diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24010 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9994890883877088,
+  "eval_steps": 500,
+  "global_step": 3424,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005838989854755127,
+      "grad_norm": 0.34797629218713766,
+      "learning_rate": 5.830903790087464e-08,
+      "loss": 0.7178,
+      "step": 1
+    },
+    {
+      "epoch": 0.0011677979709510255,
+      "grad_norm": 0.37939237582162205,
+      "learning_rate": 1.1661807580174928e-07,
+      "loss": 0.7403,
+      "step": 2
+    },
+    {
+      "epoch": 0.0017516969564265383,
+      "grad_norm": 0.3211930616851675,
+      "learning_rate": 1.7492711370262392e-07,
+      "loss": 0.6783,
+      "step": 3
+    },
+    {
+      "epoch": 0.002335595941902051,
+      "grad_norm": 0.3673418516405313,
+      "learning_rate": 2.3323615160349856e-07,
+      "loss": 0.7402,
+      "step": 4
+    },
+    {
+      "epoch": 0.0029194949273775635,
+      "grad_norm": 0.3405267268207614,
+      "learning_rate": 2.915451895043732e-07,
+      "loss": 0.7398,
+      "step": 5
+    },
+    {
+      "epoch": 0.0035033939128530766,
+      "grad_norm": 0.3397025256855228,
+      "learning_rate": 3.4985422740524783e-07,
+      "loss": 0.6877,
+      "step": 6
+    },
+    {
+      "epoch": 0.004087292898328589,
+      "grad_norm": 0.3472379235979867,
+      "learning_rate": 4.0816326530612243e-07,
+      "loss": 0.7454,
+      "step": 7
+    },
+    {
+      "epoch": 0.004671191883804102,
+      "grad_norm": 0.3561049865343007,
+      "learning_rate": 4.6647230320699713e-07,
+      "loss": 0.7248,
+      "step": 8
+    },
+    {
+      "epoch": 0.005255090869279615,
+      "grad_norm": 0.3611187777029013,
+      "learning_rate": 5.247813411078718e-07,
+      "loss": 0.7525,
+      "step": 9
+    },
+    {
+      "epoch": 0.005838989854755127,
+      "grad_norm": 0.32830328068769976,
+      "learning_rate": 5.830903790087464e-07,
+      "loss": 0.7088,
+      "step": 10
+    },
+    {
+      "epoch": 0.00642288884023064,
+      "grad_norm": 0.3774591510787734,
+      "learning_rate": 6.413994169096211e-07,
+      "loss": 0.7475,
+      "step": 11
+    },
+    {
+      "epoch": 0.007006787825706153,
+      "grad_norm": 0.32961983453550053,
+      "learning_rate": 6.997084548104957e-07,
+      "loss": 0.694,
+      "step": 12
+    },
+    {
+      "epoch": 0.007590686811181665,
+      "grad_norm": 0.29274600262109074,
+      "learning_rate": 7.580174927113704e-07,
+      "loss": 0.6624,
+      "step": 13
+    },
+    {
+      "epoch": 0.008174585796657178,
+      "grad_norm": 0.3148250084765494,
+      "learning_rate": 8.163265306122449e-07,
+      "loss": 0.742,
+      "step": 14
+    },
+    {
+      "epoch": 0.00875848478213269,
+      "grad_norm": 0.26219907486346394,
+      "learning_rate": 8.746355685131196e-07,
+      "loss": 0.6642,
+      "step": 15
+    },
+    {
+      "epoch": 0.009342383767608204,
+      "grad_norm": 0.2839610603142151,
+      "learning_rate": 9.329446064139943e-07,
+      "loss": 0.6702,
+      "step": 16
+    },
+    {
+      "epoch": 0.009926282753083717,
+      "grad_norm": 0.24426962984128664,
+      "learning_rate": 9.91253644314869e-07,
+      "loss": 0.6499,
+      "step": 17
+    },
+    {
+      "epoch": 0.01051018173855923,
+      "grad_norm": 0.2657832681030622,
+      "learning_rate": 1.0495626822157436e-06,
+      "loss": 0.6864,
+      "step": 18
+    },
+    {
+      "epoch": 0.011094080724034743,
+      "grad_norm": 0.2068277784803581,
+      "learning_rate": 1.1078717201166181e-06,
+      "loss": 0.6689,
+      "step": 19
+    },
+    {
+      "epoch": 0.011677979709510254,
+      "grad_norm": 0.17941824078321372,
+      "learning_rate": 1.1661807580174927e-06,
+      "loss": 0.6111,
+      "step": 20
+    },
+    {
+      "epoch": 0.012261878694985767,
+      "grad_norm": 0.1856709063950481,
+      "learning_rate": 1.2244897959183673e-06,
+      "loss": 0.6756,
+      "step": 21
+    },
+    {
+      "epoch": 0.01284577768046128,
+      "grad_norm": 0.17392965245871514,
+      "learning_rate": 1.2827988338192421e-06,
+      "loss": 0.6544,
+      "step": 22
+    },
+    {
+      "epoch": 0.013429676665936793,
+      "grad_norm": 0.1774155941728696,
+      "learning_rate": 1.3411078717201167e-06,
+      "loss": 0.6617,
+      "step": 23
+    },
+    {
+      "epoch": 0.014013575651412306,
+      "grad_norm": 0.1721947000290894,
+      "learning_rate": 1.3994169096209913e-06,
+      "loss": 0.6294,
+      "step": 24
+    },
+    {
+      "epoch": 0.014597474636887818,
+      "grad_norm": 0.1688002030306263,
+      "learning_rate": 1.4577259475218661e-06,
+      "loss": 0.6252,
+      "step": 25
+    },
+    {
+      "epoch": 0.01518137362236333,
+      "grad_norm": 0.16113399543130622,
+      "learning_rate": 1.5160349854227407e-06,
+      "loss": 0.6398,
+      "step": 26
+    },
+    {
+      "epoch": 0.015765272607838846,
+      "grad_norm": 0.21439948049938257,
+      "learning_rate": 1.5743440233236153e-06,
+      "loss": 0.6581,
+      "step": 27
+    },
+    {
+      "epoch": 0.016349171593314355,
+      "grad_norm": 0.2548638260595163,
+      "learning_rate": 1.6326530612244897e-06,
+      "loss": 0.6423,
+      "step": 28
+    },
+    {
+      "epoch": 0.016933070578789868,
+      "grad_norm": 0.24461697399741664,
+      "learning_rate": 1.6909620991253645e-06,
+      "loss": 0.6112,
+      "step": 29
+    },
+    {
+      "epoch": 0.01751696956426538,
+      "grad_norm": 0.25124117495401055,
+      "learning_rate": 1.7492711370262391e-06,
+      "loss": 0.6362,
+      "step": 30
+    },
+    {
+      "epoch": 0.018100868549740894,
+      "grad_norm": 0.20556287055227845,
+      "learning_rate": 1.8075801749271137e-06,
+      "loss": 0.6635,
+      "step": 31
+    },
+    {
+      "epoch": 0.018684767535216407,
+      "grad_norm": 0.20735897106303364,
+      "learning_rate": 1.8658892128279885e-06,
+      "loss": 0.6571,
+      "step": 32
+    },
+    {
+      "epoch": 0.01926866652069192,
+      "grad_norm": 0.16388959942240836,
+      "learning_rate": 1.9241982507288633e-06,
+      "loss": 0.6096,
+      "step": 33
+    },
+    {
+      "epoch": 0.019852565506167433,
+      "grad_norm": 0.1579276273599527,
+      "learning_rate": 1.982507288629738e-06,
+      "loss": 0.6076,
+      "step": 34
+    },
+    {
+      "epoch": 0.020436464491642947,
+      "grad_norm": 0.13746529563620632,
+      "learning_rate": 2.0408163265306125e-06,
+      "loss": 0.5543,
+      "step": 35
+    },
+    {
+      "epoch": 0.02102036347711846,
+      "grad_norm": 0.1409487036760126,
+      "learning_rate": 2.099125364431487e-06,
+      "loss": 0.5674,
+      "step": 36
+    },
+    {
+      "epoch": 0.021604262462593973,
+      "grad_norm": 0.14409211583110956,
+      "learning_rate": 2.1574344023323617e-06,
+      "loss": 0.5894,
+      "step": 37
+    },
+    {
+      "epoch": 0.022188161448069486,
+      "grad_norm": 0.14241102694059576,
+      "learning_rate": 2.2157434402332363e-06,
+      "loss": 0.5637,
+      "step": 38
+    },
+    {
+      "epoch": 0.022772060433544995,
+      "grad_norm": 0.15554565969343062,
+      "learning_rate": 2.274052478134111e-06,
+      "loss": 0.5851,
+      "step": 39
+    },
+    {
+      "epoch": 0.02335595941902051,
+      "grad_norm": 0.16208731365633774,
+      "learning_rate": 2.3323615160349855e-06,
+      "loss": 0.5981,
+      "step": 40
+    },
+    {
+      "epoch": 0.02393985840449602,
+      "grad_norm": 0.1455930556848085,
+      "learning_rate": 2.39067055393586e-06,
+      "loss": 0.5464,
+      "step": 41
+    },
+    {
+      "epoch": 0.024523757389971534,
+      "grad_norm": 0.13681965582701403,
+      "learning_rate": 2.4489795918367347e-06,
+      "loss": 0.5263,
+      "step": 42
+    },
+    {
+      "epoch": 0.025107656375447047,
+      "grad_norm": 0.139658747601752,
+      "learning_rate": 2.5072886297376097e-06,
+      "loss": 0.5685,
+      "step": 43
+    },
+    {
+      "epoch": 0.02569155536092256,
+      "grad_norm": 0.1501595786161326,
+      "learning_rate": 2.5655976676384843e-06,
+      "loss": 0.6392,
+      "step": 44
+    },
+    {
+      "epoch": 0.026275454346398074,
+      "grad_norm": 0.13573680413767497,
+      "learning_rate": 2.6239067055393585e-06,
+      "loss": 0.5463,
+      "step": 45
+    },
+    {
+      "epoch": 0.026859353331873587,
+      "grad_norm": 0.12566744301562577,
+      "learning_rate": 2.6822157434402335e-06,
+      "loss": 0.5441,
+      "step": 46
+    },
+    {
+      "epoch": 0.0274432523173491,
+      "grad_norm": 0.13593497372842595,
+      "learning_rate": 2.740524781341108e-06,
+      "loss": 0.628,
+      "step": 47
+    },
+    {
+      "epoch": 0.028027151302824613,
+      "grad_norm": 0.12003887105865298,
+      "learning_rate": 2.7988338192419827e-06,
+      "loss": 0.5225,
+      "step": 48
+    },
+    {
+      "epoch": 0.028611050288300126,
+      "grad_norm": 0.14097459152623956,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.5593,
+      "step": 49
+    },
+    {
+      "epoch": 0.029194949273775635,
+      "grad_norm": 0.12264626298803548,
+      "learning_rate": 2.9154518950437323e-06,
+      "loss": 0.4973,
+      "step": 50
+    },
+    {
+      "epoch": 0.02977884825925115,
+      "grad_norm": 0.11903025600332431,
+      "learning_rate": 2.9737609329446064e-06,
+      "loss": 0.5247,
+      "step": 51
+    },
+    {
+      "epoch": 0.03036274724472666,
+      "grad_norm": 0.12671999020730157,
+      "learning_rate": 3.0320699708454815e-06,
+      "loss": 0.5863,
+      "step": 52
+    },
+    {
+      "epoch": 0.030946646230202175,
+      "grad_norm": 0.1286579840506993,
+      "learning_rate": 3.090379008746356e-06,
+      "loss": 0.5043,
+      "step": 53
+    },
+    {
+      "epoch": 0.03153054521567769,
+      "grad_norm": 0.11754747996364631,
+      "learning_rate": 3.1486880466472307e-06,
+      "loss": 0.568,
+      "step": 54
+    },
+    {
+      "epoch": 0.0321144442011532,
+      "grad_norm": 0.11940416834607172,
+      "learning_rate": 3.2069970845481052e-06,
+      "loss": 0.5504,
+      "step": 55
+    },
+    {
+      "epoch": 0.03269834318662871,
+      "grad_norm": 0.11809218597240961,
+      "learning_rate": 3.2653061224489794e-06,
+      "loss": 0.5196,
+      "step": 56
+    },
+    {
+      "epoch": 0.03328224217210422,
+      "grad_norm": 0.1216179127651747,
+      "learning_rate": 3.3236151603498544e-06,
+      "loss": 0.5235,
+      "step": 57
+    },
+    {
+      "epoch": 0.033866141157579736,
+      "grad_norm": 0.11997284871696193,
+      "learning_rate": 3.381924198250729e-06,
+      "loss": 0.5357,
+      "step": 58
+    },
+    {
+      "epoch": 0.03445004014305525,
+      "grad_norm": 0.12473873373168291,
+      "learning_rate": 3.440233236151604e-06,
+      "loss": 0.5434,
+      "step": 59
+    },
+    {
+      "epoch": 0.03503393912853076,
+      "grad_norm": 0.12926928135076024,
+      "learning_rate": 3.4985422740524782e-06,
+      "loss": 0.6092,
+      "step": 60
+    },
+    {
+      "epoch": 0.035617838114006276,
+      "grad_norm": 0.121410607286127,
+      "learning_rate": 3.5568513119533532e-06,
+      "loss": 0.5357,
+      "step": 61
+    },
+    {
+      "epoch": 0.03620173709948179,
+      "grad_norm": 0.12285474731915894,
+      "learning_rate": 3.6151603498542274e-06,
+      "loss": 0.6184,
+      "step": 62
+    },
+    {
+      "epoch": 0.0367856360849573,
+      "grad_norm": 0.13380976495700578,
+      "learning_rate": 3.6734693877551024e-06,
+      "loss": 0.5959,
+      "step": 63
+    },
+    {
+      "epoch": 0.037369535070432815,
+      "grad_norm": 0.11813064010226204,
+      "learning_rate": 3.731778425655977e-06,
+      "loss": 0.5834,
+      "step": 64
+    },
+    {
+      "epoch": 0.03795343405590833,
+      "grad_norm": 0.12998391279846627,
+      "learning_rate": 3.790087463556852e-06,
+      "loss": 0.5355,
+      "step": 65
+    },
+    {
+      "epoch": 0.03853733304138384,
+      "grad_norm": 0.11770562662855222,
+      "learning_rate": 3.848396501457727e-06,
+      "loss": 0.5227,
+      "step": 66
+    },
+    {
+      "epoch": 0.039121232026859354,
+      "grad_norm": 0.13107978237306653,
+      "learning_rate": 3.906705539358601e-06,
+      "loss": 0.5389,
+      "step": 67
+    },
+    {
+      "epoch": 0.03970513101233487,
+      "grad_norm": 0.12319041424395982,
+      "learning_rate": 3.965014577259476e-06,
+      "loss": 0.5143,
+      "step": 68
+    },
+    {
+      "epoch": 0.04028902999781038,
+      "grad_norm": 0.11986337933564546,
+      "learning_rate": 4.02332361516035e-06,
+      "loss": 0.5373,
+      "step": 69
+    },
+    {
+      "epoch": 0.04087292898328589,
+      "grad_norm": 0.11592131415761107,
+      "learning_rate": 4.081632653061225e-06,
+      "loss": 0.4945,
+      "step": 70
+    },
+    {
+      "epoch": 0.041456827968761406,
+      "grad_norm": 0.11953625428732638,
+      "learning_rate": 4.139941690962099e-06,
+      "loss": 0.504,
+      "step": 71
+    },
+    {
+      "epoch": 0.04204072695423692,
+      "grad_norm": 0.11346589228092914,
+      "learning_rate": 4.198250728862974e-06,
+      "loss": 0.4728,
+      "step": 72
+    },
+    {
+      "epoch": 0.04262462593971243,
+      "grad_norm": 0.12273082976475955,
+      "learning_rate": 4.256559766763848e-06,
+      "loss": 0.509,
+      "step": 73
+    },
+    {
+      "epoch": 0.043208524925187945,
+      "grad_norm": 0.12636422024860947,
+      "learning_rate": 4.314868804664723e-06,
+      "loss": 0.5375,
+      "step": 74
+    },
+    {
+      "epoch": 0.04379242391066346,
+      "grad_norm": 0.12889896571137324,
+      "learning_rate": 4.3731778425655976e-06,
+      "loss": 0.5628,
+      "step": 75
+    },
+    {
+      "epoch": 0.04437632289613897,
+      "grad_norm": 0.1119240399609327,
+      "learning_rate": 4.431486880466473e-06,
+      "loss": 0.4855,
+      "step": 76
+    },
+    {
+      "epoch": 0.04496022188161448,
+      "grad_norm": 0.11470207936904074,
+      "learning_rate": 4.489795918367348e-06,
+      "loss": 0.5375,
+      "step": 77
+    },
+    {
+      "epoch": 0.04554412086708999,
+      "grad_norm": 0.1218766787696568,
+      "learning_rate": 4.548104956268222e-06,
+      "loss": 0.508,
+      "step": 78
+    },
+    {
+      "epoch": 0.046128019852565504,
+      "grad_norm": 0.1210080317363853,
+      "learning_rate": 4.606413994169097e-06,
+      "loss": 0.4835,
+      "step": 79
+    },
+    {
+      "epoch": 0.04671191883804102,
+      "grad_norm": 0.12105990442659609,
+      "learning_rate": 4.664723032069971e-06,
+      "loss": 0.5587,
+      "step": 80
+    },
+    {
+      "epoch": 0.04729581782351653,
+      "grad_norm": 0.11897823767402924,
+      "learning_rate": 4.723032069970846e-06,
+      "loss": 0.5966,
+      "step": 81
+    },
+    {
+      "epoch": 0.04787971680899204,
+      "grad_norm": 0.11634346204320371,
+      "learning_rate": 4.78134110787172e-06,
+      "loss": 0.5424,
+      "step": 82
+    },
+    {
+      "epoch": 0.048463615794467556,
+      "grad_norm": 0.1440848826471587,
+      "learning_rate": 4.839650145772595e-06,
+      "loss": 0.5402,
+      "step": 83
+    },
+    {
+      "epoch": 0.04904751477994307,
+      "grad_norm": 0.12012047213109964,
+      "learning_rate": 4.897959183673469e-06,
+      "loss": 0.5375,
+      "step": 84
+    },
+    {
+      "epoch": 0.04963141376541858,
+      "grad_norm": 0.12291287843723532,
+      "learning_rate": 4.956268221574344e-06,
+      "loss": 0.5539,
+      "step": 85
+    },
+    {
+      "epoch": 0.050215312750894095,
+      "grad_norm": 0.1293194792618398,
+      "learning_rate": 5.014577259475219e-06,
+      "loss": 0.5864,
+      "step": 86
+    },
+    {
+      "epoch": 0.05079921173636961,
+      "grad_norm": 0.11375996470368034,
+      "learning_rate": 5.0728862973760935e-06,
+      "loss": 0.4705,
+      "step": 87
+    },
+    {
+      "epoch": 0.05138311072184512,
+      "grad_norm": 0.12025487668180054,
+      "learning_rate": 5.1311953352769686e-06,
+      "loss": 0.5225,
+      "step": 88
+    },
+    {
+      "epoch": 0.051967009707320634,
+      "grad_norm": 0.12000690262818887,
+      "learning_rate": 5.189504373177843e-06,
+      "loss": 0.5435,
+      "step": 89
+    },
+    {
+      "epoch": 0.05255090869279615,
+      "grad_norm": 0.1288610889625379,
+      "learning_rate": 5.247813411078717e-06,
+      "loss": 0.5577,
+      "step": 90
+    },
+    {
+      "epoch": 0.05313480767827166,
+      "grad_norm": 0.12621763771355476,
+      "learning_rate": 5.306122448979593e-06,
+      "loss": 0.4993,
+      "step": 91
+    },
+    {
+      "epoch": 0.05371870666374717,
+      "grad_norm": 0.13277319165341347,
+      "learning_rate": 5.364431486880467e-06,
+      "loss": 0.5073,
+      "step": 92
+    },
+    {
+      "epoch": 0.054302605649222686,
+      "grad_norm": 0.11802197526980429,
+      "learning_rate": 5.422740524781341e-06,
+      "loss": 0.4944,
+      "step": 93
+    },
+    {
+      "epoch": 0.0548865046346982,
+      "grad_norm": 0.12233678592202392,
+      "learning_rate": 5.481049562682216e-06,
+      "loss": 0.5336,
+      "step": 94
+    },
+    {
+      "epoch": 0.05547040362017371,
+      "grad_norm": 0.11070655839309021,
+      "learning_rate": 5.539358600583091e-06,
+      "loss": 0.5102,
+      "step": 95
+    },
+    {
+      "epoch": 0.056054302605649226,
+      "grad_norm": 0.12446163883557436,
+      "learning_rate": 5.597667638483965e-06,
+      "loss": 0.495,
+      "step": 96
+    },
+    {
+      "epoch": 0.05663820159112474,
+      "grad_norm": 0.11508553084346662,
+      "learning_rate": 5.65597667638484e-06,
+      "loss": 0.5433,
+      "step": 97
+    },
+    {
+      "epoch": 0.05722210057660025,
+      "grad_norm": 0.12132804044380216,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.5041,
+      "step": 98
+    },
+    {
+      "epoch": 0.05780599956207576,
+      "grad_norm": 0.1266523570042547,
+      "learning_rate": 5.7725947521865895e-06,
+      "loss": 0.5102,
+      "step": 99
+    },
+    {
+      "epoch": 0.05838989854755127,
+      "grad_norm": 0.1337299728224099,
+      "learning_rate": 5.8309037900874645e-06,
+      "loss": 0.4632,
+      "step": 100
+    },
+    {
+      "epoch": 0.058973797533026784,
+      "grad_norm": 0.12042164825415605,
+      "learning_rate": 5.889212827988339e-06,
+      "loss": 0.5097,
+      "step": 101
+    },
+    {
+      "epoch": 0.0595576965185023,
+      "grad_norm": 0.1282189131165611,
+      "learning_rate": 5.947521865889213e-06,
+      "loss": 0.51,
+      "step": 102
+    },
+    {
+      "epoch": 0.06014159550397781,
+      "grad_norm": 0.11911278813626588,
+      "learning_rate": 6.005830903790088e-06,
+      "loss": 0.4722,
+      "step": 103
+    },
+    {
+      "epoch": 0.06072549448945332,
+      "grad_norm": 0.12468624838929701,
+      "learning_rate": 6.064139941690963e-06,
+      "loss": 0.4768,
+      "step": 104
+    },
+    {
+      "epoch": 0.061309393474928836,
+      "grad_norm": 0.11861980102083973,
+      "learning_rate": 6.122448979591837e-06,
+      "loss": 0.5104,
+      "step": 105
+    },
+    {
+      "epoch": 0.06189329246040435,
+      "grad_norm": 0.12406795405645518,
+      "learning_rate": 6.180758017492712e-06,
+      "loss": 0.478,
+      "step": 106
+    },
+    {
+      "epoch": 0.06247719144587986,
+      "grad_norm": 0.12175840167434945,
+      "learning_rate": 6.239067055393586e-06,
+      "loss": 0.4823,
+      "step": 107
+    },
+    {
+      "epoch": 0.06306109043135538,
+      "grad_norm": 0.12618202468532327,
+      "learning_rate": 6.297376093294461e-06,
+      "loss": 0.5236,
+      "step": 108
+    },
+    {
+      "epoch": 0.06364498941683089,
+      "grad_norm": 0.12587719128126854,
+      "learning_rate": 6.355685131195336e-06,
+      "loss": 0.4878,
+      "step": 109
+    },
+    {
+      "epoch": 0.0642288884023064,
+      "grad_norm": 0.1224994251821843,
+      "learning_rate": 6.4139941690962105e-06,
+      "loss": 0.4636,
+      "step": 110
+    },
+    {
+      "epoch": 0.06481278738778191,
+      "grad_norm": 0.13322676918183363,
+      "learning_rate": 6.472303206997085e-06,
+      "loss": 0.5062,
+      "step": 111
+    },
+    {
+      "epoch": 0.06539668637325742,
+      "grad_norm": 0.12774188008190585,
+      "learning_rate": 6.530612244897959e-06,
+      "loss": 0.47,
+      "step": 112
+    },
+    {
+      "epoch": 0.06598058535873294,
+      "grad_norm": 0.11785400668315196,
+      "learning_rate": 6.588921282798835e-06,
+      "loss": 0.5033,
+      "step": 113
+    },
+    {
+      "epoch": 0.06656448434420845,
+      "grad_norm": 0.12794769801804934,
+      "learning_rate": 6.647230320699709e-06,
+      "loss": 0.5178,
+      "step": 114
+    },
+    {
+      "epoch": 0.06714838332968397,
+      "grad_norm": 0.1273661611156257,
+      "learning_rate": 6.705539358600584e-06,
+      "loss": 0.474,
+      "step": 115
+    },
+    {
+      "epoch": 0.06773228231515947,
+      "grad_norm": 0.1293645211236864,
+      "learning_rate": 6.763848396501458e-06,
+      "loss": 0.508,
+      "step": 116
+    },
+    {
+      "epoch": 0.06831618130063499,
+      "grad_norm": 0.11932371311165128,
+      "learning_rate": 6.822157434402333e-06,
+      "loss": 0.537,
+      "step": 117
+    },
+    {
+      "epoch": 0.0689000802861105,
+      "grad_norm": 0.12673979514956368,
+      "learning_rate": 6.880466472303208e-06,
+      "loss": 0.5183,
+      "step": 118
+    },
+    {
+      "epoch": 0.06948397927158602,
+      "grad_norm": 0.12595136053036562,
+      "learning_rate": 6.938775510204082e-06,
+      "loss": 0.4644,
+      "step": 119
+    },
+    {
+      "epoch": 0.07006787825706152,
+      "grad_norm": 0.12299888706906272,
+      "learning_rate": 6.9970845481049564e-06,
+      "loss": 0.4919,
+      "step": 120
+    },
+    {
+      "epoch": 0.07065177724253704,
+      "grad_norm": 0.12083005613805631,
+      "learning_rate": 7.055393586005832e-06,
+      "loss": 0.5,
+      "step": 121
+    },
+    {
+      "epoch": 0.07123567622801255,
+      "grad_norm": 0.1211546204214517,
+      "learning_rate": 7.1137026239067065e-06,
+      "loss": 0.5118,
+      "step": 122
+    },
+    {
+      "epoch": 0.07181957521348807,
+      "grad_norm": 0.11611648065418555,
+      "learning_rate": 7.172011661807581e-06,
+      "loss": 0.4899,
+      "step": 123
+    },
+    {
+      "epoch": 0.07240347419896358,
+      "grad_norm": 0.13510538598848665,
+      "learning_rate": 7.230320699708455e-06,
+      "loss": 0.5402,
+      "step": 124
+    },
+    {
+      "epoch": 0.0729873731844391,
+      "grad_norm": 0.1308633680816093,
+      "learning_rate": 7.28862973760933e-06,
+      "loss": 0.5663,
+      "step": 125
+    },
+    {
+      "epoch": 0.0735712721699146,
+      "grad_norm": 0.15580025835064823,
+      "learning_rate": 7.346938775510205e-06,
+      "loss": 0.4559,
+      "step": 126
+    },
+    {
+      "epoch": 0.07415517115539012,
+      "grad_norm": 0.11393538418682475,
+      "learning_rate": 7.40524781341108e-06,
+      "loss": 0.4868,
+      "step": 127
+    },
+    {
+      "epoch": 0.07473907014086563,
+      "grad_norm": 0.11323771708880152,
+      "learning_rate": 7.463556851311954e-06,
+      "loss": 0.4818,
+      "step": 128
+    },
+    {
+      "epoch": 0.07532296912634115,
+      "grad_norm": 0.13156488258845986,
+      "learning_rate": 7.521865889212828e-06,
+      "loss": 0.5332,
+      "step": 129
+    },
+    {
+      "epoch": 0.07590686811181666,
+      "grad_norm": 0.1340083425921203,
+      "learning_rate": 7.580174927113704e-06,
+      "loss": 0.53,
+      "step": 130
+    },
+    {
+      "epoch": 0.07649076709729216,
+      "grad_norm": 0.12074585913812508,
+      "learning_rate": 7.638483965014577e-06,
+      "loss": 0.4914,
+      "step": 131
+    },
+    {
+      "epoch": 0.07707466608276768,
+      "grad_norm": 0.12363069444881924,
+      "learning_rate": 7.696793002915453e-06,
+      "loss": 0.5324,
+      "step": 132
+    },
+    {
+      "epoch": 0.07765856506824319,
+      "grad_norm": 0.12081434779965833,
+      "learning_rate": 7.755102040816327e-06,
+      "loss": 0.535,
+      "step": 133
+    },
+    {
+      "epoch": 0.07824246405371871,
+      "grad_norm": 0.11909031940429428,
+      "learning_rate": 7.813411078717202e-06,
+      "loss": 0.4709,
+      "step": 134
+    },
+    {
+      "epoch": 0.07882636303919421,
+      "grad_norm": 0.12155862255010864,
+      "learning_rate": 7.871720116618077e-06,
+      "loss": 0.4979,
+      "step": 135
+    },
+    {
+      "epoch": 0.07941026202466973,
+      "grad_norm": 0.11519739456872996,
+      "learning_rate": 7.930029154518952e-06,
+      "loss": 0.4599,
+      "step": 136
+    },
+    {
+      "epoch": 0.07999416101014524,
+      "grad_norm": 0.12713602651229577,
+      "learning_rate": 7.988338192419826e-06,
+      "loss": 0.5396,
+      "step": 137
+    },
+    {
+      "epoch": 0.08057805999562076,
+      "grad_norm": 0.12280179465979835,
+      "learning_rate": 8.0466472303207e-06,
+      "loss": 0.4986,
+      "step": 138
+    },
+    {
+      "epoch": 0.08116195898109627,
+      "grad_norm": 0.12438666320342781,
+      "learning_rate": 8.104956268221576e-06,
+      "loss": 0.5455,
+      "step": 139
+    },
+    {
+      "epoch": 0.08174585796657179,
+      "grad_norm": 0.12067104522166396,
+      "learning_rate": 8.16326530612245e-06,
+      "loss": 0.489,
+      "step": 140
+    },
+    {
+      "epoch": 0.08232975695204729,
+      "grad_norm": 0.10865884801530365,
+      "learning_rate": 8.221574344023324e-06,
+      "loss": 0.4551,
+      "step": 141
+    },
+    {
+      "epoch": 0.08291365593752281,
+      "grad_norm": 0.11273476337541347,
+      "learning_rate": 8.279883381924198e-06,
+      "loss": 0.4434,
+      "step": 142
+    },
+    {
+      "epoch": 0.08349755492299832,
+      "grad_norm": 0.1231877373706282,
+      "learning_rate": 8.338192419825074e-06,
+      "loss": 0.495,
+      "step": 143
+    },
+    {
+      "epoch": 0.08408145390847384,
+      "grad_norm": 0.12065352584250427,
+      "learning_rate": 8.396501457725948e-06,
+      "loss": 0.4948,
+      "step": 144
+    },
+    {
+      "epoch": 0.08466535289394934,
+      "grad_norm": 0.12830745282443723,
+      "learning_rate": 8.454810495626823e-06,
+      "loss": 0.4936,
+      "step": 145
+    },
+    {
+      "epoch": 0.08524925187942486,
+      "grad_norm": 0.11964018736868538,
+      "learning_rate": 8.513119533527697e-06,
+      "loss": 0.4446,
+      "step": 146
+    },
+    {
+      "epoch": 0.08583315086490037,
+      "grad_norm": 0.12678811130736195,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.4336,
+      "step": 147
+    },
+    {
+      "epoch": 0.08641704985037589,
+      "grad_norm": 0.12177495955493131,
+      "learning_rate": 8.629737609329447e-06,
+      "loss": 0.4627,
+      "step": 148
+    },
+    {
+      "epoch": 0.0870009488358514,
+      "grad_norm": 0.11882976198336359,
+      "learning_rate": 8.688046647230321e-06,
+      "loss": 0.4958,
+      "step": 149
+    },
+    {
+      "epoch": 0.08758484782132692,
+      "grad_norm": 0.12002363110252463,
+      "learning_rate": 8.746355685131195e-06,
+      "loss": 0.4383,
+      "step": 150
+    },
+    {
+      "epoch": 0.08816874680680242,
+      "grad_norm": 0.11477715855770114,
+      "learning_rate": 8.804664723032071e-06,
+      "loss": 0.4513,
+      "step": 151
+    },
+    {
+      "epoch": 0.08875264579227794,
+      "grad_norm": 0.12093079017609433,
+      "learning_rate": 8.862973760932945e-06,
+      "loss": 0.4943,
+      "step": 152
+    },
+    {
+      "epoch": 0.08933654477775345,
+      "grad_norm": 0.11804978178515525,
+      "learning_rate": 8.921282798833821e-06,
+      "loss": 0.4779,
+      "step": 153
+    },
+    {
+      "epoch": 0.08992044376322895,
+      "grad_norm": 0.13405678610400526,
+      "learning_rate": 8.979591836734695e-06,
+      "loss": 0.471,
+      "step": 154
+    },
+    {
+      "epoch": 0.09050434274870447,
+      "grad_norm": 0.11491490859650165,
+      "learning_rate": 9.03790087463557e-06,
+      "loss": 0.4639,
+      "step": 155
+    },
+    {
+      "epoch": 0.09108824173417998,
+      "grad_norm": 0.12568063398628002,
+      "learning_rate": 9.096209912536444e-06,
+      "loss": 0.4988,
+      "step": 156
+    },
+    {
+      "epoch": 0.0916721407196555,
+      "grad_norm": 0.12319313800384202,
+      "learning_rate": 9.15451895043732e-06,
+      "loss": 0.5304,
+      "step": 157
+    },
+    {
+      "epoch": 0.09225603970513101,
+      "grad_norm": 0.1699743279850764,
+      "learning_rate": 9.212827988338194e-06,
+      "loss": 0.4851,
+      "step": 158
+    },
+    {
+      "epoch": 0.09283993869060653,
+      "grad_norm": 0.11927051778915101,
+      "learning_rate": 9.271137026239068e-06,
+      "loss": 0.4971,
+      "step": 159
+    },
+    {
+      "epoch": 0.09342383767608203,
+      "grad_norm": 0.12261541717069109,
+      "learning_rate": 9.329446064139942e-06,
+      "loss": 0.5473,
+      "step": 160
+    },
+    {
+      "epoch": 0.09400773666155755,
+      "grad_norm": 0.12209484214190072,
+      "learning_rate": 9.387755102040818e-06,
+      "loss": 0.504,
+      "step": 161
+    },
+    {
+      "epoch": 0.09459163564703306,
+      "grad_norm": 0.6461953167431549,
+      "learning_rate": 9.446064139941692e-06,
+      "loss": 0.4687,
+      "step": 162
+    },
+    {
+      "epoch": 0.09517553463250858,
+      "grad_norm": 0.11874885918234361,
+      "learning_rate": 9.504373177842566e-06,
+      "loss": 0.4829,
+      "step": 163
+    },
+    {
+      "epoch": 0.09575943361798409,
+      "grad_norm": 0.13144948930395803,
+      "learning_rate": 9.56268221574344e-06,
+      "loss": 0.4624,
+      "step": 164
+    },
+    {
+      "epoch": 0.0963433326034596,
+      "grad_norm": 0.11896981551418268,
+      "learning_rate": 9.620991253644316e-06,
+      "loss": 0.4744,
+      "step": 165
+    },
+    {
+      "epoch": 0.09692723158893511,
+      "grad_norm": 0.10391157620986394,
+      "learning_rate": 9.67930029154519e-06,
+      "loss": 0.4413,
+      "step": 166
+    },
+    {
+      "epoch": 0.09751113057441063,
+      "grad_norm": 0.12177286848001329,
+      "learning_rate": 9.737609329446065e-06,
+      "loss": 0.4677,
+      "step": 167
+    },
+    {
+      "epoch": 0.09809502955988614,
+      "grad_norm": 0.1292862064569658,
+      "learning_rate": 9.795918367346939e-06,
+      "loss": 0.5062,
+      "step": 168
+    },
+    {
+      "epoch": 0.09867892854536166,
+      "grad_norm": 0.13003129200868843,
+      "learning_rate": 9.854227405247815e-06,
+      "loss": 0.4778,
+      "step": 169
+    },
+    {
+      "epoch": 0.09926282753083716,
+      "grad_norm": 0.12152742853049435,
+      "learning_rate": 9.912536443148689e-06,
+      "loss": 0.4837,
+      "step": 170
+    },
+    {
+      "epoch": 0.09984672651631268,
+      "grad_norm": 0.12095495308768837,
+      "learning_rate": 9.970845481049563e-06,
+      "loss": 0.42,
+      "step": 171
+    },
+    {
+      "epoch": 0.10043062550178819,
+      "grad_norm": 0.1176426320900958,
+      "learning_rate": 1.0029154518950439e-05,
+      "loss": 0.5267,
+      "step": 172
+    },
+    {
+      "epoch": 0.10101452448726371,
+      "grad_norm": 0.12144720417293908,
+      "learning_rate": 1.0087463556851313e-05,
+      "loss": 0.4755,
+      "step": 173
+    },
+    {
+      "epoch": 0.10159842347273922,
+      "grad_norm": 0.124849366844711,
+      "learning_rate": 1.0145772594752187e-05,
+      "loss": 0.5064,
+      "step": 174
+    },
+    {
+      "epoch": 0.10218232245821472,
+      "grad_norm": 0.11961143418277417,
+      "learning_rate": 1.0204081632653063e-05,
+      "loss": 0.4887,
+      "step": 175
+    },
+    {
+      "epoch": 0.10276622144369024,
+      "grad_norm": 0.11305302860874425,
+      "learning_rate": 1.0262390670553937e-05,
+      "loss": 0.4732,
+      "step": 176
+    },
+    {
+      "epoch": 0.10335012042916575,
+      "grad_norm": 0.12276878798868197,
+      "learning_rate": 1.0320699708454811e-05,
+      "loss": 0.5021,
+      "step": 177
+    },
+    {
+      "epoch": 0.10393401941464127,
+      "grad_norm": 0.11349763202062922,
+      "learning_rate": 1.0379008746355685e-05,
+      "loss": 0.462,
+      "step": 178
+    },
+    {
+      "epoch": 0.10451791840011677,
+      "grad_norm": 0.12320137581028516,
+      "learning_rate": 1.043731778425656e-05,
+      "loss": 0.4661,
+      "step": 179
+    },
+    {
+      "epoch": 0.1051018173855923,
+      "grad_norm": 0.11360139818063392,
+      "learning_rate": 1.0495626822157434e-05,
+      "loss": 0.477,
+      "step": 180
+    },
+    {
+      "epoch": 0.1056857163710678,
+      "grad_norm": 0.11981964875579429,
+      "learning_rate": 1.0553935860058311e-05,
+      "loss": 0.5353,
+      "step": 181
+    },
+    {
+      "epoch": 0.10626961535654332,
+      "grad_norm": 0.12183565394104451,
+      "learning_rate": 1.0612244897959186e-05,
+      "loss": 0.5104,
+      "step": 182
+    },
+    {
+      "epoch": 0.10685351434201883,
+      "grad_norm": 0.12388133833711053,
+      "learning_rate": 1.067055393586006e-05,
+      "loss": 0.4868,
+      "step": 183
+    },
+    {
+      "epoch": 0.10743741332749435,
+      "grad_norm": 0.1164264171971401,
+      "learning_rate": 1.0728862973760934e-05,
+      "loss": 0.4714,
+      "step": 184
+    },
+    {
+      "epoch": 0.10802131231296985,
+      "grad_norm": 0.1181176343568213,
+      "learning_rate": 1.0787172011661808e-05,
+      "loss": 0.4877,
+      "step": 185
+    },
+    {
+      "epoch": 0.10860521129844537,
+      "grad_norm": 0.11387598663951679,
+      "learning_rate": 1.0845481049562682e-05,
+      "loss": 0.4839,
+      "step": 186
+    },
+    {
+      "epoch": 0.10918911028392088,
+      "grad_norm": 0.12281128996524167,
+      "learning_rate": 1.0903790087463556e-05,
+      "loss": 0.4615,
+      "step": 187
+    },
+    {
+      "epoch": 0.1097730092693964,
+      "grad_norm": 0.12111501681054686,
+      "learning_rate": 1.0962099125364432e-05,
+      "loss": 0.4802,
+      "step": 188
+    },
+    {
+      "epoch": 0.1103569082548719,
+      "grad_norm": 0.12013303068905667,
+      "learning_rate": 1.1020408163265306e-05,
+      "loss": 0.4469,
+      "step": 189
+    },
+    {
+      "epoch": 0.11094080724034742,
+      "grad_norm": 0.11998714367515836,
+      "learning_rate": 1.1078717201166182e-05,
+      "loss": 0.4811,
+      "step": 190
+    },
+    {
+      "epoch": 0.11152470622582293,
+      "grad_norm": 0.12503719180886072,
+      "learning_rate": 1.1137026239067056e-05,
+      "loss": 0.5085,
+      "step": 191
+    },
+    {
+      "epoch": 0.11210860521129845,
+      "grad_norm": 0.12161353835164174,
+      "learning_rate": 1.119533527696793e-05,
+      "loss": 0.4853,
+      "step": 192
+    },
+    {
+      "epoch": 0.11269250419677396,
+      "grad_norm": 0.1193629934753903,
+      "learning_rate": 1.1253644314868807e-05,
+      "loss": 0.5039,
+      "step": 193
+    },
+    {
+      "epoch": 0.11327640318224948,
+      "grad_norm": 0.12588619686276323,
+      "learning_rate": 1.131195335276968e-05,
+      "loss": 0.4823,
+      "step": 194
+    },
+    {
+      "epoch": 0.11386030216772498,
+      "grad_norm": 0.11515327746681453,
+      "learning_rate": 1.1370262390670555e-05,
+      "loss": 0.4805,
+      "step": 195
+    },
+    {
+      "epoch": 0.1144442011532005,
+      "grad_norm": 0.11352598968281107,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.4553,
+      "step": 196
+    },
+    {
+      "epoch": 0.11502810013867601,
+      "grad_norm": 0.11645497517126423,
+      "learning_rate": 1.1486880466472303e-05,
+      "loss": 0.4451,
+      "step": 197
+    },
+    {
+      "epoch": 0.11561199912415152,
+      "grad_norm": 0.12020484531313662,
+      "learning_rate": 1.1545189504373179e-05,
+      "loss": 0.5159,
+      "step": 198
+    },
+    {
+      "epoch": 0.11619589810962704,
+      "grad_norm": 0.11100764759242791,
+      "learning_rate": 1.1603498542274055e-05,
+      "loss": 0.4447,
+      "step": 199
+    },
+    {
+      "epoch": 0.11677979709510254,
+      "grad_norm": 0.12443951867539835,
+      "learning_rate": 1.1661807580174929e-05,
+      "loss": 0.5273,
+      "step": 200
+    },
+    {
+      "epoch": 0.11736369608057806,
+      "grad_norm": 0.12474723898917135,
+      "learning_rate": 1.1720116618075803e-05,
+      "loss": 0.4659,
+      "step": 201
+    },
+    {
+      "epoch": 0.11794759506605357,
+      "grad_norm": 0.11988864151021485,
+      "learning_rate": 1.1778425655976677e-05,
+      "loss": 0.44,
+      "step": 202
+    },
+    {
+      "epoch": 0.11853149405152909,
+      "grad_norm": 0.12382641929448247,
+      "learning_rate": 1.1836734693877552e-05,
+      "loss": 0.4746,
+      "step": 203
+    },
+    {
+      "epoch": 0.1191153930370046,
+      "grad_norm": 0.1235721070065947,
+      "learning_rate": 1.1895043731778426e-05,
+      "loss": 0.4689,
+      "step": 204
+    },
+    {
+      "epoch": 0.11969929202248011,
+      "grad_norm": 0.12057949190250324,
+      "learning_rate": 1.19533527696793e-05,
+      "loss": 0.4661,
+      "step": 205
+    },
+    {
+      "epoch": 0.12028319100795562,
+      "grad_norm": 0.1079269611898759,
+      "learning_rate": 1.2011661807580176e-05,
+      "loss": 0.4273,
+      "step": 206
+    },
+    {
+      "epoch": 0.12086708999343114,
+      "grad_norm": 0.11846440506566613,
+      "learning_rate": 1.2069970845481052e-05,
+      "loss": 0.4575,
+      "step": 207
+    },
+    {
+      "epoch": 0.12145098897890665,
+      "grad_norm": 0.1186900696312661,
+      "learning_rate": 1.2128279883381926e-05,
+      "loss": 0.4259,
+      "step": 208
+    },
+    {
+      "epoch": 0.12203488796438217,
+      "grad_norm": 0.11921258585591628,
+      "learning_rate": 1.21865889212828e-05,
+      "loss": 0.4733,
+      "step": 209
+    },
+    {
+      "epoch": 0.12261878694985767,
+      "grad_norm": 0.12678296099330968,
+      "learning_rate": 1.2244897959183674e-05,
+      "loss": 0.4439,
+      "step": 210
+    },
+    {
+      "epoch": 0.12320268593533319,
+      "grad_norm": 0.13313906704082984,
+      "learning_rate": 1.2303206997084548e-05,
+      "loss": 0.464,
+      "step": 211
+    },
+    {
+      "epoch": 0.1237865849208087,
+      "grad_norm": 0.11988806633990326,
+      "learning_rate": 1.2361516034985424e-05,
+      "loss": 0.473,
+      "step": 212
+    },
+    {
+      "epoch": 0.12437048390628422,
+      "grad_norm": 0.12289136944160484,
+      "learning_rate": 1.2419825072886298e-05,
+      "loss": 0.4521,
+      "step": 213
+    },
+    {
+      "epoch": 0.12495438289175972,
+      "grad_norm": 0.11648978751406831,
+      "learning_rate": 1.2478134110787173e-05,
+      "loss": 0.4683,
+      "step": 214
+    },
+    {
+      "epoch": 0.12553828187723523,
+      "grad_norm": 0.12121715155226112,
+      "learning_rate": 1.2536443148688047e-05,
+      "loss": 0.517,
+      "step": 215
+    },
+    {
+      "epoch": 0.12612218086271076,
+      "grad_norm": 0.1229206784704676,
+      "learning_rate": 1.2594752186588923e-05,
+      "loss": 0.5345,
+      "step": 216
+    },
+    {
+      "epoch": 0.12670607984818627,
+      "grad_norm": 0.11705123876327105,
+      "learning_rate": 1.2653061224489798e-05,
+      "loss": 0.5439,
+      "step": 217
+    },
+    {
+      "epoch": 0.12728997883366178,
+      "grad_norm": 0.11571960095283303,
+      "learning_rate": 1.2711370262390673e-05,
+      "loss": 0.4871,
+      "step": 218
+    },
+    {
+      "epoch": 0.12787387781913728,
+      "grad_norm": 0.12080973442708406,
+      "learning_rate": 1.2769679300291547e-05,
+      "loss": 0.4936,
+      "step": 219
+    },
+    {
+      "epoch": 0.1284577768046128,
+      "grad_norm": 0.12345551183011412,
+      "learning_rate": 1.2827988338192421e-05,
+      "loss": 0.5039,
+      "step": 220
+    },
+    {
+      "epoch": 0.12904167579008832,
+      "grad_norm": 0.12714585450238197,
+      "learning_rate": 1.2886297376093295e-05,
+      "loss": 0.5312,
+      "step": 221
+    },
+    {
+      "epoch": 0.12962557477556383,
+      "grad_norm": 0.11988227444480619,
+      "learning_rate": 1.294460641399417e-05,
+      "loss": 0.4725,
+      "step": 222
+    },
+    {
+      "epoch": 0.13020947376103933,
+      "grad_norm": 0.12465621457555062,
+      "learning_rate": 1.3002915451895044e-05,
+      "loss": 0.4918,
+      "step": 223
+    },
+    {
+      "epoch": 0.13079337274651484,
+      "grad_norm": 0.1159935825832531,
+      "learning_rate": 1.3061224489795918e-05,
+      "loss": 0.4631,
+      "step": 224
+    },
+    {
+      "epoch": 0.13137727173199037,
+      "grad_norm": 0.11274267887775086,
+      "learning_rate": 1.3119533527696795e-05,
+      "loss": 0.4508,
+      "step": 225
+    },
+    {
+      "epoch": 0.13196117071746588,
+      "grad_norm": 0.12675512090304134,
+      "learning_rate": 1.317784256559767e-05,
+      "loss": 0.5439,
+      "step": 226
+    },
+    {
+      "epoch": 0.1325450697029414,
+      "grad_norm": 0.12131843263351978,
+      "learning_rate": 1.3236151603498544e-05,
+      "loss": 0.5051,
+      "step": 227
+    },
+    {
+      "epoch": 0.1331289686884169,
+      "grad_norm": 0.12220508753993224,
+      "learning_rate": 1.3294460641399418e-05,
+      "loss": 0.4569,
+      "step": 228
+    },
+    {
+      "epoch": 0.13371286767389243,
+      "grad_norm": 0.11946566843946793,
+      "learning_rate": 1.3352769679300292e-05,
+      "loss": 0.4953,
+      "step": 229
+    },
+    {
+      "epoch": 0.13429676665936793,
+      "grad_norm": 0.12779214444588172,
+      "learning_rate": 1.3411078717201168e-05,
+      "loss": 0.5056,
+      "step": 230
+    },
+    {
+      "epoch": 0.13488066564484344,
+      "grad_norm": 0.12799066668817458,
+      "learning_rate": 1.3469387755102042e-05,
+      "loss": 0.4993,
+      "step": 231
+    },
+    {
+      "epoch": 0.13546456463031895,
+      "grad_norm": 0.12235182069834803,
+      "learning_rate": 1.3527696793002916e-05,
+      "loss": 0.4632,
+      "step": 232
+    },
+    {
+      "epoch": 0.13604846361579448,
+      "grad_norm": 0.12526255723036422,
+      "learning_rate": 1.358600583090379e-05,
+      "loss": 0.5005,
+      "step": 233
+    },
+    {
+      "epoch": 0.13663236260126999,
+      "grad_norm": 0.1210733031506796,
+      "learning_rate": 1.3644314868804666e-05,
+      "loss": 0.4756,
+      "step": 234
+    },
+    {
+      "epoch": 0.1372162615867455,
+      "grad_norm": 0.12059584856135772,
+      "learning_rate": 1.370262390670554e-05,
+      "loss": 0.5038,
+      "step": 235
+    },
+    {
+      "epoch": 0.137800160572221,
+      "grad_norm": 0.12276174617876118,
+      "learning_rate": 1.3760932944606416e-05,
+      "loss": 0.4563,
+      "step": 236
+    },
+    {
+      "epoch": 0.13838405955769653,
+      "grad_norm": 0.11749231248282237,
+      "learning_rate": 1.381924198250729e-05,
+      "loss": 0.4852,
+      "step": 237
+    },
+    {
+      "epoch": 0.13896795854317204,
+      "grad_norm": 0.12048939153940323,
+      "learning_rate": 1.3877551020408165e-05,
+      "loss": 0.5036,
+      "step": 238
+    },
+    {
+      "epoch": 0.13955185752864754,
+      "grad_norm": 0.13089428767834732,
+      "learning_rate": 1.3935860058309039e-05,
+      "loss": 0.4533,
+      "step": 239
+    },
+    {
+      "epoch": 0.14013575651412305,
+      "grad_norm": 0.12687266022443466,
+      "learning_rate": 1.3994169096209913e-05,
+      "loss": 0.5123,
+      "step": 240
+    },
+    {
+      "epoch": 0.14071965549959856,
+      "grad_norm": 0.11269013979300824,
+      "learning_rate": 1.4052478134110787e-05,
+      "loss": 0.4925,
+      "step": 241
+    },
+    {
+      "epoch": 0.1413035544850741,
+      "grad_norm": 0.1258195333189289,
+      "learning_rate": 1.4110787172011665e-05,
+      "loss": 0.4881,
+      "step": 242
+    },
+    {
+      "epoch": 0.1418874534705496,
+      "grad_norm": 0.12592818504152503,
+      "learning_rate": 1.4169096209912539e-05,
+      "loss": 0.5049,
+      "step": 243
+    },
+    {
+      "epoch": 0.1424713524560251,
+      "grad_norm": 0.11413267743946869,
+      "learning_rate": 1.4227405247813413e-05,
+      "loss": 0.465,
+      "step": 244
+    },
+    {
+      "epoch": 0.1430552514415006,
+      "grad_norm": 0.12792861370450098,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 0.4879,
+      "step": 245
+    },
+    {
+      "epoch": 0.14363915042697614,
+      "grad_norm": 0.1301536469973807,
+      "learning_rate": 1.4344023323615161e-05,
+      "loss": 0.4648,
+      "step": 246
+    },
+    {
+      "epoch": 0.14422304941245165,
+      "grad_norm": 0.13240230288715193,
+      "learning_rate": 1.4402332361516035e-05,
+      "loss": 0.5249,
+      "step": 247
+    },
+    {
+      "epoch": 0.14480694839792715,
+      "grad_norm": 0.12347915053453282,
+      "learning_rate": 1.446064139941691e-05,
+      "loss": 0.4728,
+      "step": 248
+    },
+    {
+      "epoch": 0.14539084738340266,
+      "grad_norm": 0.11455093839376808,
+      "learning_rate": 1.4518950437317786e-05,
+      "loss": 0.4567,
+      "step": 249
+    },
+    {
+      "epoch": 0.1459747463688782,
+      "grad_norm": 0.12716101270008381,
+      "learning_rate": 1.457725947521866e-05,
+      "loss": 0.4732,
+      "step": 250
+    },
+    {
+      "epoch": 0.1465586453543537,
+      "grad_norm": 0.12508866904505608,
+      "learning_rate": 1.4635568513119536e-05,
+      "loss": 0.4795,
+      "step": 251
+    },
+    {
+      "epoch": 0.1471425443398292,
+      "grad_norm": 0.1139488478374208,
+      "learning_rate": 1.469387755102041e-05,
+      "loss": 0.4992,
+      "step": 252
+    },
+    {
+      "epoch": 0.1477264433253047,
+      "grad_norm": 0.1164406251888475,
+      "learning_rate": 1.4752186588921284e-05,
+      "loss": 0.501,
+      "step": 253
+    },
+    {
+      "epoch": 0.14831034231078025,
+      "grad_norm": 0.12155993797398704,
+      "learning_rate": 1.481049562682216e-05,
+      "loss": 0.4658,
+      "step": 254
+    },
+    {
+      "epoch": 0.14889424129625575,
+      "grad_norm": 0.12076686891831294,
+      "learning_rate": 1.4868804664723034e-05,
+      "loss": 0.4907,
+      "step": 255
+    },
+    {
+      "epoch": 0.14947814028173126,
+      "grad_norm": 0.11817153488165164,
+      "learning_rate": 1.4927113702623908e-05,
+      "loss": 0.4434,
+      "step": 256
+    },
+    {
+      "epoch": 0.15006203926720676,
+      "grad_norm": 0.11433820687060522,
+      "learning_rate": 1.4985422740524782e-05,
+      "loss": 0.4821,
+      "step": 257
+    },
+    {
+      "epoch": 0.1506459382526823,
+      "grad_norm": 0.12304539687178093,
+      "learning_rate": 1.5043731778425656e-05,
+      "loss": 0.4924,
+      "step": 258
+    },
+    {
+      "epoch": 0.1512298372381578,
+      "grad_norm": 0.10750688616443047,
+      "learning_rate": 1.510204081632653e-05,
+      "loss": 0.4316,
+      "step": 259
+    },
+    {
+      "epoch": 0.1518137362236333,
+      "grad_norm": 0.12721684015199142,
+      "learning_rate": 1.5160349854227408e-05,
+      "loss": 0.499,
+      "step": 260
+    },
+    {
+      "epoch": 0.15239763520910882,
+      "grad_norm": 0.12107331994172216,
+      "learning_rate": 1.5218658892128282e-05,
+      "loss": 0.4923,
+      "step": 261
+    },
+    {
+      "epoch": 0.15298153419458432,
+      "grad_norm": 0.11319050995259397,
+      "learning_rate": 1.5276967930029155e-05,
+      "loss": 0.4452,
+      "step": 262
+    },
+    {
