+    {
+      "epoch": 1.8348170696686004,
+      "grad_norm": 0.6129407286643982,
+      "learning_rate": 0.0001,
+      "loss": 1.3928,
+      "step": 15973
+    },
+    {
+      "epoch": 1.8349319395784276,
+      "grad_norm": 0.6540585160255432,
+      "learning_rate": 0.0001,
+      "loss": 1.701,
+      "step": 15974
+    },
+    {
+      "epoch": 1.8350468094882546,
+      "grad_norm": 0.6090661287307739,
+      "learning_rate": 0.0001,
+      "loss": 1.4129,
+      "step": 15975
+    },
+    {
+      "epoch": 1.8351616793980816,
+      "grad_norm": 0.6051062345504761,
+      "learning_rate": 0.0001,
+      "loss": 1.4449,
+      "step": 15976
+    },
+    {
+      "epoch": 1.8352765493079088,
+      "grad_norm": 0.6473617553710938,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 15977
+    },
+    {
+      "epoch": 1.835391419217736,
+      "grad_norm": 0.6278075575828552,
+      "learning_rate": 0.0001,
+      "loss": 1.3918,
+      "step": 15978
+    },
+    {
+      "epoch": 1.835506289127563,
+      "grad_norm": 0.6662875413894653,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 15979
+    },
+    {
+      "epoch": 1.83562115903739,
+      "grad_norm": 0.6201448440551758,
+      "learning_rate": 0.0001,
+      "loss": 1.4804,
+      "step": 15980
+    },
+    {
+      "epoch": 1.8357360289472173,
+      "grad_norm": 0.6352674961090088,
+      "learning_rate": 0.0001,
+      "loss": 1.4066,
+      "step": 15981
+    },
+    {
+      "epoch": 1.8358508988570446,
+      "grad_norm": 0.5949738621711731,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 15982
+    },
+    {
+      "epoch": 1.8359657687668716,
+      "grad_norm": 0.6130930781364441,
+      "learning_rate": 0.0001,
+      "loss": 1.4186,
+      "step": 15983
+    },
+    {
+      "epoch": 1.8360806386766986,
+      "grad_norm": 0.5820370316505432,
+      "learning_rate": 0.0001,
+      "loss": 1.223,
+      "step": 15984
+    },
+    {
+      "epoch": 1.8361955085865258,
+      "grad_norm": 0.6435645222663879,
+      "learning_rate": 0.0001,
+      "loss": 1.4747,
+      "step": 15985
+    },
+    {
+      "epoch": 1.836310378496353,
+      "grad_norm": 0.5986513495445251,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 15986
+    },
+    {
+      "epoch": 1.83642524840618,
+      "grad_norm": 0.6947880983352661,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 15987
+    },
+    {
+      "epoch": 1.836540118316007,
+      "grad_norm": 0.6280114054679871,
+      "learning_rate": 0.0001,
+      "loss": 1.3834,
+      "step": 15988
+    },
+    {
+      "epoch": 1.8366549882258343,
+      "grad_norm": 0.669037401676178,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 15989
+    },
+    {
+      "epoch": 1.8367698581356615,
+      "grad_norm": 0.7081412076950073,
+      "learning_rate": 0.0001,
+      "loss": 1.4418,
+      "step": 15990
+    },
+    {
+      "epoch": 1.8368847280454885,
+      "grad_norm": 0.6335996985435486,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 15991
+    },
+    {
+      "epoch": 1.8369995979553155,
+      "grad_norm": 0.68028724193573,
+      "learning_rate": 0.0001,
+      "loss": 1.4157,
+      "step": 15992
+    },
+    {
+      "epoch": 1.8371144678651428,
+      "grad_norm": 0.6293099522590637,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 15993
+    },
+    {
+      "epoch": 1.83722933777497,
+      "grad_norm": 0.6450070738792419,
+      "learning_rate": 0.0001,
+      "loss": 1.3165,
+      "step": 15994
+    },
+    {
+      "epoch": 1.837344207684797,
+      "grad_norm": 0.6043501496315002,
+      "learning_rate": 0.0001,
+      "loss": 1.3544,
+      "step": 15995
+    },
+    {
+      "epoch": 1.837459077594624,
+      "grad_norm": 0.5740723013877869,
+      "learning_rate": 0.0001,
+      "loss": 1.3516,
+      "step": 15996
+    },
+    {
+      "epoch": 1.8375739475044512,
+      "grad_norm": 0.6358054876327515,
+      "learning_rate": 0.0001,
+      "loss": 1.3855,
+      "step": 15997
+    },
+    {
+      "epoch": 1.8376888174142785,
+      "grad_norm": 0.5969364047050476,
+      "learning_rate": 0.0001,
+      "loss": 1.3998,
+      "step": 15998
+    },
+    {
+      "epoch": 1.8378036873241055,
+      "grad_norm": 0.6494911909103394,
+      "learning_rate": 0.0001,
+      "loss": 1.4437,
+      "step": 15999
+    },
+    {
+      "epoch": 1.8379185572339325,
+      "grad_norm": 0.6509250402450562,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 16000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 17410,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.4136217259933696e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}