+      "epoch": 0.15356543318005986,
+      "grad_norm": 0.1197510106306974,
+      "learning_rate": 1.533527696793003e-05,
+      "loss": 0.4829,
+      "step": 263
+    },
+    {
+      "epoch": 0.15414933216553536,
+      "grad_norm": 0.11659678920962292,
+      "learning_rate": 1.5393586005830907e-05,
+      "loss": 0.4628,
+      "step": 264
+    },
+    {
+      "epoch": 0.15473323115101087,
+      "grad_norm": 0.12101218392668313,
+      "learning_rate": 1.545189504373178e-05,
+      "loss": 0.4955,
+      "step": 265
+    },
+    {
+      "epoch": 0.15531713013648638,
+      "grad_norm": 0.10921033784845438,
+      "learning_rate": 1.5510204081632655e-05,
+      "loss": 0.4397,
+      "step": 266
+    },
+    {
+      "epoch": 0.1559010291219619,
+      "grad_norm": 0.11813323066927452,
+      "learning_rate": 1.5568513119533527e-05,
+      "loss": 0.4944,
+      "step": 267
+    },
+    {
+      "epoch": 0.15648492810743742,
+      "grad_norm": 0.10908211441102279,
+      "learning_rate": 1.5626822157434403e-05,
+      "loss": 0.4898,
+      "step": 268
+    },
+    {
+      "epoch": 0.15706882709291292,
+      "grad_norm": 0.5425118474356572,
+      "learning_rate": 1.568513119533528e-05,
+      "loss": 0.48,
+      "step": 269
+    },
+    {
+      "epoch": 0.15765272607838843,
+      "grad_norm": 0.11763036309777052,
+      "learning_rate": 1.5743440233236155e-05,
+      "loss": 0.475,
+      "step": 270
+    },
+    {
+      "epoch": 0.15823662506386396,
+      "grad_norm": 0.11293238605647272,
+      "learning_rate": 1.5801749271137027e-05,
+      "loss": 0.4894,
+      "step": 271
+    },
+    {
+      "epoch": 0.15882052404933947,
+      "grad_norm": 0.2375406845337137,
+      "learning_rate": 1.5860058309037903e-05,
+      "loss": 0.4544,
+      "step": 272
+    },
+    {
+      "epoch": 0.15940442303481497,
+      "grad_norm": 0.10988697497969815,
+      "learning_rate": 1.5918367346938776e-05,
+      "loss": 0.4839,
+      "step": 273
+    },
+    {
+      "epoch": 0.15998832202029048,
+      "grad_norm": 0.1915901031513559,
+      "learning_rate": 1.597667638483965e-05,
+      "loss": 0.4466,
+      "step": 274
+    },
+    {
+      "epoch": 0.160572221005766,
+      "grad_norm": 1.0829683897429536,
+      "learning_rate": 1.6034985422740524e-05,
+      "loss": 0.6058,
+      "step": 275
+    },
+    {
+      "epoch": 0.16115611999124152,
+      "grad_norm": 0.38124193900265224,
+      "learning_rate": 1.60932944606414e-05,
+      "loss": 0.5439,
+      "step": 276
+    },
+    {
+      "epoch": 0.16174001897671703,
+      "grad_norm": 0.19131045350581488,
+      "learning_rate": 1.6151603498542276e-05,
+      "loss": 0.4459,
+      "step": 277
+    },
+    {
+      "epoch": 0.16232391796219253,
+      "grad_norm": 0.37547972383439737,
+      "learning_rate": 1.6209912536443152e-05,
+      "loss": 0.5011,
+      "step": 278
+    },
+    {
+      "epoch": 0.16290781694766807,
+      "grad_norm": 0.1469161541494368,
+      "learning_rate": 1.6268221574344024e-05,
+      "loss": 0.4887,
+      "step": 279
+    },
+    {
+      "epoch": 0.16349171593314357,
+      "grad_norm": 0.14136715583385023,
+      "learning_rate": 1.63265306122449e-05,
+      "loss": 0.4754,
+      "step": 280
+    },
+    {
+      "epoch": 0.16407561491861908,
+      "grad_norm": 0.13151370092956538,
+      "learning_rate": 1.6384839650145773e-05,
+      "loss": 0.48,
+      "step": 281
+    },
+    {
+      "epoch": 0.16465951390409458,
+      "grad_norm": 0.14033421323078532,
+      "learning_rate": 1.644314868804665e-05,
+      "loss": 0.4478,
+      "step": 282
+    },
+    {
+      "epoch": 0.16524341288957012,
+      "grad_norm": 0.13060926540180795,
+      "learning_rate": 1.6501457725947524e-05,
+      "loss": 0.5163,
+      "step": 283
+    },
+    {
+      "epoch": 0.16582731187504562,
+      "grad_norm": 0.12983437741577833,
+      "learning_rate": 1.6559766763848397e-05,
+      "loss": 0.4631,
+      "step": 284
+    },
+    {
+      "epoch": 0.16641121086052113,
+      "grad_norm": 0.12889027889765484,
+      "learning_rate": 1.6618075801749273e-05,
+      "loss": 0.4437,
+      "step": 285
+    },
+    {
+      "epoch": 0.16699510984599664,
+      "grad_norm": 0.13207553507655626,
+      "learning_rate": 1.667638483965015e-05,
+      "loss": 0.4926,
+      "step": 286
+    },
+    {
+      "epoch": 0.16757900883147214,
+      "grad_norm": 0.12548306221530606,
+      "learning_rate": 1.673469387755102e-05,
+      "loss": 0.4768,
+      "step": 287
+    },
+    {
+      "epoch": 0.16816290781694768,
+      "grad_norm": 0.13005121286615592,
+      "learning_rate": 1.6793002915451897e-05,
+      "loss": 0.5175,
+      "step": 288
+    },
+    {
+      "epoch": 0.16874680680242318,
+      "grad_norm": 0.1205168669946125,
+      "learning_rate": 1.6851311953352773e-05,
+      "loss": 0.4851,
+      "step": 289
+    },
+    {
+      "epoch": 0.1693307057878987,
+      "grad_norm": 0.13074873988342248,
+      "learning_rate": 1.6909620991253645e-05,
+      "loss": 0.4356,
+      "step": 290
+    },
+    {
+      "epoch": 0.1699146047733742,
+      "grad_norm": 0.1270137892030576,
+      "learning_rate": 1.696793002915452e-05,
+      "loss": 0.5365,
+      "step": 291
+    },
+    {
+      "epoch": 0.17049850375884973,
+      "grad_norm": 0.13998772447354765,
+      "learning_rate": 1.7026239067055393e-05,
+      "loss": 0.5091,
+      "step": 292
+    },
+    {
+      "epoch": 0.17108240274432523,
+      "grad_norm": 0.1259767288117116,
+      "learning_rate": 1.708454810495627e-05,
+      "loss": 0.4454,
+      "step": 293
+    },
+    {
+      "epoch": 0.17166630172980074,
+      "grad_norm": 0.11976137204275838,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.4892,
+      "step": 294
+    },
+    {
+      "epoch": 0.17225020071527625,
+      "grad_norm": 0.12624932278606618,
+      "learning_rate": 1.720116618075802e-05,
+      "loss": 0.4863,
+      "step": 295
+    },
+    {
+      "epoch": 0.17283409970075178,
+      "grad_norm": 0.11765777244826535,
+      "learning_rate": 1.7259475218658894e-05,
+      "loss": 0.467,
+      "step": 296
+    },
+    {
+      "epoch": 0.1734179986862273,
+      "grad_norm": 0.1211916458277821,
+      "learning_rate": 1.731778425655977e-05,
+      "loss": 0.4408,
+      "step": 297
+    },
+    {
+      "epoch": 0.1740018976717028,
+      "grad_norm": 0.1166340204965783,
+      "learning_rate": 1.7376093294460642e-05,
+      "loss": 0.4112,
+      "step": 298
+    },
+    {
+      "epoch": 0.1745857966571783,
+      "grad_norm": 0.119543887157399,
+      "learning_rate": 1.7434402332361518e-05,
+      "loss": 0.4759,
+      "step": 299
+    },
+    {
+      "epoch": 0.17516969564265383,
+      "grad_norm": 0.11477499075911166,
+      "learning_rate": 1.749271137026239e-05,
+      "loss": 0.4733,
+      "step": 300
+    },
+    {
+      "epoch": 0.17575359462812934,
+      "grad_norm": 0.1064657376205133,
+      "learning_rate": 1.7551020408163266e-05,
+      "loss": 0.4403,
+      "step": 301
+    },
+    {
+      "epoch": 0.17633749361360485,
+      "grad_norm": 0.12386271456918148,
+      "learning_rate": 1.7609329446064142e-05,
+      "loss": 0.5267,
+      "step": 302
+    },
+    {
+      "epoch": 0.17692139259908035,
+      "grad_norm": 0.13263900171804305,
+      "learning_rate": 1.7667638483965014e-05,
+      "loss": 0.4821,
+      "step": 303
+    },
+    {
+      "epoch": 0.17750529158455589,
+      "grad_norm": 0.10849888180336344,
+      "learning_rate": 1.772594752186589e-05,
+      "loss": 0.4421,
+      "step": 304
+    },
+    {
+      "epoch": 0.1780891905700314,
+      "grad_norm": 0.13996371528380316,
+      "learning_rate": 1.7784256559766766e-05,
+      "loss": 0.4925,
+      "step": 305
+    },
+    {
+      "epoch": 0.1786730895555069,
+      "grad_norm": 0.12340728908837315,
+      "learning_rate": 1.7842565597667642e-05,
+      "loss": 0.4939,
+      "step": 306
+    },
+    {
+      "epoch": 0.1792569885409824,
+      "grad_norm": 0.12093022655658743,
+      "learning_rate": 1.7900874635568515e-05,
+      "loss": 0.4707,
+      "step": 307
+    },
+    {
+      "epoch": 0.1798408875264579,
+      "grad_norm": 0.12390015682018699,
+      "learning_rate": 1.795918367346939e-05,
+      "loss": 0.4504,
+      "step": 308
+    },
+    {
+      "epoch": 0.18042478651193344,
+      "grad_norm": 0.1142597321993865,
+      "learning_rate": 1.8017492711370263e-05,
+      "loss": 0.4478,
+      "step": 309
+    },
+    {
+      "epoch": 0.18100868549740895,
+      "grad_norm": 0.12202487228045188,
+      "learning_rate": 1.807580174927114e-05,
+      "loss": 0.4044,
+      "step": 310
+    },
+    {
+      "epoch": 0.18159258448288446,
+      "grad_norm": 0.11654450303993694,
+      "learning_rate": 1.813411078717201e-05,
+      "loss": 0.4591,
+      "step": 311
+    },
+    {
+      "epoch": 0.18217648346835996,
+      "grad_norm": 0.13971750690087242,
+      "learning_rate": 1.8192419825072887e-05,
+      "loss": 0.4792,
+      "step": 312
+    },
+    {
+      "epoch": 0.1827603824538355,
+      "grad_norm": 0.12506460178471393,
+      "learning_rate": 1.8250728862973763e-05,
+      "loss": 0.5496,
+      "step": 313
+    },
+    {
+      "epoch": 0.183344281439311,
+      "grad_norm": 0.13042197162321098,
+      "learning_rate": 1.830903790087464e-05,
+      "loss": 0.4757,
+      "step": 314
+    },
+    {
+      "epoch": 0.1839281804247865,
+      "grad_norm": 0.11289779621285932,
+      "learning_rate": 1.836734693877551e-05,
+      "loss": 0.4719,
+      "step": 315
+    },
+    {
+      "epoch": 0.18451207941026201,
+      "grad_norm": 0.12229311148485625,
+      "learning_rate": 1.8425655976676387e-05,
+      "loss": 0.465,
+      "step": 316
+    },
+    {
+      "epoch": 0.18509597839573755,
+      "grad_norm": 0.1256475189259005,
+      "learning_rate": 1.848396501457726e-05,
+      "loss": 0.4795,
+      "step": 317
+    },
+    {
+      "epoch": 0.18567987738121305,
+      "grad_norm": 0.12177444572873361,
+      "learning_rate": 1.8542274052478135e-05,
+      "loss": 0.4754,
+      "step": 318
+    },
+    {
+      "epoch": 0.18626377636668856,
+      "grad_norm": 0.12300388648329212,
+      "learning_rate": 1.8600583090379008e-05,
+      "loss": 0.4807,
+      "step": 319
+    },
+    {
+      "epoch": 0.18684767535216407,
+      "grad_norm": 0.11186674724827829,
+      "learning_rate": 1.8658892128279884e-05,
+      "loss": 0.5151,
+      "step": 320
+    },
+    {
+      "epoch": 0.1874315743376396,
+      "grad_norm": 0.10545689286739936,
+      "learning_rate": 1.871720116618076e-05,
+      "loss": 0.4736,
+      "step": 321
+    },
+    {
+      "epoch": 0.1880154733231151,
+      "grad_norm": 0.119809888903107,
+      "learning_rate": 1.8775510204081636e-05,
+      "loss": 0.4199,
+      "step": 322
+    },
+    {
+      "epoch": 0.1885993723085906,
+      "grad_norm": 0.11218428325510385,
+      "learning_rate": 1.8833819241982508e-05,
+      "loss": 0.4731,
+      "step": 323
+    },
+    {
+      "epoch": 0.18918327129406612,
+      "grad_norm": 0.10738594150825266,
+      "learning_rate": 1.8892128279883384e-05,
+      "loss": 0.4298,
+      "step": 324
+    },
+    {
+      "epoch": 0.18976717027954165,
+      "grad_norm": 0.12490192176439074,
+      "learning_rate": 1.895043731778426e-05,
+      "loss": 0.5113,
+      "step": 325
+    },
+    {
+      "epoch": 0.19035106926501716,
+      "grad_norm": 0.12110611612150933,
+      "learning_rate": 1.9008746355685132e-05,
+      "loss": 0.4821,
+      "step": 326
+    },
+    {
+      "epoch": 0.19093496825049266,
+      "grad_norm": 0.11529936265821267,
+      "learning_rate": 1.9067055393586008e-05,
+      "loss": 0.4652,
+      "step": 327
+    },
+    {
+      "epoch": 0.19151886723596817,
+      "grad_norm": 0.11389090421403425,
+      "learning_rate": 1.912536443148688e-05,
+      "loss": 0.4908,
+      "step": 328
+    },
+    {
+      "epoch": 0.19210276622144368,
+      "grad_norm": 0.12195527413916457,
+      "learning_rate": 1.9183673469387756e-05,
+      "loss": 0.4723,
+      "step": 329
+    },
+    {
+      "epoch": 0.1926866652069192,
+      "grad_norm": 0.12662065990683613,
+      "learning_rate": 1.9241982507288632e-05,
+      "loss": 0.4925,
+      "step": 330
+    },
+    {
+      "epoch": 0.19327056419239472,
+      "grad_norm": 0.12083357332091113,
+      "learning_rate": 1.9300291545189508e-05,
+      "loss": 0.5144,
+      "step": 331
+    },
+    {
+      "epoch": 0.19385446317787022,
+      "grad_norm": 0.11527668188298273,
+      "learning_rate": 1.935860058309038e-05,
+      "loss": 0.468,
+      "step": 332
+    },
+    {
+      "epoch": 0.19443836216334573,
+      "grad_norm": 0.12074217632652143,
+      "learning_rate": 1.9416909620991257e-05,
+      "loss": 0.5043,
+      "step": 333
+    },
+    {
+      "epoch": 0.19502226114882126,
+      "grad_norm": 0.12525132790065632,
+      "learning_rate": 1.947521865889213e-05,
+      "loss": 0.5194,
+      "step": 334
+    },
+    {
+      "epoch": 0.19560616013429677,
+      "grad_norm": 0.10131781634232856,
+      "learning_rate": 1.9533527696793005e-05,
+      "loss": 0.4538,
+      "step": 335
+    },
+    {
+      "epoch": 0.19619005911977228,
+      "grad_norm": 0.13067904466053795,
+      "learning_rate": 1.9591836734693877e-05,
+      "loss": 0.4578,
+      "step": 336
+    },
+    {
+      "epoch": 0.19677395810524778,
+      "grad_norm": 0.12785065582654148,
+      "learning_rate": 1.9650145772594753e-05,
+      "loss": 0.5056,
+      "step": 337
+    },
+    {
+      "epoch": 0.19735785709072332,
+      "grad_norm": 0.12011165080837728,
+      "learning_rate": 1.970845481049563e-05,
+      "loss": 0.4879,
+      "step": 338
+    },
+    {
+      "epoch": 0.19794175607619882,
+      "grad_norm": 0.1136447821176873,
+      "learning_rate": 1.9766763848396505e-05,
+      "loss": 0.47,
+      "step": 339
+    },
+    {
+      "epoch": 0.19852565506167433,
+      "grad_norm": 0.12261697055752825,
+      "learning_rate": 1.9825072886297377e-05,
+      "loss": 0.4608,
+      "step": 340
+    },
+    {
+      "epoch": 0.19910955404714983,
+      "grad_norm": 0.12349325296400492,
+      "learning_rate": 1.9883381924198253e-05,
+      "loss": 0.4884,
+      "step": 341
+    },
+    {
+      "epoch": 0.19969345303262537,
+      "grad_norm": 0.13015875125116785,
+      "learning_rate": 1.9941690962099126e-05,
+      "loss": 0.5348,
+      "step": 342
+    },
+    {
+      "epoch": 0.20027735201810087,
+      "grad_norm": 0.12544064781589404,
+      "learning_rate": 2e-05,
+      "loss": 0.5274,
+      "step": 343
+    },
+    {
+      "epoch": 0.20086125100357638,
+      "grad_norm": 0.12242820011660073,
+      "learning_rate": 1.999999480140104e-05,
+      "loss": 0.5015,
+      "step": 344
+    },
+    {
+      "epoch": 0.20144514998905189,
+      "grad_norm": 0.12071117676048071,
+      "learning_rate": 1.999997920560957e-05,
+      "loss": 0.455,
+      "step": 345
+    },
+    {
+      "epoch": 0.20202904897452742,
+      "grad_norm": 0.13681394489211746,
+      "learning_rate": 1.9999953212641804e-05,
+      "loss": 0.5587,
+      "step": 346
+    },
+    {
+      "epoch": 0.20261294796000293,
+      "grad_norm": 0.11239239981504404,
+      "learning_rate": 1.9999916822524766e-05,
+      "loss": 0.458,
+      "step": 347
+    },
+    {
+      "epoch": 0.20319684694547843,
+      "grad_norm": 0.1199213224523809,
+      "learning_rate": 1.999987003529629e-05,
+      "loss": 0.4705,
+      "step": 348
+    },
+    {
+      "epoch": 0.20378074593095394,
+      "grad_norm": 0.12146615255235108,
+      "learning_rate": 1.9999812851005024e-05,
+      "loss": 0.451,
+      "step": 349
+    },
+    {
+      "epoch": 0.20436464491642944,
+      "grad_norm": 0.12449419805381148,
+      "learning_rate": 1.9999745269710423e-05,
+      "loss": 0.4886,
+      "step": 350
+    },
+    {
+      "epoch": 0.20494854390190498,
+      "grad_norm": 0.11854589344041525,
+      "learning_rate": 1.999966729148275e-05,
+      "loss": 0.4593,
+      "step": 351
+    },
+    {
+      "epoch": 0.20553244288738048,
+      "grad_norm": 0.10901489037829362,
+      "learning_rate": 1.9999578916403086e-05,
+      "loss": 0.4629,
+      "step": 352
+    },
+    {
+      "epoch": 0.206116341872856,
+      "grad_norm": 0.11803944691819918,
+      "learning_rate": 1.9999480144563316e-05,
+      "loss": 0.4685,
+      "step": 353
+    },
+    {
+      "epoch": 0.2067002408583315,
+      "grad_norm": 0.12020190738737759,
+      "learning_rate": 1.999937097606613e-05,
+      "loss": 0.485,
+      "step": 354
+    },
+    {
+      "epoch": 0.20728413984380703,
+      "grad_norm": 0.11418107687854197,
+      "learning_rate": 1.9999251411025034e-05,
+      "loss": 0.4807,
+      "step": 355
+    },
+    {
+      "epoch": 0.20786803882928254,
+      "grad_norm": 0.1216249515272213,
+      "learning_rate": 1.9999121449564347e-05,
+      "loss": 0.4588,
+      "step": 356
+    },
+    {
+      "epoch": 0.20845193781475804,
+      "grad_norm": 0.11685759055966302,
+      "learning_rate": 1.999898109181919e-05,
+      "loss": 0.4679,
+      "step": 357
+    },
+    {
+      "epoch": 0.20903583680023355,
+      "grad_norm": 0.10964476553785171,
+      "learning_rate": 1.9998830337935488e-05,
+      "loss": 0.4674,
+      "step": 358
+    },
+    {
+      "epoch": 0.20961973578570908,
+      "grad_norm": 0.12419738483312452,
+      "learning_rate": 1.9998669188069992e-05,
+      "loss": 0.4658,
+      "step": 359
+    },
+    {
+      "epoch": 0.2102036347711846,
+      "grad_norm": 0.1084728039409969,
+      "learning_rate": 1.9998497642390255e-05,
+      "loss": 0.4147,
+      "step": 360
+    },
+    {
+      "epoch": 0.2107875337566601,
+      "grad_norm": 0.117248470764389,
+      "learning_rate": 1.9998315701074624e-05,
+      "loss": 0.4925,
+      "step": 361
+    },
+    {
+      "epoch": 0.2113714327421356,
+      "grad_norm": 0.11355025429871086,
+      "learning_rate": 1.999812336431228e-05,
+      "loss": 0.4665,
+      "step": 362
+    },
+    {
+      "epoch": 0.21195533172761113,
+      "grad_norm": 0.11368010062541409,
+      "learning_rate": 1.9997920632303192e-05,
+      "loss": 0.4522,
+      "step": 363
+    },
+    {
+      "epoch": 0.21253923071308664,
+      "grad_norm": 0.11684278244397342,
+      "learning_rate": 1.9997707505258147e-05,
+      "loss": 0.4389,
+      "step": 364
+    },
+    {
+      "epoch": 0.21312312969856215,
+      "grad_norm": 0.1160841276934915,
+      "learning_rate": 1.9997483983398736e-05,
+      "loss": 0.4722,
+      "step": 365
+    },
+    {
+      "epoch": 0.21370702868403765,
+      "grad_norm": 0.10792011037198496,
+      "learning_rate": 1.9997250066957357e-05,
+      "loss": 0.4879,
+      "step": 366
+    },
+    {
+      "epoch": 0.2142909276695132,
+      "grad_norm": 0.1180391440468137,
+      "learning_rate": 1.9997005756177228e-05,
+      "loss": 0.4932,
+      "step": 367
+    },
+    {
+      "epoch": 0.2148748266549887,
+      "grad_norm": 0.12225887949710575,
+      "learning_rate": 1.9996751051312352e-05,
+      "loss": 0.4686,
+      "step": 368
+    },
+    {
+      "epoch": 0.2154587256404642,
+      "grad_norm": 0.1177779306422269,
+      "learning_rate": 1.9996485952627554e-05,
+      "loss": 0.5305,
+      "step": 369
+    },
+    {
+      "epoch": 0.2160426246259397,
+      "grad_norm": 0.11145359962247031,
+      "learning_rate": 1.9996210460398464e-05,
+      "loss": 0.4649,
+      "step": 370
+    },
+    {
+      "epoch": 0.2166265236114152,
+      "grad_norm": 0.10585225168393766,
+      "learning_rate": 1.9995924574911516e-05,
+      "loss": 0.4482,
+      "step": 371
+    },
+    {
+      "epoch": 0.21721042259689075,
+      "grad_norm": 0.11354558811739207,
+      "learning_rate": 1.9995628296463953e-05,
+      "loss": 0.4978,
+      "step": 372
+    },
+    {
+      "epoch": 0.21779432158236625,
+      "grad_norm": 0.10850367490155878,
+      "learning_rate": 1.9995321625363814e-05,
+      "loss": 0.4403,
+      "step": 373
+    },
+    {
+      "epoch": 0.21837822056784176,
+      "grad_norm": 0.12279018391590048,
+      "learning_rate": 1.999500456192996e-05,
+      "loss": 0.5025,
+      "step": 374
+    },
+    {
+      "epoch": 0.21896211955331726,
+      "grad_norm": 0.14052675039217163,
+      "learning_rate": 1.9994677106492046e-05,
+      "loss": 0.4646,
+      "step": 375
+    },
+    {
+      "epoch": 0.2195460185387928,
+      "grad_norm": 0.11283464921934792,
+      "learning_rate": 1.999433925939053e-05,
+      "loss": 0.4728,
+      "step": 376
+    },
+    {
+      "epoch": 0.2201299175242683,
+      "grad_norm": 0.11074501573213792,
+      "learning_rate": 1.999399102097668e-05,
+      "loss": 0.4471,
+      "step": 377
+    },
+    {
+      "epoch": 0.2207138165097438,
+      "grad_norm": 0.13436987519981916,
+      "learning_rate": 1.999363239161257e-05,
+      "loss": 0.5206,
+      "step": 378
+    },
+    {
+      "epoch": 0.22129771549521932,
+      "grad_norm": 0.10753951089558575,
+      "learning_rate": 1.9993263371671067e-05,
+      "loss": 0.4595,
+      "step": 379
+    },
+    {
+      "epoch": 0.22188161448069485,
+      "grad_norm": 0.13406515973621924,
+      "learning_rate": 1.9992883961535857e-05,
+      "loss": 0.4621,
+      "step": 380
+    },
+    {
+      "epoch": 0.22246551346617036,
+      "grad_norm": 0.11532264719183574,
+      "learning_rate": 1.9992494161601414e-05,
+      "loss": 0.4516,
+      "step": 381
+    },
+    {
+      "epoch": 0.22304941245164586,
+      "grad_norm": 0.11937339356236873,
+      "learning_rate": 1.999209397227302e-05,
+      "loss": 0.4365,
+      "step": 382
+    },
+    {
+      "epoch": 0.22363331143712137,
+      "grad_norm": 0.12903622710778534,
+      "learning_rate": 1.9991683393966764e-05,
+      "loss": 0.4842,
+      "step": 383
+    },
+    {
+      "epoch": 0.2242172104225969,
+      "grad_norm": 0.11597063443745541,
+      "learning_rate": 1.9991262427109532e-05,
+      "loss": 0.4804,
+      "step": 384
+    },
+    {
+      "epoch": 0.2248011094080724,
+      "grad_norm": 0.11598497623107912,
+      "learning_rate": 1.9990831072139008e-05,
+      "loss": 0.4579,
+      "step": 385
+    },
+    {
+      "epoch": 0.22538500839354791,
+      "grad_norm": 0.1179571881754252,
+      "learning_rate": 1.9990389329503685e-05,
+      "loss": 0.4429,
+      "step": 386
+    },
+    {
+      "epoch": 0.22596890737902342,
+      "grad_norm": 0.11816468391070299,
+      "learning_rate": 1.9989937199662845e-05,
+      "loss": 0.4531,
+      "step": 387
+    },
+    {
+      "epoch": 0.22655280636449895,
+      "grad_norm": 0.11357023508805243,
+      "learning_rate": 1.998947468308658e-05,
+      "loss": 0.4217,
+      "step": 388
+    },
+    {
+      "epoch": 0.22713670534997446,
+      "grad_norm": 0.12140194288337988,
+      "learning_rate": 1.9989001780255784e-05,
+      "loss": 0.4568,
+      "step": 389
+    },
+    {
+      "epoch": 0.22772060433544997,
+      "grad_norm": 0.1234465224020816,
+      "learning_rate": 1.9988518491662134e-05,
+      "loss": 0.5316,
+      "step": 390
+    },
+    {
+      "epoch": 0.22830450332092547,
+      "grad_norm": 0.11344600107991246,
+      "learning_rate": 1.9988024817808116e-05,
+      "loss": 0.5095,
+      "step": 391
+    },
+    {
+      "epoch": 0.228888402306401,
+      "grad_norm": 0.10189655395844914,
+      "learning_rate": 1.9987520759207014e-05,
+      "loss": 0.4482,
+      "step": 392
+    },
+    {
+      "epoch": 0.2294723012918765,
+      "grad_norm": 0.11095372651393152,
+      "learning_rate": 1.9987006316382913e-05,
+      "loss": 0.4566,
+      "step": 393
+    },
+    {
+      "epoch": 0.23005620027735202,
+      "grad_norm": 0.11283371382869316,
+      "learning_rate": 1.9986481489870684e-05,
+      "loss": 0.4661,
+      "step": 394
+    },
+    {
+      "epoch": 0.23064009926282752,
+      "grad_norm": 0.11607816103323904,
+      "learning_rate": 1.9985946280215996e-05,
+      "loss": 0.4977,
+      "step": 395
+    },
+    {
+      "epoch": 0.23122399824830303,
+      "grad_norm": 0.11291483623486528,
+      "learning_rate": 1.9985400687975325e-05,
+      "loss": 0.4578,
+      "step": 396
+    },
+    {
+      "epoch": 0.23180789723377856,
+      "grad_norm": 0.12071437963333484,
+      "learning_rate": 1.998484471371593e-05,
+      "loss": 0.4658,
+      "step": 397
+    },
+    {
+      "epoch": 0.23239179621925407,
+      "grad_norm": 0.1268326440708213,
+      "learning_rate": 1.9984278358015867e-05,
+      "loss": 0.485,
+      "step": 398
+    },
+    {
+      "epoch": 0.23297569520472958,
+      "grad_norm": 0.10782883152762061,
+      "learning_rate": 1.998370162146399e-05,
+      "loss": 0.4538,
+      "step": 399
+    },
+    {
+      "epoch": 0.23355959419020508,
+      "grad_norm": 0.11505961984745687,
+      "learning_rate": 1.9983114504659943e-05,
+      "loss": 0.4929,
+      "step": 400
+    },
+    {
+      "epoch": 0.23414349317568062,
+      "grad_norm": 0.11766467224556623,
+      "learning_rate": 1.998251700821416e-05,
+      "loss": 0.5181,
+      "step": 401
+    },
+    {
+      "epoch": 0.23472739216115612,
+      "grad_norm": 0.10835589282791386,
+      "learning_rate": 1.9981909132747875e-05,
+      "loss": 0.4665,
+      "step": 402
+    },
+    {
+      "epoch": 0.23531129114663163,
+      "grad_norm": 0.11869345107281155,
+      "learning_rate": 1.9981290878893103e-05,
+      "loss": 0.4504,
+      "step": 403
+    },
+    {
+      "epoch": 0.23589519013210714,
+      "grad_norm": 0.12310844372103465,
+      "learning_rate": 1.9980662247292657e-05,
+      "loss": 0.4891,
+      "step": 404
+    },
+    {
+      "epoch": 0.23647908911758267,
+      "grad_norm": 0.1156883254381253,
+      "learning_rate": 1.998002323860014e-05,
+      "loss": 0.5028,
+      "step": 405
+    },
+    {
+      "epoch": 0.23706298810305818,
+      "grad_norm": 0.10729427727385983,
+      "learning_rate": 1.997937385347994e-05,
+      "loss": 0.457,
+      "step": 406
+    },
+    {
+      "epoch": 0.23764688708853368,
+      "grad_norm": 0.12132828507618491,
+      "learning_rate": 1.9978714092607234e-05,
+      "loss": 0.5266,
+      "step": 407
+    },
+    {
+      "epoch": 0.2382307860740092,
+      "grad_norm": 0.11781495146242855,
+      "learning_rate": 1.997804395666799e-05,
+      "loss": 0.5442,
+      "step": 408
+    },
+    {
+      "epoch": 0.23881468505948472,
+      "grad_norm": 0.11688647881256983,
+      "learning_rate": 1.997736344635896e-05,
+      "loss": 0.4835,
+      "step": 409
+    },
+    {
+      "epoch": 0.23939858404496023,
+      "grad_norm": 0.10676469089211893,
+      "learning_rate": 1.997667256238769e-05,
+      "loss": 0.4611,
+      "step": 410
+    },
+    {
+      "epoch": 0.23998248303043573,
+      "grad_norm": 0.09960902977661758,
+      "learning_rate": 1.99759713054725e-05,
+      "loss": 0.484,
+      "step": 411
+    },
+    {
+      "epoch": 0.24056638201591124,
+      "grad_norm": 0.13636583384796497,
+      "learning_rate": 1.99752596763425e-05,
+      "loss": 0.4845,
+      "step": 412
+    },
+    {
+      "epoch": 0.24115028100138677,
+      "grad_norm": 0.1132759421197816,
+      "learning_rate": 1.9974537675737587e-05,
+      "loss": 0.4533,
+      "step": 413
+    },
+    {
+      "epoch": 0.24173417998686228,
+      "grad_norm": 0.11454018023407353,
+      "learning_rate": 1.9973805304408437e-05,
+      "loss": 0.4817,
+      "step": 414
+    },
+    {
+      "epoch": 0.24231807897233779,
+      "grad_norm": 0.11078933509507415,
+      "learning_rate": 1.9973062563116515e-05,
+      "loss": 0.4425,
+      "step": 415
+    },
+    {
+      "epoch": 0.2429019779578133,
+      "grad_norm": 0.10562030485038727,
+      "learning_rate": 1.9972309452634064e-05,
+      "loss": 0.4274,
+      "step": 416
+    },
+    {
+      "epoch": 0.2434858769432888,
+      "grad_norm": 0.10983083978468834,
+      "learning_rate": 1.9971545973744102e-05,
+      "loss": 0.437,
+      "step": 417
+    },
+    {
+      "epoch": 0.24406977592876433,
+      "grad_norm": 0.11465402721132896,
+      "learning_rate": 1.997077212724044e-05,
+      "loss": 0.496,
+      "step": 418
+    },
+    {
+      "epoch": 0.24465367491423984,
+      "grad_norm": 0.10819708059126844,
+      "learning_rate": 1.9969987913927657e-05,
+      "loss": 0.4621,
+      "step": 419
+    },
+    {
+      "epoch": 0.24523757389971534,
+      "grad_norm": 0.10720294111160987,
+      "learning_rate": 1.9969193334621117e-05,
+      "loss": 0.4251,
+      "step": 420
+    },
+    {
+      "epoch": 0.24582147288519085,
+      "grad_norm": 0.12761635361687212,
+      "learning_rate": 1.996838839014696e-05,
+      "loss": 0.4858,
+      "step": 421
+    },
+    {
+      "epoch": 0.24640537187066638,
+      "grad_norm": 0.0996996360884402,
+      "learning_rate": 1.9967573081342103e-05,
+      "loss": 0.4295,
+      "step": 422
+    },
+    {
+      "epoch": 0.2469892708561419,
+      "grad_norm": 0.11268958695776551,
+      "learning_rate": 1.9966747409054235e-05,
+      "loss": 0.4665,
+      "step": 423
+    },
+    {
+      "epoch": 0.2475731698416174,
+      "grad_norm": 0.1154869666917733,
+      "learning_rate": 1.996591137414183e-05,
+      "loss": 0.4852,
+      "step": 424
+    },
+    {
+      "epoch": 0.2481570688270929,
+      "grad_norm": 0.10810167220814666,
+      "learning_rate": 1.996506497747412e-05,
+      "loss": 0.478,
+      "step": 425
+    },
+    {
+      "epoch": 0.24874096781256844,
+      "grad_norm": 0.10541284395786413,
+      "learning_rate": 1.9964208219931135e-05,
+      "loss": 0.4809,
+      "step": 426
+    },
+    {
+      "epoch": 0.24932486679804394,
+      "grad_norm": 0.11573086556337946,
+      "learning_rate": 1.9963341102403652e-05,
+      "loss": 0.4759,
+      "step": 427
+    },
+    {
+      "epoch": 0.24990876578351945,
+      "grad_norm": 0.12557548204118502,
+      "learning_rate": 1.996246362579323e-05,
+      "loss": 0.5516,
+      "step": 428
+    },
+    {
+      "epoch": 0.25049266476899495,
+      "grad_norm": 0.10745384381521049,
+      "learning_rate": 1.99615757910122e-05,
+      "loss": 0.4914,
+      "step": 429
+    },
+    {
+      "epoch": 0.25107656375447046,
+      "grad_norm": 0.1107337039692336,
+      "learning_rate": 1.9960677598983672e-05,
+      "loss": 0.4744,
+      "step": 430
+    },
+    {
+      "epoch": 0.25166046273994597,
+      "grad_norm": 0.10852079386333091,
+      "learning_rate": 1.9959769050641498e-05,
+      "loss": 0.4212,
+      "step": 431
+    },
+    {
+      "epoch": 0.25224436172542153,
+      "grad_norm": 0.10239626467035028,
+      "learning_rate": 1.9958850146930326e-05,
+      "loss": 0.4893,
+      "step": 432
+    },
+    {
+      "epoch": 0.25282826071089703,
+      "grad_norm": 0.11482631148356598,
+      "learning_rate": 1.9957920888805548e-05,
+      "loss": 0.4445,
+      "step": 433
+    },
+    {
+      "epoch": 0.25341215969637254,
+      "grad_norm": 0.11374692089385137,
+      "learning_rate": 1.9956981277233342e-05,
+      "loss": 0.4463,
+      "step": 434
+    },
+    {
+      "epoch": 0.25399605868184805,
+      "grad_norm": 0.10443129431727755,
+      "learning_rate": 1.9956031313190634e-05,
+      "loss": 0.4651,
+      "step": 435
+    },
+    {
+      "epoch": 0.25457995766732355,
+      "grad_norm": 0.10144859549961727,
+      "learning_rate": 1.9955070997665122e-05,
+      "loss": 0.4405,
+      "step": 436
+    },
+    {
+      "epoch": 0.25516385665279906,
+      "grad_norm": 0.11100613396402258,
+      "learning_rate": 1.9954100331655265e-05,
+      "loss": 0.4769,
+      "step": 437
+    },
+    {
+      "epoch": 0.25574775563827457,
+      "grad_norm": 0.10563721678375876,
+      "learning_rate": 1.9953119316170286e-05,
+      "loss": 0.4975,
+      "step": 438
+    },
+    {
+      "epoch": 0.25633165462375007,
+      "grad_norm": 0.1135865782319654,
+      "learning_rate": 1.9952127952230166e-05,
+      "loss": 0.446,
+      "step": 439
+    },
+    {
+      "epoch": 0.2569155536092256,
+      "grad_norm": 0.11147753652260058,
+      "learning_rate": 1.995112624086564e-05,
+      "loss": 0.4755,
+      "step": 440
+    },
+    {
+      "epoch": 0.25749945259470114,
+      "grad_norm": 0.10519937206283746,
+      "learning_rate": 1.9950114183118215e-05,
+      "loss": 0.4744,
+      "step": 441
+    },
+    {
+      "epoch": 0.25808335158017665,
+      "grad_norm": 0.1188640291767546,
+      "learning_rate": 1.9949091780040143e-05,
+      "loss": 0.4636,
+      "step": 442
+    },
+    {
+      "epoch": 0.25866725056565215,
+      "grad_norm": 0.10530117243698524,
+      "learning_rate": 1.9948059032694432e-05,
+      "loss": 0.45,
+      "step": 443
+    },
+    {
+      "epoch": 0.25925114955112766,
+      "grad_norm": 0.11739451212907752,
+      "learning_rate": 1.9947015942154864e-05,
+      "loss": 0.5186,
+      "step": 444
+    },
+    {
+      "epoch": 0.25983504853660316,
+      "grad_norm": 0.10693269032594271,
+      "learning_rate": 1.994596250950595e-05,
+      "loss": 0.4397,
+      "step": 445
+    },
+    {
+      "epoch": 0.26041894752207867,
+      "grad_norm": 0.11346051013565371,
+      "learning_rate": 1.9944898735842963e-05,
+      "loss": 0.4938,
+      "step": 446
+    },
+    {
+      "epoch": 0.2610028465075542,
+      "grad_norm": 0.1140707699861024,
+      "learning_rate": 1.9943824622271934e-05,
+      "loss": 0.516,
+      "step": 447
+    },
+    {
+      "epoch": 0.2615867454930297,
+      "grad_norm": 0.11154129019257217,
+      "learning_rate": 1.9942740169909643e-05,
+      "loss": 0.502,
+      "step": 448
+    },
+    {
+      "epoch": 0.26217064447850524,
+      "grad_norm": 0.10250396235439452,
+      "learning_rate": 1.9941645379883613e-05,
+      "loss": 0.4614,
+      "step": 449
+    },
+    {
+      "epoch": 0.26275454346398075,
+      "grad_norm": 0.11367899563603116,
+      "learning_rate": 1.9940540253332118e-05,
+      "loss": 0.4622,
+      "step": 450
+    },
+    {
+      "epoch": 0.26333844244945626,
+      "grad_norm": 0.12599603104469675,
+      "learning_rate": 1.993942479140418e-05,
+      "loss": 0.5493,
+      "step": 451
+    },
+    {
+      "epoch": 0.26392234143493176,
+      "grad_norm": 0.108227004404012,
+      "learning_rate": 1.993829899525957e-05,
+      "loss": 0.4745,
+      "step": 452
+    },
+    {
+      "epoch": 0.26450624042040727,
+      "grad_norm": 0.11918594725013579,
+      "learning_rate": 1.99371628660688e-05,
+      "loss": 0.5183,
+      "step": 453
+    },
+    {
+      "epoch": 0.2650901394058828,
+      "grad_norm": 0.10289972494787383,
+      "learning_rate": 1.9936016405013117e-05,
+      "loss": 0.4144,
+      "step": 454
+    },
+    {
+      "epoch": 0.2656740383913583,
+      "grad_norm": 0.11355493430316647,
+      "learning_rate": 1.9934859613284535e-05,
+      "loss": 0.4479,
+      "step": 455
+    },
+    {
+      "epoch": 0.2662579373768338,
+      "grad_norm": 0.10114934085294533,
+      "learning_rate": 1.993369249208578e-05,
+      "loss": 0.4389,
+      "step": 456
+    },
+    {
+      "epoch": 0.26684183636230935,
+      "grad_norm": 0.11679681784318303,
+      "learning_rate": 1.9932515042630335e-05,
+      "loss": 0.5342,
+      "step": 457
+    },
+    {
+      "epoch": 0.26742573534778485,
+      "grad_norm": 0.09862726287303314,
+      "learning_rate": 1.9931327266142425e-05,
+      "loss": 0.4538,
+      "step": 458
+    },
+    {
+      "epoch": 0.26800963433326036,
+      "grad_norm": 0.10894305909878735,
+      "learning_rate": 1.9930129163856992e-05,
+      "loss": 0.4597,
+      "step": 459
+    },
+    {
+      "epoch": 0.26859353331873587,
+      "grad_norm": 0.10733738422858136,
+      "learning_rate": 1.9928920737019735e-05,
+      "loss": 0.4845,
+      "step": 460
+    },
+    {
+      "epoch": 0.2691774323042114,
+      "grad_norm": 0.12202752693633889,
+      "learning_rate": 1.9927701986887077e-05,
+      "loss": 0.4542,
+      "step": 461
+    },
+    {
+      "epoch": 0.2697613312896869,
+      "grad_norm": 0.11079085534961226,
+      "learning_rate": 1.9926472914726177e-05,
+      "loss": 0.4083,
+      "step": 462
+    },
+    {
+      "epoch": 0.2703452302751624,
+      "grad_norm": 0.10626627798562037,
+      "learning_rate": 1.9925233521814926e-05,
+      "loss": 0.4665,
+      "step": 463
+    },
+    {
+      "epoch": 0.2709291292606379,
+      "grad_norm": 0.1069575247297899,
+      "learning_rate": 1.9923983809441945e-05,
+      "loss": 0.4726,
+      "step": 464
+    },
+    {
+      "epoch": 0.2715130282461134,
+      "grad_norm": 0.10918370786545561,
+      "learning_rate": 1.9922723778906583e-05,
+      "loss": 0.4677,
+      "step": 465
+    },
+    {
+      "epoch": 0.27209692723158896,
+      "grad_norm": 0.1064259449305271,
+      "learning_rate": 1.9921453431518923e-05,
+      "loss": 0.4594,
+      "step": 466
+    },
+    {
+      "epoch": 0.27268082621706446,
+      "grad_norm": 0.1071844642217332,
+      "learning_rate": 1.9920172768599763e-05,
+      "loss": 0.4732,
+      "step": 467
+    },
+    {
+      "epoch": 0.27326472520253997,
+      "grad_norm": 0.10451587859624682,
+      "learning_rate": 1.991888179148064e-05,
+      "loss": 0.4932,
+      "step": 468
+    },
+    {
+      "epoch": 0.2738486241880155,
+      "grad_norm": 0.10417683223925346,
+      "learning_rate": 1.991758050150381e-05,
+      "loss": 0.4738,
+      "step": 469
+    },
+    {
+      "epoch": 0.274432523173491,
+      "grad_norm": 0.10484032570716677,
+      "learning_rate": 1.991626890002224e-05,
+      "loss": 0.4791,
+      "step": 470
+    },
+    {
+      "epoch": 0.2750164221589665,
+      "grad_norm": 0.10343376688742935,
+      "learning_rate": 1.9914946988399636e-05,
+      "loss": 0.4581,
+      "step": 471
+    },
+    {
+      "epoch": 0.275600321144442,
+      "grad_norm": 0.10732143893074236,
+      "learning_rate": 1.9913614768010418e-05,
+      "loss": 0.4903,
+      "step": 472
+    },
+    {
+      "epoch": 0.2761842201299175,
+      "grad_norm": 0.10458036064432694,
+      "learning_rate": 1.9912272240239715e-05,
+      "loss": 0.403,
+      "step": 473
+    },
+    {
+      "epoch": 0.27676811911539306,
+      "grad_norm": 0.1104661719688372,
+      "learning_rate": 1.9910919406483384e-05,
+      "loss": 0.4593,
+      "step": 474
+    },
+    {
+      "epoch": 0.27735201810086857,
+      "grad_norm": 0.118866973934279,
+      "learning_rate": 1.9909556268147995e-05,
+      "loss": 0.4616,
+      "step": 475
+    },
+    {
+      "epoch": 0.2779359170863441,
+      "grad_norm": 0.09481653227038156,
+      "learning_rate": 1.990818282665082e-05,
+      "loss": 0.4162,
+      "step": 476
+    },
+    {
+      "epoch": 0.2785198160718196,
+      "grad_norm": 0.1155272323866598,
+      "learning_rate": 1.9906799083419865e-05,
+      "loss": 0.4482,
+      "step": 477
+    },
+    {
+      "epoch": 0.2791037150572951,
+      "grad_norm": 0.1086230510525182,
+      "learning_rate": 1.9905405039893827e-05,
+      "loss": 0.494,
+      "step": 478
+    },
+    {
+      "epoch": 0.2796876140427706,
+      "grad_norm": 0.10607918408003353,
+      "learning_rate": 1.9904000697522126e-05,
+      "loss": 0.4033,
+      "step": 479
+    },
+    {
+      "epoch": 0.2802715130282461,
+      "grad_norm": 0.11161892554518525,
+      "learning_rate": 1.9902586057764882e-05,
+      "loss": 0.4635,
+      "step": 480
+    },
+    {
+      "epoch": 0.2808554120137216,
+      "grad_norm": 0.10524589834220159,
+      "learning_rate": 1.9901161122092923e-05,
+      "loss": 0.4882,
+      "step": 481
+    },
+    {
+      "epoch": 0.2814393109991971,
+      "grad_norm": 0.10497238569726665,
+      "learning_rate": 1.9899725891987788e-05,
+      "loss": 0.4966,
+      "step": 482
+    },
+    {
+      "epoch": 0.2820232099846727,
+      "grad_norm": 0.1107298232988867,
+      "learning_rate": 1.9898280368941708e-05,
+      "loss": 0.411,
+      "step": 483
+    },
+    {
+      "epoch": 0.2826071089701482,
+      "grad_norm": 0.10454293050095116,
+      "learning_rate": 1.989682455445762e-05,
+      "loss": 0.4233,
+      "step": 484
+    },
+    {
+      "epoch": 0.2831910079556237,
+      "grad_norm": 0.10462702390416419,
+      "learning_rate": 1.9895358450049175e-05,
+      "loss": 0.4831,
+      "step": 485
+    },
+    {
+      "epoch": 0.2837749069410992,
+      "grad_norm": 0.10256787948741952,
+      "learning_rate": 1.9893882057240698e-05,
+      "loss": 0.4747,
+      "step": 486
+    },
+    {
+      "epoch": 0.2843588059265747,
+      "grad_norm": 0.10415827831616743,
+      "learning_rate": 1.989239537756723e-05,
+      "loss": 0.4286,
+      "step": 487
+    },
+    {
+      "epoch": 0.2849427049120502,
+      "grad_norm": 0.10189285568076126,
+      "learning_rate": 1.98908984125745e-05,
+      "loss": 0.4324,
+      "step": 488
+    },
+    {
+      "epoch": 0.2855266038975257,
+      "grad_norm": 0.10386873198569821,
+      "learning_rate": 1.9889391163818935e-05,
+      "loss": 0.4837,
+      "step": 489
+    },
+    {
+      "epoch": 0.2861105028830012,
+      "grad_norm": 0.10123247346576326,
+      "learning_rate": 1.9887873632867645e-05,
+      "loss": 0.4869,
+      "step": 490
+    },
+    {
+      "epoch": 0.2866944018684768,
+      "grad_norm": 0.10404039900935746,
+      "learning_rate": 1.9886345821298442e-05,
+      "loss": 0.4611,
+      "step": 491
+    },
+    {
+      "epoch": 0.2872783008539523,
+      "grad_norm": 0.09896395543347042,
+      "learning_rate": 1.988480773069982e-05,
+      "loss": 0.4697,
+      "step": 492
+    },
+    {
+      "epoch": 0.2878621998394278,
+      "grad_norm": 0.1049554089299839,
+      "learning_rate": 1.9883259362670967e-05,
+      "loss": 0.4749,
+      "step": 493
+    },
+    {
+      "epoch": 0.2884460988249033,
+      "grad_norm": 0.10773883899322642,
+      "learning_rate": 1.9881700718821744e-05,
+      "loss": 0.4496,
+      "step": 494
+    },
+    {
+      "epoch": 0.2890299978103788,
+      "grad_norm": 0.10047202217801549,
+      "learning_rate": 1.988013180077271e-05,
+      "loss": 0.4559,
+      "step": 495
+    },
+    {
+      "epoch": 0.2896138967958543,
+      "grad_norm": 0.09798189240704867,
+      "learning_rate": 1.9878552610155096e-05,
+      "loss": 0.4602,
+      "step": 496
+    },
+    {
+      "epoch": 0.2901977957813298,
+      "grad_norm": 0.10173797906627924,
+      "learning_rate": 1.987696314861082e-05,
+      "loss": 0.4703,
+      "step": 497
+    },
+    {
+      "epoch": 0.2907816947668053,
+      "grad_norm": 0.10704170932000218,
+      "learning_rate": 1.9875363417792477e-05,
+      "loss": 0.4974,
+      "step": 498
+    },
+    {
+      "epoch": 0.2913655937522809,
+      "grad_norm": 0.10481358481313462,
+      "learning_rate": 1.9873753419363336e-05,
+      "loss": 0.4716,
+      "step": 499
+    },
+    {
+      "epoch": 0.2919494927377564,
+      "grad_norm": 0.11012593323760718,
+      "learning_rate": 1.9872133154997345e-05,
+      "loss": 0.5077,
+      "step": 500
+    },
+    {
+      "epoch": 0.2925333917232319,
+      "grad_norm": 0.10322061940418992,
+      "learning_rate": 1.9870502626379127e-05,
+      "loss": 0.462,
+      "step": 501
+    },
+    {
+      "epoch": 0.2931172907087074,
+      "grad_norm": 0.1098428626309581,
+      "learning_rate": 1.986886183520398e-05,
+      "loss": 0.4752,
+      "step": 502
+    },
+    {
+      "epoch": 0.2937011896941829,
+      "grad_norm": 0.10305383250251034,
+      "learning_rate": 1.9867210783177857e-05,
+      "loss": 0.4894,
+      "step": 503
+    },
+    {
+      "epoch": 0.2942850886796584,
+      "grad_norm": 0.10499179103011382,
+      "learning_rate": 1.986554947201739e-05,
+      "loss": 0.4405,
+      "step": 504
+    },
+    {
+      "epoch": 0.2948689876651339,
+      "grad_norm": 0.09795368612018414,
+      "learning_rate": 1.9863877903449883e-05,
+      "loss": 0.4498,
+      "step": 505
+    },
+    {
+      "epoch": 0.2954528866506094,
+      "grad_norm": 0.10181161872285258,
+      "learning_rate": 1.9862196079213298e-05,
+      "loss": 0.4708,
+      "step": 506
+    },
+    {
+      "epoch": 0.29603678563608493,
+      "grad_norm": 0.10876653360791701,
+      "learning_rate": 1.986050400105626e-05,
+      "loss": 0.4577,
+      "step": 507
+    },
+    {
+      "epoch": 0.2966206846215605,
+      "grad_norm": 0.09077334188190138,
+      "learning_rate": 1.9858801670738052e-05,
+      "loss": 0.4284,
+      "step": 508
+    },
+    {
+      "epoch": 0.297204583607036,
+      "grad_norm": 0.1070935098370218,
+      "learning_rate": 1.9857089090028628e-05,
+      "loss": 0.5024,
+      "step": 509
+    },
+    {
+      "epoch": 0.2977884825925115,
+      "grad_norm": 0.10185148799988585,
+      "learning_rate": 1.9855366260708586e-05,
+      "loss": 0.4601,
+      "step": 510
+    },
+    {
+      "epoch": 0.298372381577987,
+      "grad_norm": 0.09618231116714977,
+      "learning_rate": 1.9853633184569187e-05,
+      "loss": 0.4671,
+      "step": 511
+    },
+    {
+      "epoch": 0.2989562805634625,
+      "grad_norm": 0.10689901621920607,
+      "learning_rate": 1.9851889863412347e-05,
+      "loss": 0.4945,
+      "step": 512
+    },
+    {
+      "epoch": 0.299540179548938,
+      "grad_norm": 0.10288159294767096,
+      "learning_rate": 1.985013629905063e-05,
+      "loss": 0.4472,
+      "step": 513
+    },
+    {
+      "epoch": 0.30012407853441353,
+      "grad_norm": 0.09946608447003057,
+      "learning_rate": 1.9848372493307253e-05,
+      "loss": 0.4082,
+      "step": 514
+    },
+    {
+      "epoch": 0.30070797751988904,
+      "grad_norm": 0.10771149065116077,
+      "learning_rate": 1.9846598448016077e-05,
+      "loss": 0.4657,
+      "step": 515
+    },
+    {
+      "epoch": 0.3012918765053646,
+      "grad_norm": 0.10653041719566898,
+      "learning_rate": 1.984481416502161e-05,
+      "loss": 0.457,
+      "step": 516
+    },
+    {
+      "epoch": 0.3018757754908401,
+      "grad_norm": 0.10392890731486312,
+      "learning_rate": 1.9843019646179014e-05,
+      "loss": 0.4738,
+      "step": 517
+    },
+    {
+      "epoch": 0.3024596744763156,
+      "grad_norm": 0.11095586832928041,
+      "learning_rate": 1.984121489335408e-05,
+      "loss": 0.4624,
+      "step": 518
+    },
+    {
+      "epoch": 0.3030435734617911,
+      "grad_norm": 0.10547644414386387,
+      "learning_rate": 1.9839399908423248e-05,
+      "loss": 0.475,
+      "step": 519
+    },
+    {
+      "epoch": 0.3036274724472666,
+      "grad_norm": 0.10557703469786497,
+      "learning_rate": 1.983757469327359e-05,
+      "loss": 0.4782,
+      "step": 520
+    },
+    {
+      "epoch": 0.30421137143274213,
+      "grad_norm": 0.11038640014002644,
+      "learning_rate": 1.983573924980282e-05,
+      "loss": 0.467,
+      "step": 521
+    },
+    {
+      "epoch": 0.30479527041821763,
+      "grad_norm": 0.10292954844395266,
+      "learning_rate": 1.983389357991929e-05,
+      "loss": 0.485,
+      "step": 522
+    },
+    {
+      "epoch": 0.30537916940369314,
+      "grad_norm": 0.09492161360120267,
+      "learning_rate": 1.9832037685541973e-05,
+      "loss": 0.4077,
+      "step": 523
+    },
+    {
+      "epoch": 0.30596306838916865,
+      "grad_norm": 0.1119335542917095,
+      "learning_rate": 1.983017156860048e-05,
+      "loss": 0.4784,
+      "step": 524
+    },
+    {
+      "epoch": 0.3065469673746442,
+      "grad_norm": 0.09773676307400724,
+      "learning_rate": 1.9828295231035054e-05,
+      "loss": 0.4377,
+      "step": 525
+    },
+    {
+      "epoch": 0.3071308663601197,
+      "grad_norm": 0.1106925665551254,
+      "learning_rate": 1.9826408674796552e-05,
+      "loss": 0.4754,
+      "step": 526
+    },
+    {
+      "epoch": 0.3077147653455952,
+      "grad_norm": 0.10528103581491247,
+      "learning_rate": 1.9824511901846475e-05,
+      "loss": 0.4975,
+      "step": 527
+    },
+    {
+      "epoch": 0.3082986643310707,
+      "grad_norm": 0.10126346157573958,
+      "learning_rate": 1.9822604914156927e-05,
+      "loss": 0.4706,
+      "step": 528
+    },
+    {
+      "epoch": 0.30888256331654623,
+      "grad_norm": 0.10609013260438106,
+      "learning_rate": 1.982068771371064e-05,
+      "loss": 0.5158,
+      "step": 529
+    },
+    {
+      "epoch": 0.30946646230202174,
+      "grad_norm": 0.10528349027908086,
+      "learning_rate": 1.9818760302500976e-05,
+      "loss": 0.4746,
+      "step": 530
+    },
+    {
+      "epoch": 0.31005036128749724,
+      "grad_norm": 0.09748314430770585,
+      "learning_rate": 1.9816822682531888e-05,
+      "loss": 0.4348,
+      "step": 531
+    },
+    {
+      "epoch": 0.31063426027297275,
+      "grad_norm": 0.10112576935919916,
+      "learning_rate": 1.981487485581797e-05,
+      "loss": 0.4252,
+      "step": 532
+    },
+    {
+      "epoch": 0.3112181592584483,
+      "grad_norm": 0.10563773606372161,
+      "learning_rate": 1.9812916824384406e-05,
+      "loss": 0.4095,
+      "step": 533
+    },
+    {
+      "epoch": 0.3118020582439238,
+      "grad_norm": 0.10238684133912998,
+      "learning_rate": 1.9810948590267013e-05,
+      "loss": 0.4566,
+      "step": 534
+    },
+    {
+      "epoch": 0.3123859572293993,
+      "grad_norm": 0.09945994206311794,
+      "learning_rate": 1.9808970155512187e-05,
+      "loss": 0.4662,
+      "step": 535
+    },
+    {
+      "epoch": 0.31296985621487483,
+      "grad_norm": 0.10462537167794941,
+      "learning_rate": 1.9806981522176957e-05,
+      "loss": 0.4434,
+      "step": 536
+    },
+    {
+      "epoch": 0.31355375520035034,
+      "grad_norm": 0.1018383230025197,
+      "learning_rate": 1.9804982692328944e-05,
+      "loss": 0.4331,
+      "step": 537
+    },
+    {
+      "epoch": 0.31413765418582584,
+      "grad_norm": 0.10612166595196044,
+      "learning_rate": 1.9802973668046364e-05,
+      "loss": 0.4699,
+      "step": 538
+    },
+    {
+      "epoch": 0.31472155317130135,
+      "grad_norm": 0.10287085591894089,
+      "learning_rate": 1.9800954451418044e-05,
+      "loss": 0.5165,
+      "step": 539
+    },
+    {
+      "epoch": 0.31530545215677686,
+      "grad_norm": 0.10717383750671472,
+      "learning_rate": 1.9798925044543402e-05,
+      "loss": 0.4835,
+      "step": 540
+    },
+    {
+      "epoch": 0.3158893511422524,
+      "grad_norm": 0.09849169328913834,
+      "learning_rate": 1.9796885449532456e-05,
+      "loss": 0.5,
+      "step": 541
+    },
+    {
+      "epoch": 0.3164732501277279,
+      "grad_norm": 0.10099217592847301,
+      "learning_rate": 1.979483566850581e-05,
+      "loss": 0.4591,
+      "step": 542
+    },
+    {
+      "epoch": 0.31705714911320343,
+      "grad_norm": 0.09945056831759849,
+      "learning_rate": 1.9792775703594663e-05,
+      "loss": 0.4475,
+      "step": 543
+    },
+    {
+      "epoch": 0.31764104809867894,
+      "grad_norm": 0.09862674689963181,
+      "learning_rate": 1.97907055569408e-05,
+      "loss": 0.431,
+      "step": 544
+    },
+    {
+      "epoch": 0.31822494708415444,
+      "grad_norm": 0.10306370447196317,
+      "learning_rate": 1.9788625230696596e-05,
+      "loss": 0.4459,
+      "step": 545
+    },
+    {
+      "epoch": 0.31880884606962995,
+      "grad_norm": 0.10716484174380612,
+      "learning_rate": 1.9786534727025005e-05,
+      "loss": 0.498,
+      "step": 546
+    },
+    {
+      "epoch": 0.31939274505510545,
+      "grad_norm": 0.10535045009335002,
+      "learning_rate": 1.9784434048099565e-05,
+      "loss": 0.4942,
+      "step": 547
+    },
+    {
+      "epoch": 0.31997664404058096,
+      "grad_norm": 0.09638841967133711,
+      "learning_rate": 1.9782323196104395e-05,
+      "loss": 0.4309,
+      "step": 548
+    },
+    {
+      "epoch": 0.32056054302605647,
+      "grad_norm": 0.09262032907116607,
+      "learning_rate": 1.978020217323419e-05,
+      "loss": 0.432,
+      "step": 549
+    },
+    {
+      "epoch": 0.321144442011532,
+      "grad_norm": 0.10429926504528186,
+      "learning_rate": 1.9778070981694216e-05,
+      "loss": 0.4561,
+      "step": 550
+    },
+    {
+      "epoch": 0.32172834099700753,
+      "grad_norm": 0.09756371938106655,
+      "learning_rate": 1.9775929623700318e-05,
+      "loss": 0.4822,
+      "step": 551
+    },
+    {
+      "epoch": 0.32231223998248304,
+      "grad_norm": 0.09893013972994963,
+      "learning_rate": 1.9773778101478908e-05,
+      "loss": 0.4783,
+      "step": 552
+    },
+    {
+      "epoch": 0.32289613896795855,
+      "grad_norm": 0.09654713909810991,
+      "learning_rate": 1.9771616417266966e-05,
+      "loss": 0.4261,
+      "step": 553
+    },
+    {
+      "epoch": 0.32348003795343405,
+      "grad_norm": 0.11096730212299205,
+      "learning_rate": 1.976944457331204e-05,
+      "loss": 0.5347,
+      "step": 554
+    },
+    {
+      "epoch": 0.32406393693890956,
+      "grad_norm": 0.09169964758560968,
+      "learning_rate": 1.976726257187223e-05,
+      "loss": 0.4243,
+      "step": 555
+    },
+    {
+      "epoch": 0.32464783592438506,
+      "grad_norm": 0.09385694465764807,
+      "learning_rate": 1.9765070415216218e-05,
+      "loss": 0.44,
+      "step": 556
+    },
+    {
+      "epoch": 0.32523173490986057,
+      "grad_norm": 0.0916276923694328,
+      "learning_rate": 1.976286810562323e-05,
+      "loss": 0.4346,
+      "step": 557
+    },
+    {
+      "epoch": 0.32581563389533613,
+      "grad_norm": 0.10128295744407903,
+      "learning_rate": 1.976065564538304e-05,
+      "loss": 0.4922,
+      "step": 558
+    },
+    {
+      "epoch": 0.32639953288081164,
+      "grad_norm": 0.10652643192501365,
+      "learning_rate": 1.9758433036796003e-05,
+      "loss": 0.4739,
+      "step": 559
+    },
+    {
+      "epoch": 0.32698343186628714,
+      "grad_norm": 0.09817424320710791,
+      "learning_rate": 1.9756200282173e-05,
+      "loss": 0.4321,
+      "step": 560
+    },
+    {
+      "epoch": 0.32756733085176265,
+      "grad_norm": 0.10016381192195714,
+      "learning_rate": 1.975395738383547e-05,
+      "loss": 0.444,
+      "step": 561
+    },
+    {
+      "epoch": 0.32815122983723816,
+      "grad_norm": 0.10444577297114783,
+      "learning_rate": 1.9751704344115402e-05,
+      "loss": 0.4476,
+      "step": 562
+    },
+    {
+      "epoch": 0.32873512882271366,
+      "grad_norm": 0.10039062408618125,
+      "learning_rate": 1.9749441165355322e-05,
+      "loss": 0.4297,
+      "step": 563
+    },
+    {
+      "epoch": 0.32931902780818917,
+      "grad_norm": 0.09875795451979454,
+      "learning_rate": 1.9747167849908305e-05,
+      "loss": 0.4383,
+      "step": 564
+    },
+    {
+      "epoch": 0.3299029267936647,
+      "grad_norm": 0.09666318701179007,
+      "learning_rate": 1.974488440013796e-05,
+      "loss": 0.4852,
+      "step": 565
+    },
+    {
+      "epoch": 0.33048682577914024,
+      "grad_norm": 0.10313249813674931,
+      "learning_rate": 1.9742590818418435e-05,
+      "loss": 0.4436,
+      "step": 566
+    },
+    {
+      "epoch": 0.33107072476461574,
+      "grad_norm": 0.11633929628696882,
+      "learning_rate": 1.9740287107134417e-05,
+      "loss": 0.4624,
+      "step": 567
+    },
+    {
+      "epoch": 0.33165462375009125,
+      "grad_norm": 0.09960284204588069,
+      "learning_rate": 1.9737973268681117e-05,
+      "loss": 0.4535,
+      "step": 568
+    },
+    {
+      "epoch": 0.33223852273556675,
+      "grad_norm": 0.1169583839114291,
+      "learning_rate": 1.9735649305464274e-05,
+      "loss": 0.4805,
+      "step": 569
+    },
+    {
+      "epoch": 0.33282242172104226,
+      "grad_norm": 0.1026043366273224,
+      "learning_rate": 1.9733315219900165e-05,
+      "loss": 0.4661,
+      "step": 570
+    },
+    {
+      "epoch": 0.33340632070651777,
+      "grad_norm": 0.10475575192278998,
+      "learning_rate": 1.9730971014415585e-05,
+      "loss": 0.4296,
+      "step": 571
+    },
+    {
+      "epoch": 0.3339902196919933,
+      "grad_norm": 0.11480588492894797,
+      "learning_rate": 1.9728616691447845e-05,
+      "loss": 0.4082,
+      "step": 572
+    },
+    {
+      "epoch": 0.3345741186774688,
+      "grad_norm": 0.12305621238134759,
+      "learning_rate": 1.9726252253444784e-05,
+      "loss": 0.447,
+      "step": 573
+    },
+    {
+      "epoch": 0.3351580176629443,
+      "grad_norm": 0.10699926083368476,
+      "learning_rate": 1.9723877702864758e-05,
+      "loss": 0.5098,
+      "step": 574
+    },
+    {
+      "epoch": 0.33574191664841985,
+      "grad_norm": 0.11987531948759156,
+      "learning_rate": 1.9721493042176632e-05,
+      "loss": 0.4826,
+      "step": 575
+    },
+    {
+      "epoch": 0.33632581563389535,
+      "grad_norm": 0.1280286425772078,
+      "learning_rate": 1.9719098273859782e-05,
+      "loss": 0.4946,
+      "step": 576
+    },
+    {
+      "epoch": 0.33690971461937086,
+      "grad_norm": 0.10107951754081562,
+      "learning_rate": 1.97166934004041e-05,
+      "loss": 0.4669,
+      "step": 577
+    },
+    {
+      "epoch": 0.33749361360484637,
+      "grad_norm": 0.11584970986470462,
+      "learning_rate": 1.9714278424309983e-05,
+      "loss": 0.4393,
+      "step": 578
+    },
+    {
+      "epoch": 0.33807751259032187,
+      "grad_norm": 0.10506544620725017,
+      "learning_rate": 1.971185334808832e-05,
+      "loss": 0.4404,
+      "step": 579
+    },
+    {
+      "epoch": 0.3386614115757974,
+      "grad_norm": 0.1043306665386585,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.4953,
+      "step": 580
+    },
+    {
+      "epoch": 0.3392453105612729,
+      "grad_norm": 0.12073066735328658,
+      "learning_rate": 1.970697290535848e-05,
+      "loss": 0.4596,
+      "step": 581
+    },
+    {
+      "epoch": 0.3398292095467484,
+      "grad_norm": 0.12141257050661294,
+      "learning_rate": 1.970451754392459e-05,
+      "loss": 0.4087,
+      "step": 582
+    },
+    {
+      "epoch": 0.34041310853222395,
+      "grad_norm": 0.10661325971837922,
+      "learning_rate": 1.970205209251174e-05,
+      "loss": 0.4458,
+      "step": 583
+    },
+    {
+      "epoch": 0.34099700751769946,
+      "grad_norm": 0.10641736959092607,
+      "learning_rate": 1.9699576553683308e-05,
+      "loss": 0.462,
+      "step": 584
+    },
+    {
+      "epoch": 0.34158090650317496,
+      "grad_norm": 0.11248159589690276,
+      "learning_rate": 1.9697090930013164e-05,
+      "loss": 0.4897,
+      "step": 585
+    },
+    {
+      "epoch": 0.34216480548865047,
+      "grad_norm": 0.10079856786156394,
+      "learning_rate": 1.969459522408566e-05,
+      "loss": 0.4076,
+      "step": 586
+    },
+    {
+      "epoch": 0.342748704474126,
+      "grad_norm": 0.11432689610629138,
+      "learning_rate": 1.9692089438495622e-05,
+      "loss": 0.4264,
+      "step": 587
+    },
+    {
+      "epoch": 0.3433326034596015,
+      "grad_norm": 0.1142033430751992,
+      "learning_rate": 1.9689573575848375e-05,
+      "loss": 0.5091,
+      "step": 588
+    },
+    {
+      "epoch": 0.343916502445077,
+      "grad_norm": 0.10697617680420486,
+      "learning_rate": 1.9687047638759707e-05,
+      "loss": 0.4488,
+      "step": 589
+    },
+    {
+      "epoch": 0.3445004014305525,
+      "grad_norm": 0.10680659418024134,
+      "learning_rate": 1.968451162985589e-05,
+      "loss": 0.4512,
+      "step": 590
+    },
+    {
+      "epoch": 0.345084300416028,
+      "grad_norm": 0.11396017317676128,
+      "learning_rate": 1.9681965551773653e-05,
+      "loss": 0.4557,
+      "step": 591
+    },
+    {
+      "epoch": 0.34566819940150356,
+      "grad_norm": 0.09636231708229258,
+      "learning_rate": 1.967940940716021e-05,
+      "loss": 0.4534,
+      "step": 592
+    },
+    {
+      "epoch": 0.34625209838697907,
+      "grad_norm": 0.104176024792229,
+      "learning_rate": 1.9676843198673237e-05,
+      "loss": 0.4417,
+      "step": 593
+    },
+    {
+      "epoch": 0.3468359973724546,
+      "grad_norm": 0.10129706323111469,
+      "learning_rate": 1.9674266928980863e-05,
+      "loss": 0.4519,
+      "step": 594
+    },
+    {
+      "epoch": 0.3474198963579301,
+      "grad_norm": 0.09892235339720071,
+      "learning_rate": 1.9671680600761694e-05,
+      "loss": 0.4087,
+      "step": 595
+    },
+    {
+      "epoch": 0.3480037953434056,
+      "grad_norm": 0.10042575468707035,
+      "learning_rate": 1.966908421670479e-05,
+      "loss": 0.4598,
+      "step": 596
+    },
+    {
+      "epoch": 0.3485876943288811,
+      "grad_norm": 0.10549149811716663,
+      "learning_rate": 1.9666477779509655e-05,
+      "loss": 0.4179,
+      "step": 597
+    },
+    {
+      "epoch": 0.3491715933143566,
+      "grad_norm": 0.10669702508193217,
+      "learning_rate": 1.9663861291886256e-05,
+      "loss": 0.475,
+      "step": 598
+    },
+    {
+      "epoch": 0.3497554922998321,
+      "grad_norm": 0.09225560953952555,
+      "learning_rate": 1.9661234756555006e-05,
+      "loss": 0.4082,
+      "step": 599
+    },
+    {
+      "epoch": 0.35033939128530767,
+      "grad_norm": 0.10192650465779941,
+      "learning_rate": 1.965859817624677e-05,
+      "loss": 0.4381,
+      "step": 600
+    },
+    {
+      "epoch": 0.3509232902707832,
+      "grad_norm": 0.10653469998367042,
+      "learning_rate": 1.9655951553702848e-05,
+      "loss": 0.4391,
+      "step": 601
+    },
+    {
+      "epoch": 0.3515071892562587,
+      "grad_norm": 0.1010730429278644,
+      "learning_rate": 1.965329489167499e-05,
+      "loss": 0.4506,
+      "step": 602
+    },
+    {
+      "epoch": 0.3520910882417342,
+      "grad_norm": 0.09845106155606329,
+      "learning_rate": 1.9650628192925372e-05,
+      "loss": 0.4518,
+      "step": 603
+    },
+    {
+      "epoch": 0.3526749872272097,
+      "grad_norm": 0.11014119931780321,
+      "learning_rate": 1.9647951460226622e-05,
+      "loss": 0.4795,
+      "step": 604
+    },
+    {
+      "epoch": 0.3532588862126852,
+      "grad_norm": 0.09885364770110962,
+      "learning_rate": 1.964526469636179e-05,
+      "loss": 0.4439,
+      "step": 605
+    },
+    {
+      "epoch": 0.3538427851981607,
+      "grad_norm": 0.0988432071177654,
+      "learning_rate": 1.9642567904124354e-05,
+      "loss": 0.4671,
+      "step": 606
+    },
+    {
+      "epoch": 0.3544266841836362,
+      "grad_norm": 0.10494451798383142,
+      "learning_rate": 1.963986108631823e-05,
+      "loss": 0.5173,
+      "step": 607
+    },
+    {
+      "epoch": 0.35501058316911177,
+      "grad_norm": 0.10221565330702487,
+      "learning_rate": 1.9637144245757742e-05,
+      "loss": 0.4394,
+      "step": 608
+    },
+    {
+      "epoch": 0.3555944821545873,
+      "grad_norm": 0.09279429026252851,
+      "learning_rate": 1.9634417385267643e-05,
+      "loss": 0.4379,
+      "step": 609
+    },
+    {
+      "epoch": 0.3561783811400628,
+      "grad_norm": 0.10881755831674576,
+      "learning_rate": 1.963168050768311e-05,
+      "loss": 0.4718,
+      "step": 610
+    },
+    {
+      "epoch": 0.3567622801255383,
+      "grad_norm": 0.09851011417925018,
+      "learning_rate": 1.9628933615849726e-05,
+      "loss": 0.4752,
+      "step": 611
+    },
+    {
+      "epoch": 0.3573461791110138,
+      "grad_norm": 0.094582053937154,
+      "learning_rate": 1.962617671262349e-05,
+      "loss": 0.4416,
+      "step": 612
+    },
+    {
+      "epoch": 0.3579300780964893,
+      "grad_norm": 0.10840938538701234,
+      "learning_rate": 1.9623409800870804e-05,
+      "loss": 0.4476,
+      "step": 613
+    },
+    {
+      "epoch": 0.3585139770819648,
+      "grad_norm": 0.10571175284700286,
+      "learning_rate": 1.9620632883468484e-05,
+      "loss": 0.464,
+      "step": 614
+    },
+    {
+      "epoch": 0.3590978760674403,
+      "grad_norm": 0.10128754683519471,
+      "learning_rate": 1.9617845963303744e-05,
+      "loss": 0.4483,
+      "step": 615
+    },
+    {
+      "epoch": 0.3596817750529158,
+      "grad_norm": 0.09655830156321613,
+      "learning_rate": 1.9615049043274207e-05,
+      "loss": 0.4235,
+      "step": 616
+    },
+    {
+      "epoch": 0.3602656740383914,
+      "grad_norm": 0.10804304370615793,
+      "learning_rate": 1.9612242126287876e-05,
+      "loss": 0.4526,
+      "step": 617
+    },
+    {
+      "epoch": 0.3608495730238669,
+      "grad_norm": 0.09730786315746394,
+      "learning_rate": 1.960942521526317e-05,
+      "loss": 0.4606,
+      "step": 618
+    },
+    {
+      "epoch": 0.3614334720093424,
+      "grad_norm": 0.09056602919230336,
+      "learning_rate": 1.9606598313128874e-05,
+      "loss": 0.4224,
+      "step": 619
+    },
+    {
+      "epoch": 0.3620173709948179,
+      "grad_norm": 0.10149928679795497,
+      "learning_rate": 1.9603761422824187e-05,
+      "loss": 0.445,
+      "step": 620
+    },
+    {
+      "epoch": 0.3626012699802934,
+      "grad_norm": 0.10713164173398641,
+      "learning_rate": 1.9600914547298666e-05,
+      "loss": 0.4543,
+      "step": 621
+    },
+    {
+      "epoch": 0.3631851689657689,
+      "grad_norm": 0.10027595609052473,
+      "learning_rate": 1.9598057689512277e-05,
+      "loss": 0.4471,
+      "step": 622
+    },
+    {
+      "epoch": 0.3637690679512444,
+      "grad_norm": 0.10398880481621489,
+      "learning_rate": 1.9595190852435345e-05,
+      "loss": 0.4443,
+      "step": 623
+    },
+    {
+      "epoch": 0.3643529669367199,
+      "grad_norm": 0.10738522755700622,
+      "learning_rate": 1.9592314039048575e-05,
+      "loss": 0.4177,
+      "step": 624
+    },
+    {
+      "epoch": 0.3649368659221955,
+      "grad_norm": 0.09588654491818868,
+      "learning_rate": 1.9589427252343054e-05,
+      "loss": 0.4638,
+      "step": 625
+    },
+    {
+      "epoch": 0.365520764907671,
+      "grad_norm": 0.10265160959651286,
+      "learning_rate": 1.9586530495320227e-05,
+      "loss": 0.4646,
+      "step": 626
+    },
+    {
+      "epoch": 0.3661046638931465,
+      "grad_norm": 0.10818657579881,
+      "learning_rate": 1.958362377099191e-05,
+      "loss": 0.4814,
+      "step": 627
+    },
+    {
+      "epoch": 0.366688562878622,
+      "grad_norm": 0.10068301523742364,
+      "learning_rate": 1.958070708238028e-05,
+      "loss": 0.4902,
+      "step": 628
+    },
+    {
+      "epoch": 0.3672724618640975,
+      "grad_norm": 0.0920273263546919,
+      "learning_rate": 1.957778043251788e-05,
+      "loss": 0.4358,
+      "step": 629
+    },
+    {
+      "epoch": 0.367856360849573,
+      "grad_norm": 0.09936314943348383,
+      "learning_rate": 1.9574843824447602e-05,
+      "loss": 0.4681,
+      "step": 630
+    },
+    {
+      "epoch": 0.3684402598350485,
+      "grad_norm": 0.09966687109605014,
+      "learning_rate": 1.9571897261222694e-05,
+      "loss": 0.4141,
+      "step": 631
+    },
+    {
+      "epoch": 0.36902415882052403,
+      "grad_norm": 0.09981208351049878,
+      "learning_rate": 1.9568940745906762e-05,
+      "loss": 0.4785,
+      "step": 632
+    },
+    {
+      "epoch": 0.36960805780599953,
+      "grad_norm": 0.10522330733898372,
+      "learning_rate": 1.956597428157375e-05,
+      "loss": 0.4619,
+      "step": 633
+    },
+    {
+      "epoch": 0.3701919567914751,
+      "grad_norm": 0.09991587286696334,
+      "learning_rate": 1.956299787130795e-05,
+      "loss": 0.463,
+      "step": 634
+    },
+    {
+      "epoch": 0.3707758557769506,
+      "grad_norm": 0.10870057723265572,
+      "learning_rate": 1.9560011518203996e-05,
+      "loss": 0.5213,
+      "step": 635
+    },
+    {
+      "epoch": 0.3713597547624261,
+      "grad_norm": 0.09830468988305653,
+      "learning_rate": 1.9557015225366855e-05,
+      "loss": 0.4623,
+      "step": 636
+    },
+    {
+      "epoch": 0.3719436537479016,
+      "grad_norm": 0.09762859184593747,
+      "learning_rate": 1.9554008995911837e-05,
+      "loss": 0.4412,
+      "step": 637
+    },
+    {
+      "epoch": 0.3725275527333771,
+      "grad_norm": 0.10594559752467105,
+      "learning_rate": 1.9550992832964575e-05,
+      "loss": 0.4292,
+      "step": 638
+    },
+    {
+      "epoch": 0.3731114517188526,
+      "grad_norm": 0.10197977727372742,
+      "learning_rate": 1.9547966739661032e-05,
+      "loss": 0.4449,
+      "step": 639
+    },
+    {
+      "epoch": 0.37369535070432813,
+      "grad_norm": 0.10014602324348887,
+      "learning_rate": 1.95449307191475e-05,
+      "loss": 0.4576,
+      "step": 640
+    },
+    {
+      "epoch": 0.37427924968980364,
+      "grad_norm": 0.09751360675504704,
+      "learning_rate": 1.9541884774580588e-05,
+      "loss": 0.4462,
+      "step": 641
+    },
+    {
+      "epoch": 0.3748631486752792,
+      "grad_norm": 0.1060775301734996,
+      "learning_rate": 1.953882890912723e-05,
+      "loss": 0.3884,
+      "step": 642
+    },
+    {
+      "epoch": 0.3754470476607547,
+      "grad_norm": 0.09621765349288934,
+      "learning_rate": 1.953576312596466e-05,
+      "loss": 0.4735,
+      "step": 643
+    },
+    {
+      "epoch": 0.3760309466462302,
+      "grad_norm": 0.10353302529924135,
+      "learning_rate": 1.9532687428280442e-05,
+      "loss": 0.4338,
+      "step": 644
+    },
+    {
+      "epoch": 0.3766148456317057,
+      "grad_norm": 0.09427258387881038,
+      "learning_rate": 1.952960181927243e-05,
+      "loss": 0.4245,
+      "step": 645
+    },
+    {
+      "epoch": 0.3771987446171812,
+      "grad_norm": 0.09362598136479655,
+      "learning_rate": 1.9526506302148805e-05,
+      "loss": 0.3827,
+      "step": 646
+    },
+    {
+      "epoch": 0.37778264360265673,
+      "grad_norm": 0.10159603551377956,
+      "learning_rate": 1.9523400880128032e-05,
+      "loss": 0.441,
+      "step": 647
+    },
+    {
+      "epoch": 0.37836654258813224,
+      "grad_norm": 0.10602170108431803,
+      "learning_rate": 1.952028555643888e-05,
+      "loss": 0.4684,
+      "step": 648
+    },
+    {
+      "epoch": 0.37895044157360774,
+      "grad_norm": 0.10149590806285146,
+      "learning_rate": 1.9517160334320405e-05,
+      "loss": 0.4587,
+      "step": 649
+    },
+    {
+      "epoch": 0.3795343405590833,
+      "grad_norm": 0.10094536203014397,
+      "learning_rate": 1.9514025217021976e-05,
+      "loss": 0.4693,
+      "step": 650
+    },
+    {
+      "epoch": 0.3801182395445588,
+      "grad_norm": 0.09969003497734165,
+      "learning_rate": 1.951088020780323e-05,
+      "loss": 0.4496,
+      "step": 651
+    },
+    {
+      "epoch": 0.3807021385300343,
+      "grad_norm": 0.10828597128672715,
+      "learning_rate": 1.950772530993409e-05,
+      "loss": 0.4591,
+      "step": 652
+    },
+    {
+      "epoch": 0.3812860375155098,
+      "grad_norm": 0.0914674076601711,
+      "learning_rate": 1.9504560526694773e-05,
+      "loss": 0.4396,
+      "step": 653
+    },
+    {
+      "epoch": 0.38186993650098533,
+      "grad_norm": 0.10004953582390431,
+      "learning_rate": 1.9501385861375765e-05,
+      "loss": 0.4729,
+      "step": 654
+    },
+    {
+      "epoch": 0.38245383548646084,
+      "grad_norm": 0.09665989335445614,
+      "learning_rate": 1.949820131727783e-05,
+      "loss": 0.4118,
+      "step": 655
+    },
+    {
+      "epoch": 0.38303773447193634,
+      "grad_norm": 0.10764448034584265,
+      "learning_rate": 1.9495006897711994e-05,
+      "loss": 0.4247,
+      "step": 656
+    },
+    {
+      "epoch": 0.38362163345741185,
+      "grad_norm": 0.096479965484575,
+      "learning_rate": 1.949180260599957e-05,
+      "loss": 0.4671,
+      "step": 657
+    },
+    {
+      "epoch": 0.38420553244288735,
+      "grad_norm": 0.09602419133562339,
+      "learning_rate": 1.9488588445472115e-05,
+      "loss": 0.4362,
+      "step": 658
+    },
+    {
+      "epoch": 0.3847894314283629,
+      "grad_norm": 0.09634120626569923,
+      "learning_rate": 1.9485364419471454e-05,
+      "loss": 0.4605,
+      "step": 659
+    },
+    {
+      "epoch": 0.3853733304138384,
+      "grad_norm": 0.10928277091311318,
+      "learning_rate": 1.948213053134968e-05,
+      "loss": 0.4708,
+      "step": 660
+    },
+    {
+      "epoch": 0.38595722939931393,
+      "grad_norm": 0.10114216578765685,
+      "learning_rate": 1.9478886784469124e-05,
+      "loss": 0.4735,
+      "step": 661
+    },
+    {
+      "epoch": 0.38654112838478943,
+      "grad_norm": 0.09794927887160888,
+      "learning_rate": 1.947563318220237e-05,
+      "loss": 0.4268,
+      "step": 662
+    },
+    {
+      "epoch": 0.38712502737026494,
+      "grad_norm": 0.10621316049470311,
+      "learning_rate": 1.9472369727932263e-05,
+      "loss": 0.4525,
+      "step": 663
+    },
+    {
+      "epoch": 0.38770892635574045,
+      "grad_norm": 0.10171880653919312,
+      "learning_rate": 1.9469096425051872e-05,
+      "loss": 0.4147,
+      "step": 664
+    },
+    {
+      "epoch": 0.38829282534121595,
+      "grad_norm": 0.09411072876374793,
+      "learning_rate": 1.946581327696452e-05,
+      "loss": 0.43,
+      "step": 665
+    },
+    {
+      "epoch": 0.38887672432669146,
+      "grad_norm": 0.10483338936046871,
+      "learning_rate": 1.9462520287083755e-05,
+      "loss": 0.4897,
+      "step": 666
+    },
+    {
+      "epoch": 0.389460623312167,
+      "grad_norm": 0.09941365658312598,
+      "learning_rate": 1.945921745883337e-05,
+      "loss": 0.4536,
+      "step": 667
+    },
+    {
+      "epoch": 0.3900445222976425,
+      "grad_norm": 0.10002097860267839,
+      "learning_rate": 1.945590479564738e-05,
+      "loss": 0.4462,
+      "step": 668
+    },
+    {
+      "epoch": 0.39062842128311803,
+      "grad_norm": 0.09992653219721541,
+      "learning_rate": 1.9452582300970025e-05,
+      "loss": 0.4904,
+      "step": 669
+    },
+    {
+      "epoch": 0.39121232026859354,
+      "grad_norm": 0.09986549081596101,
+      "learning_rate": 1.944924997825577e-05,
+      "loss": 0.4826,
+      "step": 670
+    },
+    {
+      "epoch": 0.39179621925406904,
+      "grad_norm": 0.11230588041282957,
+      "learning_rate": 1.944590783096929e-05,
+      "loss": 0.4155,
+      "step": 671
+    },
+    {
+      "epoch": 0.39238011823954455,
+      "grad_norm": 0.09339040842093176,
+      "learning_rate": 1.944255586258549e-05,
+      "loss": 0.4335,
+      "step": 672
+    },
+    {
+      "epoch": 0.39296401722502006,
+      "grad_norm": 0.10145417092390281,
+      "learning_rate": 1.9439194076589477e-05,
+      "loss": 0.4439,
+      "step": 673
+    },
+    {
+      "epoch": 0.39354791621049556,
+      "grad_norm": 0.10170942013292884,
+      "learning_rate": 1.9435822476476566e-05,
+      "loss": 0.4359,
+      "step": 674
+    },
+    {
+      "epoch": 0.3941318151959711,
+      "grad_norm": 0.10379268553776654,
+      "learning_rate": 1.943244106575227e-05,
+      "loss": 0.4563,
+      "step": 675
+    },
+    {
+      "epoch": 0.39471571418144663,
+      "grad_norm": 0.09625980716725219,
+      "learning_rate": 1.9429049847932317e-05,
+      "loss": 0.4498,
+      "step": 676
+    },
+    {
+      "epoch": 0.39529961316692214,
+      "grad_norm": 0.09939714653478116,
+      "learning_rate": 1.9425648826542618e-05,
+      "loss": 0.4256,
+      "step": 677
+    },
+    {
+      "epoch": 0.39588351215239764,
+      "grad_norm": 0.10061872840630152,
+      "learning_rate": 1.9422238005119287e-05,
+      "loss": 0.426,
+      "step": 678
+    },
+    {
+      "epoch": 0.39646741113787315,
+      "grad_norm": 0.09469781456291265,
+      "learning_rate": 1.9418817387208614e-05,
+      "loss": 0.4491,
+      "step": 679
+    },
+    {
+      "epoch": 0.39705131012334866,
+      "grad_norm": 0.095111535841256,
+      "learning_rate": 1.9415386976367095e-05,
+      "loss": 0.4336,
+      "step": 680
+    },
+    {
+      "epoch": 0.39763520910882416,
+      "grad_norm": 0.11922183742107853,
+      "learning_rate": 1.9411946776161388e-05,
+      "loss": 0.5349,
+      "step": 681
+    },
+    {
+      "epoch": 0.39821910809429967,
+      "grad_norm": 0.0958859632097798,
+      "learning_rate": 1.9408496790168337e-05,
+      "loss": 0.4602,
+      "step": 682
+    },
+    {
+      "epoch": 0.3988030070797752,
+      "grad_norm": 0.09253278680606843,
+      "learning_rate": 1.9405037021974965e-05,
+      "loss": 0.4438,
+      "step": 683
+    },
+    {
+      "epoch": 0.39938690606525074,
+      "grad_norm": 0.09173474947852882,
+      "learning_rate": 1.9401567475178457e-05,
+      "loss": 0.435,
+      "step": 684
+    },
+    {
+      "epoch": 0.39997080505072624,
+      "grad_norm": 0.1013319115974769,
+      "learning_rate": 1.9398088153386175e-05,
+      "loss": 0.459,
+      "step": 685
+    },
+    {
+      "epoch": 0.40055470403620175,
+      "grad_norm": 0.09275581334201348,
+      "learning_rate": 1.939459906021563e-05,
+      "loss": 0.4368,
+      "step": 686
+    },
+    {
+      "epoch": 0.40113860302167725,
+      "grad_norm": 0.09944340124938565,
+      "learning_rate": 1.939110019929451e-05,
+      "loss": 0.4244,
+      "step": 687
+    },
+    {
+      "epoch": 0.40172250200715276,
+      "grad_norm": 0.10380160030962625,
+      "learning_rate": 1.938759157426065e-05,
+      "loss": 0.4675,
+      "step": 688
+    },
+    {
+      "epoch": 0.40230640099262827,
+      "grad_norm": 0.10280101456379473,
+      "learning_rate": 1.9384073188762027e-05,
+      "loss": 0.4579,
+      "step": 689
+    },
+    {
+      "epoch": 0.40289029997810377,
+      "grad_norm": 0.09858958530462739,
+      "learning_rate": 1.9380545046456787e-05,
+      "loss": 0.4591,
+      "step": 690
+    },
+    {
+      "epoch": 0.4034741989635793,
+      "grad_norm": 0.09877598786868644,
+      "learning_rate": 1.9377007151013205e-05,
+      "loss": 0.4651,
+      "step": 691
+    },
+    {
+      "epoch": 0.40405809794905484,
+      "grad_norm": 0.10116769242424246,
+      "learning_rate": 1.93734595061097e-05,
+      "loss": 0.4368,
+      "step": 692
+    },
+    {
+      "epoch": 0.40464199693453035,
+      "grad_norm": 0.10199328435931654,
+      "learning_rate": 1.9369902115434827e-05,
+      "loss": 0.4687,
+      "step": 693
+    },
+    {
+      "epoch": 0.40522589592000585,
+      "grad_norm": 0.10449623824491154,
+      "learning_rate": 1.936633498268728e-05,
+      "loss": 0.4714,
+      "step": 694
+    },
+    {
+      "epoch": 0.40580979490548136,
+      "grad_norm": 0.09022023939477218,
+      "learning_rate": 1.9362758111575878e-05,
+      "loss": 0.3978,
+      "step": 695
+    },
+    {
+      "epoch": 0.40639369389095686,
+      "grad_norm": 0.1068005795180678,
+      "learning_rate": 1.9359171505819558e-05,
+      "loss": 0.4803,
+      "step": 696
+    },
+    {
+      "epoch": 0.40697759287643237,
+      "grad_norm": 0.09287382231141565,
+      "learning_rate": 1.935557516914739e-05,
+      "loss": 0.4528,
+      "step": 697
+    },
+    {
+      "epoch": 0.4075614918619079,
+      "grad_norm": 0.09958624755906896,
+      "learning_rate": 1.9351969105298558e-05,
+      "loss": 0.4385,
+      "step": 698
+    },
+    {
+      "epoch": 0.4081453908473834,
+      "grad_norm": 0.1177355313968971,
+      "learning_rate": 1.9348353318022353e-05,
+      "loss": 0.4576,
+      "step": 699
+    },
+    {
+      "epoch": 0.4087292898328589,
+      "grad_norm": 0.09297018177486417,
+      "learning_rate": 1.9344727811078183e-05,
+      "loss": 0.4522,
+      "step": 700
+    },
+    {
+      "epoch": 0.40931318881833445,
+      "grad_norm": 0.09910179568094346,
+      "learning_rate": 1.934109258823556e-05,
+      "loss": 0.462,
+      "step": 701
+    },
+    {
+      "epoch": 0.40989708780380996,
+      "grad_norm": 0.09703905035132251,
+      "learning_rate": 1.9337447653274097e-05,
+      "loss": 0.4163,
+      "step": 702
+    },
+    {
+      "epoch": 0.41048098678928546,
+      "grad_norm": 0.09421155359109035,
+      "learning_rate": 1.9333793009983505e-05,
+      "loss": 0.4571,
+      "step": 703
+    },
+    {
+      "epoch": 0.41106488577476097,
+      "grad_norm": 0.08934604388943311,
+      "learning_rate": 1.9330128662163588e-05,
+      "loss": 0.4573,
+      "step": 704
+    },
+    {
+      "epoch": 0.4116487847602365,
+      "grad_norm": 0.10016416257586813,
+      "learning_rate": 1.9326454613624243e-05,
+      "loss": 0.3922,
+      "step": 705
+    },
+    {
+      "epoch": 0.412232683745712,
+      "grad_norm": 0.1434323775674136,
+      "learning_rate": 1.932277086818545e-05,
+      "loss": 0.487,
+      "step": 706
+    },
+    {
+      "epoch": 0.4128165827311875,
+      "grad_norm": 0.09520933251641711,
+      "learning_rate": 1.931907742967727e-05,
+      "loss": 0.4119,
+      "step": 707
+    },
+    {
+      "epoch": 0.413400481716663,
+      "grad_norm": 0.0984831396366693,
+      "learning_rate": 1.9315374301939843e-05,
+      "loss": 0.4379,
+      "step": 708
+    },
+    {
+      "epoch": 0.41398438070213855,
+      "grad_norm": 0.09890017716431015,
+      "learning_rate": 1.9311661488823388e-05,
+      "loss": 0.4821,
+      "step": 709
+    },
+    {
+      "epoch": 0.41456827968761406,
+      "grad_norm": 0.09609458290581789,
+      "learning_rate": 1.930793899418819e-05,
+      "loss": 0.4436,
+      "step": 710
+    },
+    {
+      "epoch": 0.41515217867308957,
+      "grad_norm": 0.0948229218695425,
+      "learning_rate": 1.93042068219046e-05,
+      "loss": 0.4392,
+      "step": 711
+    },
+    {
+      "epoch": 0.4157360776585651,
+      "grad_norm": 0.10401818241961087,
+      "learning_rate": 1.9300464975853032e-05,
+      "loss": 0.4359,
+      "step": 712
+    },
+    {
+      "epoch": 0.4163199766440406,
+      "grad_norm": 0.09703193761091314,
+      "learning_rate": 1.9296713459923955e-05,
+      "loss": 0.4315,
+      "step": 713
+    },
+    {
+      "epoch": 0.4169038756295161,
+      "grad_norm": 0.09381865018550128,
+      "learning_rate": 1.9292952278017892e-05,
+      "loss": 0.4759,
+      "step": 714
+    },
+    {
+      "epoch": 0.4174877746149916,
+      "grad_norm": 0.10078797059933509,
+      "learning_rate": 1.9289181434045428e-05,
+      "loss": 0.4507,
+      "step": 715
+    },
+    {
+      "epoch": 0.4180716736004671,
+      "grad_norm": 0.0904400360070524,
+      "learning_rate": 1.9285400931927177e-05,
+      "loss": 0.4073,
+      "step": 716
+    },
+    {
+      "epoch": 0.41865557258594266,
+      "grad_norm": 0.10290839952246728,
+      "learning_rate": 1.92816107755938e-05,
+      "loss": 0.4695,
+      "step": 717
+    },
+    {
+      "epoch": 0.41923947157141817,
+      "grad_norm": 0.10688863723668592,
+      "learning_rate": 1.9277810968986004e-05,
+      "loss": 0.4807,
+      "step": 718
+    },
+    {
+      "epoch": 0.41982337055689367,
+      "grad_norm": 0.08585126006363124,
+      "learning_rate": 1.9274001516054513e-05,
+      "loss": 0.3978,
+      "step": 719
+    },
+    {
+      "epoch": 0.4204072695423692,
+      "grad_norm": 0.09095355682047139,
+      "learning_rate": 1.9270182420760104e-05,
+      "loss": 0.4264,
+      "step": 720
+    },
+    {
+      "epoch": 0.4209911685278447,
+      "grad_norm": 0.09670665397773011,
+      "learning_rate": 1.9266353687073557e-05,
+      "loss": 0.447,
+      "step": 721
+    },
+    {
+      "epoch": 0.4215750675133202,
+      "grad_norm": 0.0941987156272365,
+      "learning_rate": 1.9262515318975686e-05,
+      "loss": 0.4678,
+      "step": 722
+    },
+    {
+      "epoch": 0.4221589664987957,
+      "grad_norm": 0.10109218153870675,
+      "learning_rate": 1.9258667320457313e-05,
+      "loss": 0.4596,
+      "step": 723
+    },
+    {
+      "epoch": 0.4227428654842712,
+      "grad_norm": 0.14192260351779284,
+      "learning_rate": 1.9254809695519284e-05,
+      "loss": 0.4352,
+      "step": 724
+    },
+    {
+      "epoch": 0.4233267644697467,
+      "grad_norm": 0.09944560126958765,
+      "learning_rate": 1.9250942448172444e-05,
+      "loss": 0.4975,
+      "step": 725
+    },
+    {
+      "epoch": 0.42391066345522227,
+      "grad_norm": 0.0965260558977296,
+      "learning_rate": 1.924706558243765e-05,
+      "loss": 0.4496,
+      "step": 726
+    },
+    {
+      "epoch": 0.4244945624406978,
+      "grad_norm": 0.11497888101167388,
+      "learning_rate": 1.9243179102345753e-05,
+      "loss": 0.5094,
+      "step": 727
+    },
+    {
+      "epoch": 0.4250784614261733,
+      "grad_norm": 0.10304609895088898,
+      "learning_rate": 1.923928301193761e-05,
+      "loss": 0.4614,
+      "step": 728
+    },
+    {
+      "epoch": 0.4256623604116488,
+      "grad_norm": 0.09725739165713139,
+      "learning_rate": 1.923537731526405e-05,
+      "loss": 0.4241,
+      "step": 729
+    },
+    {
+      "epoch": 0.4262462593971243,
+      "grad_norm": 0.09958820080309132,
+      "learning_rate": 1.9231462016385917e-05,
+      "loss": 0.465,
+      "step": 730
+    },
+    {
+      "epoch": 0.4268301583825998,
+      "grad_norm": 0.09569904964151116,
+      "learning_rate": 1.9227537119374017e-05,
+      "loss": 0.4376,
+      "step": 731
+    },
+    {
+      "epoch": 0.4274140573680753,
+      "grad_norm": 0.10557190352647561,
+      "learning_rate": 1.9223602628309144e-05,
+      "loss": 0.4704,
+      "step": 732
+    },
+    {
+      "epoch": 0.4279979563535508,
+      "grad_norm": 0.09607431501889971,
+      "learning_rate": 1.921965854728207e-05,
+      "loss": 0.4366,
+      "step": 733
+    },
+    {
+      "epoch": 0.4285818553390264,
+      "grad_norm": 0.09583176907311895,
+      "learning_rate": 1.9215704880393527e-05,
+      "loss": 0.4562,
+      "step": 734
+    },
+    {
+      "epoch": 0.4291657543245019,
+      "grad_norm": 0.09988337547136068,
+      "learning_rate": 1.9211741631754228e-05,
+      "loss": 0.4464,
+      "step": 735
+    },
+    {
+      "epoch": 0.4297496533099774,
+      "grad_norm": 0.09925979048178446,
+      "learning_rate": 1.9207768805484838e-05,
+      "loss": 0.459,
+      "step": 736
+    },
+    {
+      "epoch": 0.4303335522954529,
+      "grad_norm": 0.10554650459096544,
+      "learning_rate": 1.9203786405715984e-05,
+      "loss": 0.4322,
+      "step": 737
+    },
+    {
+      "epoch": 0.4309174512809284,
+      "grad_norm": 0.09262431261817664,
+      "learning_rate": 1.9199794436588244e-05,
+      "loss": 0.4523,
+      "step": 738
+    },
+    {
+      "epoch": 0.4315013502664039,
+      "grad_norm": 0.09343887714326786,
+      "learning_rate": 1.9195792902252148e-05,
+      "loss": 0.4685,
+      "step": 739
+    },
+    {
+      "epoch": 0.4320852492518794,
+      "grad_norm": 0.09419248598946861,
+      "learning_rate": 1.9191781806868172e-05,
+      "loss": 0.4107,
+      "step": 740
+    },
+    {
+      "epoch": 0.4326691482373549,
+      "grad_norm": 0.10126632427696626,
+      "learning_rate": 1.918776115460673e-05,
+      "loss": 0.5147,
+      "step": 741
+    },
+    {
+      "epoch": 0.4332530472228304,
+      "grad_norm": 0.10432266343527202,
+      "learning_rate": 1.9183730949648173e-05,
+      "loss": 0.4594,
+      "step": 742
+    },
+    {
+      "epoch": 0.433836946208306,
+      "grad_norm": 0.10097575624199039,
+      "learning_rate": 1.9179691196182782e-05,
+      "loss": 0.4169,
+      "step": 743
+    },
+    {
+      "epoch": 0.4344208451937815,
+      "grad_norm": 0.09995499911885283,
+      "learning_rate": 1.917564189841078e-05,
+      "loss": 0.4261,
+      "step": 744
+    },
+    {
+      "epoch": 0.435004744179257,
+      "grad_norm": 0.09411516711788596,
+      "learning_rate": 1.9171583060542288e-05,
+      "loss": 0.4261,
+      "step": 745
+    },
+    {
+      "epoch": 0.4355886431647325,
+      "grad_norm": 0.0948292619632234,
+      "learning_rate": 1.916751468679737e-05,
+      "loss": 0.4051,
+      "step": 746
+    },
+    {
+      "epoch": 0.436172542150208,
+      "grad_norm": 0.10365755120995163,
+      "learning_rate": 1.9163436781405992e-05,
+      "loss": 0.4547,
+      "step": 747
+    },
+    {
+      "epoch": 0.4367564411356835,
+      "grad_norm": 0.09156620850848936,
+      "learning_rate": 1.915934934860803e-05,
+      "loss": 0.3941,
+      "step": 748
+    },
+    {
+      "epoch": 0.437340340121159,
+      "grad_norm": 0.08950215301722389,
+      "learning_rate": 1.915525239265327e-05,
+      "loss": 0.4109,
+      "step": 749
+    },
+    {
+      "epoch": 0.4379242391066345,
+      "grad_norm": 0.10134268373292632,
+      "learning_rate": 1.91511459178014e-05,
+      "loss": 0.4085,
+      "step": 750
+    },
+    {
+      "epoch": 0.4385081380921101,
+      "grad_norm": 0.10181475335036737,
+      "learning_rate": 1.9147029928322002e-05,
+      "loss": 0.4635,
+      "step": 751
+    },
+    {
+      "epoch": 0.4390920370775856,
+      "grad_norm": 0.10244603206410163,
+      "learning_rate": 1.9142904428494554e-05,
+      "loss": 0.4481,
+      "step": 752
+    },
+    {
+      "epoch": 0.4396759360630611,
+      "grad_norm": 0.09247507843367718,
+      "learning_rate": 1.9138769422608413e-05,
+      "loss": 0.4023,
+      "step": 753
+    },
+    {
+      "epoch": 0.4402598350485366,
+      "grad_norm": 0.10406961083866023,
+      "learning_rate": 1.9134624914962835e-05,
+      "loss": 0.4681,
+      "step": 754
+    },
+    {
+      "epoch": 0.4408437340340121,
+      "grad_norm": 0.09763677191889047,
+      "learning_rate": 1.9130470909866943e-05,
+      "loss": 0.4283,
+      "step": 755
+    },
+    {
+      "epoch": 0.4414276330194876,
+      "grad_norm": 0.1020799048265384,
+      "learning_rate": 1.9126307411639736e-05,
+      "loss": 0.4614,
+      "step": 756
+    },
+    {
+      "epoch": 0.4420115320049631,
+      "grad_norm": 0.0984182803603719,
+      "learning_rate": 1.912213442461009e-05,
+      "loss": 0.43,
+      "step": 757
+    },
+    {
+      "epoch": 0.44259543099043863,
+      "grad_norm": 0.09696235927070691,
+      "learning_rate": 1.9117951953116737e-05,
+      "loss": 0.4755,
+      "step": 758
+    },
+    {
+      "epoch": 0.4431793299759142,
+      "grad_norm": 0.09729660247470492,
+      "learning_rate": 1.911376000150828e-05,
+      "loss": 0.5107,
+      "step": 759
+    },
+    {
+      "epoch": 0.4437632289613897,
+      "grad_norm": 0.09590558296578128,
+      "learning_rate": 1.9109558574143173e-05,
+      "loss": 0.443,
+      "step": 760
+    },
+    {
+      "epoch": 0.4443471279468652,
+      "grad_norm": 0.1000803436925568,
+      "learning_rate": 1.9105347675389723e-05,
+      "loss": 0.4695,
+      "step": 761
+    },
+    {
+      "epoch": 0.4449310269323407,
+      "grad_norm": 0.10630023534703796,
+      "learning_rate": 1.9101127309626083e-05,
+      "loss": 0.4556,
+      "step": 762
+    },
+    {
+      "epoch": 0.4455149259178162,
+      "grad_norm": 0.10227603487246416,
+      "learning_rate": 1.909689748124025e-05,
+      "loss": 0.4296,
+      "step": 763
+    },
+    {
+      "epoch": 0.4460988249032917,
+      "grad_norm": 0.09941465452601822,
+      "learning_rate": 1.9092658194630065e-05,
+      "loss": 0.4419,
+      "step": 764
+    },
+    {
+      "epoch": 0.44668272388876723,
+      "grad_norm": 0.10525272116736917,
+      "learning_rate": 1.9088409454203196e-05,
+      "loss": 0.463,
+      "step": 765
+    },
+    {
+      "epoch": 0.44726662287424274,
+      "grad_norm": 0.09635561772232379,
+      "learning_rate": 1.908415126437714e-05,
+      "loss": 0.4562,
+      "step": 766
+    },
+    {
+      "epoch": 0.44785052185971824,
+      "grad_norm": 0.10937703408291156,
+      "learning_rate": 1.9079883629579224e-05,
+      "loss": 0.4586,
+      "step": 767
+    },
+    {
+      "epoch": 0.4484344208451938,
+      "grad_norm": 0.10839508070847423,
+      "learning_rate": 1.9075606554246594e-05,
+      "loss": 0.4925,
+      "step": 768
+    },
+    {
+      "epoch": 0.4490183198306693,
+      "grad_norm": 0.08614461923531828,
+      "learning_rate": 1.9071320042826206e-05,
+      "loss": 0.4288,
+      "step": 769
+    },
+    {
+      "epoch": 0.4496022188161448,
+      "grad_norm": 0.09076556662250398,
+      "learning_rate": 1.9067024099774828e-05,
+      "loss": 0.4344,
+      "step": 770
+    },
+    {
+      "epoch": 0.4501861178016203,
+      "grad_norm": 0.10054248865789397,
+      "learning_rate": 1.9062718729559048e-05,
+      "loss": 0.4551,
+      "step": 771
+    },
+    {
+      "epoch": 0.45077001678709583,
+      "grad_norm": 0.09926652638632685,
+      "learning_rate": 1.9058403936655235e-05,
+      "loss": 0.4639,
+      "step": 772
+    },
+    {
+      "epoch": 0.45135391577257133,
+      "grad_norm": 0.09666025425957689,
+      "learning_rate": 1.9054079725549565e-05,
+      "loss": 0.4469,
+      "step": 773
+    },
+    {
+      "epoch": 0.45193781475804684,
+      "grad_norm": 0.09866803718017458,
+      "learning_rate": 1.9049746100738012e-05,
+      "loss": 0.4697,
+      "step": 774
+    },
+    {
+      "epoch": 0.45252171374352235,
+      "grad_norm": 0.0992277254408262,
+      "learning_rate": 1.9045403066726325e-05,
+      "loss": 0.5019,
+      "step": 775
+    },
+    {
+      "epoch": 0.4531056127289979,
+      "grad_norm": 0.10694798542504211,
+      "learning_rate": 1.904105062803005e-05,
+      "loss": 0.4769,
+      "step": 776
+    },
+    {
+      "epoch": 0.4536895117144734,
+      "grad_norm": 0.1137480126583914,
+      "learning_rate": 1.9036688789174496e-05,
+      "loss": 0.4794,
+      "step": 777
+    },
+    {
+      "epoch": 0.4542734106999489,
+      "grad_norm": 0.10258010513070626,
+      "learning_rate": 1.9032317554694756e-05,
+      "loss": 0.4613,
+      "step": 778
+    },
+    {
+      "epoch": 0.4548573096854244,
+      "grad_norm": 0.10278423277607618,
+      "learning_rate": 1.9027936929135688e-05,
+      "loss": 0.4579,
+      "step": 779
+    },
+    {
+      "epoch": 0.45544120867089993,
+      "grad_norm": 0.08781746264612454,
+      "learning_rate": 1.9023546917051917e-05,
+      "loss": 0.4017,
+      "step": 780
+    },
+    {
+      "epoch": 0.45602510765637544,
+      "grad_norm": 0.09570775478332648,
+      "learning_rate": 1.901914752300783e-05,
+      "loss": 0.4758,
+      "step": 781
+    },
+    {
+      "epoch": 0.45660900664185095,
+      "grad_norm": 0.09758546193516619,
+      "learning_rate": 1.9014738751577552e-05,
+      "loss": 0.4065,
+      "step": 782
+    },
+    {
+      "epoch": 0.45719290562732645,
+      "grad_norm": 0.09904102245970194,
+      "learning_rate": 1.901032060734498e-05,
+      "loss": 0.4258,
+      "step": 783
+    },
+    {
+      "epoch": 0.457776804612802,
+      "grad_norm": 0.09754314399273714,
+      "learning_rate": 1.900589309490374e-05,
+      "loss": 0.4456,
+      "step": 784
+    },
+    {
+      "epoch": 0.4583607035982775,
+      "grad_norm": 0.09568926260019013,
+      "learning_rate": 1.9001456218857207e-05,
+      "loss": 0.4575,
+      "step": 785
+    },
+    {
+      "epoch": 0.458944602583753,
+      "grad_norm": 0.11267818144105222,
+      "learning_rate": 1.899700998381849e-05,
+      "loss": 0.4859,
+      "step": 786
+    },
+    {
+      "epoch": 0.45952850156922853,
+      "grad_norm": 0.09897165223165426,
+      "learning_rate": 1.899255439441043e-05,
+      "loss": 0.52,
+      "step": 787
+    },
+    {
+      "epoch": 0.46011240055470404,
+      "grad_norm": 0.10540432016760225,
+      "learning_rate": 1.8988089455265585e-05,
+      "loss": 0.4645,
+      "step": 788
+    },
+    {
+      "epoch": 0.46069629954017954,
+      "grad_norm": 0.09677210849429799,
+      "learning_rate": 1.898361517102624e-05,
+      "loss": 0.4393,
+      "step": 789
+    },
+    {
+      "epoch": 0.46128019852565505,
+      "grad_norm": 0.09101282251180395,
+      "learning_rate": 1.8979131546344404e-05,
+      "loss": 0.442,
+      "step": 790
+    },
+    {
+      "epoch": 0.46186409751113056,
+      "grad_norm": 0.10384333994102946,
+      "learning_rate": 1.8974638585881787e-05,
+      "loss": 0.4121,
+      "step": 791
+    },
+    {
+      "epoch": 0.46244799649660606,
+      "grad_norm": 0.09355108046152415,
+      "learning_rate": 1.8970136294309805e-05,
+      "loss": 0.4376,
+      "step": 792
+    },
+    {
+      "epoch": 0.4630318954820816,
+      "grad_norm": 0.09682495382779813,
+      "learning_rate": 1.896562467630959e-05,
+      "loss": 0.4744,
+      "step": 793
+    },
+    {
+      "epoch": 0.46361579446755713,
+      "grad_norm": 0.09955669065328376,
+      "learning_rate": 1.896110373657195e-05,
+      "loss": 0.4354,
+      "step": 794
+    },
+    {
+      "epoch": 0.46419969345303264,
+      "grad_norm": 0.09833610024205068,
+      "learning_rate": 1.89565734797974e-05,
+      "loss": 0.4316,
+      "step": 795
+    },
+    {
+      "epoch": 0.46478359243850814,
+      "grad_norm": 0.09685465933827658,
+      "learning_rate": 1.895203391069613e-05,
+      "loss": 0.4858,
+      "step": 796
+    },
+    {
+      "epoch": 0.46536749142398365,
+      "grad_norm": 0.1172262028520278,
+      "learning_rate": 1.8947485033988034e-05,
+      "loss": 0.5084,
+      "step": 797
+    },
+    {
+      "epoch": 0.46595139040945915,
+      "grad_norm": 0.09640341837167511,
+      "learning_rate": 1.894292685440266e-05,
+      "loss": 0.4833,
+      "step": 798
+    },
+    {
+      "epoch": 0.46653528939493466,
+      "grad_norm": 0.0884006053961318,
+      "learning_rate": 1.893835937667924e-05,
+      "loss": 0.4062,
+      "step": 799
+    },
+    {
+      "epoch": 0.46711918838041017,
+      "grad_norm": 0.1061643802732,
+      "learning_rate": 1.8933782605566672e-05,
+      "loss": 0.4609,
+      "step": 800
+    },
+    {
+      "epoch": 0.46770308736588573,
+      "grad_norm": 0.10380648201407856,
+      "learning_rate": 1.8929196545823512e-05,
+      "loss": 0.466,
+      "step": 801
+    },
+    {
+      "epoch": 0.46828698635136123,
+      "grad_norm": 0.09190094435079603,
+      "learning_rate": 1.8924601202217977e-05,
+      "loss": 0.4277,
+      "step": 802
+    },
+    {
+      "epoch": 0.46887088533683674,
+      "grad_norm": 0.10171570476843911,
+      "learning_rate": 1.8919996579527943e-05,
+      "loss": 0.4162,
+      "step": 803
+    },
+    {
+      "epoch": 0.46945478432231225,
+      "grad_norm": 0.10717426306813022,
+      "learning_rate": 1.891538268254092e-05,
+      "loss": 0.4676,
+      "step": 804
+    },
+    {
+      "epoch": 0.47003868330778775,
+      "grad_norm": 0.10450786235935891,
+      "learning_rate": 1.8910759516054074e-05,
+      "loss": 0.4385,
+      "step": 805
+    },
+    {
+      "epoch": 0.47062258229326326,
+      "grad_norm": 0.10109893514158874,
+      "learning_rate": 1.8906127084874198e-05,
+      "loss": 0.4197,
+      "step": 806
+    },
+    {
+      "epoch": 0.47120648127873876,
+      "grad_norm": 0.09836641693967965,
+      "learning_rate": 1.8901485393817724e-05,
+      "loss": 0.4007,
+      "step": 807
+    },
+    {
+      "epoch": 0.47179038026421427,
+      "grad_norm": 0.09909660340929992,
+      "learning_rate": 1.889683444771071e-05,
+      "loss": 0.4432,
+      "step": 808
+    },
+    {
+      "epoch": 0.4723742792496898,
+      "grad_norm": 0.1044105086186372,
+      "learning_rate": 1.889217425138884e-05,
+      "loss": 0.4927,
+      "step": 809
+    },
+    {
+      "epoch": 0.47295817823516534,
+      "grad_norm": 0.09682514565658916,
+      "learning_rate": 1.8887504809697405e-05,
+      "loss": 0.4208,
+      "step": 810
+    },
+    {
+      "epoch": 0.47354207722064084,
+      "grad_norm": 0.09452230204247015,
+      "learning_rate": 1.888282612749132e-05,
+      "loss": 0.4432,
+      "step": 811
+    },
+    {
+      "epoch": 0.47412597620611635,
+      "grad_norm": 0.09317601813856818,
+      "learning_rate": 1.8878138209635107e-05,
+      "loss": 0.4166,
+      "step": 812
+    },
+    {
+      "epoch": 0.47470987519159186,
+      "grad_norm": 0.09238779021394822,
+      "learning_rate": 1.887344106100288e-05,
+      "loss": 0.4223,
+      "step": 813
+    },
+    {
+      "epoch": 0.47529377417706736,
+      "grad_norm": 0.10754520881862152,
+      "learning_rate": 1.886873468647836e-05,
+      "loss": 0.4056,
+      "step": 814
+    },
+    {
+      "epoch": 0.47587767316254287,
+      "grad_norm": 0.10147587541966865,
+      "learning_rate": 1.8864019090954865e-05,
+      "loss": 0.4688,
+      "step": 815
+    },
+    {
+      "epoch": 0.4764615721480184,
+      "grad_norm": 0.09122697455586583,
+      "learning_rate": 1.8859294279335285e-05,
+      "loss": 0.4022,
+      "step": 816
+    },
+    {
+      "epoch": 0.4770454711334939,
+      "grad_norm": 0.09446699542169681,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.4378,
+      "step": 817
+    },
+    {
+      "epoch": 0.47762937011896944,
+      "grad_norm": 0.09976908712550324,
+      "learning_rate": 1.884981702746737e-05,
+      "loss": 0.4813,
+      "step": 818
+    },
+    {
+      "epoch": 0.47821326910444495,
+      "grad_norm": 0.10214474642280319,
+      "learning_rate": 1.8845064597072723e-05,
+      "loss": 0.5069,
+      "step": 819
+    },
+    {
+      "epoch": 0.47879716808992046,
+      "grad_norm": 0.09330401096751005,
+      "learning_rate": 1.884030297028936e-05,
+      "loss": 0.4247,
+      "step": 820
+    },
+    {
+      "epoch": 0.47938106707539596,
+      "grad_norm": 0.09959991084125033,
+      "learning_rate": 1.8835532152068025e-05,
+      "loss": 0.4223,
+      "step": 821
+    },
+    {
+      "epoch": 0.47996496606087147,
+      "grad_norm": 0.0924072517336489,
+      "learning_rate": 1.8830752147369047e-05,
+      "loss": 0.4618,
+      "step": 822
+    },
+    {
+      "epoch": 0.480548865046347,
+      "grad_norm": 0.10048714081392035,
+      "learning_rate": 1.8825962961162284e-05,
+      "loss": 0.4372,
+      "step": 823
+    },
+    {
+      "epoch": 0.4811327640318225,
+      "grad_norm": 0.09763696793474341,
+      "learning_rate": 1.8821164598427148e-05,
+      "loss": 0.4335,
+      "step": 824
+    },
+    {
+      "epoch": 0.481716663017298,
+      "grad_norm": 0.09380435988794988,
+      "learning_rate": 1.8816357064152596e-05,
+      "loss": 0.4308,
+      "step": 825
+    },
+    {
+      "epoch": 0.48230056200277355,
+      "grad_norm": 0.1057510867908029,
+      "learning_rate": 1.8811540363337107e-05,
+      "loss": 0.4325,
+      "step": 826
+    },
+    {
+      "epoch": 0.48288446098824905,
+      "grad_norm": 0.09665424847318177,
+      "learning_rate": 1.880671450098871e-05,
+      "loss": 0.4634,
+      "step": 827
+    },
+    {
+      "epoch": 0.48346835997372456,
+      "grad_norm": 0.09094778325592885,
+      "learning_rate": 1.880187948212495e-05,
+      "loss": 0.4632,
+      "step": 828
+    },
+    {
+      "epoch": 0.48405225895920007,
+      "grad_norm": 0.09822367173641883,
+      "learning_rate": 1.8797035311772884e-05,
+      "loss": 0.4344,
+      "step": 829
+    },
+    {
+      "epoch": 0.48463615794467557,
+      "grad_norm": 0.09523482637906457,
+      "learning_rate": 1.8792181994969095e-05,
+      "loss": 0.5184,
+      "step": 830
+    },
+    {
+      "epoch": 0.4852200569301511,
+      "grad_norm": 0.09452117235943411,
+      "learning_rate": 1.8787319536759677e-05,
+      "loss": 0.444,
+      "step": 831
+    },
+    {
+      "epoch": 0.4858039559156266,
+      "grad_norm": 0.100978651296685,
+      "learning_rate": 1.878244794220022e-05,
+      "loss": 0.5098,
+      "step": 832
+    },
+    {
+      "epoch": 0.4863878549011021,
+      "grad_norm": 0.09185565093558858,
+      "learning_rate": 1.8777567216355814e-05,
+      "loss": 0.4628,
+      "step": 833
+    },
+    {
+      "epoch": 0.4869717538865776,
+      "grad_norm": 0.0956994456069338,
+      "learning_rate": 1.8772677364301052e-05,
+      "loss": 0.4187,
+      "step": 834
+    },
+    {
+      "epoch": 0.48755565287205316,
+      "grad_norm": 0.0956607338262737,
+      "learning_rate": 1.8767778391120008e-05,
+      "loss": 0.4735,
+      "step": 835
+    },
+    {
+      "epoch": 0.48813955185752866,
+      "grad_norm": 0.09453854449024554,
+      "learning_rate": 1.876287030190624e-05,
+      "loss": 0.4287,
+      "step": 836
+    },
+    {
+      "epoch": 0.48872345084300417,
+      "grad_norm": 0.09474513072974634,
+      "learning_rate": 1.8757953101762786e-05,
+      "loss": 0.4782,
+      "step": 837
+    },
+    {
+      "epoch": 0.4893073498284797,
+      "grad_norm": 0.09391778748265113,
+      "learning_rate": 1.8753026795802158e-05,
+      "loss": 0.4127,
+      "step": 838
+    },
+    {
+      "epoch": 0.4898912488139552,
+      "grad_norm": 0.0969487135782114,
+      "learning_rate": 1.8748091389146336e-05,
+      "loss": 0.4634,
+      "step": 839
+    },
+    {
+      "epoch": 0.4904751477994307,
+      "grad_norm": 0.09262915901649006,
+      "learning_rate": 1.8743146886926755e-05,
+      "loss": 0.4666,
+      "step": 840
+    },
+    {
+      "epoch": 0.4910590467849062,
+      "grad_norm": 0.09248867997580001,
+      "learning_rate": 1.8738193294284312e-05,
+      "loss": 0.469,
+      "step": 841
+    },
+    {
+      "epoch": 0.4916429457703817,
+      "grad_norm": 0.10189727269747245,
+      "learning_rate": 1.873323061636936e-05,
+      "loss": 0.4422,
+      "step": 842
+    },
+    {
+      "epoch": 0.49222684475585726,
+      "grad_norm": 0.10471368885720961,
+      "learning_rate": 1.8728258858341684e-05,
+      "loss": 0.436,
+      "step": 843
+    },
+    {
+      "epoch": 0.49281074374133277,
+      "grad_norm": 0.09003608157172086,
+      "learning_rate": 1.872327802537053e-05,
+      "loss": 0.4091,
+      "step": 844
+    },
+    {
+      "epoch": 0.4933946427268083,
+      "grad_norm": 0.08883999211901127,
+      "learning_rate": 1.8718288122634566e-05,
+      "loss": 0.3638,
+      "step": 845
+    },
+    {
+      "epoch": 0.4939785417122838,
+      "grad_norm": 0.09487305155158705,
+      "learning_rate": 1.8713289155321888e-05,
+      "loss": 0.4551,
+      "step": 846
+    },
+    {
+      "epoch": 0.4945624406977593,
+      "grad_norm": 0.09046191531868158,
+      "learning_rate": 1.8708281128630023e-05,
+      "loss": 0.4101,
+      "step": 847
+    },
+    {
+      "epoch": 0.4951463396832348,
+      "grad_norm": 0.10649673342317059,
+      "learning_rate": 1.870326404776592e-05,
+      "loss": 0.4809,
+      "step": 848
+    },
+    {
+      "epoch": 0.4957302386687103,
+      "grad_norm": 0.10013957700017202,
+      "learning_rate": 1.869823791794593e-05,
+      "loss": 0.4393,
+      "step": 849
+    },
+    {
+      "epoch": 0.4963141376541858,
+      "grad_norm": 0.09809723057680672,
+      "learning_rate": 1.869320274439583e-05,
+      "loss": 0.4509,
+      "step": 850
+    },
+    {
+      "epoch": 0.4968980366396613,
+      "grad_norm": 0.10077738695580205,
+      "learning_rate": 1.8688158532350775e-05,
+      "loss": 0.4676,
+      "step": 851
+    },
+    {
+      "epoch": 0.4974819356251369,
+      "grad_norm": 0.09621955132423932,
+      "learning_rate": 1.8683105287055344e-05,
+      "loss": 0.4745,
+      "step": 852
+    },
+    {
+      "epoch": 0.4980658346106124,
+      "grad_norm": 0.09677725648919984,
+      "learning_rate": 1.8678043013763493e-05,
+      "loss": 0.4522,
+      "step": 853
+    },
+    {
+      "epoch": 0.4986497335960879,
+      "grad_norm": 0.09992938635514091,
+      "learning_rate": 1.8672971717738565e-05,
+      "loss": 0.4312,
+      "step": 854
+    },
+    {
+      "epoch": 0.4992336325815634,
+      "grad_norm": 0.09497788731083842,
+      "learning_rate": 1.866789140425329e-05,
+      "loss": 0.4292,
+      "step": 855
+    },
+    {
+      "epoch": 0.4998175315670389,
+      "grad_norm": 0.10054746011614274,
+      "learning_rate": 1.866280207858977e-05,
+      "loss": 0.4573,
+      "step": 856
+    },
+    {
+      "epoch": 0.5004014305525144,
+      "grad_norm": 0.08859341626120505,
+      "learning_rate": 1.865770374603948e-05,
+      "loss": 0.4501,
+      "step": 857
+    },
+    {
+      "epoch": 0.5009853295379899,
+      "grad_norm": 0.10903067044298262,
+      "learning_rate": 1.865259641190325e-05,
+      "loss": 0.4329,
+      "step": 858
+    },
+    {
+      "epoch": 0.5015692285234654,
+      "grad_norm": 0.1010374876453612,
+      "learning_rate": 1.864748008149128e-05,
+      "loss": 0.4483,
+      "step": 859
+    },
+    {
+      "epoch": 0.5021531275089409,
+      "grad_norm": 0.09372122761375737,
+      "learning_rate": 1.8642354760123122e-05,
+      "loss": 0.4144,
+      "step": 860
+    },
+    {
+      "epoch": 0.5027370264944164,
+      "grad_norm": 0.11427347902895961,
+      "learning_rate": 1.8637220453127675e-05,
+      "loss": 0.4308,
+      "step": 861
+    },
+    {
+      "epoch": 0.5033209254798919,
+      "grad_norm": 0.10131621870479911,
+      "learning_rate": 1.8632077165843174e-05,
+      "loss": 0.4363,
+      "step": 862
+    },
+    {
+      "epoch": 0.5039048244653674,
+      "grad_norm": 0.09381325326901896,
+      "learning_rate": 1.86269249036172e-05,
+      "loss": 0.3993,
+      "step": 863
+    },
+    {
+      "epoch": 0.5044887234508431,
+      "grad_norm": 0.0953655973949474,
+      "learning_rate": 1.8621763671806663e-05,
+      "loss": 0.4327,
+      "step": 864
+    },
+    {
+      "epoch": 0.5050726224363186,
+      "grad_norm": 0.08917169712859266,
+      "learning_rate": 1.8616593475777795e-05,
+      "loss": 0.4627,
+      "step": 865
+    },
+    {
+      "epoch": 0.5056565214217941,
+      "grad_norm": 0.09669901749057389,
+      "learning_rate": 1.8611414320906155e-05,
+      "loss": 0.4965,
+      "step": 866
+    },
+    {
+      "epoch": 0.5062404204072696,
+      "grad_norm": 0.10055205321817864,
+      "learning_rate": 1.8606226212576612e-05,
+      "loss": 0.4491,
+      "step": 867
+    },
+    {
+      "epoch": 0.5068243193927451,
+      "grad_norm": 0.10383070424084935,
+      "learning_rate": 1.860102915618334e-05,
+      "loss": 0.462,
+      "step": 868
+    },
+    {
+      "epoch": 0.5074082183782206,
+      "grad_norm": 0.09404516590501995,
+      "learning_rate": 1.8595823157129828e-05,
+      "loss": 0.4273,
+      "step": 869
+    },
+    {
+      "epoch": 0.5079921173636961,
+      "grad_norm": 0.08896690599491935,
+      "learning_rate": 1.8590608220828855e-05,
+      "loss": 0.4383,
+      "step": 870
+    },
+    {
+      "epoch": 0.5085760163491716,
+      "grad_norm": 0.0885095042594782,
+      "learning_rate": 1.8585384352702486e-05,
+      "loss": 0.4557,
+      "step": 871
+    },
+    {
+      "epoch": 0.5091599153346471,
+      "grad_norm": 0.0910159150167828,
+      "learning_rate": 1.8580151558182093e-05,
+      "loss": 0.3867,
+      "step": 872
+    },
+    {
+      "epoch": 0.5097438143201226,
+      "grad_norm": 0.101701747549505,
+      "learning_rate": 1.85749098427083e-05,
+      "loss": 0.499,
+      "step": 873
+    },
+    {
+      "epoch": 0.5103277133055981,
+      "grad_norm": 0.08852021265841592,
+      "learning_rate": 1.856965921173104e-05,
+      "loss": 0.42,
+      "step": 874
+    },
+    {
+      "epoch": 0.5109116122910736,
+      "grad_norm": 0.09495123331363481,
+      "learning_rate": 1.8564399670709482e-05,
+      "loss": 0.412,
+      "step": 875
+    },
+    {
+      "epoch": 0.5114955112765491,
+      "grad_norm": 0.09434198623204824,
+      "learning_rate": 1.8559131225112085e-05,
+      "loss": 0.4481,
+      "step": 876
+    },
+    {
+      "epoch": 0.5120794102620246,
+      "grad_norm": 0.09137717678725185,
+      "learning_rate": 1.8553853880416555e-05,
+      "loss": 0.4379,
+      "step": 877
+    },
+    {
+      "epoch": 0.5126633092475001,
+      "grad_norm": 0.08998884222216091,
+      "learning_rate": 1.8548567642109847e-05,
+      "loss": 0.4304,
+      "step": 878
+    },
+    {
+      "epoch": 0.5132472082329756,
+      "grad_norm": 0.09379088443222235,
+      "learning_rate": 1.8543272515688172e-05,
+      "loss": 0.4305,
+      "step": 879
+    },
+    {
+      "epoch": 0.5138311072184512,
+      "grad_norm": 0.09267173816797518,
+      "learning_rate": 1.8537968506656976e-05,
+      "loss": 0.4124,
+      "step": 880
+    },
+    {
+      "epoch": 0.5144150062039268,
+      "grad_norm": 0.0866560053725094,
+      "learning_rate": 1.8532655620530943e-05,
+      "loss": 0.4223,
+      "step": 881
+    },
+    {
+      "epoch": 0.5149989051894023,
+      "grad_norm": 0.09504651440846842,
+      "learning_rate": 1.8527333862833986e-05,
+      "loss": 0.4428,
+      "step": 882
+    },
+    {
+      "epoch": 0.5155828041748778,
+      "grad_norm": 0.09727359039507599,
+      "learning_rate": 1.852200323909924e-05,
+      "loss": 0.5004,
+      "step": 883
+    },
+    {
+      "epoch": 0.5161667031603533,
+      "grad_norm": 0.1025538463239529,
+      "learning_rate": 1.851666375486906e-05,
+      "loss": 0.461,
+      "step": 884
+    },
+    {
+      "epoch": 0.5167506021458288,
+      "grad_norm": 0.09608739165219653,
+      "learning_rate": 1.8511315415695013e-05,
+      "loss": 0.493,
+      "step": 885
+    },
+    {
+      "epoch": 0.5173345011313043,
+      "grad_norm": 0.09325686656551307,
+      "learning_rate": 1.8505958227137875e-05,
+      "loss": 0.4246,
+      "step": 886
+    },
+    {
+      "epoch": 0.5179184001167798,
+      "grad_norm": 0.08851593829815257,
+      "learning_rate": 1.8500592194767625e-05,
+      "loss": 0.3888,
+      "step": 887
+    },
+    {
+      "epoch": 0.5185022991022553,
+      "grad_norm": 0.08889801010030103,
+      "learning_rate": 1.8495217324163428e-05,
+      "loss": 0.4459,
+      "step": 888
+    },
+    {
+      "epoch": 0.5190861980877308,
+      "grad_norm": 0.08914592998382691,
+      "learning_rate": 1.8489833620913644e-05,
+      "loss": 0.3828,
+      "step": 889
+    },
+    {
+      "epoch": 0.5196700970732063,
+      "grad_norm": 0.10579328171991495,
+      "learning_rate": 1.848444109061581e-05,
+      "loss": 0.4517,
+      "step": 890
+    },
+    {
+      "epoch": 0.5202539960586818,
+      "grad_norm": 0.10726899292748011,
+      "learning_rate": 1.847903973887666e-05,
+      "loss": 0.5537,
+      "step": 891
+    },
+    {
+      "epoch": 0.5208378950441573,
+      "grad_norm": 0.08737703850880596,
+      "learning_rate": 1.8473629571312073e-05,
+      "loss": 0.4265,
+      "step": 892
+    },
+    {
+      "epoch": 0.5214217940296328,
+      "grad_norm": 0.10592420225355896,
+      "learning_rate": 1.8468210593547114e-05,
+      "loss": 0.5343,
+      "step": 893
+    },
+    {
+      "epoch": 0.5220056930151084,
+      "grad_norm": 0.10520530529447254,
+      "learning_rate": 1.8462782811216e-05,
+      "loss": 0.4443,
+      "step": 894
+    },
+    {
+      "epoch": 0.5225895920005839,
+      "grad_norm": 0.10247906125137146,
+      "learning_rate": 1.8457346229962106e-05,
+      "loss": 0.4651,
+      "step": 895
+    },
+    {
+      "epoch": 0.5231734909860594,
+      "grad_norm": 0.09889211729317403,
+      "learning_rate": 1.845190085543795e-05,
+      "loss": 0.4379,
+      "step": 896
+    },
+    {
+      "epoch": 0.5237573899715349,
+      "grad_norm": 0.09612566146840282,
+      "learning_rate": 1.8446446693305194e-05,
+      "loss": 0.4636,
+      "step": 897
+    },
+    {
+      "epoch": 0.5243412889570105,
+      "grad_norm": 0.09978791343194193,
+      "learning_rate": 1.8440983749234647e-05,
+      "loss": 0.4347,
+      "step": 898
+    },
+    {
+      "epoch": 0.524925187942486,
+      "grad_norm": 0.09414951120948468,
+      "learning_rate": 1.8435512028906232e-05,
+      "loss": 0.4387,
+      "step": 899
+    },
+    {
+      "epoch": 0.5255090869279615,
+      "grad_norm": 0.10409557091946874,
+      "learning_rate": 1.8430031538009005e-05,
+      "loss": 0.5234,
+      "step": 900
+    },
+    {
+      "epoch": 0.526092985913437,
+      "grad_norm": 0.1002497927993902,
+      "learning_rate": 1.8424542282241144e-05,
+      "loss": 0.4752,
+      "step": 901
+    },
+    {
+      "epoch": 0.5266768848989125,
+      "grad_norm": 0.09849493036595028,
+      "learning_rate": 1.841904426730994e-05,
+      "loss": 0.4467,
+      "step": 902
+    },
+    {
+      "epoch": 0.527260783884388,
+      "grad_norm": 0.09707776956583683,
+      "learning_rate": 1.8413537498931778e-05,
+      "loss": 0.4095,
+      "step": 903
+    },
+    {
+      "epoch": 0.5278446828698635,
+      "grad_norm": 0.10085942722818439,
+      "learning_rate": 1.840802198283216e-05,
+      "loss": 0.4554,
+      "step": 904
+    },
+    {
+      "epoch": 0.528428581855339,
+      "grad_norm": 0.09595690822918276,
+      "learning_rate": 1.840249772474568e-05,
+      "loss": 0.4294,
+      "step": 905
+    },
+    {
+      "epoch": 0.5290124808408145,
+      "grad_norm": 0.0997591835299673,
+      "learning_rate": 1.8396964730416014e-05,
+      "loss": 0.4297,
+      "step": 906
+    },
+    {
+      "epoch": 0.52959637982629,
+      "grad_norm": 0.09530444705021095,
+      "learning_rate": 1.8391423005595928e-05,
+      "loss": 0.4572,
+      "step": 907
+    },
+    {
+      "epoch": 0.5301802788117655,
+      "grad_norm": 0.08971743957912894,
+      "learning_rate": 1.8385872556047263e-05,
+      "loss": 0.4079,
+      "step": 908
+    },
+    {
+      "epoch": 0.530764177797241,
+      "grad_norm": 0.09196513450814672,
+      "learning_rate": 1.8380313387540928e-05,
+      "loss": 0.4223,
+      "step": 909
+    },
+    {
+      "epoch": 0.5313480767827166,
+      "grad_norm": 0.10129328412223565,
+      "learning_rate": 1.8374745505856904e-05,
+      "loss": 0.4988,
+      "step": 910
+    },
+    {
+      "epoch": 0.5319319757681921,
+      "grad_norm": 0.09866081856492116,
+      "learning_rate": 1.836916891678423e-05,
+      "loss": 0.467,
+      "step": 911
+    },
+    {
+      "epoch": 0.5325158747536676,
+      "grad_norm": 0.09286182745534963,
+      "learning_rate": 1.836358362612099e-05,
+      "loss": 0.4224,
+      "step": 912
+    },
+    {
+      "epoch": 0.5330997737391431,
+      "grad_norm": 0.10112591274950722,
+      "learning_rate": 1.8357989639674324e-05,
+      "loss": 0.4347,
+      "step": 913
+    },
+    {
+      "epoch": 0.5336836727246187,
+      "grad_norm": 0.12317236618933393,
+      "learning_rate": 1.835238696326041e-05,
+      "loss": 0.4382,
+      "step": 914
+    },
+    {
+      "epoch": 0.5342675717100942,
+      "grad_norm": 0.09926221706174522,
+      "learning_rate": 1.8346775602704464e-05,
+      "loss": 0.4513,
+      "step": 915
+    },
+    {
+      "epoch": 0.5348514706955697,
+      "grad_norm": 0.09514596990663036,
+      "learning_rate": 1.8341155563840726e-05,
+      "loss": 0.4665,
+      "step": 916
+    },
+    {
+      "epoch": 0.5354353696810452,
+      "grad_norm": 0.09657505855455074,
+      "learning_rate": 1.833552685251246e-05,
+      "loss": 0.4372,
+      "step": 917
+    },
+    {
+      "epoch": 0.5360192686665207,
+      "grad_norm": 0.09591512286627799,
+      "learning_rate": 1.8329889474571952e-05,
+      "loss": 0.3899,
+      "step": 918
+    },
+    {
+      "epoch": 0.5366031676519962,
+      "grad_norm": 0.09231622677724466,
+      "learning_rate": 1.83242434358805e-05,
+      "loss": 0.4199,
+      "step": 919
+    },
+    {
+      "epoch": 0.5371870666374717,
+      "grad_norm": 0.10599033743785997,
+      "learning_rate": 1.8318588742308387e-05,
+      "loss": 0.4692,
+      "step": 920
+    },
+    {
+      "epoch": 0.5377709656229472,
+      "grad_norm": 0.10341767068246,
+      "learning_rate": 1.8312925399734923e-05,
+      "loss": 0.486,
+      "step": 921
+    },
+    {
+      "epoch": 0.5383548646084227,
+      "grad_norm": 0.1010384987654491,
+      "learning_rate": 1.8307253414048395e-05,
+      "loss": 0.4495,
+      "step": 922
+    },
+    {
+      "epoch": 0.5389387635938983,
+      "grad_norm": 0.1033928639451908,
+      "learning_rate": 1.8301572791146077e-05,
+      "loss": 0.434,
+      "step": 923
+    },
+    {
+      "epoch": 0.5395226625793738,
+      "grad_norm": 0.09206283002063882,
+      "learning_rate": 1.8295883536934228e-05,
+      "loss": 0.4233,
+      "step": 924
+    },
+    {
+      "epoch": 0.5401065615648493,
+      "grad_norm": 0.09461625969675637,
+      "learning_rate": 1.8290185657328073e-05,
+      "loss": 0.5013,
+      "step": 925
+    },
+    {
+      "epoch": 0.5406904605503248,
+      "grad_norm": 0.1034244058618936,
+      "learning_rate": 1.8284479158251813e-05,
+      "loss": 0.4284,
+      "step": 926
+    },
+    {
+      "epoch": 0.5412743595358003,
+      "grad_norm": 0.09431694661274341,
+      "learning_rate": 1.827876404563861e-05,
+      "loss": 0.3829,
+      "step": 927
+    },
+    {
+      "epoch": 0.5418582585212758,
+      "grad_norm": 0.09418895369363826,
+      "learning_rate": 1.8273040325430575e-05,
+      "loss": 0.4255,
+      "step": 928
+    },
+    {
+      "epoch": 0.5424421575067513,
+      "grad_norm": 0.09301558522079134,
+      "learning_rate": 1.8267308003578774e-05,
+      "loss": 0.4344,
+      "step": 929
+    },
+    {
+      "epoch": 0.5430260564922268,
+      "grad_norm": 0.0997029556179698,
+      "learning_rate": 1.826156708604322e-05,
+      "loss": 0.4277,
+      "step": 930
+    },
+    {
+      "epoch": 0.5436099554777024,
+      "grad_norm": 0.08935163917705494,
+      "learning_rate": 1.8255817578792858e-05,
+      "loss": 0.3609,
+      "step": 931
+    },
+    {
+      "epoch": 0.5441938544631779,
+      "grad_norm": 0.09451139410234458,
+      "learning_rate": 1.825005948780556e-05,
+      "loss": 0.4279,
+      "step": 932
+    },
+    {
+      "epoch": 0.5447777534486534,
+      "grad_norm": 0.08608506704836422,
+      "learning_rate": 1.824429281906813e-05,
+      "loss": 0.4392,
+      "step": 933
+    },
+    {
+      "epoch": 0.5453616524341289,
+      "grad_norm": 0.09578510233516845,
+      "learning_rate": 1.8238517578576288e-05,
+      "loss": 0.4096,
+      "step": 934
+    },
+    {
+      "epoch": 0.5459455514196044,
+      "grad_norm": 0.098988829178566,
+      "learning_rate": 1.8232733772334663e-05,
+      "loss": 0.4136,
+      "step": 935
+    },
+    {
+      "epoch": 0.5465294504050799,
+      "grad_norm": 0.0937211762805789,
+      "learning_rate": 1.8226941406356794e-05,
+      "loss": 0.4658,
+      "step": 936
+    },
+    {
+      "epoch": 0.5471133493905554,
+      "grad_norm": 0.09961538742450378,
+      "learning_rate": 1.8221140486665125e-05,
+      "loss": 0.4597,
+      "step": 937
+    },
+    {
+      "epoch": 0.547697248376031,
+      "grad_norm": 0.09658142362389026,
+      "learning_rate": 1.8215331019290975e-05,
+      "loss": 0.4385,
+      "step": 938
+    },
+    {
+      "epoch": 0.5482811473615065,
+      "grad_norm": 0.11390802028457178,
+      "learning_rate": 1.8209513010274572e-05,
+      "loss": 0.467,
+      "step": 939
+    },
+    {
+      "epoch": 0.548865046346982,
+      "grad_norm": 0.09936600696472352,
+      "learning_rate": 1.820368646566501e-05,
+      "loss": 0.4519,
+      "step": 940
+    },
+    {
+      "epoch": 0.5494489453324575,
+      "grad_norm": 0.09493911114045901,
+      "learning_rate": 1.8197851391520265e-05,
+      "loss": 0.4541,
+      "step": 941
+    },
+    {
+      "epoch": 0.550032844317933,
+      "grad_norm": 0.09240295666014679,
+      "learning_rate": 1.8192007793907177e-05,
+      "loss": 0.4334,
+      "step": 942
+    },
+    {
+      "epoch": 0.5506167433034085,
+      "grad_norm": 0.09788998834203151,
+      "learning_rate": 1.8186155678901457e-05,
+      "loss": 0.4548,
+      "step": 943
+    },
+    {
+      "epoch": 0.551200642288884,
+      "grad_norm": 0.09063192952262786,
+      "learning_rate": 1.8180295052587653e-05,
+      "loss": 0.4233,
+      "step": 944
+    },
+    {
+      "epoch": 0.5517845412743595,
+      "grad_norm": 0.08692926898480201,
+      "learning_rate": 1.8174425921059183e-05,
+      "loss": 0.4158,
+      "step": 945
+    },
+    {
+      "epoch": 0.552368440259835,
+      "grad_norm": 0.09087820694480891,
+      "learning_rate": 1.81685482904183e-05,
+      "loss": 0.4389,
+      "step": 946
+    },
+    {
+      "epoch": 0.5529523392453105,
+      "grad_norm": 0.10120795540765723,
+      "learning_rate": 1.8162662166776085e-05,
+      "loss": 0.4994,
+      "step": 947
+    },
+    {
+      "epoch": 0.5535362382307861,
+      "grad_norm": 0.1013505444213902,
+      "learning_rate": 1.8156767556252464e-05,
+      "loss": 0.4218,
+      "step": 948
+    },
+    {
+      "epoch": 0.5541201372162616,
+      "grad_norm": 0.08997274911995008,
+      "learning_rate": 1.815086446497618e-05,
+      "loss": 0.3779,
+      "step": 949
+    },
+    {
+      "epoch": 0.5547040362017371,
+      "grad_norm": 0.09103042499731277,
+      "learning_rate": 1.8144952899084787e-05,
+      "loss": 0.4357,
+      "step": 950
+    },
+    {
+      "epoch": 0.5552879351872126,
+      "grad_norm": 0.10557869732470204,
+      "learning_rate": 1.8139032864724665e-05,
+      "loss": 0.5279,
+      "step": 951
+    },
+    {
+      "epoch": 0.5558718341726882,
+      "grad_norm": 0.0969702944736897,
+      "learning_rate": 1.813310436805099e-05,
+      "loss": 0.5004,
+      "step": 952
+    },
+    {
+      "epoch": 0.5564557331581637,
+      "grad_norm": 0.08763260290940755,
+      "learning_rate": 1.8127167415227736e-05,
+      "loss": 0.3894,
+      "step": 953
+    },
+    {
+      "epoch": 0.5570396321436392,
+      "grad_norm": 0.09312761076500888,
+      "learning_rate": 1.8121222012427666e-05,
+      "loss": 0.4316,
+      "step": 954
+    },
+    {
+      "epoch": 0.5576235311291147,
+      "grad_norm": 0.09027049103660485,
+      "learning_rate": 1.8115268165832336e-05,
+      "loss": 0.4553,
+      "step": 955
+    },
+    {
+      "epoch": 0.5582074301145902,
+      "grad_norm": 0.09540093321206308,
+      "learning_rate": 1.810930588163208e-05,
+      "loss": 0.4842,
+      "step": 956
+    },
+    {
+      "epoch": 0.5587913291000657,
+      "grad_norm": 0.09870825042866023,
+      "learning_rate": 1.8103335166026002e-05,
+      "loss": 0.4437,
+      "step": 957
+    },
+    {
+      "epoch": 0.5593752280855412,
+      "grad_norm": 0.08566092255820369,
+      "learning_rate": 1.8097356025221975e-05,
+      "loss": 0.4169,
+      "step": 958
+    },
+    {
+      "epoch": 0.5599591270710167,
+      "grad_norm": 0.09208068448303999,
+      "learning_rate": 1.8091368465436626e-05,
+      "loss": 0.4564,
+      "step": 959
+    },
+    {
+      "epoch": 0.5605430260564922,
+      "grad_norm": 0.0900554510673484,
+      "learning_rate": 1.8085372492895338e-05,
+      "loss": 0.4134,
+      "step": 960
+    },
+    {
+      "epoch": 0.5611269250419677,
+      "grad_norm": 0.09857270563853884,
+      "learning_rate": 1.807936811383225e-05,
+      "loss": 0.4489,
+      "step": 961
+    },
+    {
+      "epoch": 0.5617108240274432,
+      "grad_norm": 0.09594170206055125,
+      "learning_rate": 1.8073355334490227e-05,
+      "loss": 0.4602,
+      "step": 962
+    },
+    {
+      "epoch": 0.5622947230129187,
+      "grad_norm": 0.08782348677371765,
+      "learning_rate": 1.806733416112088e-05,
+      "loss": 0.4312,
+      "step": 963
+    },
+    {
+      "epoch": 0.5628786219983942,
+      "grad_norm": 0.10266794513538419,
+      "learning_rate": 1.8061304599984537e-05,
+      "loss": 0.4248,
+      "step": 964
+    },
+    {
+      "epoch": 0.5634625209838698,
+      "grad_norm": 0.09319926116735038,
+      "learning_rate": 1.8055266657350256e-05,
+      "loss": 0.4557,
+      "step": 965
+    },
+    {
+      "epoch": 0.5640464199693453,
+      "grad_norm": 0.09493511735660687,
+      "learning_rate": 1.8049220339495797e-05,
+      "loss": 0.4642,
+      "step": 966
+    },
+    {
+      "epoch": 0.5646303189548209,
+      "grad_norm": 0.08794476333877156,
+      "learning_rate": 1.804316565270765e-05,
+      "loss": 0.3947,
+      "step": 967
+    },
+    {
+      "epoch": 0.5652142179402964,
+      "grad_norm": 0.09584897022930479,
+      "learning_rate": 1.8037102603280984e-05,
+      "loss": 0.4479,
+      "step": 968
+    },
+    {
+      "epoch": 0.5657981169257719,
+      "grad_norm": 0.09345250609468499,
+      "learning_rate": 1.8031031197519673e-05,
+      "loss": 0.4185,
+      "step": 969
+    },
+    {
+      "epoch": 0.5663820159112474,
+      "grad_norm": 0.09747131403965749,
+      "learning_rate": 1.8024951441736275e-05,
+      "loss": 0.4232,
+      "step": 970
+    },
+    {
+      "epoch": 0.5669659148967229,
+      "grad_norm": 0.09092753314349322,
+      "learning_rate": 1.8018863342252038e-05,
+      "loss": 0.4366,
+      "step": 971
+    },
+    {
+      "epoch": 0.5675498138821984,
+      "grad_norm": 0.09067933084709702,
+      "learning_rate": 1.801276690539688e-05,
+      "loss": 0.3874,
+      "step": 972
+    },
+    {
+      "epoch": 0.5681337128676739,
+      "grad_norm": 0.10088398815224028,
+      "learning_rate": 1.800666213750938e-05,
+      "loss": 0.4896,
+      "step": 973
+    },
+    {
+      "epoch": 0.5687176118531494,
+      "grad_norm": 0.08646278337712668,
+      "learning_rate": 1.800054904493679e-05,
+      "loss": 0.3915,
+      "step": 974
+    },
+    {
+      "epoch": 0.5693015108386249,
+      "grad_norm": 0.09614637115760741,
+      "learning_rate": 1.7994427634035016e-05,
+      "loss": 0.424,
+      "step": 975
+    },
+    {
+      "epoch": 0.5698854098241004,
+      "grad_norm": 0.09150453883936621,
+      "learning_rate": 1.7988297911168602e-05,
+      "loss": 0.3928,
+      "step": 976
+    },
+    {
+      "epoch": 0.5704693088095759,
+      "grad_norm": 0.08688233350517048,
+      "learning_rate": 1.798215988271075e-05,
+      "loss": 0.4126,
+      "step": 977
+    },
+    {
+      "epoch": 0.5710532077950514,
+      "grad_norm": 0.09532039074812583,
+      "learning_rate": 1.7976013555043286e-05,
+      "loss": 0.4201,
+      "step": 978
+    },
+    {
+      "epoch": 0.5716371067805269,
+      "grad_norm": 0.09235136275358072,
+      "learning_rate": 1.7969858934556676e-05,
+      "loss": 0.4301,
+      "step": 979
+    },
+    {
+      "epoch": 0.5722210057660024,
+      "grad_norm": 0.09943850901480054,
+      "learning_rate": 1.796369602764999e-05,
+      "loss": 0.4436,
+      "step": 980
+    },
+    {
+      "epoch": 0.5728049047514779,
+      "grad_norm": 0.08775548636161967,
+      "learning_rate": 1.7957524840730925e-05,
+      "loss": 0.4136,
+      "step": 981
+    },
+    {
+      "epoch": 0.5733888037369536,
+      "grad_norm": 0.08755422169362043,
+      "learning_rate": 1.7951345380215795e-05,
+      "loss": 0.4164,
+      "step": 982
+    },
+    {
+      "epoch": 0.5739727027224291,
+      "grad_norm": 0.10007393302510446,
+      "learning_rate": 1.79451576525295e-05,
+      "loss": 0.4483,
+      "step": 983
+    },
+    {
+      "epoch": 0.5745566017079046,
+      "grad_norm": 0.10309236024564333,
+      "learning_rate": 1.7938961664105546e-05,
+      "loss": 0.5024,
+      "step": 984
+    },
+    {
+      "epoch": 0.5751405006933801,
+      "grad_norm": 0.08765422102235457,
+      "learning_rate": 1.793275742138602e-05,
+      "loss": 0.4336,
+      "step": 985
+    },
+    {
+      "epoch": 0.5757243996788556,
+      "grad_norm": 0.09316659724176794,
+      "learning_rate": 1.7926544930821608e-05,
+      "loss": 0.4246,
+      "step": 986
+    },
+    {
+      "epoch": 0.5763082986643311,
+      "grad_norm": 0.10362901270995364,
+      "learning_rate": 1.7920324198871546e-05,
+      "loss": 0.4846,
+      "step": 987
+    },
+    {
+      "epoch": 0.5768921976498066,
+      "grad_norm": 0.09666748211699074,
+      "learning_rate": 1.791409523200366e-05,
+      "loss": 0.4447,
+      "step": 988
+    },
+    {
+      "epoch": 0.5774760966352821,
+      "grad_norm": 0.09116224023634917,
+      "learning_rate": 1.7907858036694325e-05,
+      "loss": 0.416,
+      "step": 989
+    },
+    {
+      "epoch": 0.5780599956207576,
+      "grad_norm": 0.09327311494130844,
+      "learning_rate": 1.790161261942848e-05,
+      "loss": 0.4567,
+      "step": 990
+    },
+    {
+      "epoch": 0.5786438946062331,
+      "grad_norm": 0.09757893842534339,
+      "learning_rate": 1.7895358986699607e-05,
+      "loss": 0.4319,
+      "step": 991
+    },
+    {
+      "epoch": 0.5792277935917086,
+      "grad_norm": 0.09331016061021268,
+      "learning_rate": 1.7889097145009736e-05,
+      "loss": 0.4545,
+      "step": 992
+    },
+    {
+      "epoch": 0.5798116925771841,
+      "grad_norm": 0.0907331564391952,
+      "learning_rate": 1.788282710086942e-05,
+      "loss": 0.4189,
+      "step": 993
+    },
+    {
+      "epoch": 0.5803955915626596,
+      "grad_norm": 0.1003881374586604,
+      "learning_rate": 1.7876548860797756e-05,
+      "loss": 0.4512,
+      "step": 994
+    },
+    {
+      "epoch": 0.5809794905481351,
+      "grad_norm": 0.08934637418005388,
+      "learning_rate": 1.787026243132235e-05,
+      "loss": 0.4071,
+      "step": 995
+    },
+    {
+      "epoch": 0.5815633895336106,
+      "grad_norm": 0.09758860399619078,
+      "learning_rate": 1.7863967818979328e-05,
+      "loss": 0.4508,
+      "step": 996
+    },
+    {
+      "epoch": 0.5821472885190861,
+      "grad_norm": 0.09544092806826925,
+      "learning_rate": 1.785766503031332e-05,
+      "loss": 0.4524,
+      "step": 997
+    },
+    {
+      "epoch": 0.5827311875045618,
+      "grad_norm": 0.09322154851525331,
+      "learning_rate": 1.785135407187747e-05,
+      "loss": 0.443,
+      "step": 998
+    },
+    {
+      "epoch": 0.5833150864900373,
+      "grad_norm": 0.10454424047737375,
+      "learning_rate": 1.7845034950233394e-05,
+      "loss": 0.4817,
+      "step": 999
+    },
+    {
+      "epoch": 0.5838989854755128,
+      "grad_norm": 0.09414148969099152,
+      "learning_rate": 1.7838707671951215e-05,
+      "loss": 0.4169,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5844828844609883,
+      "grad_norm": 0.08855322823777409,
+      "learning_rate": 1.7832372243609527e-05,
+      "loss": 0.4499,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5850667834464638,
+      "grad_norm": 0.09944085358526374,
+      "learning_rate": 1.78260286717954e-05,
+      "loss": 0.4757,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5856506824319393,
+      "grad_norm": 0.09579688476732336,
+      "learning_rate": 1.781967696310437e-05,
+      "loss": 0.4388,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5862345814174148,
+      "grad_norm": 0.09018554080088893,
+      "learning_rate": 1.7813317124140445e-05,
+      "loss": 0.4475,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5868184804028903,
+      "grad_norm": 0.09074306031957446,
+      "learning_rate": 1.7806949161516062e-05,
+      "loss": 0.4466,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5874023793883658,
+      "grad_norm": 0.09891567293014034,
+      "learning_rate": 1.7800573081852124e-05,
+      "loss": 0.4187,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5879862783738413,
+      "grad_norm": 0.09565477927288923,
+      "learning_rate": 1.7794188891777964e-05,
+      "loss": 0.4323,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5885701773593168,
+      "grad_norm": 0.09003552402970175,
+      "learning_rate": 1.7787796597931354e-05,
+      "loss": 0.3862,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5891540763447923,
+      "grad_norm": 0.09915374630001458,
+      "learning_rate": 1.7781396206958485e-05,
+      "loss": 0.4495,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5897379753302678,
+      "grad_norm": 0.0945365629310544,
+      "learning_rate": 1.7774987725513975e-05,
+      "loss": 0.4179,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5903218743157433,
+      "grad_norm": 0.09673883279410804,
+      "learning_rate": 1.7768571160260845e-05,
+      "loss": 0.4145,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5909057733012189,
+      "grad_norm": 0.09549911441241216,
+      "learning_rate": 1.7762146517870526e-05,
+      "loss": 0.4817,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5914896722866944,
+      "grad_norm": 0.09263634258616428,
+      "learning_rate": 1.7755713805022846e-05,
+      "loss": 0.4377,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5920735712721699,
+      "grad_norm": 0.08512533745534888,
+      "learning_rate": 1.7749273028406025e-05,
+      "loss": 0.3968,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5926574702576455,
+      "grad_norm": 0.09817084699292195,
+      "learning_rate": 1.7742824194716664e-05,
+      "loss": 0.5104,
+      "step": 1015
+    },
+    {
+      "epoch": 0.593241369243121,
+      "grad_norm": 0.0823680754502438,
+      "learning_rate": 1.7736367310659743e-05,
+      "loss": 0.429,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5938252682285965,
+      "grad_norm": 0.0840678918052975,
+      "learning_rate": 1.7729902382948617e-05,
+      "loss": 0.4153,
+      "step": 1017
+    },
+    {
+      "epoch": 0.594409167214072,
+      "grad_norm": 0.08786900942424455,
+      "learning_rate": 1.772342941830499e-05,
+      "loss": 0.4297,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5949930661995475,
+      "grad_norm": 0.08974648821956324,
+      "learning_rate": 1.771694842345894e-05,
+      "loss": 0.4384,
+      "step": 1019
+    },
+    {
+      "epoch": 0.595576965185023,
+      "grad_norm": 0.09161956946443398,
+      "learning_rate": 1.771045940514888e-05,
+      "loss": 0.4151,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5961608641704985,
+      "grad_norm": 0.08977274274143883,
+      "learning_rate": 1.7703962370121575e-05,
+      "loss": 0.4274,
+      "step": 1021
+    },
+    {
+      "epoch": 0.596744763155974,
+      "grad_norm": 0.09419158492640331,
+      "learning_rate": 1.769745732513212e-05,
+      "loss": 0.4593,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5973286621414495,
+      "grad_norm": 0.09288795923337607,
+      "learning_rate": 1.7690944276943935e-05,
+      "loss": 0.4129,
+      "step": 1023
+    },
+    {
+      "epoch": 0.597912561126925,
+      "grad_norm": 0.08707857505942682,
+      "learning_rate": 1.768442323232877e-05,
+      "loss": 0.4499,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5984964601124005,
+      "grad_norm": 0.09086795023229922,
+      "learning_rate": 1.767789419806668e-05,
+      "loss": 0.4545,
+      "step": 1025
+    },
+    {
+      "epoch": 0.599080359097876,
+      "grad_norm": 0.08963629694477594,
+      "learning_rate": 1.7671357180946035e-05,
+      "loss": 0.402,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5996642580833516,
+      "grad_norm": 0.08372777496814697,
+      "learning_rate": 1.76648121877635e-05,
+      "loss": 0.4042,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6002481570688271,
+      "grad_norm": 0.08183505710947246,
+      "learning_rate": 1.7658259225324036e-05,
+      "loss": 0.4271,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6008320560543026,
+      "grad_norm": 0.09310272821291517,
+      "learning_rate": 1.765169830044088e-05,
+      "loss": 0.4826,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6014159550397781,
+      "grad_norm": 0.09298824684497155,
+      "learning_rate": 1.7645129419935565e-05,
+      "loss": 0.4242,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6019998540252536,
+      "grad_norm": 0.08374136579623451,
+      "learning_rate": 1.763855259063788e-05,
+      "loss": 0.3697,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6025837530107292,
+      "grad_norm": 0.09310681376155815,
+      "learning_rate": 1.7631967819385883e-05,
+      "loss": 0.4267,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6031676519962047,
+      "grad_norm": 0.08511304053336544,
+      "learning_rate": 1.76253751130259e-05,
+      "loss": 0.3972,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6037515509816802,
+      "grad_norm": 0.09278902306590756,
+      "learning_rate": 1.761877447841249e-05,
+      "loss": 0.4302,
+      "step": 1034
+    },
+    {
+      "epoch": 0.6043354499671557,
+      "grad_norm": 0.09048834111859336,
+      "learning_rate": 1.7612165922408463e-05,
+      "loss": 0.4242,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6049193489526312,
+      "grad_norm": 0.09006690439926274,
+      "learning_rate": 1.760554945188487e-05,
+      "loss": 0.4326,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6055032479381067,
+      "grad_norm": 0.10055053068810905,
+      "learning_rate": 1.759892507372099e-05,
+      "loss": 0.4594,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6060871469235822,
+      "grad_norm": 0.09235398784246698,
+      "learning_rate": 1.759229279480431e-05,
+      "loss": 0.4244,
+      "step": 1038
+    },
+    {
+      "epoch": 0.6066710459090577,
+      "grad_norm": 0.0843238581691318,
+      "learning_rate": 1.758565262203055e-05,
+      "loss": 0.4359,
+      "step": 1039
+    },
+    {
+      "epoch": 0.6072549448945332,
+      "grad_norm": 0.09163218002949744,
+      "learning_rate": 1.757900456230362e-05,
+      "loss": 0.4273,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6078388438800088,
+      "grad_norm": 0.0881653369897237,
+      "learning_rate": 1.757234862253565e-05,
+      "loss": 0.442,
+      "step": 1041
+    },
+    {
+      "epoch": 0.6084227428654843,
+      "grad_norm": 0.08868885817938024,
+      "learning_rate": 1.7565684809646946e-05,
+      "loss": 0.4168,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6090066418509598,
+      "grad_norm": 0.08846014003644663,
+      "learning_rate": 1.7559013130566003e-05,
+      "loss": 0.406,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6095905408364353,
+      "grad_norm": 0.09694004061406085,
+      "learning_rate": 1.755233359222951e-05,
+      "loss": 0.4264,
+      "step": 1044
+    },
+    {
+      "epoch": 0.6101744398219108,
+      "grad_norm": 0.09199448538507488,
+      "learning_rate": 1.7545646201582304e-05,
+      "loss": 0.4248,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6107583388073863,
+      "grad_norm": 0.0969549371597795,
+      "learning_rate": 1.75389509655774e-05,
+      "loss": 0.4345,
+      "step": 1046
+    },
+    {
+      "epoch": 0.6113422377928618,
+      "grad_norm": 0.08998036983842693,
+      "learning_rate": 1.7532247891175968e-05,
+      "loss": 0.4233,
+      "step": 1047
+    },
+    {
+      "epoch": 0.6119261367783373,
+      "grad_norm": 0.0904716904612348,
+      "learning_rate": 1.7525536985347328e-05,
+      "loss": 0.4937,
+      "step": 1048
+    },
+    {
+      "epoch": 0.6125100357638129,
+      "grad_norm": 0.09226977721467346,
+      "learning_rate": 1.751881825506894e-05,
+      "loss": 0.5158,
+      "step": 1049
+    },
+    {
+      "epoch": 0.6130939347492884,
+      "grad_norm": 0.089606513763851,
+      "learning_rate": 1.7512091707326403e-05,
+      "loss": 0.4223,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6136778337347639,
+      "grad_norm": 0.08512150877113146,
+      "learning_rate": 1.750535734911344e-05,
+      "loss": 0.4129,
+      "step": 1051
+    },
+    {
+      "epoch": 0.6142617327202394,
+      "grad_norm": 0.08805797211332342,
+      "learning_rate": 1.7498615187431894e-05,
+      "loss": 0.444,
+      "step": 1052
+    },
+    {
+      "epoch": 0.6148456317057149,
+      "grad_norm": 0.09330844331427343,
+      "learning_rate": 1.7491865229291733e-05,
+      "loss": 0.439,
+      "step": 1053
+    },
+    {
+      "epoch": 0.6154295306911904,
+      "grad_norm": 0.09471812518051606,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.4912,
+      "step": 1054
+    },
+    {
+      "epoch": 0.616013429676666,
+      "grad_norm": 0.08458194863552877,
+      "learning_rate": 1.74783419517159e-05,
+      "loss": 0.4132,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6165973286621415,
+      "grad_norm": 0.08984709466957168,
+      "learning_rate": 1.7471568646340653e-05,
+      "loss": 0.4029,
+      "step": 1056
+    },
+    {
+      "epoch": 0.617181227647617,
+      "grad_norm": 0.09308953785398552,
+      "learning_rate": 1.746478757262761e-05,
+      "loss": 0.4505,
+      "step": 1057
+    },
+    {
+      "epoch": 0.6177651266330925,
+      "grad_norm": 0.08266036459353078,
+      "learning_rate": 1.7457998737627183e-05,
+      "loss": 0.449,
+      "step": 1058
+    },
+    {
+      "epoch": 0.618349025618568,
+      "grad_norm": 0.09367401787259014,
+      "learning_rate": 1.745120214839786e-05,
+      "loss": 0.4334,
+      "step": 1059
+    },
+    {
+      "epoch": 0.6189329246040435,
+      "grad_norm": 0.10727638726566346,
+      "learning_rate": 1.7444397812006194e-05,
+      "loss": 0.4603,
+      "step": 1060
+    },
+    {
+      "epoch": 0.619516823589519,
+      "grad_norm": 0.09501560905806689,
+      "learning_rate": 1.7437585735526785e-05,
+      "loss": 0.4559,
+      "step": 1061
+    },
+    {
+      "epoch": 0.6201007225749945,
+      "grad_norm": 0.09933126329807934,
+      "learning_rate": 1.7430765926042287e-05,
+      "loss": 0.5159,
+      "step": 1062
+    },
+    {
+      "epoch": 0.62068462156047,
+      "grad_norm": 0.09188967274811534,
+      "learning_rate": 1.7423938390643384e-05,
+      "loss": 0.4252,
+      "step": 1063
+    },
+    {
+      "epoch": 0.6212685205459455,
+      "grad_norm": 0.10372561659381267,
+      "learning_rate": 1.7417103136428806e-05,
+      "loss": 0.4533,
+      "step": 1064
+    },
+    {
+      "epoch": 0.6218524195314211,
+      "grad_norm": 0.08623435328581035,
+      "learning_rate": 1.74102601705053e-05,
+      "loss": 0.4455,
+      "step": 1065
+    },
+    {
+      "epoch": 0.6224363185168966,
+      "grad_norm": 0.08979763664356266,
+      "learning_rate": 1.7403409499987633e-05,
+      "loss": 0.442,
+      "step": 1066
+    },
+    {
+      "epoch": 0.6230202175023721,
+      "grad_norm": 0.09287918509624764,
+      "learning_rate": 1.739655113199858e-05,
+      "loss": 0.4146,
+      "step": 1067
+    },
+    {
+      "epoch": 0.6236041164878476,
+      "grad_norm": 0.2019244377936218,
+      "learning_rate": 1.7389685073668925e-05,
+      "loss": 0.4527,
+      "step": 1068
+    },
+    {
+      "epoch": 0.6241880154733231,
+      "grad_norm": 0.09353779053945302,
+      "learning_rate": 1.7382811332137444e-05,
+      "loss": 0.3644,
+      "step": 1069
+    },
+    {
+      "epoch": 0.6247719144587986,
+      "grad_norm": 0.09882394693475512,
+      "learning_rate": 1.7375929914550906e-05,
+      "loss": 0.457,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6253558134442742,
+      "grad_norm": 0.09276655979863291,
+      "learning_rate": 1.7369040828064046e-05,
+      "loss": 0.4022,
+      "step": 1071
+    },
+    {
+      "epoch": 0.6259397124297497,
+      "grad_norm": 0.09627737468433518,
+      "learning_rate": 1.73621440798396e-05,
+      "loss": 0.4356,
+      "step": 1072
+    },
+    {
+      "epoch": 0.6265236114152252,
+      "grad_norm": 0.0975486353038753,
+      "learning_rate": 1.7355239677048237e-05,
+      "loss": 0.4673,
+      "step": 1073
+    },
+    {
+      "epoch": 0.6271075104007007,
+      "grad_norm": 0.0869307548353545,
+      "learning_rate": 1.734832762686861e-05,
+      "loss": 0.4422,
+      "step": 1074
+    },
+    {
+      "epoch": 0.6276914093861762,
+      "grad_norm": 0.08654520091120523,
+      "learning_rate": 1.7341407936487316e-05,
+      "loss": 0.4341,
+      "step": 1075
+    },
+    {
+      "epoch": 0.6282753083716517,
+      "grad_norm": 0.09738635621198366,
+      "learning_rate": 1.7334480613098893e-05,
+      "loss": 0.4575,
+      "step": 1076
+    },
+    {
+      "epoch": 0.6288592073571272,
+      "grad_norm": 0.08475960191980164,
+      "learning_rate": 1.7327545663905813e-05,
+      "loss": 0.3993,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6294431063426027,
+      "grad_norm": 0.09068613829566882,
+      "learning_rate": 1.7320603096118476e-05,
+      "loss": 0.4368,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6300270053280782,
+      "grad_norm": 0.08862724193973477,
+      "learning_rate": 1.731365291695522e-05,
+      "loss": 0.4293,
+      "step": 1079
+    },
+    {
+      "epoch": 0.6306109043135537,
+      "grad_norm": 0.09503497169894197,
+      "learning_rate": 1.730669513364227e-05,
+      "loss": 0.4134,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6311948032990292,
+      "grad_norm": 0.09031186716963921,
+      "learning_rate": 1.7299729753413783e-05,
+      "loss": 0.4343,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6317787022845048,
+      "grad_norm": 0.08837664288300213,
+      "learning_rate": 1.7292756783511793e-05,
+      "loss": 0.4436,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6323626012699803,
+      "grad_norm": 0.0852607030293982,
+      "learning_rate": 1.728577623118624e-05,
+      "loss": 0.4167,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6329465002554558,
+      "grad_norm": 0.08989887067633713,
+      "learning_rate": 1.7278788103694944e-05,
+      "loss": 0.3907,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6335303992409314,
+      "grad_norm": 0.09433293307855042,
+      "learning_rate": 1.7271792408303593e-05,
+      "loss": 0.4698,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6341142982264069,
+      "grad_norm": 0.08657071138254312,
+      "learning_rate": 1.7264789152285754e-05,
+      "loss": 0.3883,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6346981972118824,
+      "grad_norm": 0.09273200459294935,
+      "learning_rate": 1.7257778342922853e-05,
+      "loss": 0.439,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6352820961973579,
+      "grad_norm": 0.09009715234195098,
+      "learning_rate": 1.7250759987504165e-05,
+      "loss": 0.4152,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6358659951828334,
+      "grad_norm": 0.09094556123161698,
+      "learning_rate": 1.724373409332681e-05,
+      "loss": 0.4433,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6364498941683089,
+      "grad_norm": 0.09084850101725897,
+      "learning_rate": 1.7236700667695754e-05,
+      "loss": 0.4335,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6370337931537844,
+      "grad_norm": 0.0914158414655426,
+      "learning_rate": 1.7229659717923784e-05,
+      "loss": 0.4509,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6376176921392599,
+      "grad_norm": 0.09107857512749057,
+      "learning_rate": 1.722261125133152e-05,
+      "loss": 0.4276,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6382015911247354,
+      "grad_norm": 0.09551550458507074,
+      "learning_rate": 1.721555527524739e-05,
+      "loss": 0.4026,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6387854901102109,
+      "grad_norm": 0.08961858238387461,
+      "learning_rate": 1.7208491797007634e-05,
+      "loss": 0.4455,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6393693890956864,
+      "grad_norm": 0.09054634359123176,
+      "learning_rate": 1.7201420823956286e-05,
+      "loss": 0.4221,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6399532880811619,
+      "grad_norm": 0.08648544277575235,
+      "learning_rate": 1.719434236344518e-05,
+      "loss": 0.4041,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6405371870666374,
+      "grad_norm": 0.08424949364546344,
+      "learning_rate": 1.7187256422833928e-05,
+      "loss": 0.44,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6411210860521129,
+      "grad_norm": 0.09727372722963082,
+      "learning_rate": 1.7180163009489924e-05,
+      "loss": 0.464,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6417049850375885,
+      "grad_norm": 0.08372587387784208,
+      "learning_rate": 1.7173062130788337e-05,
+      "loss": 0.4012,
+      "step": 1099
+    },
+    {
+      "epoch": 0.642288884023064,
+      "grad_norm": 0.09124111852726635,
+      "learning_rate": 1.716595379411208e-05,
+      "loss": 0.4948,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6428727830085396,
+      "grad_norm": 0.09189369804402228,
+      "learning_rate": 1.715883800685184e-05,
+      "loss": 0.4294,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6434566819940151,
+      "grad_norm": 0.0854633570398992,
+      "learning_rate": 1.7151714776406034e-05,
+      "loss": 0.4047,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6440405809794906,
+      "grad_norm": 0.08960507132851385,
+      "learning_rate": 1.7144584110180834e-05,
+      "loss": 0.436,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6446244799649661,
+      "grad_norm": 0.10065698334089884,
+      "learning_rate": 1.7137446015590128e-05,
+      "loss": 0.4209,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6452083789504416,
+      "grad_norm": 0.08674032989300252,
+      "learning_rate": 1.7130300500055537e-05,
+      "loss": 0.4004,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6457922779359171,
+      "grad_norm": 0.1022891952031832,
+      "learning_rate": 1.7123147571006398e-05,
+      "loss": 0.469,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6463761769213926,
+      "grad_norm": 0.0909317793743503,
+      "learning_rate": 1.711598723587975e-05,
+      "loss": 0.4871,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6469600759068681,
+      "grad_norm": 0.0957290651843512,
+      "learning_rate": 1.710881950212033e-05,
+      "loss": 0.454,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6475439748923436,
+      "grad_norm": 0.0921966480373317,
+      "learning_rate": 1.7101644377180586e-05,
+      "loss": 0.4325,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6481278738778191,
+      "grad_norm": 0.09327407471041978,
+      "learning_rate": 1.7094461868520625e-05,
+      "loss": 0.4386,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6487117728632946,
+      "grad_norm": 0.09424397394127361,
+      "learning_rate": 1.708727198360825e-05,
+      "loss": 0.4297,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6492956718487701,
+      "grad_norm": 0.09308840225331715,
+      "learning_rate": 1.7080074729918918e-05,
+      "loss": 0.452,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6498795708342456,
+      "grad_norm": 0.09265512272026641,
+      "learning_rate": 1.7072870114935766e-05,
+      "loss": 0.4074,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6504634698197211,
+      "grad_norm": 0.08655884943934565,
+      "learning_rate": 1.7065658146149572e-05,
+      "loss": 0.3804,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6510473688051966,
+      "grad_norm": 0.09074478138858898,
+      "learning_rate": 1.7058438831058763e-05,
+      "loss": 0.444,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6516312677906723,
+      "grad_norm": 0.09773677339157741,
+      "learning_rate": 1.70512121771694e-05,
+      "loss": 0.502,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6522151667761478,
+      "grad_norm": 0.08290477692306944,
+      "learning_rate": 1.7043978191995177e-05,
+      "loss": 0.3895,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6527990657616233,
+      "grad_norm": 0.11024766171933363,
+      "learning_rate": 1.7036736883057422e-05,
+      "loss": 0.4556,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6533829647470988,
+      "grad_norm": 0.09734434101982897,
+      "learning_rate": 1.702948825788506e-05,
+      "loss": 0.4602,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6539668637325743,
+      "grad_norm": 0.0992999357867832,
+      "learning_rate": 1.7022232324014628e-05,
+      "loss": 0.4372,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6545507627180498,
+      "grad_norm": 0.08776694518910083,
+      "learning_rate": 1.7014969088990265e-05,
+      "loss": 0.3746,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6551346617035253,
+      "grad_norm": 0.0923207279090065,
+      "learning_rate": 1.7007698560363704e-05,
+      "loss": 0.3989,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6557185606890008,
+      "grad_norm": 0.10024703320991771,
+      "learning_rate": 1.7000420745694256e-05,
+      "loss": 0.4107,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6563024596744763,
+      "grad_norm": 0.09106743908835577,
+      "learning_rate": 1.6993135652548803e-05,
+      "loss": 0.4126,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6568863586599518,
+      "grad_norm": 0.09092765723442198,
+      "learning_rate": 1.6985843288501814e-05,
+      "loss": 0.4089,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6574702576454273,
+      "grad_norm": 0.0960399477490767,
+      "learning_rate": 1.697854366113529e-05,
+      "loss": 0.3921,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6580541566309028,
+      "grad_norm": 0.09970176147270751,
+      "learning_rate": 1.6971236778038806e-05,
+      "loss": 0.4363,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6586380556163783,
+      "grad_norm": 0.0874684387227263,
+      "learning_rate": 1.6963922646809475e-05,
+      "loss": 0.3908,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6592219546018538,
+      "grad_norm": 0.0911606180175274,
+      "learning_rate": 1.6956601275051933e-05,
+      "loss": 0.415,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6598058535873293,
+      "grad_norm": 0.09368550518563384,
+      "learning_rate": 1.694927267037837e-05,
+      "loss": 0.4421,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6603897525728049,
+      "grad_norm": 0.08734434760749583,
+      "learning_rate": 1.6941936840408465e-05,
+      "loss": 0.4062,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6609736515582805,
+      "grad_norm": 0.09234675102332189,
+      "learning_rate": 1.6934593792769435e-05,
+      "loss": 0.4218,
+      "step": 1132
+    },
+    {
+      "epoch": 0.661557550543756,
+      "grad_norm": 0.0960207591914515,
+      "learning_rate": 1.6927243535095995e-05,
+      "loss": 0.4953,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6621414495292315,
+      "grad_norm": 0.08306661422907055,
+      "learning_rate": 1.691988607503035e-05,
+      "loss": 0.4293,
+      "step": 1134
+    },
+    {
+      "epoch": 0.662725348514707,
+      "grad_norm": 0.09822813992720224,
+      "learning_rate": 1.691252142022219e-05,
+      "loss": 0.4267,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6633092475001825,
+      "grad_norm": 0.08787091196690758,
+      "learning_rate": 1.6905149578328705e-05,
+      "loss": 0.3788,
+      "step": 1136
+    },
+    {
+      "epoch": 0.663893146485658,
+      "grad_norm": 0.09690452114785926,
+      "learning_rate": 1.6897770557014535e-05,
+      "loss": 0.4675,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6644770454711335,
+      "grad_norm": 0.09925630645608022,
+      "learning_rate": 1.6890384363951802e-05,
+      "loss": 0.4252,
+      "step": 1138
+    },
+    {
+      "epoch": 0.665060944456609,
+      "grad_norm": 0.08788815262352959,
+      "learning_rate": 1.688299100682007e-05,
+      "loss": 0.4345,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6656448434420845,
+      "grad_norm": 0.09316906821485002,
+      "learning_rate": 1.687559049330636e-05,
+      "loss": 0.4172,
+      "step": 1140
+    },
+    {
+      "epoch": 0.66622874242756,
+      "grad_norm": 0.08868280106383877,
+      "learning_rate": 1.686818283110514e-05,
+      "loss": 0.4198,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6668126414130355,
+      "grad_norm": 0.08801283790222092,
+      "learning_rate": 1.6860768027918293e-05,
+      "loss": 0.4264,
+      "step": 1142
+    },
+    {
+      "epoch": 0.667396540398511,
+      "grad_norm": 0.08099908935354623,
+      "learning_rate": 1.6853346091455143e-05,
+      "loss": 0.3831,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6679804393839865,
+      "grad_norm": 0.08588203167405431,
+      "learning_rate": 1.684591702943242e-05,
+      "loss": 0.4331,
+      "step": 1144
+    },
+    {
+      "epoch": 0.668564338369462,
+      "grad_norm": 0.09365457716424158,
+      "learning_rate": 1.683848084957427e-05,
+      "loss": 0.3874,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6691482373549376,
+      "grad_norm": 0.08368613509525503,
+      "learning_rate": 1.6831037559612235e-05,
+      "loss": 0.4012,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6697321363404131,
+      "grad_norm": 0.08771287155713048,
+      "learning_rate": 1.682358716728525e-05,
+      "loss": 0.4185,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6703160353258886,
+      "grad_norm": 0.09725850348341017,
+      "learning_rate": 1.681612968033964e-05,
+      "loss": 0.4267,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6708999343113642,
+      "grad_norm": 0.07856673440836127,
+      "learning_rate": 1.6808665106529096e-05,
+      "loss": 0.3994,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6714838332968397,
+      "grad_norm": 0.0905549531250247,
+      "learning_rate": 1.6801193453614683e-05,
+      "loss": 0.4301,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6720677322823152,
+      "grad_norm": 0.08967676840466045,
+      "learning_rate": 1.679371472936483e-05,
+      "loss": 0.4572,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6726516312677907,
+      "grad_norm": 0.08504090087725076,
+      "learning_rate": 1.6786228941555318e-05,
+      "loss": 0.4277,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6732355302532662,
+      "grad_norm": 0.09242703790956032,
+      "learning_rate": 1.6778736097969258e-05,
+      "loss": 0.44,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6738194292387417,
+      "grad_norm": 0.08477722245915649,
+      "learning_rate": 1.6771236206397123e-05,
+      "loss": 0.4384,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6744033282242172,
+      "grad_norm": 0.0883539103328718,
+      "learning_rate": 1.676372927463668e-05,
+      "loss": 0.4328,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6749872272096927,
+      "grad_norm": 0.09185873507428662,
+      "learning_rate": 1.675621531049305e-05,
+      "loss": 0.4317,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6755711261951682,
+      "grad_norm": 0.09155377755545487,
+      "learning_rate": 1.674869432177864e-05,
+      "loss": 0.4712,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6761550251806437,
+      "grad_norm": 0.08019865946308939,
+      "learning_rate": 1.674116631631318e-05,
+      "loss": 0.419,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6767389241661192,
+      "grad_norm": 0.08164093760183552,
+      "learning_rate": 1.6733631301923678e-05,
+      "loss": 0.3601,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6773228231515948,
+      "grad_norm": 0.08758314009161688,
+      "learning_rate": 1.672608928644444e-05,
+      "loss": 0.4297,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6779067221370703,
+      "grad_norm": 0.09332800656578472,
+      "learning_rate": 1.6718540277717057e-05,
+      "loss": 0.4248,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6784906211225458,
+      "grad_norm": 0.09171057889763837,
+      "learning_rate": 1.671098428359037e-05,
+      "loss": 0.4629,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6790745201080213,
+      "grad_norm": 0.10674289030041303,
+      "learning_rate": 1.67034213119205e-05,
+      "loss": 0.4476,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6796584190934968,
+      "grad_norm": 0.09236804036989342,
+      "learning_rate": 1.6695851370570822e-05,
+      "loss": 0.4492,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6802423180789723,
+      "grad_norm": 0.08090697812666334,
+      "learning_rate": 1.6688274467411953e-05,
+      "loss": 0.3843,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6808262170644479,
+      "grad_norm": 0.09134016281324955,
+      "learning_rate": 1.6680690610321747e-05,
+      "loss": 0.4477,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6814101160499234,
+      "grad_norm": 0.09495763391793081,
+      "learning_rate": 1.667309980718529e-05,
+      "loss": 0.4355,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6819940150353989,
+      "grad_norm": 0.08399984337693617,
+      "learning_rate": 1.666550206589489e-05,
+      "loss": 0.4513,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6825779140208744,
+      "grad_norm": 0.08355252481507748,
+      "learning_rate": 1.6657897394350073e-05,
+      "loss": 0.4069,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6831618130063499,
+      "grad_norm": 0.08972300705271437,
+      "learning_rate": 1.665028580045756e-05,
+      "loss": 0.4125,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6837457119918254,
+      "grad_norm": 0.09329585528226102,
+      "learning_rate": 1.664266729213128e-05,
+      "loss": 0.4077,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6843296109773009,
+      "grad_norm": 0.08784603902556752,
+      "learning_rate": 1.6635041877292354e-05,
+      "loss": 0.4481,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6849135099627764,
+      "grad_norm": 0.09861263255563057,
+      "learning_rate": 1.662740956386906e-05,
+      "loss": 0.4732,
+      "step": 1173
+    },
+    {
+      "epoch": 0.685497408948252,
+      "grad_norm": 0.09369911466850389,
+      "learning_rate": 1.661977035979688e-05,
+      "loss": 0.4107,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6860813079337275,
+      "grad_norm": 0.0856628360850009,
+      "learning_rate": 1.661212427301844e-05,
+      "loss": 0.4259,
+      "step": 1175
+    },
+    {
+      "epoch": 0.686665206919203,
+      "grad_norm": 0.09882802938627022,
+      "learning_rate": 1.6604471311483526e-05,
+      "loss": 0.4701,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6872491059046785,
+      "grad_norm": 0.0965511312247176,
+      "learning_rate": 1.6596811483149077e-05,
+      "loss": 0.4136,
+      "step": 1177
+    },
+    {
+      "epoch": 0.687833004890154,
+      "grad_norm": 0.07837825965784558,
+      "learning_rate": 1.6589144795979165e-05,
+      "loss": 0.3932,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6884169038756295,
+      "grad_norm": 0.08445987058004721,
+      "learning_rate": 1.6581471257944996e-05,
+      "loss": 0.3896,
+      "step": 1179
+    },
+    {
+      "epoch": 0.689000802861105,
+      "grad_norm": 0.0958856133753873,
+      "learning_rate": 1.6573790877024903e-05,
+      "loss": 0.432,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6895847018465805,
+      "grad_norm": 0.08151760986818869,
+      "learning_rate": 1.656610366120433e-05,
+      "loss": 0.3853,
+      "step": 1181
+    },
+    {
+      "epoch": 0.690168600832056,
+      "grad_norm": 0.09192740838941628,
+      "learning_rate": 1.6558409618475826e-05,
+      "loss": 0.4422,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6907524998175316,
+      "grad_norm": 0.09243784833714043,
+      "learning_rate": 1.655070875683904e-05,
+      "loss": 0.4477,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6913363988030071,
+      "grad_norm": 0.08677529446766526,
+      "learning_rate": 1.6543001084300703e-05,
+      "loss": 0.4429,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6919202977884826,
+      "grad_norm": 0.08816896346761367,
+      "learning_rate": 1.653528660887465e-05,
+      "loss": 0.4478,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6925041967739581,
+      "grad_norm": 0.08259822999372791,
+      "learning_rate": 1.652756533858176e-05,
+      "loss": 0.3976,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6930880957594336,
+      "grad_norm": 0.08722276860925642,
+      "learning_rate": 1.651983728145e-05,
+      "loss": 0.4536,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6936719947449091,
+      "grad_norm": 0.09510864292916613,
+      "learning_rate": 1.6512102445514376e-05,
+      "loss": 0.431,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6942558937303847,
+      "grad_norm": 0.08851203610134378,
+      "learning_rate": 1.6504360838816956e-05,
+      "loss": 0.4106,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6948397927158602,
+      "grad_norm": 0.08836951826933792,
+      "learning_rate": 1.6496612469406835e-05,
+      "loss": 0.4677,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6954236917013357,
+      "grad_norm": 0.09351756757650727,
+      "learning_rate": 1.648885734534015e-05,
+      "loss": 0.4057,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6960075906868112,
+      "grad_norm": 0.08295798746984066,
+      "learning_rate": 1.6481095474680062e-05,
+      "loss": 0.4174,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6965914896722867,
+      "grad_norm": 0.09075608853330418,
+      "learning_rate": 1.6473326865496736e-05,
+      "loss": 0.4395,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6971753886577622,
+      "grad_norm": 0.0907479964279005,
+      "learning_rate": 1.6465551525867347e-05,
+      "loss": 0.432,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6977592876432377,
+      "grad_norm": 0.08184684057458651,
+      "learning_rate": 1.6457769463876078e-05,
+      "loss": 0.4189,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6983431866287132,
+      "grad_norm": 0.0911308926664145,
+      "learning_rate": 1.644998068761408e-05,
+      "loss": 0.442,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6989270856141887,
+      "grad_norm": 0.08407423198995699,
+      "learning_rate": 1.6442185205179507e-05,
+      "loss": 0.4,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6995109845996642,
+      "grad_norm": 0.08596313858783154,
+      "learning_rate": 1.6434383024677475e-05,
+      "loss": 0.4176,
+      "step": 1198
+    },
+    {
+      "epoch": 0.7000948835851397,
+      "grad_norm": 0.08292737262303802,
+      "learning_rate": 1.6426574154220066e-05,
+      "loss": 0.4593,
+      "step": 1199
+    },
+    {
+      "epoch": 0.7006787825706153,
+      "grad_norm": 0.08539561765885754,
+      "learning_rate": 1.6418758601926313e-05,
+      "loss": 0.3982,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7012626815560908,
+      "grad_norm": 0.08912413150934746,
+      "learning_rate": 1.64109363759222e-05,
+      "loss": 0.422,
+      "step": 1201
+    },
+    {
+      "epoch": 0.7018465805415663,
+      "grad_norm": 0.08982451897672145,
+      "learning_rate": 1.640310748434066e-05,
+      "loss": 0.4305,
+      "step": 1202
+    },
+    {
+      "epoch": 0.7024304795270419,
+      "grad_norm": 0.08209989164705718,
+      "learning_rate": 1.639527193532154e-05,
+      "loss": 0.4727,
+      "step": 1203
+    },
+    {
+      "epoch": 0.7030143785125174,
+      "grad_norm": 0.08935195563283709,
+      "learning_rate": 1.6387429737011612e-05,
+      "loss": 0.4377,
+      "step": 1204
+    },
+    {
+      "epoch": 0.7035982774979929,
+      "grad_norm": 0.08174099854905177,
+      "learning_rate": 1.6379580897564568e-05,
+      "loss": 0.4006,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7041821764834684,
+      "grad_norm": 0.08601574410273431,
+      "learning_rate": 1.6371725425141e-05,
+      "loss": 0.4059,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7047660754689439,
+      "grad_norm": 0.09479900581101951,
+      "learning_rate": 1.6363863327908405e-05,
+      "loss": 0.4752,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7053499744544194,
+      "grad_norm": 0.09302586669989119,
+      "learning_rate": 1.6355994614041154e-05,
+      "loss": 0.4796,
+      "step": 1208
+    },
+    {
+      "epoch": 0.7059338734398949,
+      "grad_norm": 0.08381395879481965,
+      "learning_rate": 1.6348119291720504e-05,
+      "loss": 0.4051,
+      "step": 1209
+    },
+    {
+      "epoch": 0.7065177724253704,
+      "grad_norm": 0.09250409886196691,
+      "learning_rate": 1.634023736913459e-05,
+      "loss": 0.4357,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7071016714108459,
+      "grad_norm": 0.09701599034462823,
+      "learning_rate": 1.6332348854478398e-05,
+      "loss": 0.451,
+      "step": 1211
+    },
+    {
+      "epoch": 0.7076855703963214,
+      "grad_norm": 0.09996490043248872,
+      "learning_rate": 1.6324453755953772e-05,
+      "loss": 0.4372,
+      "step": 1212
+    },
+    {
+      "epoch": 0.7082694693817969,
+      "grad_norm": 0.08638337651416796,
+      "learning_rate": 1.6316552081769404e-05,
+      "loss": 0.4121,
+      "step": 1213
+    },
+    {
+      "epoch": 0.7088533683672724,
+      "grad_norm": 0.10290298605955167,
+      "learning_rate": 1.630864384014083e-05,
+      "loss": 0.4973,
+      "step": 1214
+    },
+    {
+      "epoch": 0.7094372673527479,
+      "grad_norm": 0.09710064387163618,
+      "learning_rate": 1.6300729039290386e-05,
+      "loss": 0.4478,
+      "step": 1215
+    },
+    {
+      "epoch": 0.7100211663382235,
+      "grad_norm": 0.0922893148207513,
+      "learning_rate": 1.6292807687447258e-05,
+      "loss": 0.4321,
+      "step": 1216
+    },
+    {
+      "epoch": 0.710605065323699,
+      "grad_norm": 0.08402375639199999,
+      "learning_rate": 1.6284879792847433e-05,
+      "loss": 0.3872,
+      "step": 1217
+    },
+    {
+      "epoch": 0.7111889643091746,
+      "grad_norm": 0.09781728368936367,
+      "learning_rate": 1.62769453637337e-05,
+      "loss": 0.4797,
+      "step": 1218
+    },
+    {
+      "epoch": 0.7117728632946501,
+      "grad_norm": 0.10210312663401595,
+      "learning_rate": 1.626900440835564e-05,
+      "loss": 0.4604,
+      "step": 1219
+    },
+    {
+      "epoch": 0.7123567622801256,
+      "grad_norm": 0.08923738474981838,
+      "learning_rate": 1.6261056934969626e-05,
+      "loss": 0.4144,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7129406612656011,
+      "grad_norm": 0.0961526304338111,
+      "learning_rate": 1.6253102951838794e-05,
+      "loss": 0.4174,
+      "step": 1221
+    },
+    {
+      "epoch": 0.7135245602510766,
+      "grad_norm": 0.09558520586303418,
+      "learning_rate": 1.6245142467233067e-05,
+      "loss": 0.408,
+      "step": 1222
+    },
+    {
+      "epoch": 0.7141084592365521,
+      "grad_norm": 0.08398861509146499,
+      "learning_rate": 1.6237175489429114e-05,
+      "loss": 0.3939,
+      "step": 1223
+    },
+    {
+      "epoch": 0.7146923582220276,
+      "grad_norm": 0.09590974255359404,
+      "learning_rate": 1.6229202026710356e-05,
+      "loss": 0.4416,
+      "step": 1224
+    },
+    {
+      "epoch": 0.7152762572075031,
+      "grad_norm": 0.0853938632533378,
+      "learning_rate": 1.622122208736697e-05,
+      "loss": 0.3821,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7158601561929786,
+      "grad_norm": 0.08641520596986882,
+      "learning_rate": 1.6213235679695847e-05,
+      "loss": 0.3995,
+      "step": 1226
+    },
+    {
+      "epoch": 0.7164440551784541,
+      "grad_norm": 0.0924650830189144,
+      "learning_rate": 1.620524281200062e-05,
+      "loss": 0.4132,
+      "step": 1227
+    },
+    {
+      "epoch": 0.7170279541639296,
+      "grad_norm": 0.08718641423957069,
+      "learning_rate": 1.6197243492591627e-05,
+      "loss": 0.4178,
+      "step": 1228
+    },
+    {
+      "epoch": 0.7176118531494051,
+      "grad_norm": 0.09786371055395311,
+      "learning_rate": 1.618923772978592e-05,
+      "loss": 0.4555,
+      "step": 1229
+    },
+    {
+      "epoch": 0.7181957521348806,
+      "grad_norm": 0.08320203015753497,
+      "learning_rate": 1.618122553190725e-05,
+      "loss": 0.3892,
+      "step": 1230
+    },
+    {
+      "epoch": 0.7187796511203561,
+      "grad_norm": 0.09737195028493588,
+      "learning_rate": 1.617320690728606e-05,
+      "loss": 0.4307,
+      "step": 1231
+    },
+    {
+      "epoch": 0.7193635501058316,
+      "grad_norm": 0.08530531048644806,
+      "learning_rate": 1.6165181864259463e-05,
+      "loss": 0.4447,
+      "step": 1232
+    },
+    {
+      "epoch": 0.7199474490913073,
+      "grad_norm": 0.0876034802983154,
+      "learning_rate": 1.6157150411171268e-05,
+      "loss": 0.4272,
+      "step": 1233
+    },
+    {
+      "epoch": 0.7205313480767828,
+      "grad_norm": 0.08623710218708526,
+      "learning_rate": 1.614911255637193e-05,
+      "loss": 0.4174,
+      "step": 1234
+    },
+    {
+      "epoch": 0.7211152470622583,
+      "grad_norm": 0.09488808809079846,
+      "learning_rate": 1.6141068308218565e-05,
+      "loss": 0.3992,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7216991460477338,
+      "grad_norm": 0.08918652565741696,
+      "learning_rate": 1.6133017675074935e-05,
+      "loss": 0.4153,
+      "step": 1236
+    },
+    {
+      "epoch": 0.7222830450332093,
+      "grad_norm": 0.08551704210722187,
+      "learning_rate": 1.6124960665311447e-05,
+      "loss": 0.4134,
+      "step": 1237
+    },
+    {
+      "epoch": 0.7228669440186848,
+      "grad_norm": 0.0803986624296922,
+      "learning_rate": 1.6116897287305132e-05,
+      "loss": 0.4033,
+      "step": 1238
+    },
+    {
+      "epoch": 0.7234508430041603,
+      "grad_norm": 0.1000469760257329,
+      "learning_rate": 1.6108827549439642e-05,
+      "loss": 0.481,
+      "step": 1239
+    },
+    {
+      "epoch": 0.7240347419896358,
+      "grad_norm": 0.09991049415758912,
+      "learning_rate": 1.6100751460105244e-05,
+      "loss": 0.4621,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7246186409751113,
+      "grad_norm": 0.09038163061510153,
+      "learning_rate": 1.6092669027698812e-05,
+      "loss": 0.4273,
+      "step": 1241
+    },
+    {
+      "epoch": 0.7252025399605868,
+      "grad_norm": 0.08150547573087052,
+      "learning_rate": 1.6084580260623805e-05,
+      "loss": 0.4261,
+      "step": 1242
+    },
+    {
+      "epoch": 0.7257864389460623,
+      "grad_norm": 0.08705772317012889,
+      "learning_rate": 1.6076485167290278e-05,
+      "loss": 0.4057,
+      "step": 1243
+    },
+    {
+      "epoch": 0.7263703379315378,
+      "grad_norm": 0.08817458392997114,
+      "learning_rate": 1.6068383756114857e-05,
+      "loss": 0.4422,
+      "step": 1244
+    },
+    {
+      "epoch": 0.7269542369170133,
+      "grad_norm": 0.09366227874279444,
+      "learning_rate": 1.606027603552074e-05,
+      "loss": 0.426,
+      "step": 1245
+    },
+    {
+      "epoch": 0.7275381359024888,
+      "grad_norm": 0.08993597694006221,
+      "learning_rate": 1.6052162013937688e-05,
+      "loss": 0.4445,
+      "step": 1246
+    },
+    {
+      "epoch": 0.7281220348879643,
+      "grad_norm": 0.08699273125590415,
+      "learning_rate": 1.6044041699802005e-05,
+      "loss": 0.4092,
+      "step": 1247
+    },
+    {
+      "epoch": 0.7287059338734398,
+      "grad_norm": 0.08804728749248808,
+      "learning_rate": 1.6035915101556544e-05,
+      "loss": 0.426,
+      "step": 1248
+    },
+    {
+      "epoch": 0.7292898328589154,
+      "grad_norm": 0.09100025255879178,
+      "learning_rate": 1.6027782227650696e-05,
+      "loss": 0.4336,
+      "step": 1249
+    },
+    {
+      "epoch": 0.729873731844391,
+      "grad_norm": 0.08831630586289121,
+      "learning_rate": 1.601964308654036e-05,
+      "loss": 0.442,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7304576308298665,
+      "grad_norm": 0.09117134601106955,
+      "learning_rate": 1.601149768668797e-05,
+      "loss": 0.4201,
+      "step": 1251
+    },
+    {
+      "epoch": 0.731041529815342,
+      "grad_norm": 0.0929859845865332,
+      "learning_rate": 1.6003346036562457e-05,
+      "loss": 0.4721,
+      "step": 1252
+    },
+    {
+      "epoch": 0.7316254288008175,
+      "grad_norm": 0.09716258496229974,
+      "learning_rate": 1.599518814463925e-05,
+      "loss": 0.4613,
+      "step": 1253
+    },
+    {
+      "epoch": 0.732209327786293,
+      "grad_norm": 0.09254159015928828,
+      "learning_rate": 1.598702401940028e-05,
+      "loss": 0.4528,
+      "step": 1254
+    },
+    {
+      "epoch": 0.7327932267717685,
+      "grad_norm": 0.08981863068889674,
+      "learning_rate": 1.5978853669333938e-05,
+      "loss": 0.4386,
+      "step": 1255
+    },
+    {
+      "epoch": 0.733377125757244,
+      "grad_norm": 0.0969066369136787,
+      "learning_rate": 1.597067710293511e-05,
+      "loss": 0.4408,
+      "step": 1256
+    },
+    {
+      "epoch": 0.7339610247427195,
+      "grad_norm": 0.08970247516171302,
+      "learning_rate": 1.5962494328705123e-05,
+      "loss": 0.3839,
+      "step": 1257
+    },
+    {
+      "epoch": 0.734544923728195,
+      "grad_norm": 0.09246677511237618,
+      "learning_rate": 1.5954305355151775e-05,
+      "loss": 0.4614,
+      "step": 1258
+    },
+    {
+      "epoch": 0.7351288227136705,
+      "grad_norm": 0.09068691494719908,
+      "learning_rate": 1.5946110190789306e-05,
+      "loss": 0.4458,
+      "step": 1259
+    },
+    {
+      "epoch": 0.735712721699146,
+      "grad_norm": 0.09556752148183027,
+      "learning_rate": 1.5937908844138386e-05,
+      "loss": 0.472,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7362966206846215,
+      "grad_norm": 0.09147134869788584,
+      "learning_rate": 1.5929701323726113e-05,
+      "loss": 0.4504,
+      "step": 1261
+    },
+    {
+      "epoch": 0.736880519670097,
+      "grad_norm": 0.08883323843599066,
+      "learning_rate": 1.5921487638086024e-05,
+      "loss": 0.4288,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7374644186555726,
+      "grad_norm": 0.08497959894690894,
+      "learning_rate": 1.5913267795758037e-05,
+      "loss": 0.4132,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7380483176410481,
+      "grad_norm": 0.09518231451855877,
+      "learning_rate": 1.590504180528849e-05,
+      "loss": 0.4548,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7386322166265236,
+      "grad_norm": 0.09065954797102842,
+      "learning_rate": 1.5896809675230106e-05,
+      "loss": 0.4202,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7392161156119991,
+      "grad_norm": 0.08484196161131463,
+      "learning_rate": 1.5888571414141997e-05,
+      "loss": 0.3951,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7398000145974747,
+      "grad_norm": 0.09192186327021566,
+      "learning_rate": 1.588032703058964e-05,
+      "loss": 0.4053,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7403839135829502,
+      "grad_norm": 0.08668077160830087,
+      "learning_rate": 1.587207653314489e-05,
+      "loss": 0.4187,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7409678125684257,
+      "grad_norm": 0.0921172767329433,
+      "learning_rate": 1.586381993038595e-05,
+      "loss": 0.4479,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7415517115539012,
+      "grad_norm": 0.08946975515016198,
+      "learning_rate": 1.5855557230897373e-05,
+      "loss": 0.4249,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7421356105393767,
+      "grad_norm": 0.09461668014907235,
+      "learning_rate": 1.584728844327005e-05,
+      "loss": 0.401,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7427195095248522,
+      "grad_norm": 0.09084318608847822,
+      "learning_rate": 1.5839013576101206e-05,
+      "loss": 0.4284,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7433034085103277,
+      "grad_norm": 0.09162843532111148,
+      "learning_rate": 1.5830732637994382e-05,
+      "loss": 0.4444,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7438873074958032,
+      "grad_norm": 0.08906945689899146,
+      "learning_rate": 1.5822445637559435e-05,
+      "loss": 0.4327,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7444712064812787,
+      "grad_norm": 0.09084520748081751,
+      "learning_rate": 1.581415258341252e-05,
+      "loss": 0.4242,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7450551054667542,
+      "grad_norm": 0.08770156253896053,
+      "learning_rate": 1.5805853484176093e-05,
+      "loss": 0.4309,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7456390044522297,
+      "grad_norm": 0.08946315892753577,
+      "learning_rate": 1.5797548348478893e-05,
+      "loss": 0.416,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7462229034377053,
+      "grad_norm": 0.0919971973152717,
+      "learning_rate": 1.578923718495593e-05,
+      "loss": 0.4378,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7468068024231808,
+      "grad_norm": 0.08894648538278113,
+      "learning_rate": 1.5780920002248484e-05,
+      "loss": 0.4482,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7473907014086563,
+      "grad_norm": 0.07931274724671593,
+      "learning_rate": 1.5772596809004103e-05,
+      "loss": 0.4026,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7479746003941318,
+      "grad_norm": 0.08818667111690112,
+      "learning_rate": 1.5764267613876565e-05,
+      "loss": 0.4156,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7485584993796073,
+      "grad_norm": 0.08850607912849418,
+      "learning_rate": 1.5755932425525907e-05,
+      "loss": 0.4214,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7491423983650829,
+      "grad_norm": 0.08540105668148078,
+      "learning_rate": 1.574759125261838e-05,
+      "loss": 0.4198,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7497262973505584,
+      "grad_norm": 0.09161359336642806,
+      "learning_rate": 1.573924410382648e-05,
+      "loss": 0.415,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7503101963360339,
+      "grad_norm": 0.09956734779585424,
+      "learning_rate": 1.5730890987828893e-05,
+      "loss": 0.4619,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7508940953215094,
+      "grad_norm": 0.08604248632515632,
+      "learning_rate": 1.5722531913310523e-05,
+      "loss": 0.454,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7514779943069849,
+      "grad_norm": 0.08516091791390863,
+      "learning_rate": 1.571416688896246e-05,
+      "loss": 0.3699,
+      "step": 1287
+    },
+    {
+      "epoch": 0.7520618932924604,
+      "grad_norm": 0.08657085696409755,
+      "learning_rate": 1.5705795923481995e-05,
+      "loss": 0.3575,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7526457922779359,
+      "grad_norm": 0.09024035340255226,
+      "learning_rate": 1.5697419025572577e-05,
+      "loss": 0.4555,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7532296912634114,
+      "grad_norm": 0.08891845662343668,
+      "learning_rate": 1.5689036203943836e-05,
+      "loss": 0.4486,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7538135902488869,
+      "grad_norm": 0.08814717314313003,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.3949,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7543974892343625,
+      "grad_norm": 0.0995478466035641,
+      "learning_rate": 1.5672252824397683e-05,
+      "loss": 0.4294,
+      "step": 1292
+    },
+    {
+      "epoch": 0.754981388219838,
+      "grad_norm": 0.09875897195786154,
+      "learning_rate": 1.5663852283930275e-05,
+      "loss": 0.4415,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7555652872053135,
+      "grad_norm": 0.09668446994182185,
+      "learning_rate": 1.5655445854643554e-05,
+      "loss": 0.499,
+      "step": 1294
+    },
+    {
+      "epoch": 0.756149186190789,
+      "grad_norm": 0.08938861505498946,
+      "learning_rate": 1.5647033545277847e-05,
+      "loss": 0.4399,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7567330851762645,
+      "grad_norm": 0.08621365897822844,
+      "learning_rate": 1.56386153645796e-05,
+      "loss": 0.4317,
+      "step": 1296
+    },
+    {
+      "epoch": 0.75731698416174,
+      "grad_norm": 0.08805910066848739,
+      "learning_rate": 1.563019132130136e-05,
+      "loss": 0.3964,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7579008831472155,
+      "grad_norm": 0.08321262999738621,
+      "learning_rate": 1.562176142420177e-05,
+      "loss": 0.4108,
+      "step": 1298
+    },
+    {
+      "epoch": 0.758484782132691,
+      "grad_norm": 0.09037258601407526,
+      "learning_rate": 1.5613325682045563e-05,
+      "loss": 0.4329,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7590686811181666,
+      "grad_norm": 0.08467100658167585,
+      "learning_rate": 1.5604884103603547e-05,
+      "loss": 0.4193,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7596525801036421,
+      "grad_norm": 0.08453136693956839,
+      "learning_rate": 1.55964366976526e-05,
+      "loss": 0.4158,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7602364790891176,
+      "grad_norm": 0.08734167832986295,
+      "learning_rate": 1.5587983472975653e-05,
+      "loss": 0.4106,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7608203780745931,
+      "grad_norm": 0.08758245816480005,
+      "learning_rate": 1.5579524438361693e-05,
+      "loss": 0.3882,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7614042770600686,
+      "grad_norm": 0.09074955582235424,
+      "learning_rate": 1.5571059602605746e-05,
+      "loss": 0.4402,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7619881760455441,
+      "grad_norm": 0.0846874802248585,
+      "learning_rate": 1.556258897450887e-05,
+      "loss": 0.451,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7625720750310196,
+      "grad_norm": 0.08575911326999842,
+      "learning_rate": 1.5554112562878144e-05,
+      "loss": 0.3871,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7631559740164952,
+      "grad_norm": 0.08982036656153952,
+      "learning_rate": 1.5545630376526665e-05,
+      "loss": 0.4228,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7637398730019707,
+      "grad_norm": 0.09203980641154026,
+      "learning_rate": 1.553714242427352e-05,
+      "loss": 0.4259,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7643237719874462,
+      "grad_norm": 0.09287967132645603,
+      "learning_rate": 1.5528648714943807e-05,
+      "loss": 0.4457,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7649076709729217,
+      "grad_norm": 0.0860068949678233,
+      "learning_rate": 1.5520149257368608e-05,
+      "loss": 0.4235,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7654915699583972,
+      "grad_norm": 0.08494700597712204,
+      "learning_rate": 1.5511644060384968e-05,
+      "loss": 0.4082,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7660754689438727,
+      "grad_norm": 0.08616900714291478,
+      "learning_rate": 1.5503133132835916e-05,
+      "loss": 0.4213,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7666593679293482,
+      "grad_norm": 0.08613809547025654,
+      "learning_rate": 1.5494616483570428e-05,
+      "loss": 0.4154,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7672432669148237,
+      "grad_norm": 0.0837406710056406,
+      "learning_rate": 1.5486094121443434e-05,
+      "loss": 0.3936,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7678271659002992,
+      "grad_norm": 0.08688951323952172,
+      "learning_rate": 1.5477566055315808e-05,
+      "loss": 0.4189,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7684110648857747,
+      "grad_norm": 0.08852377490382674,
+      "learning_rate": 1.5469032294054336e-05,
+      "loss": 0.4521,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7689949638712503,
+      "grad_norm": 0.08227315393123309,
+      "learning_rate": 1.5460492846531748e-05,
+      "loss": 0.3984,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7695788628567258,
+      "grad_norm": 0.08225570997416073,
+      "learning_rate": 1.5451947721626676e-05,
+      "loss": 0.4077,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7701627618422013,
+      "grad_norm": 0.08459113862127972,
+      "learning_rate": 1.5443396928223655e-05,
+      "loss": 0.4372,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7707466608276768,
+      "grad_norm": 0.08937392889428214,
+      "learning_rate": 1.5434840475213113e-05,
+      "loss": 0.41,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7713305598131523,
+      "grad_norm": 0.09061326145441673,
+      "learning_rate": 1.5426278371491363e-05,
+      "loss": 0.4144,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7719144587986279,
+      "grad_norm": 0.08454009654537566,
+      "learning_rate": 1.5417710625960598e-05,
+      "loss": 0.415,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7724983577841034,
+      "grad_norm": 0.0906703302393027,
+      "learning_rate": 1.5409137247528868e-05,
+      "loss": 0.4134,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7730822567695789,
+      "grad_norm": 0.08700571764635147,
+      "learning_rate": 1.5400558245110083e-05,
+      "loss": 0.386,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7736661557550544,
+      "grad_norm": 0.08947539144982224,
+      "learning_rate": 1.5391973627624004e-05,
+      "loss": 0.4211,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7742500547405299,
+      "grad_norm": 0.0896607713946373,
+      "learning_rate": 1.538338340399623e-05,
+      "loss": 0.4022,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7748339537260054,
+      "grad_norm": 0.08501201276744527,
+      "learning_rate": 1.5374787583158188e-05,
+      "loss": 0.4468,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7754178527114809,
+      "grad_norm": 0.09217256435507994,
+      "learning_rate": 1.5366186174047114e-05,
+      "loss": 0.4471,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7760017516969564,
+      "grad_norm": 0.08798899098610476,
+      "learning_rate": 1.535757918560607e-05,
+      "loss": 0.4396,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7765856506824319,
+      "grad_norm": 0.11711224913670701,
+      "learning_rate": 1.534896662678391e-05,
+      "loss": 0.4911,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7771695496679074,
+      "grad_norm": 0.08900880722615902,
+      "learning_rate": 1.534034850653528e-05,
+      "loss": 0.4272,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7777534486533829,
+      "grad_norm": 0.0917561490675343,
+      "learning_rate": 1.533172483382062e-05,
+      "loss": 0.3936,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7783373476388584,
+      "grad_norm": 0.08552671016681809,
+      "learning_rate": 1.532309561760612e-05,
+      "loss": 0.4156,
+      "step": 1333
+    },
+    {
+      "epoch": 0.778921246624334,
+      "grad_norm": 0.08850201131674452,
+      "learning_rate": 1.5314460866863758e-05,
+      "loss": 0.4676,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7795051456098095,
+      "grad_norm": 0.08275541286583958,
+      "learning_rate": 1.530582059057125e-05,
+      "loss": 0.3827,
+      "step": 1335
+    },
+    {
+      "epoch": 0.780089044595285,
+      "grad_norm": 0.09812999273262774,
+      "learning_rate": 1.5297174797712057e-05,
+      "loss": 0.4727,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7806729435807606,
+      "grad_norm": 0.09324335462091086,
+      "learning_rate": 1.5288523497275392e-05,
+      "loss": 0.4129,
+      "step": 1337
+    },
+    {
+      "epoch": 0.7812568425662361,
+      "grad_norm": 0.09592462559031027,
+      "learning_rate": 1.5279866698256177e-05,
+      "loss": 0.4706,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7818407415517116,
+      "grad_norm": 0.08314225476851099,
+      "learning_rate": 1.5271204409655055e-05,
+      "loss": 0.3875,
+      "step": 1339
+    },
+    {
+      "epoch": 0.7824246405371871,
+      "grad_norm": 0.09171763397454903,
+      "learning_rate": 1.5262536640478386e-05,
+      "loss": 0.417,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7830085395226626,
+      "grad_norm": 0.08732713817990942,
+      "learning_rate": 1.5253863399738218e-05,
+      "loss": 0.4198,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7835924385081381,
+      "grad_norm": 0.08113686715451263,
+      "learning_rate": 1.5245184696452286e-05,
+      "loss": 0.4231,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7841763374936136,
+      "grad_norm": 0.09331287606911522,
+      "learning_rate": 1.5236500539644015e-05,
+      "loss": 0.4924,
+      "step": 1343
+    },
+    {
+      "epoch": 0.7847602364790891,
+      "grad_norm": 0.08915904955849609,
+      "learning_rate": 1.5227810938342493e-05,
+      "loss": 0.4373,
+      "step": 1344
+    },
+    {
+      "epoch": 0.7853441354645646,
+      "grad_norm": 0.08313964318544646,
+      "learning_rate": 1.5219115901582471e-05,
+      "loss": 0.4413,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7859280344500401,
+      "grad_norm": 0.08619376473465097,
+      "learning_rate": 1.5210415438404354e-05,
+      "loss": 0.447,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7865119334355156,
+      "grad_norm": 0.08728490607424277,
+      "learning_rate": 1.5201709557854178e-05,
+      "loss": 0.4709,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7870958324209911,
+      "grad_norm": 0.0862061320672254,
+      "learning_rate": 1.5192998268983625e-05,
+      "loss": 0.44,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7876797314064666,
+      "grad_norm": 0.08093460476160619,
+      "learning_rate": 1.5184281580849999e-05,
+      "loss": 0.4134,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7882636303919422,
+      "grad_norm": 0.08871021625242784,
+      "learning_rate": 1.51755595025162e-05,
+      "loss": 0.424,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7888475293774178,
+      "grad_norm": 0.08221116529811918,
+      "learning_rate": 1.5166832043050757e-05,
+      "loss": 0.3993,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7894314283628933,
+      "grad_norm": 0.08480451298168319,
+      "learning_rate": 1.5158099211527776e-05,
+      "loss": 0.4123,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7900153273483688,
+      "grad_norm": 0.0795889927213935,
+      "learning_rate": 1.5149361017026957e-05,
+      "loss": 0.4174,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7905992263338443,
+      "grad_norm": 0.08592216954200851,
+      "learning_rate": 1.5140617468633579e-05,
+      "loss": 0.4299,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7911831253193198,
+      "grad_norm": 0.09431799568912874,
+      "learning_rate": 1.513186857543847e-05,
+      "loss": 0.4448,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7917670243047953,
+      "grad_norm": 0.0868691355990168,
+      "learning_rate": 1.5123114346538037e-05,
+      "loss": 0.4148,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7923509232902708,
+      "grad_norm": 0.08004297342242324,
+      "learning_rate": 1.5114354791034225e-05,
+      "loss": 0.4178,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7929348222757463,
+      "grad_norm": 0.08946896916428285,
+      "learning_rate": 1.5105589918034511e-05,
+      "loss": 0.3991,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7935187212612218,
+      "grad_norm": 0.09276454978737551,
+      "learning_rate": 1.5096819736651913e-05,
+      "loss": 0.4424,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7941026202466973,
+      "grad_norm": 0.08614512623105408,
+      "learning_rate": 1.5088044256004958e-05,
+      "loss": 0.4302,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7946865192321728,
+      "grad_norm": 0.0915332435108306,
+      "learning_rate": 1.5079263485217693e-05,
+      "loss": 0.464,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7952704182176483,
+      "grad_norm": 0.08128136201222123,
+      "learning_rate": 1.507047743341965e-05,
+      "loss": 0.4098,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7958543172031238,
+      "grad_norm": 0.08969220861684299,
+      "learning_rate": 1.506168610974587e-05,
+      "loss": 0.4259,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7964382161885993,
+      "grad_norm": 0.09027220284407747,
+      "learning_rate": 1.505288952333686e-05,
+      "loss": 0.4188,
+      "step": 1364
+    },
+    {
+      "epoch": 0.7970221151740748,
+      "grad_norm": 0.08760538602208404,
+      "learning_rate": 1.5044087683338609e-05,
+      "loss": 0.4444,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7976060141595503,
+      "grad_norm": 0.08321545010450747,
+      "learning_rate": 1.5035280598902557e-05,
+      "loss": 0.3951,
+      "step": 1366
+    },
+    {
+      "epoch": 0.798189913145026,
+      "grad_norm": 0.08081809704403048,
+      "learning_rate": 1.5026468279185615e-05,
+      "loss": 0.414,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7987738121305015,
+      "grad_norm": 0.08950260764471552,
+      "learning_rate": 1.5017650733350121e-05,
+      "loss": 0.4376,
+      "step": 1368
+    },
+    {
+      "epoch": 0.799357711115977,
+      "grad_norm": 0.09734577740355589,
+      "learning_rate": 1.5008827970563848e-05,
+      "loss": 0.512,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7999416101014525,
+      "grad_norm": 0.08701137258419459,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.4363,
+      "step": 1370
+    },
+    {
+      "epoch": 0.800525509086928,
+      "grad_norm": 0.08927530115830506,
+      "learning_rate": 1.4991166830837198e-05,
+      "loss": 0.4154,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8011094080724035,
+      "grad_norm": 0.08542649098860353,
+      "learning_rate": 1.4982328472259453e-05,
+      "loss": 0.3885,
+      "step": 1372
+    },
+    {
+      "epoch": 0.801693307057879,
+      "grad_norm": 0.09702722333413816,
+      "learning_rate": 1.4973484933456191e-05,
+      "loss": 0.4337,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8022772060433545,
+      "grad_norm": 0.09769084516159367,
+      "learning_rate": 1.4964636223622206e-05,
+      "loss": 0.4825,
+      "step": 1374
+    },
+    {
+      "epoch": 0.80286110502883,
+      "grad_norm": 0.08205800844336521,
+      "learning_rate": 1.4955782351957681e-05,
+      "loss": 0.4146,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8034450040143055,
+      "grad_norm": 0.09547941184327449,
+      "learning_rate": 1.4946923327668164e-05,
+      "loss": 0.3986,
+      "step": 1376
+    },
+    {
+      "epoch": 0.804028902999781,
+      "grad_norm": 0.08817037122819231,
+      "learning_rate": 1.4938059159964555e-05,
+      "loss": 0.3878,
+      "step": 1377
+    },
+    {
+      "epoch": 0.8046128019852565,
+      "grad_norm": 0.08619961401244314,
+      "learning_rate": 1.4929189858063103e-05,
+      "loss": 0.3837,
+      "step": 1378
+    },
+    {
+      "epoch": 0.805196700970732,
+      "grad_norm": 0.0889582827254665,
+      "learning_rate": 1.4920315431185398e-05,
+      "loss": 0.4203,
+      "step": 1379
+    },
+    {
+      "epoch": 0.8057805999562075,
+      "grad_norm": 0.09434637312001887,
+      "learning_rate": 1.4911435888558356e-05,
+      "loss": 0.4414,
+      "step": 1380
+    },
+    {
+      "epoch": 0.806364498941683,
+      "grad_norm": 0.08499264531205213,
+      "learning_rate": 1.4902551239414218e-05,
+      "loss": 0.4283,
+      "step": 1381
+    },
+    {
+      "epoch": 0.8069483979271586,
+      "grad_norm": 0.0831669689421554,
+      "learning_rate": 1.4893661492990527e-05,
+      "loss": 0.4375,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8075322969126341,
+      "grad_norm": 0.08891687439274013,
+      "learning_rate": 1.4884766658530126e-05,
+      "loss": 0.4084,
+      "step": 1383
+    },
+    {
+      "epoch": 0.8081161958981097,
+      "grad_norm": 0.08286954302485844,
+      "learning_rate": 1.487586674528115e-05,
+      "loss": 0.3786,
+      "step": 1384
+    },
+    {
+      "epoch": 0.8087000948835852,
+      "grad_norm": 0.09113178252518411,
+      "learning_rate": 1.4866961762497018e-05,
+      "loss": 0.4337,
+      "step": 1385
+    },
+    {
+      "epoch": 0.8092839938690607,
+      "grad_norm": 0.08658788245598888,
+      "learning_rate": 1.4858051719436418e-05,
+      "loss": 0.3989,
+      "step": 1386
+    },
+    {
+      "epoch": 0.8098678928545362,
+      "grad_norm": 0.08384416264595368,
+      "learning_rate": 1.4849136625363297e-05,
+      "loss": 0.3721,
+      "step": 1387
+    },
+    {
+      "epoch": 0.8104517918400117,
+      "grad_norm": 0.08919742970575643,
+      "learning_rate": 1.484021648954685e-05,
+      "loss": 0.4125,
+      "step": 1388
+    },
+    {
+      "epoch": 0.8110356908254872,
+      "grad_norm": 0.09586990535846085,
+      "learning_rate": 1.4831291321261523e-05,
+      "loss": 0.4803,
+      "step": 1389
+    },
+    {
+      "epoch": 0.8116195898109627,
+      "grad_norm": 0.08829188968242362,
+      "learning_rate": 1.4822361129786992e-05,
+      "loss": 0.4485,
+      "step": 1390
+    },
+    {
+      "epoch": 0.8122034887964382,
+      "grad_norm": 0.08610110175024323,
+      "learning_rate": 1.4813425924408151e-05,
+      "loss": 0.3954,
+      "step": 1391
+    },
+    {
+      "epoch": 0.8127873877819137,
+      "grad_norm": 0.08276002839694131,
+      "learning_rate": 1.480448571441511e-05,
+      "loss": 0.4112,
+      "step": 1392
+    },
+    {
+      "epoch": 0.8133712867673892,
+      "grad_norm": 0.08545870174731693,
+      "learning_rate": 1.4795540509103182e-05,
+      "loss": 0.4494,
+      "step": 1393
+    },
+    {
+      "epoch": 0.8139551857528647,
+      "grad_norm": 0.08141907944209822,
+      "learning_rate": 1.4786590317772875e-05,
+      "loss": 0.3886,
+      "step": 1394
+    },
+    {
+      "epoch": 0.8145390847383402,
+      "grad_norm": 0.08602182981301064,
+      "learning_rate": 1.4777635149729878e-05,
+      "loss": 0.4108,
+      "step": 1395
+    },
+    {
+      "epoch": 0.8151229837238158,
+      "grad_norm": 0.08144142019395251,
+      "learning_rate": 1.4768675014285063e-05,
+      "loss": 0.3923,
+      "step": 1396
+    },
+    {
+      "epoch": 0.8157068827092913,
+      "grad_norm": 0.08023195785348626,
+      "learning_rate": 1.4759709920754453e-05,
+      "loss": 0.446,
+      "step": 1397
+    },
+    {
+      "epoch": 0.8162907816947668,
+      "grad_norm": 0.0921140190415617,
+      "learning_rate": 1.4750739878459233e-05,
+      "loss": 0.4238,
+      "step": 1398
+    },
+    {
+      "epoch": 0.8168746806802423,
+      "grad_norm": 0.08323385912103386,
+      "learning_rate": 1.4741764896725736e-05,
+      "loss": 0.4134,
+      "step": 1399
+    },
+    {
+      "epoch": 0.8174585796657178,
+      "grad_norm": 0.08331428919814646,
+      "learning_rate": 1.473278498488543e-05,
+      "loss": 0.4194,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8180424786511934,
+      "grad_norm": 0.08972140410254771,
+      "learning_rate": 1.4723800152274905e-05,
+      "loss": 0.4255,
+      "step": 1401
+    },
+    {
+      "epoch": 0.8186263776366689,
+      "grad_norm": 0.08170388116189564,
+      "learning_rate": 1.471481040823587e-05,
+      "loss": 0.4042,
+      "step": 1402
+    },
+    {
+      "epoch": 0.8192102766221444,
+      "grad_norm": 0.079466122770402,
+      "learning_rate": 1.4705815762115138e-05,
+      "loss": 0.4043,
+      "step": 1403
+    },
+    {
+      "epoch": 0.8197941756076199,
+      "grad_norm": 0.07924132398697102,
+      "learning_rate": 1.4696816223264622e-05,
+      "loss": 0.4027,
+      "step": 1404
+    },
+    {
+      "epoch": 0.8203780745930954,
+      "grad_norm": 0.08236672403715109,
+      "learning_rate": 1.4687811801041323e-05,
+      "loss": 0.3897,
+      "step": 1405
+    },
+    {
+      "epoch": 0.8209619735785709,
+      "grad_norm": 0.08336187514859361,
+      "learning_rate": 1.4678802504807313e-05,
+      "loss": 0.4074,
+      "step": 1406
+    },
+    {
+      "epoch": 0.8215458725640464,
+      "grad_norm": 0.08420762082295365,
+      "learning_rate": 1.4669788343929736e-05,
+      "loss": 0.4271,
+      "step": 1407
+    },
+    {
+      "epoch": 0.8221297715495219,
+      "grad_norm": 0.07922384191348895,
+      "learning_rate": 1.4660769327780796e-05,
+      "loss": 0.3947,
+      "step": 1408
+    },
+    {
+      "epoch": 0.8227136705349974,
+      "grad_norm": 0.09421565559486329,
+      "learning_rate": 1.465174546573774e-05,
+      "loss": 0.4596,
+      "step": 1409
+    },
+    {
+      "epoch": 0.823297569520473,
+      "grad_norm": 0.0849448939916547,
+      "learning_rate": 1.4642716767182858e-05,
+      "loss": 0.39,
+      "step": 1410
+    },
+    {
+      "epoch": 0.8238814685059485,
+      "grad_norm": 0.09012394598909018,
+      "learning_rate": 1.4633683241503464e-05,
+      "loss": 0.4224,
+      "step": 1411
+    },
+    {
+      "epoch": 0.824465367491424,
+      "grad_norm": 0.08785820749554267,
+      "learning_rate": 1.4624644898091898e-05,
+      "loss": 0.3771,
+      "step": 1412
+    },
+    {
+      "epoch": 0.8250492664768995,
+      "grad_norm": 0.08956544254056811,
+      "learning_rate": 1.4615601746345501e-05,
+      "loss": 0.4136,
+      "step": 1413
+    },
+    {
+      "epoch": 0.825633165462375,
+      "grad_norm": 0.08190894183584209,
+      "learning_rate": 1.4606553795666616e-05,
+      "loss": 0.3807,
+      "step": 1414
+    },
+    {
+      "epoch": 0.8262170644478505,
+      "grad_norm": 0.08886283594452578,
+      "learning_rate": 1.4597501055462577e-05,
+      "loss": 0.4274,
+      "step": 1415
+    },
+    {
+      "epoch": 0.826800963433326,
+      "grad_norm": 0.0887535969639663,
+      "learning_rate": 1.45884435351457e-05,
+      "loss": 0.4031,
+      "step": 1416
+    },
+    {
+      "epoch": 0.8273848624188015,
+      "grad_norm": 0.09050286027978477,
+      "learning_rate": 1.4579381244133265e-05,
+      "loss": 0.4462,
+      "step": 1417
+    },
+    {
+      "epoch": 0.8279687614042771,
+      "grad_norm": 0.09150790697022396,
+      "learning_rate": 1.457031419184752e-05,
+      "loss": 0.4598,
+      "step": 1418
+    },
+    {
+      "epoch": 0.8285526603897526,
+      "grad_norm": 0.07928772942246552,
+      "learning_rate": 1.4561242387715652e-05,
+      "loss": 0.39,
+      "step": 1419
+    },
+    {
+      "epoch": 0.8291365593752281,
+      "grad_norm": 0.0829510797861744,
+      "learning_rate": 1.45521658411698e-05,
+      "loss": 0.4351,
+      "step": 1420
+    },
+    {
+      "epoch": 0.8297204583607036,
+      "grad_norm": 0.08517974667840758,
+      "learning_rate": 1.4543084561647028e-05,
+      "loss": 0.4424,
+      "step": 1421
+    },
+    {
+      "epoch": 0.8303043573461791,
+      "grad_norm": 0.08719246340462658,
+      "learning_rate": 1.4533998558589319e-05,
+      "loss": 0.3978,
+      "step": 1422
+    },
+    {
+      "epoch": 0.8308882563316546,
+      "grad_norm": 0.08819106522723598,
+      "learning_rate": 1.4524907841443576e-05,
+      "loss": 0.4021,
+      "step": 1423
+    },
+    {
+      "epoch": 0.8314721553171301,
+      "grad_norm": 0.08623976961294613,
+      "learning_rate": 1.4515812419661595e-05,
+      "loss": 0.3828,
+      "step": 1424
+    },
+    {
+      "epoch": 0.8320560543026057,
+      "grad_norm": 0.08249079811337974,
+      "learning_rate": 1.4506712302700064e-05,
+      "loss": 0.4131,
+      "step": 1425
+    },
+    {
+      "epoch": 0.8326399532880812,
+      "grad_norm": 0.09218042415956439,
+      "learning_rate": 1.4497607500020556e-05,
+      "loss": 0.3972,
+      "step": 1426
+    },
+    {
+      "epoch": 0.8332238522735567,
+      "grad_norm": 0.09246961518159985,
+      "learning_rate": 1.4488498021089514e-05,
+      "loss": 0.4071,
+      "step": 1427
+    },
+    {
+      "epoch": 0.8338077512590322,
+      "grad_norm": 0.08087224512157197,
+      "learning_rate": 1.4479383875378245e-05,
+      "loss": 0.4252,
+      "step": 1428
+    },
+    {
+      "epoch": 0.8343916502445077,
+      "grad_norm": 0.09011730860037306,
+      "learning_rate": 1.4470265072362906e-05,
+      "loss": 0.4116,
+      "step": 1429
+    },
+    {
+      "epoch": 0.8349755492299832,
+      "grad_norm": 0.08368200881424352,
+      "learning_rate": 1.4461141621524498e-05,
+      "loss": 0.4487,
+      "step": 1430
+    },
+    {
+      "epoch": 0.8355594482154587,
+      "grad_norm": 0.08880649072194767,
+      "learning_rate": 1.4452013532348849e-05,
+      "loss": 0.4756,
+      "step": 1431
+    },
+    {
+      "epoch": 0.8361433472009342,
+      "grad_norm": 0.08279845925885351,
+      "learning_rate": 1.444288081432662e-05,
+      "loss": 0.3979,
+      "step": 1432
+    },
+    {
+      "epoch": 0.8367272461864097,
+      "grad_norm": 0.0877586077838818,
+      "learning_rate": 1.443374347695328e-05,
+      "loss": 0.455,
+      "step": 1433
+    },
+    {
+      "epoch": 0.8373111451718853,
+      "grad_norm": 0.08632908506942297,
+      "learning_rate": 1.442460152972909e-05,
+      "loss": 0.4034,
+      "step": 1434
+    },
+    {
+      "epoch": 0.8378950441573608,
+      "grad_norm": 0.08489354700796757,
+      "learning_rate": 1.4415454982159121e-05,
+      "loss": 0.4065,
+      "step": 1435
+    },
+    {
+      "epoch": 0.8384789431428363,
+      "grad_norm": 0.08447590199710801,
+      "learning_rate": 1.4406303843753215e-05,
+      "loss": 0.4397,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8390628421283118,
+      "grad_norm": 0.08687818790002634,
+      "learning_rate": 1.4397148124025997e-05,
+      "loss": 0.4568,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8396467411137873,
+      "grad_norm": 0.08611709585511437,
+      "learning_rate": 1.4387987832496848e-05,
+      "loss": 0.4179,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8402306400992628,
+      "grad_norm": 0.08196123165176804,
+      "learning_rate": 1.4378822978689901e-05,
+      "loss": 0.3857,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8408145390847384,
+      "grad_norm": 0.08480688455869545,
+      "learning_rate": 1.436965357213404e-05,
+      "loss": 0.4016,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8413984380702139,
+      "grad_norm": 0.0880525371274207,
+      "learning_rate": 1.4360479622362877e-05,
+      "loss": 0.4197,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8419823370556894,
+      "grad_norm": 0.08890342362296252,
+      "learning_rate": 1.4351301138914749e-05,
+      "loss": 0.408,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8425662360411649,
+      "grad_norm": 0.08310223205305926,
+      "learning_rate": 1.4342118131332704e-05,
+      "loss": 0.3804,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8431501350266404,
+      "grad_norm": 0.08205788472812424,
+      "learning_rate": 1.4332930609164503e-05,
+      "loss": 0.3837,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8437340340121159,
+      "grad_norm": 0.08169737833460683,
+      "learning_rate": 1.4323738581962593e-05,
+      "loss": 0.4138,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8443179329975914,
+      "grad_norm": 0.08443820458366996,
+      "learning_rate": 1.4314542059284102e-05,
+      "loss": 0.4217,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8449018319830669,
+      "grad_norm": 0.08490122335349844,
+      "learning_rate": 1.4305341050690845e-05,
+      "loss": 0.4569,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8454857309685424,
+      "grad_norm": 0.09925405535460585,
+      "learning_rate": 1.429613556574928e-05,
+      "loss": 0.4655,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8460696299540179,
+      "grad_norm": 0.09070106218221316,
+      "learning_rate": 1.4286925614030542e-05,
+      "loss": 0.4525,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8466535289394934,
+      "grad_norm": 0.08917123357248109,
+      "learning_rate": 1.4277711205110398e-05,
+      "loss": 0.426,
+      "step": 1450
+    },
+    {
+      "epoch": 0.847237427924969,
+      "grad_norm": 0.08052016424561877,
+      "learning_rate": 1.4268492348569252e-05,
+      "loss": 0.4131,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8478213269104445,
+      "grad_norm": 0.08993504153740149,
+      "learning_rate": 1.425926905399213e-05,
+      "loss": 0.4232,
+      "step": 1452
+    },
+    {
+      "epoch": 0.84840522589592,
+      "grad_norm": 0.08893917613221719,
+      "learning_rate": 1.4250041330968674e-05,
+      "loss": 0.4231,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8489891248813956,
+      "grad_norm": 0.09105914443704308,
+      "learning_rate": 1.424080918909313e-05,
+      "loss": 0.4324,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8495730238668711,
+      "grad_norm": 0.08758108698081693,
+      "learning_rate": 1.4231572637964338e-05,
+      "loss": 0.4346,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8501569228523466,
+      "grad_norm": 0.09199069747343194,
+      "learning_rate": 1.4222331687185723e-05,
+      "loss": 0.4132,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8507408218378221,
+      "grad_norm": 0.09220747835692521,
+      "learning_rate": 1.421308634636529e-05,
+      "loss": 0.3879,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8513247208232976,
+      "grad_norm": 0.09490216000654761,
+      "learning_rate": 1.4203836625115595e-05,
+      "loss": 0.4334,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8519086198087731,
+      "grad_norm": 0.09060617484948602,
+      "learning_rate": 1.419458253305376e-05,
+      "loss": 0.3958,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8524925187942486,
+      "grad_norm": 0.09119765081176508,
+      "learning_rate": 1.4185324079801447e-05,
+      "loss": 0.437,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8530764177797241,
+      "grad_norm": 0.08414729670525932,
+      "learning_rate": 1.4176061274984858e-05,
+      "loss": 0.429,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8536603167651996,
+      "grad_norm": 0.09643363365363182,
+      "learning_rate": 1.4166794128234705e-05,
+      "loss": 0.418,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8542442157506751,
+      "grad_norm": 0.09037229126018019,
+      "learning_rate": 1.415752264918623e-05,
+      "loss": 0.4132,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8548281147361506,
+      "grad_norm": 0.08830812426428669,
+      "learning_rate": 1.4148246847479173e-05,
+      "loss": 0.4206,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8554120137216261,
+      "grad_norm": 0.09227898660241378,
+      "learning_rate": 1.4138966732757766e-05,
+      "loss": 0.4218,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8559959127071016,
+      "grad_norm": 0.08741610015244673,
+      "learning_rate": 1.4129682314670731e-05,
+      "loss": 0.4275,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8565798116925771,
+      "grad_norm": 0.083389536659968,
+      "learning_rate": 1.412039360287126e-05,
+      "loss": 0.4151,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8571637106780527,
+      "grad_norm": 0.08808993226010912,
+      "learning_rate": 1.411110060701701e-05,
+      "loss": 0.4331,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8577476096635283,
+      "grad_norm": 0.08824338078841779,
+      "learning_rate": 1.4101803336770092e-05,
+      "loss": 0.3962,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8583315086490038,
+      "grad_norm": 0.08746499990706068,
+      "learning_rate": 1.4092501801797063e-05,
+      "loss": 0.3971,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8589154076344793,
+      "grad_norm": 0.08557901156829577,
+      "learning_rate": 1.4083196011768913e-05,
+      "loss": 0.4218,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8594993066199548,
+      "grad_norm": 0.08665276555062452,
+      "learning_rate": 1.4073885976361056e-05,
+      "loss": 0.4315,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8600832056054303,
+      "grad_norm": 0.08548711704460585,
+      "learning_rate": 1.4064571705253323e-05,
+      "loss": 0.3932,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8606671045909058,
+      "grad_norm": 0.0815281746925988,
+      "learning_rate": 1.405525320812994e-05,
+      "loss": 0.4102,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8612510035763813,
+      "grad_norm": 0.07981574079771336,
+      "learning_rate": 1.4045930494679538e-05,
+      "loss": 0.376,
+      "step": 1475
+    },
+    {
+      "epoch": 0.8618349025618568,
+      "grad_norm": 0.08929058848675574,
+      "learning_rate": 1.4036603574595122e-05,
+      "loss": 0.4382,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8624188015473323,
+      "grad_norm": 0.08075200387106833,
+      "learning_rate": 1.4027272457574082e-05,
+      "loss": 0.3961,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8630027005328078,
+      "grad_norm": 0.08388680146821,
+      "learning_rate": 1.4017937153318157e-05,
+      "loss": 0.4431,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8635865995182833,
+      "grad_norm": 0.0815847092455702,
+      "learning_rate": 1.4008597671533455e-05,
+      "loss": 0.4032,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8641704985037588,
+      "grad_norm": 0.08469195338809236,
+      "learning_rate": 1.3999254021930416e-05,
+      "loss": 0.3943,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8647543974892343,
+      "grad_norm": 0.08923456845771494,
+      "learning_rate": 1.3989906214223817e-05,
+      "loss": 0.3875,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8653382964747098,
+      "grad_norm": 0.09043188169628925,
+      "learning_rate": 1.3980554258132761e-05,
+      "loss": 0.4386,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8659221954601853,
+      "grad_norm": 0.08627101305504592,
+      "learning_rate": 1.3971198163380659e-05,
+      "loss": 0.4944,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8665060944456608,
+      "grad_norm": 0.08294431639512238,
+      "learning_rate": 1.3961837939695231e-05,
+      "loss": 0.441,
+      "step": 1484
+    },
+    {
+      "epoch": 0.8670899934311365,
+      "grad_norm": 0.08521248916502198,
+      "learning_rate": 1.3952473596808485e-05,
+      "loss": 0.4099,
+      "step": 1485
+    },
+    {
+      "epoch": 0.867673892416612,
+      "grad_norm": 0.08511402710583015,
+      "learning_rate": 1.3943105144456715e-05,
+      "loss": 0.4031,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8682577914020875,
+      "grad_norm": 0.08670955084059544,
+      "learning_rate": 1.3933732592380485e-05,
+      "loss": 0.4232,
+      "step": 1487
+    },
+    {
+      "epoch": 0.868841690387563,
+      "grad_norm": 0.0895822904133851,
+      "learning_rate": 1.3924355950324623e-05,
+      "loss": 0.4184,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8694255893730385,
+      "grad_norm": 0.08517248533920482,
+      "learning_rate": 1.391497522803821e-05,
+      "loss": 0.4114,
+      "step": 1489
+    },
+    {
+      "epoch": 0.870009488358514,
+      "grad_norm": 0.09454216572523248,
+      "learning_rate": 1.390559043527457e-05,
+      "loss": 0.4202,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8705933873439895,
+      "grad_norm": 0.09356661430585539,
+      "learning_rate": 1.3896201581791253e-05,
+      "loss": 0.4171,
+      "step": 1491
+    },
+    {
+      "epoch": 0.871177286329465,
+      "grad_norm": 0.08726063151100519,
+      "learning_rate": 1.388680867735004e-05,
+      "loss": 0.4458,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8717611853149405,
+      "grad_norm": 0.08278313273373822,
+      "learning_rate": 1.3877411731716917e-05,
+      "loss": 0.416,
+      "step": 1493
+    },
+    {
+      "epoch": 0.872345084300416,
+      "grad_norm": 0.09217212583675392,
+      "learning_rate": 1.3868010754662077e-05,
+      "loss": 0.4329,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8729289832858915,
+      "grad_norm": 0.08182784537434606,
+      "learning_rate": 1.3858605755959902e-05,
+      "loss": 0.4337,
+      "step": 1495
+    },
+    {
+      "epoch": 0.873512882271367,
+      "grad_norm": 0.08289507569497061,
+      "learning_rate": 1.3849196745388953e-05,
+      "loss": 0.3814,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8740967812568425,
+      "grad_norm": 0.08405861282088051,
+      "learning_rate": 1.3839783732731966e-05,
+      "loss": 0.4262,
+      "step": 1497
+    },
+    {
+      "epoch": 0.874680680242318,
+      "grad_norm": 0.07655458074827988,
+      "learning_rate": 1.3830366727775835e-05,
+      "loss": 0.4033,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8752645792277935,
+      "grad_norm": 0.08949123965309178,
+      "learning_rate": 1.3820945740311609e-05,
+      "loss": 0.4093,
+      "step": 1499
+    },
+    {
+      "epoch": 0.875848478213269,
+      "grad_norm": 0.08648880013942892,
+      "learning_rate": 1.3811520780134471e-05,
+      "loss": 0.407,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8764323771987447,
+      "grad_norm": 0.08926114344743293,
+      "learning_rate": 1.3802091857043745e-05,
+      "loss": 0.4595,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8770162761842202,
+      "grad_norm": 0.0759036319896071,
+      "learning_rate": 1.3792658980842861e-05,
+      "loss": 0.3859,
+      "step": 1502
+    },
+    {
+      "epoch": 0.8776001751696957,
+      "grad_norm": 0.0952497963215699,
+      "learning_rate": 1.3783222161339375e-05,
+      "loss": 0.468,
+      "step": 1503
+    },
+    {
+      "epoch": 0.8781840741551712,
+      "grad_norm": 0.08926257916768073,
+      "learning_rate": 1.3773781408344931e-05,
+      "loss": 0.4091,
+      "step": 1504
+    },
+    {
+      "epoch": 0.8787679731406467,
+      "grad_norm": 0.09001023046815096,
+      "learning_rate": 1.3764336731675266e-05,
+      "loss": 0.4474,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8793518721261222,
+      "grad_norm": 0.08811338682640284,
+      "learning_rate": 1.3754888141150197e-05,
+      "loss": 0.439,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8799357711115977,
+      "grad_norm": 0.08447557451148596,
+      "learning_rate": 1.3745435646593613e-05,
+      "loss": 0.4004,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8805196700970732,
+      "grad_norm": 0.08120582633607253,
+      "learning_rate": 1.373597925783346e-05,
+      "loss": 0.3791,
+      "step": 1508
+    },
+    {
+      "epoch": 0.8811035690825487,
+      "grad_norm": 0.08469559266495609,
+      "learning_rate": 1.3726518984701731e-05,
+      "loss": 0.4631,
+      "step": 1509
+    },
+    {
+      "epoch": 0.8816874680680242,
+      "grad_norm": 0.08701911018860554,
+      "learning_rate": 1.3717054837034459e-05,
+      "loss": 0.432,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8822713670534997,
+      "grad_norm": 0.08281436736360163,
+      "learning_rate": 1.3707586824671703e-05,
+      "loss": 0.4161,
+      "step": 1511
+    },
+    {
+      "epoch": 0.8828552660389752,
+      "grad_norm": 0.07929546477360602,
+      "learning_rate": 1.369811495745755e-05,
+      "loss": 0.3818,
+      "step": 1512
+    },
+    {
+      "epoch": 0.8834391650244507,
+      "grad_norm": 0.07834217638592637,
+      "learning_rate": 1.3688639245240078e-05,
+      "loss": 0.4245,
+      "step": 1513
+    },
+    {
+      "epoch": 0.8840230640099263,
+      "grad_norm": 0.09192308236398318,
+      "learning_rate": 1.3679159697871383e-05,
+      "loss": 0.5281,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8846069629954018,
+      "grad_norm": 0.08944985093404384,
+      "learning_rate": 1.3669676325207531e-05,
+      "loss": 0.4308,
+      "step": 1515
+    },
+    {
+      "epoch": 0.8851908619808773,
+      "grad_norm": 0.08350401810312967,
+      "learning_rate": 1.3660189137108578e-05,
+      "loss": 0.3969,
+      "step": 1516
+    },
+    {
+      "epoch": 0.8857747609663528,
+      "grad_norm": 0.08424835102439512,
+      "learning_rate": 1.3650698143438534e-05,
+      "loss": 0.3958,
+      "step": 1517
+    },
+    {
+      "epoch": 0.8863586599518284,
+      "grad_norm": 0.09153618204871851,
+      "learning_rate": 1.3641203354065378e-05,
+      "loss": 0.4023,
+      "step": 1518
+    },
+    {
+      "epoch": 0.8869425589373039,
+      "grad_norm": 0.08774708316992592,
+      "learning_rate": 1.3631704778861028e-05,
+      "loss": 0.5136,
+      "step": 1519
+    },
+    {
+      "epoch": 0.8875264579227794,
+      "grad_norm": 0.08110344632980092,
+      "learning_rate": 1.3622202427701344e-05,
+      "loss": 0.3914,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8881103569082549,
+      "grad_norm": 0.08359675817510623,
+      "learning_rate": 1.3612696310466103e-05,
+      "loss": 0.409,
+      "step": 1521
+    },
+    {
+      "epoch": 0.8886942558937304,
+      "grad_norm": 0.08475069099601605,
+      "learning_rate": 1.360318643703901e-05,
+      "loss": 0.3945,
+      "step": 1522
+    },
+    {
+      "epoch": 0.8892781548792059,
+      "grad_norm": 0.08841184016255975,
+      "learning_rate": 1.3593672817307661e-05,
+      "loss": 0.3835,
+      "step": 1523
+    },
+    {
+      "epoch": 0.8898620538646814,
+      "grad_norm": 0.08840641469326788,
+      "learning_rate": 1.3584155461163562e-05,
+      "loss": 0.4879,
+      "step": 1524
+    },
+    {
+      "epoch": 0.8904459528501569,
+      "grad_norm": 0.07936790789581172,
+      "learning_rate": 1.3574634378502092e-05,
+      "loss": 0.4232,
+      "step": 1525
+    },
+    {
+      "epoch": 0.8910298518356324,
+      "grad_norm": 0.08103478115238165,
+      "learning_rate": 1.3565109579222511e-05,
+      "loss": 0.3956,
+      "step": 1526
+    },
+    {
+      "epoch": 0.8916137508211079,
+      "grad_norm": 0.07819996486132182,
+      "learning_rate": 1.3555581073227942e-05,
+      "loss": 0.4105,
+      "step": 1527
+    },
+    {
+      "epoch": 0.8921976498065834,
+      "grad_norm": 0.07893687290066324,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.3905,
+      "step": 1528
+    },
+    {
+      "epoch": 0.892781548792059,
+      "grad_norm": 0.08322470136831914,
+      "learning_rate": 1.353651298072558e-05,
+      "loss": 0.3768,
+      "step": 1529
+    },
+    {
+      "epoch": 0.8933654477775345,
+      "grad_norm": 0.09199331086271972,
+      "learning_rate": 1.3526973414043263e-05,
+      "loss": 0.4291,
+      "step": 1530
+    },
+    {
+      "epoch": 0.89394934676301,
+      "grad_norm": 0.09193994118306378,
+      "learning_rate": 1.3517430180296886e-05,
+      "loss": 0.4017,
+      "step": 1531
+    },
+    {
+      "epoch": 0.8945332457484855,
+      "grad_norm": 0.08933314799027446,
+      "learning_rate": 1.3507883289408732e-05,
+      "loss": 0.4191,
+      "step": 1532
+    },
+    {
+      "epoch": 0.895117144733961,
+      "grad_norm": 0.08047011259479905,
+      "learning_rate": 1.3498332751304895e-05,
+      "loss": 0.4007,
+      "step": 1533
+    },
+    {
+      "epoch": 0.8957010437194365,
+      "grad_norm": 0.08533785875999435,
+      "learning_rate": 1.3488778575915258e-05,
+      "loss": 0.4089,
+      "step": 1534
+    },
+    {
+      "epoch": 0.8962849427049121,
+      "grad_norm": 0.0815460916512401,
+      "learning_rate": 1.3479220773173485e-05,
+      "loss": 0.4014,
+      "step": 1535
+    },
+    {
+      "epoch": 0.8968688416903876,
+      "grad_norm": 0.07961474765712534,
+      "learning_rate": 1.3469659353017019e-05,
+      "loss": 0.3812,
+      "step": 1536
+    },
+    {
+      "epoch": 0.8974527406758631,
+      "grad_norm": 0.08531916009615587,
+      "learning_rate": 1.346009432538705e-05,
+      "loss": 0.454,
+      "step": 1537
+    },
+    {
+      "epoch": 0.8980366396613386,
+      "grad_norm": 0.07582665056212996,
+      "learning_rate": 1.345052570022853e-05,
+      "loss": 0.3419,
+      "step": 1538
+    },
+    {
+      "epoch": 0.8986205386468141,
+      "grad_norm": 0.07871651702462862,
+      "learning_rate": 1.3440953487490145e-05,
+      "loss": 0.3787,
+      "step": 1539
+    },
+    {
+      "epoch": 0.8992044376322896,
+      "grad_norm": 0.07853115465064762,
+      "learning_rate": 1.343137769712432e-05,
+      "loss": 0.4027,
+      "step": 1540
+    },
+    {
+      "epoch": 0.8997883366177651,
+      "grad_norm": 0.0807644920021475,
+      "learning_rate": 1.342179833908719e-05,
+      "loss": 0.3537,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9003722356032406,
+      "grad_norm": 0.07819285123105765,
+      "learning_rate": 1.3412215423338601e-05,
+      "loss": 0.3416,
+      "step": 1542
+    },
+    {
+      "epoch": 0.9009561345887162,
+      "grad_norm": 0.0863984388822104,
+      "learning_rate": 1.3402628959842106e-05,
+      "loss": 0.4348,
+      "step": 1543
+    },
+    {
+      "epoch": 0.9015400335741917,
+      "grad_norm": 0.08272909133506265,
+      "learning_rate": 1.3393038958564934e-05,
+      "loss": 0.4041,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9021239325596672,
+      "grad_norm": 0.08980866397155235,
+      "learning_rate": 1.3383445429478008e-05,
+      "loss": 0.4804,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9027078315451427,
+      "grad_norm": 0.08283486773411809,
+      "learning_rate": 1.33738483825559e-05,
+      "loss": 0.389,
+      "step": 1546
+    },
+    {
+      "epoch": 0.9032917305306182,
+      "grad_norm": 0.09652079756286676,
+      "learning_rate": 1.3364247827776854e-05,
+      "loss": 0.5007,
+      "step": 1547
+    },
+    {
+      "epoch": 0.9038756295160937,
+      "grad_norm": 0.08116085565801566,
+      "learning_rate": 1.3354643775122762e-05,
+      "loss": 0.4029,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9044595285015692,
+      "grad_norm": 0.0887495149192001,
+      "learning_rate": 1.3345036234579138e-05,
+      "loss": 0.4256,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9050434274870447,
+      "grad_norm": 0.08215451889801811,
+      "learning_rate": 1.333542521613514e-05,
+      "loss": 0.4029,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9056273264725202,
+      "grad_norm": 0.08632426712137557,
+      "learning_rate": 1.332581072978353e-05,
+      "loss": 0.3959,
+      "step": 1551
+    },
+    {
+      "epoch": 0.9062112254579958,
+      "grad_norm": 0.07986634490604924,
+      "learning_rate": 1.331619278552068e-05,
+      "loss": 0.3867,
+      "step": 1552
+    },
+    {
+      "epoch": 0.9067951244434713,
+      "grad_norm": 0.09088478121590726,
+      "learning_rate": 1.3306571393346557e-05,
+      "loss": 0.4512,
+      "step": 1553
+    },
+    {
+      "epoch": 0.9073790234289468,
+      "grad_norm": 0.08163990923626306,
+      "learning_rate": 1.3296946563264715e-05,
+      "loss": 0.3932,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9079629224144223,
+      "grad_norm": 0.09711481211168883,
+      "learning_rate": 1.3287318305282277e-05,
+      "loss": 0.4387,
+      "step": 1555
+    },
+    {
+      "epoch": 0.9085468213998978,
+      "grad_norm": 0.08560477508319146,
+      "learning_rate": 1.3277686629409936e-05,
+      "loss": 0.4126,
+      "step": 1556
+    },
+    {
+      "epoch": 0.9091307203853733,
+      "grad_norm": 0.08751644897306579,
+      "learning_rate": 1.3268051545661937e-05,
+      "loss": 0.4259,
+      "step": 1557
+    },
+    {
+      "epoch": 0.9097146193708489,
+      "grad_norm": 0.0938614729135381,
+      "learning_rate": 1.3258413064056066e-05,
+      "loss": 0.4586,
+      "step": 1558
+    },
+    {
+      "epoch": 0.9102985183563244,
+      "grad_norm": 0.08241429220909723,
+      "learning_rate": 1.3248771194613641e-05,
+      "loss": 0.3816,
+      "step": 1559
+    },
+    {
+      "epoch": 0.9108824173417999,
+      "grad_norm": 0.08087439260454915,
+      "learning_rate": 1.3239125947359506e-05,
+      "loss": 0.4409,
+      "step": 1560
+    },
+    {
+      "epoch": 0.9114663163272754,
+      "grad_norm": 0.08391079904876426,
+      "learning_rate": 1.3229477332322016e-05,
+      "loss": 0.4165,
+      "step": 1561
+    },
+    {
+      "epoch": 0.9120502153127509,
+      "grad_norm": 0.0793722870430516,
+      "learning_rate": 1.3219825359533025e-05,
+      "loss": 0.4043,
+      "step": 1562
+    },
+    {
+      "epoch": 0.9126341142982264,
+      "grad_norm": 0.08792912559853962,
+      "learning_rate": 1.3210170039027886e-05,
+      "loss": 0.4148,
+      "step": 1563
+    },
+    {
+      "epoch": 0.9132180132837019,
+      "grad_norm": 0.0794531584292294,
+      "learning_rate": 1.320051138084542e-05,
+      "loss": 0.4586,
+      "step": 1564
+    },
+    {
+      "epoch": 0.9138019122691774,
+      "grad_norm": 0.08296247656089574,
+      "learning_rate": 1.3190849395027926e-05,
+      "loss": 0.433,
+      "step": 1565
+    },
+    {
+      "epoch": 0.9143858112546529,
+      "grad_norm": 0.086976433430398,
+      "learning_rate": 1.3181184091621165e-05,
+      "loss": 0.4192,
+      "step": 1566
+    },
+    {
+      "epoch": 0.9149697102401284,
+      "grad_norm": 0.09048828328427744,
+      "learning_rate": 1.3171515480674342e-05,
+      "loss": 0.446,
+      "step": 1567
+    },
+    {
+      "epoch": 0.915553609225604,
+      "grad_norm": 0.0835939990385343,
+      "learning_rate": 1.3161843572240107e-05,
+      "loss": 0.3951,
+      "step": 1568
+    },
+    {
+      "epoch": 0.9161375082110795,
+      "grad_norm": 0.08071437371744158,
+      "learning_rate": 1.3152168376374528e-05,
+      "loss": 0.3933,
+      "step": 1569
+    },
+    {
+      "epoch": 0.916721407196555,
+      "grad_norm": 0.09563752156903395,
+      "learning_rate": 1.3142489903137101e-05,
+      "loss": 0.4471,
+      "step": 1570
+    },
+    {
+      "epoch": 0.9173053061820305,
+      "grad_norm": 0.09125571164511431,
+      "learning_rate": 1.313280816259073e-05,
+      "loss": 0.4387,
+      "step": 1571
+    },
+    {
+      "epoch": 0.917889205167506,
+      "grad_norm": 0.07791565092965531,
+      "learning_rate": 1.3123123164801706e-05,
+      "loss": 0.3884,
+      "step": 1572
+    },
+    {
+      "epoch": 0.9184731041529816,
+      "grad_norm": 0.08541601702544663,
+      "learning_rate": 1.3113434919839715e-05,
+      "loss": 0.4281,
+      "step": 1573
+    },
+    {
+      "epoch": 0.9190570031384571,
+      "grad_norm": 0.08151541779773863,
+      "learning_rate": 1.310374343777782e-05,
+      "loss": 0.4306,
+      "step": 1574
+    },
+    {
+      "epoch": 0.9196409021239326,
+      "grad_norm": 0.08681479205617093,
+      "learning_rate": 1.3094048728692443e-05,
+      "loss": 0.4108,
+      "step": 1575
+    },
+    {
+      "epoch": 0.9202248011094081,
+      "grad_norm": 0.08238242339029798,
+      "learning_rate": 1.3084350802663365e-05,
+      "loss": 0.3972,
+      "step": 1576
+    },
+    {
+      "epoch": 0.9208087000948836,
+      "grad_norm": 0.08493934891225856,
+      "learning_rate": 1.3074649669773716e-05,
+      "loss": 0.4233,
+      "step": 1577
+    },
+    {
+      "epoch": 0.9213925990803591,
+      "grad_norm": 0.08928313772404654,
+      "learning_rate": 1.306494534010995e-05,
+      "loss": 0.4284,
+      "step": 1578
+    },
+    {
+      "epoch": 0.9219764980658346,
+      "grad_norm": 0.0839828111935972,
+      "learning_rate": 1.3055237823761855e-05,
+      "loss": 0.4166,
+      "step": 1579
+    },
+    {
+      "epoch": 0.9225603970513101,
+      "grad_norm": 0.08204918329353646,
+      "learning_rate": 1.3045527130822524e-05,
+      "loss": 0.4141,
+      "step": 1580
+    },
+    {
+      "epoch": 0.9231442960367856,
+      "grad_norm": 0.08480225008701794,
+      "learning_rate": 1.303581327138836e-05,
+      "loss": 0.4255,
+      "step": 1581
+    },
+    {
+      "epoch": 0.9237281950222611,
+      "grad_norm": 0.08631616871313282,
+      "learning_rate": 1.3026096255559055e-05,
+      "loss": 0.492,
+      "step": 1582
+    },
+    {
+      "epoch": 0.9243120940077366,
+      "grad_norm": 0.08005259512940054,
+      "learning_rate": 1.3016376093437577e-05,
+      "loss": 0.3754,
+      "step": 1583
+    },
+    {
+      "epoch": 0.9248959929932121,
+      "grad_norm": 0.08622140974535683,
+      "learning_rate": 1.3006652795130179e-05,
+      "loss": 0.4319,
+      "step": 1584
+    },
+    {
+      "epoch": 0.9254798919786877,
+      "grad_norm": 0.08381345173263231,
+      "learning_rate": 1.299692637074636e-05,
+      "loss": 0.422,
+      "step": 1585
+    },
+    {
+      "epoch": 0.9260637909641632,
+      "grad_norm": 0.079956727059352,
+      "learning_rate": 1.2987196830398884e-05,
+      "loss": 0.3532,
+      "step": 1586
+    },
+    {
+      "epoch": 0.9266476899496388,
+      "grad_norm": 0.08390878106630571,
+      "learning_rate": 1.297746418420374e-05,
+      "loss": 0.3994,
+      "step": 1587
+    },
+    {
+      "epoch": 0.9272315889351143,
+      "grad_norm": 0.08038792208408477,
+      "learning_rate": 1.2967728442280154e-05,
+      "loss": 0.3747,
+      "step": 1588
+    },
+    {
+      "epoch": 0.9278154879205898,
+      "grad_norm": 0.07556167150978939,
+      "learning_rate": 1.2957989614750569e-05,
+      "loss": 0.3768,
+      "step": 1589
+    },
+    {
+      "epoch": 0.9283993869060653,
+      "grad_norm": 0.09071362594877536,
+      "learning_rate": 1.2948247711740638e-05,
+      "loss": 0.4712,
+      "step": 1590
+    },
+    {
+      "epoch": 0.9289832858915408,
+      "grad_norm": 0.09010808236245998,
+      "learning_rate": 1.2938502743379212e-05,
+      "loss": 0.4443,
+      "step": 1591
+    },
+    {
+      "epoch": 0.9295671848770163,
+      "grad_norm": 0.08310538968516451,
+      "learning_rate": 1.2928754719798324e-05,
+      "loss": 0.4319,
+      "step": 1592
+    },
+    {
+      "epoch": 0.9301510838624918,
+      "grad_norm": 0.08918056584241667,
+      "learning_rate": 1.291900365113319e-05,
+      "loss": 0.4401,
+      "step": 1593
+    },
+    {
+      "epoch": 0.9307349828479673,
+      "grad_norm": 0.0781249207560133,
+      "learning_rate": 1.2909249547522184e-05,
+      "loss": 0.4027,
+      "step": 1594
+    },
+    {
+      "epoch": 0.9313188818334428,
+      "grad_norm": 0.09134810961711312,
+      "learning_rate": 1.2899492419106848e-05,
+      "loss": 0.4356,
+      "step": 1595
+    },
+    {
+      "epoch": 0.9319027808189183,
+      "grad_norm": 0.09022110746633594,
+      "learning_rate": 1.2889732276031856e-05,
+      "loss": 0.4179,
+      "step": 1596
+    },
+    {
+      "epoch": 0.9324866798043938,
+      "grad_norm": 0.08310132672769759,
+      "learning_rate": 1.2879969128445025e-05,
+      "loss": 0.3834,
+      "step": 1597
+    },
+    {
+      "epoch": 0.9330705787898693,
+      "grad_norm": 0.08140967533600656,
+      "learning_rate": 1.2870202986497291e-05,
+      "loss": 0.3872,
+      "step": 1598
+    },
+    {
+      "epoch": 0.9336544777753448,
+      "grad_norm": 0.08078750023737453,
+      "learning_rate": 1.2860433860342705e-05,
+      "loss": 0.4145,
+      "step": 1599
+    },
+    {
+      "epoch": 0.9342383767608203,
+      "grad_norm": 0.09092545330432761,
+      "learning_rate": 1.2850661760138423e-05,
+      "loss": 0.4187,
+      "step": 1600
+    },
+    {
+      "epoch": 0.9348222757462958,
+      "grad_norm": 0.09022016159388037,
+      "learning_rate": 1.284088669604469e-05,
+      "loss": 0.416,
+      "step": 1601
+    },
+    {
+      "epoch": 0.9354061747317715,
+      "grad_norm": 0.08804888730918083,
+      "learning_rate": 1.283110867822483e-05,
+      "loss": 0.3905,
+      "step": 1602
+    },
+    {
+      "epoch": 0.935990073717247,
+      "grad_norm": 0.08749548503715492,
+      "learning_rate": 1.2821327716845246e-05,
+      "loss": 0.4033,
+      "step": 1603
+    },
+    {
+      "epoch": 0.9365739727027225,
+      "grad_norm": 0.09079736430173337,
+      "learning_rate": 1.2811543822075396e-05,
+      "loss": 0.4647,
+      "step": 1604
+    },
+    {
+      "epoch": 0.937157871688198,
+      "grad_norm": 0.08531614284801951,
+      "learning_rate": 1.2801757004087792e-05,
+      "loss": 0.3897,
+      "step": 1605
+    },
+    {
+      "epoch": 0.9377417706736735,
+      "grad_norm": 0.08780101119684026,
+      "learning_rate": 1.2791967273057978e-05,
+      "loss": 0.456,
+      "step": 1606
+    },
+    {
+      "epoch": 0.938325669659149,
+      "grad_norm": 0.07681824551242163,
+      "learning_rate": 1.2782174639164528e-05,
+      "loss": 0.3875,
+      "step": 1607
+    },
+    {
+      "epoch": 0.9389095686446245,
+      "grad_norm": 0.07712000972553822,
+      "learning_rate": 1.2772379112589043e-05,
+      "loss": 0.3864,
+      "step": 1608
+    },
+    {
+      "epoch": 0.9394934676301,
+      "grad_norm": 0.07966119679469448,
+      "learning_rate": 1.2762580703516127e-05,
+      "loss": 0.3911,
+      "step": 1609
+    },
+    {
+      "epoch": 0.9400773666155755,
+      "grad_norm": 0.08902134842693428,
+      "learning_rate": 1.2752779422133377e-05,
+      "loss": 0.3835,
+      "step": 1610
+    },
+    {
+      "epoch": 0.940661265601051,
+      "grad_norm": 0.08044474687892239,
+      "learning_rate": 1.2742975278631378e-05,
+      "loss": 0.3856,
+      "step": 1611
+    },
+    {
+      "epoch": 0.9412451645865265,
+      "grad_norm": 0.07855363785819411,
+      "learning_rate": 1.2733168283203692e-05,
+      "loss": 0.3564,
+      "step": 1612
+    },
+    {
+      "epoch": 0.941829063572002,
+      "grad_norm": 0.08234623274736898,
+      "learning_rate": 1.272335844604685e-05,
+      "loss": 0.425,
+      "step": 1613
+    },
+    {
+      "epoch": 0.9424129625574775,
+      "grad_norm": 0.08498265570700943,
+      "learning_rate": 1.2713545777360334e-05,
+      "loss": 0.4303,
+      "step": 1614
+    },
+    {
+      "epoch": 0.942996861542953,
+      "grad_norm": 0.08161740378984432,
+      "learning_rate": 1.2703730287346565e-05,
+      "loss": 0.3831,
+      "step": 1615
+    },
+    {
+      "epoch": 0.9435807605284285,
+      "grad_norm": 0.08377470772939512,
+      "learning_rate": 1.2693911986210905e-05,
+      "loss": 0.3695,
+      "step": 1616
+    },
+    {
+      "epoch": 0.944164659513904,
+      "grad_norm": 0.08572821733646184,
+      "learning_rate": 1.2684090884161636e-05,
+      "loss": 0.4196,
+      "step": 1617
+    },
+    {
+      "epoch": 0.9447485584993796,
+      "grad_norm": 0.08990732612977403,
+      "learning_rate": 1.2674266991409949e-05,
+      "loss": 0.475,
+      "step": 1618
+    },
+    {
+      "epoch": 0.9453324574848552,
+      "grad_norm": 0.08329935608929678,
+      "learning_rate": 1.2664440318169949e-05,
+      "loss": 0.4006,
+      "step": 1619
+    },
+    {
+      "epoch": 0.9459163564703307,
+      "grad_norm": 0.0801541639436299,
+      "learning_rate": 1.265461087465861e-05,
+      "loss": 0.4234,
+      "step": 1620
+    },
+    {
+      "epoch": 0.9465002554558062,
+      "grad_norm": 0.08430240949504242,
+      "learning_rate": 1.2644778671095808e-05,
+      "loss": 0.4022,
+      "step": 1621
+    },
+    {
+      "epoch": 0.9470841544412817,
+      "grad_norm": 0.0820184019447128,
+      "learning_rate": 1.2634943717704275e-05,
+      "loss": 0.4111,
+      "step": 1622
+    },
+    {
+      "epoch": 0.9476680534267572,
+      "grad_norm": 0.08014321788071689,
+      "learning_rate": 1.262510602470961e-05,
+      "loss": 0.4105,
+      "step": 1623
+    },
+    {
+      "epoch": 0.9482519524122327,
+      "grad_norm": 0.08190969710595589,
+      "learning_rate": 1.2615265602340259e-05,
+      "loss": 0.3993,
+      "step": 1624
+    },
+    {
+      "epoch": 0.9488358513977082,
+      "grad_norm": 0.07750869241360776,
+      "learning_rate": 1.2605422460827494e-05,
+      "loss": 0.3828,
+      "step": 1625
+    },
+    {
+      "epoch": 0.9494197503831837,
+      "grad_norm": 0.09035434396377277,
+      "learning_rate": 1.2595576610405436e-05,
+      "loss": 0.4137,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9500036493686592,
+      "grad_norm": 0.08262292055375683,
+      "learning_rate": 1.2585728061311003e-05,
+      "loss": 0.3862,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9505875483541347,
+      "grad_norm": 0.08534046652155597,
+      "learning_rate": 1.257587682378393e-05,
+      "loss": 0.425,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9511714473396102,
+      "grad_norm": 0.08334330834449981,
+      "learning_rate": 1.256602290806674e-05,
+      "loss": 0.3702,
+      "step": 1629
+    },
+    {
+      "epoch": 0.9517553463250857,
+      "grad_norm": 0.08570377614862874,
+      "learning_rate": 1.2556166324404747e-05,
+      "loss": 0.4043,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9523392453105612,
+      "grad_norm": 0.08320503848675599,
+      "learning_rate": 1.2546307083046037e-05,
+      "loss": 0.42,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9529231442960368,
+      "grad_norm": 0.08370166024471179,
+      "learning_rate": 1.2536445194241455e-05,
+      "loss": 0.4068,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9535070432815123,
+      "grad_norm": 0.08560583681848602,
+      "learning_rate": 1.2526580668244607e-05,
+      "loss": 0.4333,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9540909422669878,
+      "grad_norm": 0.0899612520065487,
+      "learning_rate": 1.2516713515311832e-05,
+      "loss": 0.4328,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9546748412524633,
+      "grad_norm": 0.07768452869452072,
+      "learning_rate": 1.2506843745702204e-05,
+      "loss": 0.4124,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9552587402379389,
+      "grad_norm": 0.0805585964479371,
+      "learning_rate": 1.2496971369677518e-05,
+      "loss": 0.4184,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9558426392234144,
+      "grad_norm": 0.09254740415651047,
+      "learning_rate": 1.248709639750228e-05,
+      "loss": 0.489,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9564265382088899,
+      "grad_norm": 0.07962114886139027,
+      "learning_rate": 1.2477218839443694e-05,
+      "loss": 0.4038,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9570104371943654,
+      "grad_norm": 0.09305513638644114,
+      "learning_rate": 1.246733870577165e-05,
+      "loss": 0.4496,
+      "step": 1639
+    },
+    {
+      "epoch": 0.9575943361798409,
+      "grad_norm": 0.07821801609666267,
+      "learning_rate": 1.2457456006758722e-05,
+      "loss": 0.3927,
+      "step": 1640
+    },
+    {
+      "epoch": 0.9581782351653164,
+      "grad_norm": 0.07846262068745843,
+      "learning_rate": 1.2447570752680147e-05,
+      "loss": 0.3814,
+      "step": 1641
+    },
+    {
+      "epoch": 0.9587621341507919,
+      "grad_norm": 0.09416913348867216,
+      "learning_rate": 1.243768295381382e-05,
+      "loss": 0.4979,
+      "step": 1642
+    },
+    {
+      "epoch": 0.9593460331362674,
+      "grad_norm": 0.09242076917880761,
+      "learning_rate": 1.242779262044028e-05,
+      "loss": 0.4278,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9599299321217429,
+      "grad_norm": 0.07541963460143082,
+      "learning_rate": 1.24178997628427e-05,
+      "loss": 0.3915,
+      "step": 1644
+    },
+    {
+      "epoch": 0.9605138311072184,
+      "grad_norm": 0.08101399139582169,
+      "learning_rate": 1.2408004391306883e-05,
+      "loss": 0.411,
+      "step": 1645
+    },
+    {
+      "epoch": 0.961097730092694,
+      "grad_norm": 0.08050788395520199,
+      "learning_rate": 1.2398106516121243e-05,
+      "loss": 0.3954,
+      "step": 1646
+    },
+    {
+      "epoch": 0.9616816290781695,
+      "grad_norm": 0.08306726000340682,
+      "learning_rate": 1.2388206147576796e-05,
+      "loss": 0.4188,
+      "step": 1647
+    },
+    {
+      "epoch": 0.962265528063645,
+      "grad_norm": 0.07851595411076907,
+      "learning_rate": 1.2378303295967147e-05,
+      "loss": 0.3834,
+      "step": 1648
+    },
+    {
+      "epoch": 0.9628494270491205,
+      "grad_norm": 0.08093589272623701,
+      "learning_rate": 1.2368397971588493e-05,
+      "loss": 0.3832,
+      "step": 1649
+    },
+    {
+      "epoch": 0.963433326034596,
+      "grad_norm": 0.08319404672716485,
+      "learning_rate": 1.2358490184739593e-05,
+      "loss": 0.4584,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9640172250200715,
+      "grad_norm": 0.08223446422389624,
+      "learning_rate": 1.2348579945721769e-05,
+      "loss": 0.3886,
+      "step": 1651
+    },
+    {
+      "epoch": 0.9646011240055471,
+      "grad_norm": 0.07972472168907593,
+      "learning_rate": 1.2338667264838895e-05,
+      "loss": 0.392,
+      "step": 1652
+    },
+    {
+      "epoch": 0.9651850229910226,
+      "grad_norm": 0.08367656017044026,
+      "learning_rate": 1.2328752152397373e-05,
+      "loss": 0.3817,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9657689219764981,
+      "grad_norm": 0.08136686807372888,
+      "learning_rate": 1.2318834618706154e-05,
+      "loss": 0.4228,
+      "step": 1654
+    },
+    {
+      "epoch": 0.9663528209619736,
+      "grad_norm": 0.08227201010715726,
+      "learning_rate": 1.2308914674076687e-05,
+      "loss": 0.4305,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9669367199474491,
+      "grad_norm": 0.08090092132794494,
+      "learning_rate": 1.2298992328822937e-05,
+      "loss": 0.4491,
+      "step": 1656
+    },
+    {
+      "epoch": 0.9675206189329246,
+      "grad_norm": 0.08182726132604191,
+      "learning_rate": 1.2289067593261358e-05,
+      "loss": 0.4036,
+      "step": 1657
+    },
+    {
+      "epoch": 0.9681045179184001,
+      "grad_norm": 0.08285628276119505,
+      "learning_rate": 1.2279140477710902e-05,
+      "loss": 0.4103,
+      "step": 1658
+    },
+    {
+      "epoch": 0.9686884169038756,
+      "grad_norm": 0.07577609663970522,
+      "learning_rate": 1.2269210992492982e-05,
+      "loss": 0.3658,
+      "step": 1659
+    },
+    {
+      "epoch": 0.9692723158893511,
+      "grad_norm": 0.07957193340961839,
+      "learning_rate": 1.2259279147931479e-05,
+      "loss": 0.4137,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9698562148748266,
+      "grad_norm": 0.08577469425898346,
+      "learning_rate": 1.2249344954352735e-05,
+      "loss": 0.446,
+      "step": 1661
+    },
+    {
+      "epoch": 0.9704401138603022,
+      "grad_norm": 0.08662128547989922,
+      "learning_rate": 1.2239408422085518e-05,
+      "loss": 0.4774,
+      "step": 1662
+    },
+    {
+      "epoch": 0.9710240128457777,
+      "grad_norm": 0.09042105547729617,
+      "learning_rate": 1.2229469561461046e-05,
+      "loss": 0.4556,
+      "step": 1663
+    },
+    {
+      "epoch": 0.9716079118312532,
+      "grad_norm": 0.09056832064703556,
+      "learning_rate": 1.2219528382812946e-05,
+      "loss": 0.4514,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9721918108167287,
+      "grad_norm": 0.08346942320207329,
+      "learning_rate": 1.2209584896477258e-05,
+      "loss": 0.394,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9727757098022042,
+      "grad_norm": 0.08389476876648999,
+      "learning_rate": 1.2199639112792423e-05,
+      "loss": 0.3984,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9733596087876797,
+      "grad_norm": 0.08815523491226017,
+      "learning_rate": 1.2189691042099265e-05,
+      "loss": 0.4306,
+      "step": 1667
+    },
+    {
+      "epoch": 0.9739435077731552,
+      "grad_norm": 0.08372558220525145,
+      "learning_rate": 1.2179740694740993e-05,
+      "loss": 0.4308,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9745274067586308,
+      "grad_norm": 0.08013915855496855,
+      "learning_rate": 1.2169788081063181e-05,
+      "loss": 0.4348,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9751113057441063,
+      "grad_norm": 0.08207841704487195,
+      "learning_rate": 1.2159833211413759e-05,
+      "loss": 0.4252,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9756952047295818,
+      "grad_norm": 0.08330948153643103,
+      "learning_rate": 1.2149876096142998e-05,
+      "loss": 0.4188,
+      "step": 1671
+    },
+    {
+      "epoch": 0.9762791037150573,
+      "grad_norm": 0.07826531737970567,
+      "learning_rate": 1.2139916745603509e-05,
+      "loss": 0.4091,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9768630027005328,
+      "grad_norm": 0.08435778675934305,
+      "learning_rate": 1.2129955170150228e-05,
+      "loss": 0.4178,
+      "step": 1673
+    },
+    {
+      "epoch": 0.9774469016860083,
+      "grad_norm": 0.08366203918840605,
+      "learning_rate": 1.21199913801404e-05,
+      "loss": 0.4308,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9780308006714838,
+      "grad_norm": 0.0825963820143715,
+      "learning_rate": 1.2110025385933582e-05,
+      "loss": 0.4236,
+      "step": 1675
+    },
+    {
+      "epoch": 0.9786146996569594,
+      "grad_norm": 0.0811412700162295,
+      "learning_rate": 1.2100057197891601e-05,
+      "loss": 0.3588,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9791985986424349,
+      "grad_norm": 0.08000574841522486,
+      "learning_rate": 1.209008682637859e-05,
+      "loss": 0.4275,
+      "step": 1677
+    },
+    {
+      "epoch": 0.9797824976279104,
+      "grad_norm": 0.08169959062492475,
+      "learning_rate": 1.2080114281760942e-05,
+      "loss": 0.4098,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9803663966133859,
+      "grad_norm": 0.09055514715700597,
+      "learning_rate": 1.2070139574407302e-05,
+      "loss": 0.4453,
+      "step": 1679
+    },
+    {
+      "epoch": 0.9809502955988614,
+      "grad_norm": 0.09157420581377164,
+      "learning_rate": 1.2060162714688582e-05,
+      "loss": 0.4186,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9815341945843369,
+      "grad_norm": 0.08281402774041656,
+      "learning_rate": 1.2050183712977903e-05,
+      "loss": 0.4222,
+      "step": 1681
+    },
+    {
+      "epoch": 0.9821180935698124,
+      "grad_norm": 0.08440402239568318,
+      "learning_rate": 1.2040202579650649e-05,
+      "loss": 0.4584,
+      "step": 1682
+    },
+    {
+      "epoch": 0.9827019925552879,
+      "grad_norm": 0.08466514915805529,
+      "learning_rate": 1.2030219325084388e-05,
+      "loss": 0.4782,
+      "step": 1683
+    },
+    {
+      "epoch": 0.9832858915407634,
+      "grad_norm": 0.08014144495087902,
+      "learning_rate": 1.2020233959658918e-05,
+      "loss": 0.3971,
+      "step": 1684
+    },
+    {
+      "epoch": 0.9838697905262389,
+      "grad_norm": 0.07712097401990997,
+      "learning_rate": 1.2010246493756215e-05,
+      "loss": 0.3808,
+      "step": 1685
+    },
+    {
+      "epoch": 0.9844536895117145,
+      "grad_norm": 0.08330106304858904,
+      "learning_rate": 1.2000256937760446e-05,
+      "loss": 0.4253,
+      "step": 1686
+    },
+    {
+      "epoch": 0.98503758849719,
+      "grad_norm": 0.08856439750878418,
+      "learning_rate": 1.1990265302057948e-05,
+      "loss": 0.3955,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9856214874826655,
+      "grad_norm": 0.08425172172700846,
+      "learning_rate": 1.1980271597037228e-05,
+      "loss": 0.3905,
+      "step": 1688
+    },
+    {
+      "epoch": 0.986205386468141,
+      "grad_norm": 0.08059490178699054,
+      "learning_rate": 1.1970275833088936e-05,
+      "loss": 0.403,
+      "step": 1689
+    },
+    {
+      "epoch": 0.9867892854536165,
+      "grad_norm": 0.07999207600573359,
+      "learning_rate": 1.1960278020605861e-05,
+      "loss": 0.4233,
+      "step": 1690
+    },
+    {
+      "epoch": 0.987373184439092,
+      "grad_norm": 0.08277173263397818,
+      "learning_rate": 1.1950278169982934e-05,
+      "loss": 0.3853,
+      "step": 1691
+    },
+    {
+      "epoch": 0.9879570834245676,
+      "grad_norm": 0.08231937879367386,
+      "learning_rate": 1.1940276291617192e-05,
+      "loss": 0.37,
+      "step": 1692
+    },
+    {
+      "epoch": 0.9885409824100431,
+      "grad_norm": 0.07881303343198738,
+      "learning_rate": 1.1930272395907789e-05,
+      "loss": 0.3828,
+      "step": 1693
+    },
+    {
+      "epoch": 0.9891248813955186,
+      "grad_norm": 0.08446528604939832,
+      "learning_rate": 1.1920266493255976e-05,
+      "loss": 0.3907,
+      "step": 1694
+    },
+    {
+      "epoch": 0.9897087803809941,
+      "grad_norm": 0.08610756912692748,
+      "learning_rate": 1.1910258594065079e-05,
+      "loss": 0.4184,
+      "step": 1695
+    },
+    {
+      "epoch": 0.9902926793664696,
+      "grad_norm": 0.08332393667913181,
+      "learning_rate": 1.1900248708740515e-05,
+      "loss": 0.416,
+      "step": 1696
+    },
+    {
+      "epoch": 0.9908765783519451,
+      "grad_norm": 0.08667373269848362,
+      "learning_rate": 1.1890236847689762e-05,
+      "loss": 0.4177,
+      "step": 1697
+    },
+    {
+      "epoch": 0.9914604773374206,
+      "grad_norm": 0.08141091304512686,
+      "learning_rate": 1.1880223021322348e-05,
+      "loss": 0.4249,
+      "step": 1698
+    },
+    {
+      "epoch": 0.9920443763228961,
+      "grad_norm": 0.08272698658109609,
+      "learning_rate": 1.1870207240049845e-05,
+      "loss": 0.3728,
+      "step": 1699
+    },
+    {
+      "epoch": 0.9926282753083716,
+      "grad_norm": 0.07886843051175789,
+      "learning_rate": 1.1860189514285858e-05,
+      "loss": 0.3717,
+      "step": 1700
+    },
+    {
+      "epoch": 0.9932121742938471,
+      "grad_norm": 0.08805322206189176,
+      "learning_rate": 1.185016985444602e-05,
+      "loss": 0.4067,
+      "step": 1701
+    },
+    {
+      "epoch": 0.9937960732793226,
+      "grad_norm": 0.08019950575324196,
+      "learning_rate": 1.1840148270947962e-05,
+      "loss": 0.3735,
+      "step": 1702
+    },
+    {
+      "epoch": 0.9943799722647982,
+      "grad_norm": 0.08780073185384403,
+      "learning_rate": 1.183012477421133e-05,
+      "loss": 0.3948,
+      "step": 1703
+    },
+    {
+      "epoch": 0.9949638712502737,
+      "grad_norm": 0.08378652407871891,
+      "learning_rate": 1.1820099374657748e-05,
+      "loss": 0.387,
+      "step": 1704
+    },
+    {
+      "epoch": 0.9955477702357493,
+      "grad_norm": 0.08558109018073358,
+      "learning_rate": 1.1810072082710823e-05,
+      "loss": 0.4026,
+      "step": 1705
+    },
+    {
+      "epoch": 0.9961316692212248,
+      "grad_norm": 0.09047159468458155,
+      "learning_rate": 1.180004290879613e-05,
+      "loss": 0.4238,
+      "step": 1706
+    },
+    {
+      "epoch": 0.9967155682067003,
+      "grad_norm": 0.08083355958596425,
+      "learning_rate": 1.1790011863341197e-05,
+      "loss": 0.3741,
+      "step": 1707
+    },
+    {
+      "epoch": 0.9972994671921758,
+      "grad_norm": 0.08399270381890485,
+      "learning_rate": 1.1779978956775507e-05,
+      "loss": 0.4195,
+      "step": 1708
+    },
+    {
+      "epoch": 0.9978833661776513,
+      "grad_norm": 0.08503587350290584,
+      "learning_rate": 1.1769944199530458e-05,
+      "loss": 0.3766,
+      "step": 1709
+    },
+    {
+      "epoch": 0.9984672651631268,
+      "grad_norm": 0.08275774487229641,
+      "learning_rate": 1.17599076020394e-05,
+      "loss": 0.3977,
+      "step": 1710
+    },
+    {
+      "epoch": 0.9990511641486023,
+      "grad_norm": 0.07579348599909388,
+      "learning_rate": 1.1749869174737575e-05,
+      "loss": 0.3967,
+      "step": 1711
+    },
+    {
+      "epoch": 0.9996350631340778,
+      "grad_norm": 0.08450272511158562,
+      "learning_rate": 1.173982892806214e-05,
+      "loss": 0.4174,
+      "step": 1712
+    },
+    {
+      "epoch": 1.0004379242391066,
+      "grad_norm": 0.1912467125681184,
+      "learning_rate": 1.172978687245213e-05,
+      "loss": 0.637,
+      "step": 1713
+    },
+    {
+      "epoch": 1.0010218232245822,
+      "grad_norm": 0.08697580322760777,
+      "learning_rate": 1.1719743018348477e-05,
+      "loss": 0.3237,
+      "step": 1714
+    },
+    {
+      "epoch": 1.0016057222100576,
+      "grad_norm": 0.0805064059267385,
+      "learning_rate": 1.1709697376193967e-05,
+      "loss": 0.3281,
+      "step": 1715
+    },
+    {
+      "epoch": 1.0021896211955332,
+      "grad_norm": 0.08517965615115591,
+      "learning_rate": 1.169964995643326e-05,
+      "loss": 0.309,
+      "step": 1716
+    },
+    {
+      "epoch": 1.0027735201810086,
+      "grad_norm": 0.0915142865313435,
+      "learning_rate": 1.1689600769512855e-05,
+      "loss": 0.3099,
+      "step": 1717
+    },
+    {
+      "epoch": 1.0033574191664842,
+      "grad_norm": 0.0993255729223709,
+      "learning_rate": 1.1679549825881087e-05,
+      "loss": 0.3414,
+      "step": 1718
+    },
+    {
+      "epoch": 1.0039413181519596,
+      "grad_norm": 0.11527759701646147,
+      "learning_rate": 1.1669497135988127e-05,
+      "loss": 0.3778,
+      "step": 1719
+    },
+    {
+      "epoch": 1.0045252171374353,
+      "grad_norm": 0.09291453716801787,
+      "learning_rate": 1.1659442710285948e-05,
+      "loss": 0.321,
+      "step": 1720
+    },
+    {
+      "epoch": 1.0051091161229107,
+      "grad_norm": 0.0770329072994594,
+      "learning_rate": 1.1649386559228342e-05,
+      "loss": 0.2728,
+      "step": 1721
+    },
+    {
+      "epoch": 1.0056930151083863,
+      "grad_norm": 0.10091667186053087,
+      "learning_rate": 1.1639328693270887e-05,
+      "loss": 0.3089,
+      "step": 1722
+    },
+    {
+      "epoch": 1.0062769140938617,
+      "grad_norm": 0.09952190807605586,
+      "learning_rate": 1.1629269122870942e-05,
+      "loss": 0.3502,
+      "step": 1723
+    },
+    {
+      "epoch": 1.0068608130793373,
+      "grad_norm": 0.08931006168728403,
+      "learning_rate": 1.1619207858487646e-05,
+      "loss": 0.3438,
+      "step": 1724
+    },
+    {
+      "epoch": 1.007444712064813,
+      "grad_norm": 0.08880395841091963,
+      "learning_rate": 1.1609144910581891e-05,
+      "loss": 0.2973,
+      "step": 1725
+    },
+    {
+      "epoch": 1.0080286110502883,
+      "grad_norm": 0.09556210570444573,
+      "learning_rate": 1.1599080289616329e-05,
+      "loss": 0.3772,
+      "step": 1726
+    },
+    {
+      "epoch": 1.008612510035764,
+      "grad_norm": 0.10156389639784554,
+      "learning_rate": 1.1589014006055337e-05,
+      "loss": 0.3664,
+      "step": 1727
+    },
+    {
+      "epoch": 1.0091964090212393,
+      "grad_norm": 0.08788545035869486,
+      "learning_rate": 1.1578946070365035e-05,
+      "loss": 0.3291,
+      "step": 1728
+    },
+    {
+      "epoch": 1.009780308006715,
+      "grad_norm": 0.10089499605574208,
+      "learning_rate": 1.1568876493013255e-05,
+      "loss": 0.3867,
+      "step": 1729
+    },
+    {
+      "epoch": 1.0103642069921903,
+      "grad_norm": 0.08893955892414476,
+      "learning_rate": 1.1558805284469533e-05,
+      "loss": 0.3202,
+      "step": 1730
+    },
+    {
+      "epoch": 1.010948105977666,
+      "grad_norm": 0.08872269674357061,
+      "learning_rate": 1.1548732455205105e-05,
+      "loss": 0.3404,
+      "step": 1731
+    },
+    {
+      "epoch": 1.0115320049631413,
+      "grad_norm": 0.08686408307194382,
+      "learning_rate": 1.1538658015692892e-05,
+      "loss": 0.3077,
+      "step": 1732
+    },
+    {
+      "epoch": 1.012115903948617,
+      "grad_norm": 0.09894446111477613,
+      "learning_rate": 1.1528581976407485e-05,
+      "loss": 0.3162,
+      "step": 1733
+    },
+    {
+      "epoch": 1.0126998029340923,
+      "grad_norm": 0.09451219332737623,
+      "learning_rate": 1.1518504347825146e-05,
+      "loss": 0.3135,
+      "step": 1734
+    },
+    {
+      "epoch": 1.013283701919568,
+      "grad_norm": 0.08452960812026944,
+      "learning_rate": 1.1508425140423782e-05,
+      "loss": 0.3079,
+      "step": 1735
+    },
+    {
+      "epoch": 1.0138676009050434,
+      "grad_norm": 0.08634422922248457,
+      "learning_rate": 1.1498344364682948e-05,
+      "loss": 0.3253,
+      "step": 1736
+    },
+    {
+      "epoch": 1.014451499890519,
+      "grad_norm": 0.09026571498898218,
+      "learning_rate": 1.1488262031083816e-05,
+      "loss": 0.2987,
+      "step": 1737
+    },
+    {
+      "epoch": 1.0150353988759944,
+      "grad_norm": 0.10393657329828083,
+      "learning_rate": 1.14781781501092e-05,
+      "loss": 0.3648,
+      "step": 1738
+    },
+    {
+      "epoch": 1.01561929786147,
+      "grad_norm": 0.09879951606135828,
+      "learning_rate": 1.1468092732243506e-05,
+      "loss": 0.3579,
+      "step": 1739
+    },
+    {
+      "epoch": 1.0162031968469454,
+      "grad_norm": 0.09381033803237881,
+      "learning_rate": 1.1458005787972743e-05,
+      "loss": 0.3492,
+      "step": 1740
+    },
+    {
+      "epoch": 1.016787095832421,
+      "grad_norm": 0.08991595151945608,
+      "learning_rate": 1.1447917327784504e-05,
+      "loss": 0.3259,
+      "step": 1741
+    },
+    {
+      "epoch": 1.0173709948178966,
+      "grad_norm": 0.08800332252792094,
+      "learning_rate": 1.143782736216796e-05,
+      "loss": 0.3352,
+      "step": 1742
+    },
+    {
+      "epoch": 1.017954893803372,
+      "grad_norm": 0.08719776357966119,
+      "learning_rate": 1.1427735901613854e-05,
+      "loss": 0.3571,
+      "step": 1743
+    },
+    {
+      "epoch": 1.0185387927888476,
+      "grad_norm": 0.08452237604412646,
+      "learning_rate": 1.1417642956614474e-05,
+      "loss": 0.3255,
+      "step": 1744
+    },
+    {
+      "epoch": 1.019122691774323,
+      "grad_norm": 0.09585559433037179,
+      "learning_rate": 1.1407548537663655e-05,
+      "loss": 0.3051,
+      "step": 1745
+    },
+    {
+      "epoch": 1.0197065907597986,
+      "grad_norm": 0.0910477784920818,
+      "learning_rate": 1.1397452655256762e-05,
+      "loss": 0.3411,
+      "step": 1746
+    },
+    {
+      "epoch": 1.020290489745274,
+      "grad_norm": 0.09014096347293853,
+      "learning_rate": 1.1387355319890685e-05,
+      "loss": 0.3504,
+      "step": 1747
+    },
+    {
+      "epoch": 1.0208743887307496,
+      "grad_norm": 0.09750486112651656,
+      "learning_rate": 1.1377256542063822e-05,
+      "loss": 0.3611,
+      "step": 1748
+    },
+    {
+      "epoch": 1.021458287716225,
+      "grad_norm": 0.08968578959019927,
+      "learning_rate": 1.1367156332276077e-05,
+      "loss": 0.318,
+      "step": 1749
+    },
+    {
+      "epoch": 1.0220421867017007,
+      "grad_norm": 0.10154361572330561,
+      "learning_rate": 1.1357054701028836e-05,
+      "loss": 0.3654,
+      "step": 1750
+    },
+    {
+      "epoch": 1.022626085687176,
+      "grad_norm": 0.08732904125831531,
+      "learning_rate": 1.1346951658824958e-05,
+      "loss": 0.3163,
+      "step": 1751
+    },
+    {
+      "epoch": 1.0232099846726517,
+      "grad_norm": 0.0882720645895556,
+      "learning_rate": 1.1336847216168785e-05,
+      "loss": 0.3401,
+      "step": 1752
+    },
+    {
+      "epoch": 1.023793883658127,
+      "grad_norm": 0.08804888936565881,
+      "learning_rate": 1.1326741383566102e-05,
+      "loss": 0.3316,
+      "step": 1753
+    },
+    {
+      "epoch": 1.0243777826436027,
+      "grad_norm": 0.08594064650925629,
+      "learning_rate": 1.1316634171524147e-05,
+      "loss": 0.3406,
+      "step": 1754
+    },
+    {
+      "epoch": 1.024961681629078,
+      "grad_norm": 0.09799570231667268,
+      "learning_rate": 1.1306525590551585e-05,
+      "loss": 0.3762,
+      "step": 1755
+    },
+    {
+      "epoch": 1.0255455806145537,
+      "grad_norm": 0.08862689741909448,
+      "learning_rate": 1.1296415651158506e-05,
+      "loss": 0.3305,
+      "step": 1756
+    },
+    {
+      "epoch": 1.026129479600029,
+      "grad_norm": 0.08529375170771417,
+      "learning_rate": 1.1286304363856418e-05,
+      "loss": 0.3219,
+      "step": 1757
+    },
+    {
+      "epoch": 1.0267133785855047,
+      "grad_norm": 0.08254822563085104,
+      "learning_rate": 1.1276191739158222e-05,
+      "loss": 0.3127,
+      "step": 1758
+    },
+    {
+      "epoch": 1.0272972775709803,
+      "grad_norm": 0.1364423558928236,
+      "learning_rate": 1.126607778757822e-05,
+      "loss": 0.3344,
+      "step": 1759
+    },
+    {
+      "epoch": 1.0278811765564557,
+      "grad_norm": 0.08816696562154225,
+      "learning_rate": 1.1255962519632082e-05,
+      "loss": 0.2891,
+      "step": 1760
+    },
+    {
+      "epoch": 1.0284650755419313,
+      "grad_norm": 0.08601739586915863,
+      "learning_rate": 1.1245845945836855e-05,
+      "loss": 0.3198,
+      "step": 1761
+    },
+    {
+      "epoch": 1.0290489745274067,
+      "grad_norm": 0.08871916134623715,
+      "learning_rate": 1.123572807671094e-05,
+      "loss": 0.3297,
+      "step": 1762
+    },
+    {
+      "epoch": 1.0296328735128824,
+      "grad_norm": 0.08661990673935958,
+      "learning_rate": 1.122560892277409e-05,
+      "loss": 0.335,
+      "step": 1763
+    },
+    {
+      "epoch": 1.0302167724983577,
+      "grad_norm": 0.08489079190594304,
+      "learning_rate": 1.1215488494547384e-05,
+      "loss": 0.3095,
+      "step": 1764
+    },
+    {
+      "epoch": 1.0308006714838334,
+      "grad_norm": 0.09208731138753518,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.3323,
+      "step": 1765
+    },
+    {
+      "epoch": 1.0313845704693088,
+      "grad_norm": 0.0870864328636266,
+      "learning_rate": 1.1195243857315358e-05,
+      "loss": 0.3071,
+      "step": 1766
+    },
+    {
+      "epoch": 1.0319684694547844,
+      "grad_norm": 0.08498691537178311,
+      "learning_rate": 1.1185119669358792e-05,
+      "loss": 0.3585,
+      "step": 1767
+    },
+    {
+      "epoch": 1.0325523684402598,
+      "grad_norm": 0.09397767561847145,
+      "learning_rate": 1.1174994249209852e-05,
+      "loss": 0.3611,
+      "step": 1768
+    },
+    {
+      "epoch": 1.0331362674257354,
+      "grad_norm": 0.09109064770085072,
+      "learning_rate": 1.1164867607396136e-05,
+      "loss": 0.329,
+      "step": 1769
+    },
+    {
+      "epoch": 1.0337201664112108,
+      "grad_norm": 0.08762096153670203,
+      "learning_rate": 1.115473975444651e-05,
+      "loss": 0.3187,
+      "step": 1770
+    },
+    {
+      "epoch": 1.0343040653966864,
+      "grad_norm": 0.09133811720265025,
+      "learning_rate": 1.1144610700891108e-05,
+      "loss": 0.3078,
+      "step": 1771
+    },
+    {
+      "epoch": 1.0348879643821618,
+      "grad_norm": 0.09588764598147553,
+      "learning_rate": 1.1134480457261308e-05,
+      "loss": 0.3441,
+      "step": 1772
+    },
+    {
+      "epoch": 1.0354718633676374,
+      "grad_norm": 0.08223225650222007,
+      "learning_rate": 1.1124349034089724e-05,
+      "loss": 0.2796,
+      "step": 1773
+    },
+    {
+      "epoch": 1.0360557623531128,
+      "grad_norm": 0.09505321894521367,
+      "learning_rate": 1.1114216441910195e-05,
+      "loss": 0.3323,
+      "step": 1774
+    },
+    {
+      "epoch": 1.0366396613385884,
+      "grad_norm": 0.08806537226079642,
+      "learning_rate": 1.1104082691257778e-05,
+      "loss": 0.3175,
+      "step": 1775
+    },
+    {
+      "epoch": 1.037223560324064,
+      "grad_norm": 0.08908658788218432,
+      "learning_rate": 1.1093947792668735e-05,
+      "loss": 0.3057,
+      "step": 1776
+    },
+    {
+      "epoch": 1.0378074593095394,
+      "grad_norm": 0.09038765830567978,
+      "learning_rate": 1.1083811756680523e-05,
+      "loss": 0.2941,
+      "step": 1777
+    },
+    {
+      "epoch": 1.038391358295015,
+      "grad_norm": 0.09306521738752088,
+      "learning_rate": 1.1073674593831778e-05,
+      "loss": 0.3516,
+      "step": 1778
+    },
+    {
+      "epoch": 1.0389752572804904,
+      "grad_norm": 0.08807988352826143,
+      "learning_rate": 1.1063536314662301e-05,
+      "loss": 0.3336,
+      "step": 1779
+    },
+    {
+      "epoch": 1.039559156265966,
+      "grad_norm": 0.09011203337598298,
+      "learning_rate": 1.1053396929713076e-05,
+      "loss": 0.3087,
+      "step": 1780
+    },
+    {
+      "epoch": 1.0401430552514415,
+      "grad_norm": 0.10157517854174045,
+      "learning_rate": 1.1043256449526214e-05,
+      "loss": 0.3246,
+      "step": 1781
+    },
+    {
+      "epoch": 1.040726954236917,
+      "grad_norm": 0.0890466690273685,
+      "learning_rate": 1.103311488464497e-05,
+      "loss": 0.3225,
+      "step": 1782
+    },
+    {
+      "epoch": 1.0413108532223925,
+      "grad_norm": 0.09267728695025493,
+      "learning_rate": 1.1022972245613735e-05,
+      "loss": 0.3062,
+      "step": 1783
+    },
+    {
+      "epoch": 1.041894752207868,
+      "grad_norm": 0.08212474940586123,
+      "learning_rate": 1.101282854297801e-05,
+      "loss": 0.2816,
+      "step": 1784
+    },
+    {
+      "epoch": 1.0424786511933435,
+      "grad_norm": 0.08965593176323144,
+      "learning_rate": 1.1002683787284403e-05,
+      "loss": 0.3259,
+      "step": 1785
+    },
+    {
+      "epoch": 1.043062550178819,
+      "grad_norm": 0.08706569075857526,
+      "learning_rate": 1.0992537989080618e-05,
+      "loss": 0.3239,
+      "step": 1786
+    },
+    {
+      "epoch": 1.0436464491642945,
+      "grad_norm": 0.09462190361543978,
+      "learning_rate": 1.0982391158915441e-05,
+      "loss": 0.3648,
+      "step": 1787
+    },
+    {
+      "epoch": 1.0442303481497701,
+      "grad_norm": 0.08667189603934797,
+      "learning_rate": 1.0972243307338733e-05,
+      "loss": 0.3125,
+      "step": 1788
+    },
+    {
+      "epoch": 1.0448142471352455,
+      "grad_norm": 0.09030038635394509,
+      "learning_rate": 1.0962094444901416e-05,
+      "loss": 0.3476,
+      "step": 1789
+    },
+    {
+      "epoch": 1.0453981461207211,
+      "grad_norm": 0.08317020274142514,
+      "learning_rate": 1.0951944582155463e-05,
+      "loss": 0.2837,
+      "step": 1790
+    },
+    {
+      "epoch": 1.0459820451061965,
+      "grad_norm": 0.0849162982968745,
+      "learning_rate": 1.094179372965389e-05,
+      "loss": 0.3041,
+      "step": 1791
+    },
+    {
+      "epoch": 1.0465659440916721,
+      "grad_norm": 0.09593282545181771,
+      "learning_rate": 1.0931641897950733e-05,
+      "loss": 0.3423,
+      "step": 1792
+    },
+    {
+      "epoch": 1.0471498430771478,
+      "grad_norm": 0.08703457855863093,
+      "learning_rate": 1.0921489097601054e-05,
+      "loss": 0.3068,
+      "step": 1793
+    },
+    {
+      "epoch": 1.0477337420626232,
+      "grad_norm": 0.08991516778826956,
+      "learning_rate": 1.0911335339160924e-05,
+      "loss": 0.3291,
+      "step": 1794
+    },
+    {
+      "epoch": 1.0483176410480988,
+      "grad_norm": 0.0836878109101553,
+      "learning_rate": 1.090118063318741e-05,
+      "loss": 0.2863,
+      "step": 1795
+    },
+    {
+      "epoch": 1.0489015400335742,
+      "grad_norm": 0.09317004178051563,
+      "learning_rate": 1.089102499023855e-05,
+      "loss": 0.3134,
+      "step": 1796
+    },
+    {
+      "epoch": 1.0494854390190498,
+      "grad_norm": 0.08783545331915901,
+      "learning_rate": 1.0880868420873375e-05,
+      "loss": 0.316,
+      "step": 1797
+    },
+    {
+      "epoch": 1.0500693380045252,
+      "grad_norm": 0.09717075051160355,
+      "learning_rate": 1.0870710935651868e-05,
+      "loss": 0.3415,
+      "step": 1798
+    },
+    {
+      "epoch": 1.0506532369900008,
+      "grad_norm": 0.10273082196167733,
+      "learning_rate": 1.086055254513497e-05,
+      "loss": 0.3701,
+      "step": 1799
+    },
+    {
+      "epoch": 1.0512371359754762,
+      "grad_norm": 0.08522548627924706,
+      "learning_rate": 1.085039325988456e-05,
+      "loss": 0.3065,
+      "step": 1800
+    },
+    {
+      "epoch": 1.0518210349609518,
+      "grad_norm": 0.09681180812067436,
+      "learning_rate": 1.0840233090463443e-05,
+      "loss": 0.3131,
+      "step": 1801
+    },
+    {
+      "epoch": 1.0524049339464272,
+      "grad_norm": 0.08175894892981807,
+      "learning_rate": 1.0830072047435354e-05,
+      "loss": 0.266,
+      "step": 1802
+    },
+    {
+      "epoch": 1.0529888329319028,
+      "grad_norm": 0.09339176575831291,
+      "learning_rate": 1.0819910141364929e-05,
+      "loss": 0.3197,
+      "step": 1803
+    },
+    {
+      "epoch": 1.0535727319173782,
+      "grad_norm": 0.09828543270335088,
+      "learning_rate": 1.0809747382817702e-05,
+      "loss": 0.3554,
+      "step": 1804
+    },
+    {
+      "epoch": 1.0541566309028538,
+      "grad_norm": 0.08663168394232607,
+      "learning_rate": 1.0799583782360097e-05,
+      "loss": 0.3176,
+      "step": 1805
+    },
+    {
+      "epoch": 1.0547405298883292,
+      "grad_norm": 0.09571089473313907,
+      "learning_rate": 1.0789419350559407e-05,
+      "loss": 0.3536,
+      "step": 1806
+    },
+    {
+      "epoch": 1.0553244288738048,
+      "grad_norm": 0.09395012528800536,
+      "learning_rate": 1.0779254097983788e-05,
+      "loss": 0.3054,
+      "step": 1807
+    },
+    {
+      "epoch": 1.0559083278592802,
+      "grad_norm": 0.09072998024826581,
+      "learning_rate": 1.0769088035202268e-05,
+      "loss": 0.3002,
+      "step": 1808
+    },
+    {
+      "epoch": 1.0564922268447559,
+      "grad_norm": 0.08733377425243394,
+      "learning_rate": 1.0758921172784696e-05,
+      "loss": 0.2985,
+      "step": 1809
+    },
+    {
+      "epoch": 1.0570761258302315,
+      "grad_norm": 0.0886473357807676,
+      "learning_rate": 1.0748753521301758e-05,
+      "loss": 0.3225,
+      "step": 1810
+    },
+    {
+      "epoch": 1.0576600248157069,
+      "grad_norm": 0.09289495932444,
+      "learning_rate": 1.0738585091324966e-05,
+      "loss": 0.3126,
+      "step": 1811
+    },
+    {
+      "epoch": 1.0582439238011825,
+      "grad_norm": 0.09690286135450574,
+      "learning_rate": 1.0728415893426636e-05,
+      "loss": 0.3458,
+      "step": 1812
+    },
+    {
+      "epoch": 1.0588278227866579,
+      "grad_norm": 0.08242696402852,
+      "learning_rate": 1.0718245938179886e-05,
+      "loss": 0.3028,
+      "step": 1813
+    },
+    {
+      "epoch": 1.0594117217721335,
+      "grad_norm": 0.08435811625341777,
+      "learning_rate": 1.0708075236158617e-05,
+      "loss": 0.3116,
+      "step": 1814
+    },
+    {
+      "epoch": 1.059995620757609,
+      "grad_norm": 0.09654781881156577,
+      "learning_rate": 1.0697903797937513e-05,
+      "loss": 0.3315,
+      "step": 1815
+    },
+    {
+      "epoch": 1.0605795197430845,
+      "grad_norm": 0.08714167970200515,
+      "learning_rate": 1.0687731634092016e-05,
+      "loss": 0.3194,
+      "step": 1816
+    },
+    {
+      "epoch": 1.06116341872856,
+      "grad_norm": 0.08766492206596804,
+      "learning_rate": 1.0677558755198327e-05,
+      "loss": 0.3133,
+      "step": 1817
+    },
+    {
+      "epoch": 1.0617473177140355,
+      "grad_norm": 0.09748307650178671,
+      "learning_rate": 1.0667385171833391e-05,
+      "loss": 0.3658,
+      "step": 1818
+    },
+    {
+      "epoch": 1.062331216699511,
+      "grad_norm": 0.08742633000244313,
+      "learning_rate": 1.0657210894574885e-05,
+      "loss": 0.3,
+      "step": 1819
+    },
+    {
+      "epoch": 1.0629151156849865,
+      "grad_norm": 0.09167938272772368,
+      "learning_rate": 1.0647035934001202e-05,
+      "loss": 0.3317,
+      "step": 1820
+    },
+    {
+      "epoch": 1.063499014670462,
+      "grad_norm": 0.0882520234788795,
+      "learning_rate": 1.0636860300691452e-05,
+      "loss": 0.345,
+      "step": 1821
+    },
+    {
+      "epoch": 1.0640829136559375,
+      "grad_norm": 0.09113695040025108,
+      "learning_rate": 1.0626684005225443e-05,
+      "loss": 0.3539,
+      "step": 1822
+    },
+    {
+      "epoch": 1.064666812641413,
+      "grad_norm": 0.0888576436164498,
+      "learning_rate": 1.0616507058183674e-05,
+      "loss": 0.3142,
+      "step": 1823
+    },
+    {
+      "epoch": 1.0652507116268886,
+      "grad_norm": 0.09039169147789959,
+      "learning_rate": 1.0606329470147313e-05,
+      "loss": 0.3325,
+      "step": 1824
+    },
+    {
+      "epoch": 1.065834610612364,
+      "grad_norm": 0.085874986699489,
+      "learning_rate": 1.05961512516982e-05,
+      "loss": 0.3205,
+      "step": 1825
+    },
+    {
+      "epoch": 1.0664185095978396,
+      "grad_norm": 0.08973004220517503,
+      "learning_rate": 1.0585972413418833e-05,
+      "loss": 0.3448,
+      "step": 1826
+    },
+    {
+      "epoch": 1.0670024085833152,
+      "grad_norm": 0.09059541291862992,
+      "learning_rate": 1.0575792965892349e-05,
+      "loss": 0.3494,
+      "step": 1827
+    },
+    {
+      "epoch": 1.0675863075687906,
+      "grad_norm": 0.0873605095928246,
+      "learning_rate": 1.0565612919702527e-05,
+      "loss": 0.3372,
+      "step": 1828
+    },
+    {
+      "epoch": 1.0681702065542662,
+      "grad_norm": 0.08969804637379235,
+      "learning_rate": 1.0555432285433754e-05,
+      "loss": 0.3062,
+      "step": 1829
+    },
+    {
+      "epoch": 1.0687541055397416,
+      "grad_norm": 0.09252899217343594,
+      "learning_rate": 1.0545251073671041e-05,
+      "loss": 0.3414,
+      "step": 1830
+    },
+    {
+      "epoch": 1.0693380045252172,
+      "grad_norm": 0.09631983925332553,
+      "learning_rate": 1.0535069294999995e-05,
+      "loss": 0.3898,
+      "step": 1831
+    },
+    {
+      "epoch": 1.0699219035106926,
+      "grad_norm": 0.08835403763009797,
+      "learning_rate": 1.0524886960006813e-05,
+      "loss": 0.3322,
+      "step": 1832
+    },
+    {
+      "epoch": 1.0705058024961682,
+      "grad_norm": 0.08771104024935827,
+      "learning_rate": 1.0514704079278273e-05,
+      "loss": 0.2895,
+      "step": 1833
+    },
+    {
+      "epoch": 1.0710897014816436,
+      "grad_norm": 0.08287677743267324,
+      "learning_rate": 1.0504520663401714e-05,
+      "loss": 0.3297,
+      "step": 1834
+    },
+    {
+      "epoch": 1.0716736004671192,
+      "grad_norm": 0.09699554024446415,
+      "learning_rate": 1.049433672296503e-05,
+      "loss": 0.3566,
+      "step": 1835
+    },
+    {
+      "epoch": 1.0722574994525946,
+      "grad_norm": 0.08177825551159992,
+      "learning_rate": 1.0484152268556677e-05,
+      "loss": 0.2851,
+      "step": 1836
+    },
+    {
+      "epoch": 1.0728413984380702,
+      "grad_norm": 0.0956503688128656,
+      "learning_rate": 1.0473967310765629e-05,
+      "loss": 0.3388,
+      "step": 1837
+    },
+    {
+      "epoch": 1.0734252974235456,
+      "grad_norm": 0.09080536792183554,
+      "learning_rate": 1.0463781860181385e-05,
+      "loss": 0.3243,
+      "step": 1838
+    },
+    {
+      "epoch": 1.0740091964090213,
+      "grad_norm": 0.08426945320777579,
+      "learning_rate": 1.0453595927393962e-05,
+      "loss": 0.3388,
+      "step": 1839
+    },
+    {
+      "epoch": 1.0745930953944967,
+      "grad_norm": 0.09731394618333872,
+      "learning_rate": 1.0443409522993877e-05,
+      "loss": 0.3372,
+      "step": 1840
+    },
+    {
+      "epoch": 1.0751769943799723,
+      "grad_norm": 0.08443950252827141,
+      "learning_rate": 1.0433222657572135e-05,
+      "loss": 0.3134,
+      "step": 1841
+    },
+    {
+      "epoch": 1.0757608933654477,
+      "grad_norm": 0.09421739659020999,
+      "learning_rate": 1.0423035341720222e-05,
+      "loss": 0.3654,
+      "step": 1842
+    },
+    {
+      "epoch": 1.0763447923509233,
+      "grad_norm": 0.09430754325293,
+      "learning_rate": 1.041284758603009e-05,
+      "loss": 0.3196,
+      "step": 1843
+    },
+    {
+      "epoch": 1.076928691336399,
+      "grad_norm": 0.08615392814860187,
+      "learning_rate": 1.0402659401094154e-05,
+      "loss": 0.3233,
+      "step": 1844
+    },
+    {
+      "epoch": 1.0775125903218743,
+      "grad_norm": 0.09045613864888608,
+      "learning_rate": 1.0392470797505268e-05,
+      "loss": 0.3495,
+      "step": 1845
+    },
+    {
+      "epoch": 1.07809648930735,
+      "grad_norm": 0.09056304052095641,
+      "learning_rate": 1.0382281785856725e-05,
+      "loss": 0.3284,
+      "step": 1846
+    },
+    {
+      "epoch": 1.0786803882928253,
+      "grad_norm": 0.08875320198350746,
+      "learning_rate": 1.0372092376742247e-05,
+      "loss": 0.3079,
+      "step": 1847
+    },
+    {
+      "epoch": 1.079264287278301,
+      "grad_norm": 0.0871461376369554,
+      "learning_rate": 1.0361902580755955e-05,
+      "loss": 0.3436,
+      "step": 1848
+    },
+    {
+      "epoch": 1.0798481862637763,
+      "grad_norm": 0.09671071157889524,
+      "learning_rate": 1.035171240849239e-05,
+      "loss": 0.371,
+      "step": 1849
+    },
+    {
+      "epoch": 1.080432085249252,
+      "grad_norm": 0.07371087797691109,
+      "learning_rate": 1.0341521870546472e-05,
+      "loss": 0.2808,
+      "step": 1850
+    },
+    {
+      "epoch": 1.0810159842347273,
+      "grad_norm": 0.08271718707257504,
+      "learning_rate": 1.033133097751351e-05,
+      "loss": 0.2947,
+      "step": 1851
+    },
+    {
+      "epoch": 1.081599883220203,
+      "grad_norm": 0.08773407067130308,
+      "learning_rate": 1.0321139739989167e-05,
+      "loss": 0.3073,
+      "step": 1852
+    },
+    {
+      "epoch": 1.0821837822056783,
+      "grad_norm": 0.0836832674492028,
+      "learning_rate": 1.0310948168569483e-05,
+      "loss": 0.3074,
+      "step": 1853
+    },
+    {
+      "epoch": 1.082767681191154,
+      "grad_norm": 0.09574137055711428,
+      "learning_rate": 1.0300756273850837e-05,
+      "loss": 0.3743,
+      "step": 1854
+    },
+    {
+      "epoch": 1.0833515801766294,
+      "grad_norm": 0.08714788969917496,
+      "learning_rate": 1.0290564066429935e-05,
+      "loss": 0.305,
+      "step": 1855
+    },
+    {
+      "epoch": 1.083935479162105,
+      "grad_norm": 0.09069074915162091,
+      "learning_rate": 1.0280371556903827e-05,
+      "loss": 0.3447,
+      "step": 1856
+    },
+    {
+      "epoch": 1.0845193781475804,
+      "grad_norm": 0.09722103494438748,
+      "learning_rate": 1.0270178755869861e-05,
+      "loss": 0.3474,
+      "step": 1857
+    },
+    {
+      "epoch": 1.085103277133056,
+      "grad_norm": 0.09531750293753882,
+      "learning_rate": 1.0259985673925694e-05,
+      "loss": 0.3392,
+      "step": 1858
+    },
+    {
+      "epoch": 1.0856871761185314,
+      "grad_norm": 0.09368848222408105,
+      "learning_rate": 1.0249792321669276e-05,
+      "loss": 0.2882,
+      "step": 1859
+    },
+    {
+      "epoch": 1.086271075104007,
+      "grad_norm": 0.08353102797194219,
+      "learning_rate": 1.0239598709698839e-05,
+      "loss": 0.3176,
+      "step": 1860
+    },
+    {
+      "epoch": 1.0868549740894826,
+      "grad_norm": 0.09351048885672311,
+      "learning_rate": 1.0229404848612882e-05,
+      "loss": 0.3266,
+      "step": 1861
+    },
+    {
+      "epoch": 1.087438873074958,
+      "grad_norm": 0.08499479077618188,
+      "learning_rate": 1.0219210749010162e-05,
+      "loss": 0.2932,
+      "step": 1862
+    },
+    {
+      "epoch": 1.0880227720604336,
+      "grad_norm": 0.09970674460961475,
+      "learning_rate": 1.0209016421489685e-05,
+      "loss": 0.3478,
+      "step": 1863
+    },
+    {
+      "epoch": 1.088606671045909,
+      "grad_norm": 0.09414889824643435,
+      "learning_rate": 1.0198821876650702e-05,
+      "loss": 0.3556,
+      "step": 1864
+    },
+    {
+      "epoch": 1.0891905700313846,
+      "grad_norm": 0.0789813923003969,
+      "learning_rate": 1.0188627125092678e-05,
+      "loss": 0.3053,
+      "step": 1865
+    },
+    {
+      "epoch": 1.08977446901686,
+      "grad_norm": 0.0872935488534086,
+      "learning_rate": 1.0178432177415298e-05,
+      "loss": 0.296,
+      "step": 1866
+    },
+    {
+      "epoch": 1.0903583680023357,
+      "grad_norm": 0.08841991436888534,
+      "learning_rate": 1.0168237044218452e-05,
+      "loss": 0.3348,
+      "step": 1867
+    },
+    {
+      "epoch": 1.090942266987811,
+      "grad_norm": 0.08894846649372935,
+      "learning_rate": 1.0158041736102221e-05,
+      "loss": 0.3093,
+      "step": 1868
+    },
+    {
+      "epoch": 1.0915261659732867,
+      "grad_norm": 0.09263387337502617,
+      "learning_rate": 1.014784626366687e-05,
+      "loss": 0.3422,
+      "step": 1869
+    },
+    {
+      "epoch": 1.092110064958762,
+      "grad_norm": 0.09259206745122081,
+      "learning_rate": 1.0137650637512835e-05,
+      "loss": 0.3345,
+      "step": 1870
+    },
+    {
+      "epoch": 1.0926939639442377,
+      "grad_norm": 0.09684669904616808,
+      "learning_rate": 1.0127454868240702e-05,
+      "loss": 0.3597,
+      "step": 1871
+    },
+    {
+      "epoch": 1.093277862929713,
+      "grad_norm": 0.0940315524073284,
+      "learning_rate": 1.0117258966451224e-05,
+      "loss": 0.3259,
+      "step": 1872
+    },
+    {
+      "epoch": 1.0938617619151887,
+      "grad_norm": 0.0994432088105281,
+      "learning_rate": 1.0107062942745276e-05,
+      "loss": 0.3475,
+      "step": 1873
+    },
+    {
+      "epoch": 1.094445660900664,
+      "grad_norm": 0.08504678089182068,
+      "learning_rate": 1.0096866807723868e-05,
+      "loss": 0.3196,
+      "step": 1874
+    },
+    {
+      "epoch": 1.0950295598861397,
+      "grad_norm": 0.0920914280188783,
+      "learning_rate": 1.0086670571988124e-05,
+      "loss": 0.3016,
+      "step": 1875
+    },
+    {
+      "epoch": 1.095613458871615,
+      "grad_norm": 0.08818180709472363,
+      "learning_rate": 1.0076474246139272e-05,
+      "loss": 0.3386,
+      "step": 1876
+    },
+    {
+      "epoch": 1.0961973578570907,
+      "grad_norm": 0.08640841207114716,
+      "learning_rate": 1.0066277840778626e-05,
+      "loss": 0.2846,
+      "step": 1877
+    },
+    {
+      "epoch": 1.0967812568425663,
+      "grad_norm": 0.09938767875633556,
+      "learning_rate": 1.0056081366507602e-05,
+      "loss": 0.4006,
+      "step": 1878
+    },
+    {
+      "epoch": 1.0973651558280417,
+      "grad_norm": 0.08844541434224838,
+      "learning_rate": 1.0045884833927673e-05,
+      "loss": 0.342,
+      "step": 1879
+    },
+    {
+      "epoch": 1.0979490548135173,
+      "grad_norm": 0.08056941050404398,
+      "learning_rate": 1.0035688253640372e-05,
+      "loss": 0.2669,
+      "step": 1880
+    },
+    {
+      "epoch": 1.0985329537989927,
+      "grad_norm": 0.10175876785807082,
+      "learning_rate": 1.0025491636247287e-05,
+      "loss": 0.3932,
+      "step": 1881
+    },
+    {
+      "epoch": 1.0991168527844684,
+      "grad_norm": 0.09327238101571621,
+      "learning_rate": 1.0015294992350044e-05,
+      "loss": 0.3056,
+      "step": 1882
+    },
+    {
+      "epoch": 1.0997007517699438,
+      "grad_norm": 0.18028170566437413,
+      "learning_rate": 1.0005098332550293e-05,
+      "loss": 0.3072,
+      "step": 1883
+    },
+    {
+      "epoch": 1.1002846507554194,
+      "grad_norm": 0.09070186951488145,
+      "learning_rate": 9.994901667449708e-06,
+      "loss": 0.3265,
+      "step": 1884
+    },
+    {
+      "epoch": 1.1008685497408948,
+      "grad_norm": 0.08978000654495079,
+      "learning_rate": 9.98470500764996e-06,
+      "loss": 0.3558,
+      "step": 1885
+    },
+    {
+      "epoch": 1.1014524487263704,
+      "grad_norm": 0.08428114404996495,
+      "learning_rate": 9.974508363752715e-06,
+      "loss": 0.3199,
+      "step": 1886
+    },
+    {
+      "epoch": 1.1020363477118458,
+      "grad_norm": 0.08862504364072923,
+      "learning_rate": 9.964311746359631e-06,
+      "loss": 0.3293,
+      "step": 1887
+    },
+    {
+      "epoch": 1.1026202466973214,
+      "grad_norm": 0.08429772938298266,
+      "learning_rate": 9.95411516607233e-06,
+      "loss": 0.3396,
+      "step": 1888
+    },
+    {
+      "epoch": 1.1032041456827968,
+      "grad_norm": 0.09045746522829,
+      "learning_rate": 9.943918633492401e-06,
+      "loss": 0.321,
+      "step": 1889
+    },
+    {
+      "epoch": 1.1037880446682724,
+      "grad_norm": 0.09705061227806401,
+      "learning_rate": 9.933722159221375e-06,
+      "loss": 0.3603,
+      "step": 1890
+    },
+    {
+      "epoch": 1.1043719436537478,
+      "grad_norm": 0.102880570894225,
+      "learning_rate": 9.923525753860735e-06,
+      "loss": 0.361,
+      "step": 1891
+    },
+    {
+      "epoch": 1.1049558426392234,
+      "grad_norm": 0.0777586869393865,
+      "learning_rate": 9.91332942801188e-06,
+      "loss": 0.2607,
+      "step": 1892
+    },
+    {
+      "epoch": 1.1055397416246988,
+      "grad_norm": 0.09746982361230387,
+      "learning_rate": 9.903133192276134e-06,
+      "loss": 0.3833,
+      "step": 1893
+    },
+    {
+      "epoch": 1.1061236406101744,
+      "grad_norm": 0.08690288581808088,
+      "learning_rate": 9.892937057254729e-06,
+      "loss": 0.2949,
+      "step": 1894
+    },
+    {
+      "epoch": 1.10670753959565,
+      "grad_norm": 0.08852659644049855,
+      "learning_rate": 9.882741033548781e-06,
+      "loss": 0.3378,
+      "step": 1895
+    },
+    {
+      "epoch": 1.1072914385811254,
+      "grad_norm": 0.08583633378702922,
+      "learning_rate": 9.872545131759301e-06,
+      "loss": 0.3258,
+      "step": 1896
+    },
+    {
+      "epoch": 1.107875337566601,
+      "grad_norm": 0.09112238219634203,
+      "learning_rate": 9.862349362487172e-06,
+      "loss": 0.3469,
+      "step": 1897
+    },
+    {
+      "epoch": 1.1084592365520765,
+      "grad_norm": 0.08483948193741526,
+      "learning_rate": 9.85215373633313e-06,
+      "loss": 0.2981,
+      "step": 1898
+    },
+    {
+      "epoch": 1.109043135537552,
+      "grad_norm": 0.09302903037936554,
+      "learning_rate": 9.841958263897779e-06,
+      "loss": 0.3494,
+      "step": 1899
+    },
+    {
+      "epoch": 1.1096270345230275,
+      "grad_norm": 0.08859663112894897,
+      "learning_rate": 9.831762955781548e-06,
+      "loss": 0.3065,
+      "step": 1900
+    },
+    {
+      "epoch": 1.110210933508503,
+      "grad_norm": 0.09172246095771235,
+      "learning_rate": 9.8215678225847e-06,
+      "loss": 0.351,
+      "step": 1901
+    },
+    {
+      "epoch": 1.1107948324939785,
+      "grad_norm": 0.088054954001908,
+      "learning_rate": 9.811372874907323e-06,
+      "loss": 0.3073,
+      "step": 1902
+    },
+    {
+      "epoch": 1.111378731479454,
+      "grad_norm": 0.08791581711020026,
+      "learning_rate": 9.801178123349298e-06,
+      "loss": 0.339,
+      "step": 1903
+    },
+    {
+      "epoch": 1.1119626304649295,
+      "grad_norm": 0.08640705542477047,
+      "learning_rate": 9.790983578510315e-06,
+      "loss": 0.3601,
+      "step": 1904
+    },
+    {
+      "epoch": 1.112546529450405,
+      "grad_norm": 0.08380502075589352,
+      "learning_rate": 9.780789250989841e-06,
+      "loss": 0.2899,
+      "step": 1905
+    },
+    {
+      "epoch": 1.1131304284358805,
+      "grad_norm": 0.08925971511966083,
+      "learning_rate": 9.77059515138712e-06,
+      "loss": 0.3223,
+      "step": 1906
+    },
+    {
+      "epoch": 1.1137143274213561,
+      "grad_norm": 0.088343586723838,
+      "learning_rate": 9.760401290301164e-06,
+      "loss": 0.3122,
+      "step": 1907
+    },
+    {
+      "epoch": 1.1142982264068315,
+      "grad_norm": 0.08915953446483597,
+      "learning_rate": 9.750207678330726e-06,
+      "loss": 0.3291,
+      "step": 1908
+    },
+    {
+      "epoch": 1.1148821253923071,
+      "grad_norm": 0.08682473639111132,
+      "learning_rate": 9.740014326074308e-06,
+      "loss": 0.3414,
+      "step": 1909
+    },
+    {
+      "epoch": 1.1154660243777825,
+      "grad_norm": 0.09635014073443268,
+      "learning_rate": 9.729821244130142e-06,
+      "loss": 0.3383,
+      "step": 1910
+    },
+    {
+      "epoch": 1.1160499233632581,
+      "grad_norm": 0.0904156955850072,
+      "learning_rate": 9.719628443096175e-06,
+      "loss": 0.3248,
+      "step": 1911
+    },
+    {
+      "epoch": 1.1166338223487338,
+      "grad_norm": 0.1087989679663815,
+      "learning_rate": 9.709435933570068e-06,
+      "loss": 0.3294,
+      "step": 1912
+    },
+    {
+      "epoch": 1.1172177213342092,
+      "grad_norm": 0.08880186264897724,
+      "learning_rate": 9.699243726149168e-06,
+      "loss": 0.3103,
+      "step": 1913
+    },
+    {
+      "epoch": 1.1178016203196848,
+      "grad_norm": 0.092009531549557,
+      "learning_rate": 9.689051831430518e-06,
+      "loss": 0.3378,
+      "step": 1914
+    },
+    {
+      "epoch": 1.1183855193051602,
+      "grad_norm": 0.08619360789218218,
+      "learning_rate": 9.678860260010834e-06,
+      "loss": 0.3177,
+      "step": 1915
+    },
+    {
+      "epoch": 1.1189694182906358,
+      "grad_norm": 0.10041207014821413,
+      "learning_rate": 9.668669022486495e-06,
+      "loss": 0.315,
+      "step": 1916
+    },
+    {
+      "epoch": 1.1195533172761112,
+      "grad_norm": 0.09101334099888968,
+      "learning_rate": 9.658478129453532e-06,
+      "loss": 0.3237,
+      "step": 1917
+    },
+    {
+      "epoch": 1.1201372162615868,
+      "grad_norm": 0.1010590248594578,
+      "learning_rate": 9.648287591507613e-06,
+      "loss": 0.3388,
+      "step": 1918
+    },
+    {
+      "epoch": 1.1207211152470622,
+      "grad_norm": 0.0947771555775023,
+      "learning_rate": 9.638097419244048e-06,
+      "loss": 0.3439,
+      "step": 1919
+    },
+    {
+      "epoch": 1.1213050142325378,
+      "grad_norm": 0.09454938287749626,
+      "learning_rate": 9.627907623257758e-06,
+      "loss": 0.343,
+      "step": 1920
+    },
+    {
+      "epoch": 1.1218889132180132,
+      "grad_norm": 0.0873728463819213,
+      "learning_rate": 9.617718214143279e-06,
+      "loss": 0.3204,
+      "step": 1921
+    },
+    {
+      "epoch": 1.1224728122034888,
+      "grad_norm": 0.0849570869035982,
+      "learning_rate": 9.607529202494739e-06,
+      "loss": 0.2788,
+      "step": 1922
+    },
+    {
+      "epoch": 1.1230567111889642,
+      "grad_norm": 0.0980269660398303,
+      "learning_rate": 9.597340598905851e-06,
+      "loss": 0.3539,
+      "step": 1923
+    },
+    {
+      "epoch": 1.1236406101744398,
+      "grad_norm": 0.08517626542221636,
+      "learning_rate": 9.587152413969915e-06,
+      "loss": 0.3171,
+      "step": 1924
+    },
+    {
+      "epoch": 1.1242245091599152,
+      "grad_norm": 0.08432732122696855,
+      "learning_rate": 9.576964658279783e-06,
+      "loss": 0.3134,
+      "step": 1925
+    },
+    {
+      "epoch": 1.1248084081453908,
+      "grad_norm": 0.09200144477854494,
+      "learning_rate": 9.566777342427867e-06,
+      "loss": 0.3115,
+      "step": 1926
+    },
+    {
+      "epoch": 1.1253923071308662,
+      "grad_norm": 0.09516905753073161,
+      "learning_rate": 9.556590477006123e-06,
+      "loss": 0.3577,
+      "step": 1927
+    },
+    {
+      "epoch": 1.1259762061163419,
+      "grad_norm": 0.08435813330880894,
+      "learning_rate": 9.546404072606038e-06,
+      "loss": 0.3148,
+      "step": 1928
+    },
+    {
+      "epoch": 1.1265601051018175,
+      "grad_norm": 0.08693897975156312,
+      "learning_rate": 9.536218139818615e-06,
+      "loss": 0.3179,
+      "step": 1929
+    },
+    {
+      "epoch": 1.1271440040872929,
+      "grad_norm": 0.08938292194605624,
+      "learning_rate": 9.526032689234374e-06,
+      "loss": 0.356,
+      "step": 1930
+    },
+    {
+      "epoch": 1.1277279030727685,
+      "grad_norm": 0.08975699018166146,
+      "learning_rate": 9.515847731443324e-06,
+      "loss": 0.2824,
+      "step": 1931
+    },
+    {
+      "epoch": 1.1283118020582439,
+      "grad_norm": 0.08677476716375034,
+      "learning_rate": 9.50566327703497e-06,
+      "loss": 0.3313,
+      "step": 1932
+    },
+    {
+      "epoch": 1.1288957010437195,
+      "grad_norm": 0.0934846770218073,
+      "learning_rate": 9.49547933659829e-06,
+      "loss": 0.338,
+      "step": 1933
+    },
+    {
+      "epoch": 1.129479600029195,
+      "grad_norm": 0.09535834071304641,
+      "learning_rate": 9.48529592072173e-06,
+      "loss": 0.3255,
+      "step": 1934
+    },
+    {
+      "epoch": 1.1300634990146705,
+      "grad_norm": 0.08373780308668563,
+      "learning_rate": 9.475113039993188e-06,
+      "loss": 0.3099,
+      "step": 1935
+    },
+    {
+      "epoch": 1.130647398000146,
+      "grad_norm": 0.08880348763938184,
+      "learning_rate": 9.464930705000008e-06,
+      "loss": 0.3284,
+      "step": 1936
+    },
+    {
+      "epoch": 1.1312312969856215,
+      "grad_norm": 0.08726186138351529,
+      "learning_rate": 9.454748926328962e-06,
+      "loss": 0.339,
+      "step": 1937
+    },
+    {
+      "epoch": 1.131815195971097,
+      "grad_norm": 0.09160235853781779,
+      "learning_rate": 9.44456771456625e-06,
+      "loss": 0.331,
+      "step": 1938
+    },
+    {
+      "epoch": 1.1323990949565725,
+      "grad_norm": 0.08892237647615919,
+      "learning_rate": 9.434387080297477e-06,
+      "loss": 0.3051,
+      "step": 1939
+    },
+    {
+      "epoch": 1.132982993942048,
+      "grad_norm": 0.08881787361266907,
+      "learning_rate": 9.424207034107653e-06,
+      "loss": 0.3429,
+      "step": 1940
+    },
+    {
+      "epoch": 1.1335668929275236,
+      "grad_norm": 0.08919889604244581,
+      "learning_rate": 9.41402758658117e-06,
+      "loss": 0.3306,
+      "step": 1941
+    },
+    {
+      "epoch": 1.1341507919129992,
+      "grad_norm": 0.07948538437174345,
+      "learning_rate": 9.403848748301802e-06,
+      "loss": 0.2746,
+      "step": 1942
+    },
+    {
+      "epoch": 1.1347346908984746,
+      "grad_norm": 0.0870705209931933,
+      "learning_rate": 9.39367052985269e-06,
+      "loss": 0.3312,
+      "step": 1943
+    },
+    {
+      "epoch": 1.13531858988395,
+      "grad_norm": 0.08590897395490216,
+      "learning_rate": 9.38349294181633e-06,
+      "loss": 0.2949,
+      "step": 1944
+    },
+    {
+      "epoch": 1.1359024888694256,
+      "grad_norm": 0.08836448430790411,
+      "learning_rate": 9.373315994774558e-06,
+      "loss": 0.2959,
+      "step": 1945
+    },
+    {
+      "epoch": 1.1364863878549012,
+      "grad_norm": 0.09131976053092578,
+      "learning_rate": 9.363139699308552e-06,
+      "loss": 0.3483,
+      "step": 1946
+    },
+    {
+      "epoch": 1.1370702868403766,
+      "grad_norm": 0.09258515169007508,
+      "learning_rate": 9.352964065998801e-06,
+      "loss": 0.3597,
+      "step": 1947
+    },
+    {
+      "epoch": 1.1376541858258522,
+      "grad_norm": 0.09097399967573525,
+      "learning_rate": 9.34278910542512e-06,
+      "loss": 0.3289,
+      "step": 1948
+    },
+    {
+      "epoch": 1.1382380848113276,
+      "grad_norm": 0.0945875751571609,
+      "learning_rate": 9.332614828166612e-06,
+      "loss": 0.3668,
+      "step": 1949
+    },
+    {
+      "epoch": 1.1388219837968032,
+      "grad_norm": 0.09575849200393413,
+      "learning_rate": 9.322441244801678e-06,
+      "loss": 0.3499,
+      "step": 1950
+    },
+    {
+      "epoch": 1.1394058827822786,
+      "grad_norm": 0.08650842200385905,
+      "learning_rate": 9.312268365907989e-06,
+      "loss": 0.3074,
+      "step": 1951
+    },
+    {
+      "epoch": 1.1399897817677542,
+      "grad_norm": 0.08621223810015469,
+      "learning_rate": 9.302096202062492e-06,
+      "loss": 0.306,
+      "step": 1952
+    },
+    {
+      "epoch": 1.1405736807532296,
+      "grad_norm": 0.08595533671801123,
+      "learning_rate": 9.291924763841387e-06,
+      "loss": 0.3096,
+      "step": 1953
+    },
+    {
+      "epoch": 1.1411575797387052,
+      "grad_norm": 0.09426752662234629,
+      "learning_rate": 9.281754061820116e-06,
+      "loss": 0.3725,
+      "step": 1954
+    },
+    {
+      "epoch": 1.1417414787241806,
+      "grad_norm": 0.08643369803889435,
+      "learning_rate": 9.271584106573364e-06,
+      "loss": 0.2948,
+      "step": 1955
+    },
+    {
+      "epoch": 1.1423253777096563,
+      "grad_norm": 0.1005955770004014,
+      "learning_rate": 9.261414908675036e-06,
+      "loss": 0.3372,
+      "step": 1956
+    },
+    {
+      "epoch": 1.1429092766951316,
+      "grad_norm": 0.0862947548551247,
+      "learning_rate": 9.251246478698242e-06,
+      "loss": 0.3059,
+      "step": 1957
+    },
+    {
+      "epoch": 1.1434931756806073,
+      "grad_norm": 0.09184596871807768,
+      "learning_rate": 9.241078827215305e-06,
+      "loss": 0.3452,
+      "step": 1958
+    },
+    {
+      "epoch": 1.1440770746660829,
+      "grad_norm": 0.09852145711878951,
+      "learning_rate": 9.230911964797734e-06,
+      "loss": 0.343,
+      "step": 1959
+    },
+    {
+      "epoch": 1.1446609736515583,
+      "grad_norm": 0.09926248683189054,
+      "learning_rate": 9.22074590201621e-06,
+      "loss": 0.3808,
+      "step": 1960
+    },
+    {
+      "epoch": 1.1452448726370337,
+      "grad_norm": 0.08771971889319342,
+      "learning_rate": 9.210580649440598e-06,
+      "loss": 0.3014,
+      "step": 1961
+    },
+    {
+      "epoch": 1.1458287716225093,
+      "grad_norm": 0.08949775815978268,
+      "learning_rate": 9.200416217639906e-06,
+      "loss": 0.3205,
+      "step": 1962
+    },
+    {
+      "epoch": 1.146412670607985,
+      "grad_norm": 0.08896033123747935,
+      "learning_rate": 9.190252617182301e-06,
+      "loss": 0.288,
+      "step": 1963
+    },
+    {
+      "epoch": 1.1469965695934603,
+      "grad_norm": 0.08896183681066128,
+      "learning_rate": 9.180089858635075e-06,
+      "loss": 0.3235,
+      "step": 1964
+    },
+    {
+      "epoch": 1.147580468578936,
+      "grad_norm": 0.09747608961742624,
+      "learning_rate": 9.169927952564649e-06,
+      "loss": 0.3338,
+      "step": 1965
+    },
+    {
+      "epoch": 1.1481643675644113,
+      "grad_norm": 0.09298165503788856,
+      "learning_rate": 9.159766909536559e-06,
+      "loss": 0.3547,
+      "step": 1966
+    },
+    {
+      "epoch": 1.148748266549887,
+      "grad_norm": 0.0908188860401638,
+      "learning_rate": 9.149606740115444e-06,
+      "loss": 0.3343,
+      "step": 1967
+    },
+    {
+      "epoch": 1.1493321655353623,
+      "grad_norm": 0.0861836945900596,
+      "learning_rate": 9.139447454865034e-06,
+      "loss": 0.3102,
+      "step": 1968
+    },
+    {
+      "epoch": 1.149916064520838,
+      "grad_norm": 0.09029540639823912,
+      "learning_rate": 9.129289064348135e-06,
+      "loss": 0.2992,
+      "step": 1969
+    },
+    {
+      "epoch": 1.1504999635063133,
+      "grad_norm": 0.09995011046150573,
+      "learning_rate": 9.119131579126628e-06,
+      "loss": 0.3346,
+      "step": 1970
+    },
+    {
+      "epoch": 1.151083862491789,
+      "grad_norm": 0.09158793378418532,
+      "learning_rate": 9.108975009761452e-06,
+      "loss": 0.3533,
+      "step": 1971
+    },
+    {
+      "epoch": 1.1516677614772644,
+      "grad_norm": 0.09125663055414374,
+      "learning_rate": 9.098819366812594e-06,
+      "loss": 0.3017,
+      "step": 1972
+    },
+    {
+      "epoch": 1.15225166046274,
+      "grad_norm": 0.08938165001533294,
+      "learning_rate": 9.088664660839078e-06,
+      "loss": 0.3505,
+      "step": 1973
+    },
+    {
+      "epoch": 1.1528355594482154,
+      "grad_norm": 0.08267540275062915,
+      "learning_rate": 9.078510902398948e-06,
+      "loss": 0.2995,
+      "step": 1974
+    },
+    {
+      "epoch": 1.153419458433691,
+      "grad_norm": 0.08993220574898021,
+      "learning_rate": 9.068358102049272e-06,
+      "loss": 0.3088,
+      "step": 1975
+    },
+    {
+      "epoch": 1.1540033574191666,
+      "grad_norm": 0.09004732078123168,
+      "learning_rate": 9.058206270346115e-06,
+      "loss": 0.3262,
+      "step": 1976
+    },
+    {
+      "epoch": 1.154587256404642,
+      "grad_norm": 0.08912931620337301,
+      "learning_rate": 9.04805541784454e-06,
+      "loss": 0.3281,
+      "step": 1977
+    },
+    {
+      "epoch": 1.1551711553901174,
+      "grad_norm": 0.08746820547571517,
+      "learning_rate": 9.037905555098589e-06,
+      "loss": 0.3119,
+      "step": 1978
+    },
+    {
+      "epoch": 1.155755054375593,
+      "grad_norm": 0.08608907764987926,
+      "learning_rate": 9.027756692661272e-06,
+      "loss": 0.2901,
+      "step": 1979
+    },
+    {
+      "epoch": 1.1563389533610686,
+      "grad_norm": 0.09482233806402172,
+      "learning_rate": 9.017608841084564e-06,
+      "loss": 0.334,
+      "step": 1980
+    },
+    {
+      "epoch": 1.156922852346544,
+      "grad_norm": 0.08534555546619706,
+      "learning_rate": 9.007462010919387e-06,
+      "loss": 0.3035,
+      "step": 1981
+    },
+    {
+      "epoch": 1.1575067513320196,
+      "grad_norm": 0.10500178545217707,
+      "learning_rate": 8.997316212715599e-06,
+      "loss": 0.3699,
+      "step": 1982
+    },
+    {
+      "epoch": 1.158090650317495,
+      "grad_norm": 0.1074787274537364,
+      "learning_rate": 8.987171457021992e-06,
+      "loss": 0.3691,
+      "step": 1983
+    },
+    {
+      "epoch": 1.1586745493029706,
+      "grad_norm": 0.09544275650760421,
+      "learning_rate": 8.977027754386267e-06,
+      "loss": 0.3195,
+      "step": 1984
+    },
+    {
+      "epoch": 1.159258448288446,
+      "grad_norm": 0.0944451732492147,
+      "learning_rate": 8.966885115355033e-06,
+      "loss": 0.3462,
+      "step": 1985
+    },
+    {
+      "epoch": 1.1598423472739217,
+      "grad_norm": 0.08598765088078299,
+      "learning_rate": 8.95674355047379e-06,
+      "loss": 0.3097,
+      "step": 1986
+    },
+    {
+      "epoch": 1.160426246259397,
+      "grad_norm": 0.09161341391249907,
+      "learning_rate": 8.946603070286926e-06,
+      "loss": 0.3315,
+      "step": 1987
+    },
+    {
+      "epoch": 1.1610101452448727,
+      "grad_norm": 0.08770138003898323,
+      "learning_rate": 8.936463685337697e-06,
+      "loss": 0.3011,
+      "step": 1988
+    },
+    {
+      "epoch": 1.161594044230348,
+      "grad_norm": 0.09175396798451009,
+      "learning_rate": 8.926325406168225e-06,
+      "loss": 0.3246,
+      "step": 1989
+    },
+    {
+      "epoch": 1.1621779432158237,
+      "grad_norm": 0.0902096185320303,
+      "learning_rate": 8.91618824331948e-06,
+      "loss": 0.3149,
+      "step": 1990
+    },
+    {
+      "epoch": 1.162761842201299,
+      "grad_norm": 0.09484902000145408,
+      "learning_rate": 8.906052207331268e-06,
+      "loss": 0.3517,
+      "step": 1991
+    },
+    {
+      "epoch": 1.1633457411867747,
+      "grad_norm": 0.08907618867840278,
+      "learning_rate": 8.895917308742224e-06,
+      "loss": 0.3083,
+      "step": 1992
+    },
+    {
+      "epoch": 1.1639296401722503,
+      "grad_norm": 0.09317831007559728,
+      "learning_rate": 8.88578355808981e-06,
+      "loss": 0.3298,
+      "step": 1993
+    },
+    {
+      "epoch": 1.1645135391577257,
+      "grad_norm": 0.09599714098106968,
+      "learning_rate": 8.87565096591028e-06,
+      "loss": 0.3541,
+      "step": 1994
+    },
+    {
+      "epoch": 1.165097438143201,
+      "grad_norm": 0.08955981699811257,
+      "learning_rate": 8.865519542738696e-06,
+      "loss": 0.327,
+      "step": 1995
+    },
+    {
+      "epoch": 1.1656813371286767,
+      "grad_norm": 0.08982999411151038,
+      "learning_rate": 8.855389299108894e-06,
+      "loss": 0.3612,
+      "step": 1996
+    },
+    {
+      "epoch": 1.1662652361141523,
+      "grad_norm": 0.0833838384495267,
+      "learning_rate": 8.845260245553493e-06,
+      "loss": 0.278,
+      "step": 1997
+    },
+    {
+      "epoch": 1.1668491350996277,
+      "grad_norm": 0.08385010237826855,
+      "learning_rate": 8.83513239260387e-06,
+      "loss": 0.2894,
+      "step": 1998
+    },
+    {
+      "epoch": 1.1674330340851033,
+      "grad_norm": 0.08621799342958701,
+      "learning_rate": 8.82500575079015e-06,
+      "loss": 0.3427,
+      "step": 1999
+    },
+    {
+      "epoch": 1.1680169330705787,
+      "grad_norm": 0.0890684291496028,
+      "learning_rate": 8.81488033064121e-06,
+      "loss": 0.2918,
+      "step": 2000
+    },
+    {
+      "epoch": 1.1686008320560544,
+      "grad_norm": 0.0871124778468499,
+      "learning_rate": 8.804756142684644e-06,
+      "loss": 0.3254,
+      "step": 2001
+    },
+    {
+      "epoch": 1.1691847310415298,
+      "grad_norm": 0.10411271786364035,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.3552,
+      "step": 2002
+    },
+    {
+      "epoch": 1.1697686300270054,
+      "grad_norm": 0.09702612357261878,
+      "learning_rate": 8.78451150545262e-06,
+      "loss": 0.3397,
+      "step": 2003
+    },
+    {
+      "epoch": 1.1703525290124808,
+      "grad_norm": 0.0816762402267171,
+      "learning_rate": 8.774391077225914e-06,
+      "loss": 0.2791,
+      "step": 2004
+    },
+    {
+      "epoch": 1.1709364279979564,
+      "grad_norm": 0.08852720569172234,
+      "learning_rate": 8.764271923289064e-06,
+      "loss": 0.3353,
+      "step": 2005
+    },
+    {
+      "epoch": 1.1715203269834318,
+      "grad_norm": 0.08696167334803934,
+      "learning_rate": 8.754154054163148e-06,
+      "loss": 0.3086,
+      "step": 2006
+    },
+    {
+      "epoch": 1.1721042259689074,
+      "grad_norm": 0.09272839269144532,
+      "learning_rate": 8.744037480367922e-06,
+      "loss": 0.3225,
+      "step": 2007
+    },
+    {
+      "epoch": 1.1726881249543828,
+      "grad_norm": 0.10517543683672191,
+      "learning_rate": 8.733922212421785e-06,
+      "loss": 0.3249,
+      "step": 2008
+    },
+    {
+      "epoch": 1.1732720239398584,
+      "grad_norm": 0.09356460631582644,
+      "learning_rate": 8.723808260841781e-06,
+      "loss": 0.3584,
+      "step": 2009
+    },
+    {
+      "epoch": 1.173855922925334,
+      "grad_norm": 0.08990121085249254,
+      "learning_rate": 8.713695636143584e-06,
+      "loss": 0.3163,
+      "step": 2010
+    },
+    {
+      "epoch": 1.1744398219108094,
+      "grad_norm": 0.08147032437566584,
+      "learning_rate": 8.703584348841494e-06,
+      "loss": 0.3029,
+      "step": 2011
+    },
+    {
+      "epoch": 1.1750237208962848,
+      "grad_norm": 0.09078659754900648,
+      "learning_rate": 8.693474409448416e-06,
+      "loss": 0.3231,
+      "step": 2012
+    },
+    {
+      "epoch": 1.1756076198817604,
+      "grad_norm": 0.09035001265413253,
+      "learning_rate": 8.683365828475855e-06,
+      "loss": 0.3318,
+      "step": 2013
+    },
+    {
+      "epoch": 1.176191518867236,
+      "grad_norm": 0.08529127043629561,
+      "learning_rate": 8.673258616433898e-06,
+      "loss": 0.3033,
+      "step": 2014
+    },
+    {
+      "epoch": 1.1767754178527114,
+      "grad_norm": 0.09459634006667002,
+      "learning_rate": 8.663152783831215e-06,
+      "loss": 0.3464,
+      "step": 2015
+    },
+    {
+      "epoch": 1.177359316838187,
+      "grad_norm": 0.08292448764532521,
+      "learning_rate": 8.653048341175044e-06,
+      "loss": 0.2998,
+      "step": 2016
+    },
+    {
+      "epoch": 1.1779432158236625,
+      "grad_norm": 0.08792986996834888,
+      "learning_rate": 8.642945298971168e-06,
+      "loss": 0.3299,
+      "step": 2017
+    },
+    {
+      "epoch": 1.178527114809138,
+      "grad_norm": 0.08479884200098901,
+      "learning_rate": 8.632843667723927e-06,
+      "loss": 0.3035,
+      "step": 2018
+    },
+    {
+      "epoch": 1.1791110137946135,
+      "grad_norm": 0.08633896624455162,
+      "learning_rate": 8.62274345793618e-06,
+      "loss": 0.3379,
+      "step": 2019
+    },
+    {
+      "epoch": 1.179694912780089,
+      "grad_norm": 0.08525196975812484,
+      "learning_rate": 8.61264468010932e-06,
+      "loss": 0.3157,
+      "step": 2020
+    },
+    {
+      "epoch": 1.1802788117655645,
+      "grad_norm": 0.08699284053455525,
+      "learning_rate": 8.602547344743241e-06,
+      "loss": 0.3533,
+      "step": 2021
+    },
+    {
+      "epoch": 1.18086271075104,
+      "grad_norm": 0.07996307706711965,
+      "learning_rate": 8.592451462336348e-06,
+      "loss": 0.2964,
+      "step": 2022
+    },
+    {
+      "epoch": 1.1814466097365155,
+      "grad_norm": 0.0882197149990018,
+      "learning_rate": 8.582357043385529e-06,
+      "loss": 0.3125,
+      "step": 2023
+    },
+    {
+      "epoch": 1.1820305087219911,
+      "grad_norm": 0.09997463755802412,
+      "learning_rate": 8.572264098386149e-06,
+      "loss": 0.3728,
+      "step": 2024
+    },
+    {
+      "epoch": 1.1826144077074665,
+      "grad_norm": 0.09659845585201246,
+      "learning_rate": 8.562172637832041e-06,
+      "loss": 0.3304,
+      "step": 2025
+    },
+    {
+      "epoch": 1.1831983066929421,
+      "grad_norm": 0.08720306201454747,
+      "learning_rate": 8.5520826722155e-06,
+      "loss": 0.3071,
+      "step": 2026
+    },
+    {
+      "epoch": 1.1837822056784177,
+      "grad_norm": 0.10244863345228425,
+      "learning_rate": 8.54199421202726e-06,
+      "loss": 0.3575,
+      "step": 2027
+    },
+    {
+      "epoch": 1.1843661046638931,
+      "grad_norm": 0.0916847889891718,
+      "learning_rate": 8.531907267756498e-06,
+      "loss": 0.3157,
+      "step": 2028
+    },
+    {
+      "epoch": 1.1849500036493688,
+      "grad_norm": 0.0933475825740031,
+      "learning_rate": 8.521821849890802e-06,
+      "loss": 0.3475,
+      "step": 2029
+    },
+    {
+      "epoch": 1.1855339026348442,
+      "grad_norm": 0.09339275535460952,
+      "learning_rate": 8.511737968916185e-06,
+      "loss": 0.3636,
+      "step": 2030
+    },
+    {
+      "epoch": 1.1861178016203198,
+      "grad_norm": 0.09387278848695595,
+      "learning_rate": 8.50165563531706e-06,
+      "loss": 0.349,
+      "step": 2031
+    },
+    {
+      "epoch": 1.1867017006057952,
+      "grad_norm": 0.08262725254447023,
+      "learning_rate": 8.491574859576222e-06,
+      "loss": 0.3024,
+      "step": 2032
+    },
+    {
+      "epoch": 1.1872855995912708,
+      "grad_norm": 0.08911417916831633,
+      "learning_rate": 8.481495652174859e-06,
+      "loss": 0.3152,
+      "step": 2033
+    },
+    {
+      "epoch": 1.1878694985767462,
+      "grad_norm": 0.0893778079209687,
+      "learning_rate": 8.47141802359252e-06,
+      "loss": 0.3081,
+      "step": 2034
+    },
+    {
+      "epoch": 1.1884533975622218,
+      "grad_norm": 0.08454948743519716,
+      "learning_rate": 8.461341984307115e-06,
+      "loss": 0.3067,
+      "step": 2035
+    },
+    {
+      "epoch": 1.1890372965476972,
+      "grad_norm": 0.0842897617397049,
+      "learning_rate": 8.4512675447949e-06,
+      "loss": 0.3069,
+      "step": 2036
+    },
+    {
+      "epoch": 1.1896211955331728,
+      "grad_norm": 0.08650331743328575,
+      "learning_rate": 8.441194715530472e-06,
+      "loss": 0.3269,
+      "step": 2037
+    },
+    {
+      "epoch": 1.1902050945186482,
+      "grad_norm": 0.08193208871471183,
+      "learning_rate": 8.431123506986747e-06,
+      "loss": 0.3127,
+      "step": 2038
+    },
+    {
+      "epoch": 1.1907889935041238,
+      "grad_norm": 0.0862561662582518,
+      "learning_rate": 8.421053929634966e-06,
+      "loss": 0.3277,
+      "step": 2039
+    },
+    {
+      "epoch": 1.1913728924895992,
+      "grad_norm": 0.089404984535366,
+      "learning_rate": 8.410985993944663e-06,
+      "loss": 0.3187,
+      "step": 2040
+    },
+    {
+      "epoch": 1.1919567914750748,
+      "grad_norm": 0.09118742063959628,
+      "learning_rate": 8.400919710383673e-06,
+      "loss": 0.3427,
+      "step": 2041
+    },
+    {
+      "epoch": 1.1925406904605502,
+      "grad_norm": 0.0821731713961786,
+      "learning_rate": 8.390855089418109e-06,
+      "loss": 0.3065,
+      "step": 2042
+    },
+    {
+      "epoch": 1.1931245894460258,
+      "grad_norm": 0.08277597815631672,
+      "learning_rate": 8.380792141512355e-06,
+      "loss": 0.3051,
+      "step": 2043
+    },
+    {
+      "epoch": 1.1937084884315015,
+      "grad_norm": 0.08150736873982867,
+      "learning_rate": 8.37073087712906e-06,
+      "loss": 0.2999,
+      "step": 2044
+    },
+    {
+      "epoch": 1.1942923874169769,
+      "grad_norm": 0.08491777649147374,
+      "learning_rate": 8.360671306729114e-06,
+      "loss": 0.298,
+      "step": 2045
+    },
+    {
+      "epoch": 1.1948762864024525,
+      "grad_norm": 0.09699847205456576,
+      "learning_rate": 8.350613440771661e-06,
+      "loss": 0.2981,
+      "step": 2046
+    },
+    {
+      "epoch": 1.1954601853879279,
+      "grad_norm": 0.08599379261928622,
+      "learning_rate": 8.340557289714055e-06,
+      "loss": 0.2987,
+      "step": 2047
+    },
+    {
+      "epoch": 1.1960440843734035,
+      "grad_norm": 0.09341099824999334,
+      "learning_rate": 8.330502864011878e-06,
+      "loss": 0.3121,
+      "step": 2048
+    },
+    {
+      "epoch": 1.1966279833588789,
+      "grad_norm": 0.09116276211263707,
+      "learning_rate": 8.320450174118914e-06,
+      "loss": 0.2951,
+      "step": 2049
+    },
+    {
+      "epoch": 1.1972118823443545,
+      "grad_norm": 0.09444967046674042,
+      "learning_rate": 8.310399230487148e-06,
+      "loss": 0.3482,
+      "step": 2050
+    },
+    {
+      "epoch": 1.19779578132983,
+      "grad_norm": 0.08757136262031544,
+      "learning_rate": 8.300350043566742e-06,
+      "loss": 0.3311,
+      "step": 2051
+    },
+    {
+      "epoch": 1.1983796803153055,
+      "grad_norm": 0.08925112389614931,
+      "learning_rate": 8.290302623806035e-06,
+      "loss": 0.2911,
+      "step": 2052
+    },
+    {
+      "epoch": 1.198963579300781,
+      "grad_norm": 0.08680916205060658,
+      "learning_rate": 8.280256981651527e-06,
+      "loss": 0.3301,
+      "step": 2053
+    },
+    {
+      "epoch": 1.1995474782862565,
+      "grad_norm": 0.08902411311623801,
+      "learning_rate": 8.270213127547871e-06,
+      "loss": 0.3336,
+      "step": 2054
+    },
+    {
+      "epoch": 1.200131377271732,
+      "grad_norm": 0.09108679234205203,
+      "learning_rate": 8.260171071937863e-06,
+      "loss": 0.3263,
+      "step": 2055
+    },
+    {
+      "epoch": 1.2007152762572075,
+      "grad_norm": 0.0882178501612453,
+      "learning_rate": 8.250130825262426e-06,
+      "loss": 0.3036,
+      "step": 2056
+    },
+    {
+      "epoch": 1.201299175242683,
+      "grad_norm": 0.0952015044222706,
+      "learning_rate": 8.240092397960601e-06,
+      "loss": 0.3284,
+      "step": 2057
+    },
+    {
+      "epoch": 1.2018830742281585,
+      "grad_norm": 0.09098292945835666,
+      "learning_rate": 8.230055800469543e-06,
+      "loss": 0.3234,
+      "step": 2058
+    },
+    {
+      "epoch": 1.202466973213634,
+      "grad_norm": 0.0815057667238275,
+      "learning_rate": 8.2200210432245e-06,
+      "loss": 0.2751,
+      "step": 2059
+    },
+    {
+      "epoch": 1.2030508721991096,
+      "grad_norm": 0.08865728041620005,
+      "learning_rate": 8.209988136658805e-06,
+      "loss": 0.3071,
+      "step": 2060
+    },
+    {
+      "epoch": 1.2036347711845852,
+      "grad_norm": 0.0880921071349412,
+      "learning_rate": 8.199957091203876e-06,
+      "loss": 0.3011,
+      "step": 2061
+    },
+    {
+      "epoch": 1.2042186701700606,
+      "grad_norm": 0.09074699704560706,
+      "learning_rate": 8.189927917289182e-06,
+      "loss": 0.3127,
+      "step": 2062
+    },
+    {
+      "epoch": 1.2048025691555362,
+      "grad_norm": 0.09685631331636614,
+      "learning_rate": 8.179900625342256e-06,
+      "loss": 0.3601,
+      "step": 2063
+    },
+    {
+      "epoch": 1.2053864681410116,
+      "grad_norm": 0.08804293000683427,
+      "learning_rate": 8.169875225788675e-06,
+      "loss": 0.3423,
+      "step": 2064
+    },
+    {
+      "epoch": 1.2059703671264872,
+      "grad_norm": 0.08113571715821154,
+      "learning_rate": 8.159851729052041e-06,
+      "loss": 0.2956,
+      "step": 2065
+    },
+    {
+      "epoch": 1.2065542661119626,
+      "grad_norm": 0.0901854920301078,
+      "learning_rate": 8.149830145553982e-06,
+      "loss": 0.37,
+      "step": 2066
+    },
+    {
+      "epoch": 1.2071381650974382,
+      "grad_norm": 0.08999821622843408,
+      "learning_rate": 8.139810485714142e-06,
+      "loss": 0.3181,
+      "step": 2067
+    },
+    {
+      "epoch": 1.2077220640829136,
+      "grad_norm": 0.08966681654957225,
+      "learning_rate": 8.129792759950157e-06,
+      "loss": 0.3552,
+      "step": 2068
+    },
+    {
+      "epoch": 1.2083059630683892,
+      "grad_norm": 0.08668530846923038,
+      "learning_rate": 8.119776978677655e-06,
+      "loss": 0.3119,
+      "step": 2069
+    },
+    {
+      "epoch": 1.2088898620538646,
+      "grad_norm": 0.09465133974139303,
+      "learning_rate": 8.10976315231024e-06,
+      "loss": 0.3707,
+      "step": 2070
+    },
+    {
+      "epoch": 1.2094737610393402,
+      "grad_norm": 0.09301162805693843,
+      "learning_rate": 8.099751291259485e-06,
+      "loss": 0.3321,
+      "step": 2071
+    },
+    {
+      "epoch": 1.2100576600248156,
+      "grad_norm": 0.08265490337305426,
+      "learning_rate": 8.089741405934923e-06,
+      "loss": 0.2787,
+      "step": 2072
+    },
+    {
+      "epoch": 1.2106415590102912,
+      "grad_norm": 0.09278192088044818,
+      "learning_rate": 8.079733506744027e-06,
+      "loss": 0.3304,
+      "step": 2073
+    },
+    {
+      "epoch": 1.2112254579957666,
+      "grad_norm": 0.09516501175769654,
+      "learning_rate": 8.069727604092213e-06,
+      "loss": 0.3212,
+      "step": 2074
+    },
+    {
+      "epoch": 1.2118093569812423,
+      "grad_norm": 0.08314460361310731,
+      "learning_rate": 8.05972370838281e-06,
+      "loss": 0.2994,
+      "step": 2075
+    },
+    {
+      "epoch": 1.2123932559667177,
+      "grad_norm": 0.09130339189427457,
+      "learning_rate": 8.04972183001707e-06,
+      "loss": 0.2919,
+      "step": 2076
+    },
+    {
+      "epoch": 1.2129771549521933,
+      "grad_norm": 0.0929061164299365,
+      "learning_rate": 8.03972197939414e-06,
+      "loss": 0.3408,
+      "step": 2077
+    },
+    {
+      "epoch": 1.2135610539376689,
+      "grad_norm": 0.09351545195498219,
+      "learning_rate": 8.029724166911069e-06,
+      "loss": 0.3747,
+      "step": 2078
+    },
+    {
+      "epoch": 1.2141449529231443,
+      "grad_norm": 0.08235993512704419,
+      "learning_rate": 8.019728402962776e-06,
+      "loss": 0.2896,
+      "step": 2079
+    },
+    {
+      "epoch": 1.21472885190862,
+      "grad_norm": 0.09341856674537213,
+      "learning_rate": 8.009734697942054e-06,
+      "loss": 0.3375,
+      "step": 2080
+    },
+    {
+      "epoch": 1.2153127508940953,
+      "grad_norm": 0.08515847717506528,
+      "learning_rate": 7.999743062239557e-06,
+      "loss": 0.2978,
+      "step": 2081
+    },
+    {
+      "epoch": 1.215896649879571,
+      "grad_norm": 0.08847765298551571,
+      "learning_rate": 7.989753506243787e-06,
+      "loss": 0.3233,
+      "step": 2082
+    },
+    {
+      "epoch": 1.2164805488650463,
+      "grad_norm": 0.08224536670740276,
+      "learning_rate": 7.979766040341084e-06,
+      "loss": 0.3049,
+      "step": 2083
+    },
+    {
+      "epoch": 1.217064447850522,
+      "grad_norm": 0.08110200884370508,
+      "learning_rate": 7.969780674915613e-06,
+      "loss": 0.3349,
+      "step": 2084
+    },
+    {
+      "epoch": 1.2176483468359973,
+      "grad_norm": 0.08346171488615804,
+      "learning_rate": 7.959797420349356e-06,
+      "loss": 0.3034,
+      "step": 2085
+    },
+    {
+      "epoch": 1.218232245821473,
+      "grad_norm": 0.09698501058006724,
+      "learning_rate": 7.949816287022098e-06,
+      "loss": 0.3293,
+      "step": 2086
+    },
+    {
+      "epoch": 1.2188161448069483,
+      "grad_norm": 0.10202568606531373,
+      "learning_rate": 7.939837285311425e-06,
+      "loss": 0.3848,
+      "step": 2087
+    },
+    {
+      "epoch": 1.219400043792424,
+      "grad_norm": 0.09004069923623871,
+      "learning_rate": 7.9298604255927e-06,
+      "loss": 0.351,
+      "step": 2088
+    },
+    {
+      "epoch": 1.2199839427778993,
+      "grad_norm": 0.08443762489989477,
+      "learning_rate": 7.919885718239063e-06,
+      "loss": 0.324,
+      "step": 2089
+    },
+    {
+      "epoch": 1.220567841763375,
+      "grad_norm": 0.09788630197896316,
+      "learning_rate": 7.909913173621413e-06,
+      "loss": 0.3682,
+      "step": 2090
+    },
+    {
+      "epoch": 1.2211517407488504,
+      "grad_norm": 0.09220981868213748,
+      "learning_rate": 7.899942802108402e-06,
+      "loss": 0.3445,
+      "step": 2091
+    },
+    {
+      "epoch": 1.221735639734326,
+      "grad_norm": 0.08948282593130057,
+      "learning_rate": 7.889974614066425e-06,
+      "loss": 0.3168,
+      "step": 2092
+    },
+    {
+      "epoch": 1.2223195387198014,
+      "grad_norm": 0.08586562532834194,
+      "learning_rate": 7.880008619859601e-06,
+      "loss": 0.3339,
+      "step": 2093
+    },
+    {
+      "epoch": 1.222903437705277,
+      "grad_norm": 0.09131654767194591,
+      "learning_rate": 7.870044829849772e-06,
+      "loss": 0.3095,
+      "step": 2094
+    },
+    {
+      "epoch": 1.2234873366907526,
+      "grad_norm": 0.0916914604581565,
+      "learning_rate": 7.860083254396491e-06,
+      "loss": 0.3225,
+      "step": 2095
+    },
+    {
+      "epoch": 1.224071235676228,
+      "grad_norm": 0.0958971636675563,
+      "learning_rate": 7.850123903857004e-06,
+      "loss": 0.3485,
+      "step": 2096
+    },
+    {
+      "epoch": 1.2246551346617036,
+      "grad_norm": 0.08756828574516562,
+      "learning_rate": 7.840166788586244e-06,
+      "loss": 0.3033,
+      "step": 2097
+    },
+    {
+      "epoch": 1.225239033647179,
+      "grad_norm": 0.08404617221924698,
+      "learning_rate": 7.83021191893682e-06,
+      "loss": 0.3263,
+      "step": 2098
+    },
+    {
+      "epoch": 1.2258229326326546,
+      "grad_norm": 0.08718485786899144,
+      "learning_rate": 7.820259305259009e-06,
+      "loss": 0.2947,
+      "step": 2099
+    },
+    {
+      "epoch": 1.22640683161813,
+      "grad_norm": 0.08810520476634816,
+      "learning_rate": 7.810308957900736e-06,
+      "loss": 0.3162,
+      "step": 2100
+    },
+    {
+      "epoch": 1.2269907306036056,
+      "grad_norm": 0.09501032308937611,
+      "learning_rate": 7.800360887207579e-06,
+      "loss": 0.3606,
+      "step": 2101
+    },
+    {
+      "epoch": 1.227574629589081,
+      "grad_norm": 0.09231313444341066,
+      "learning_rate": 7.790415103522744e-06,
+      "loss": 0.3267,
+      "step": 2102
+    },
+    {
+      "epoch": 1.2281585285745567,
+      "grad_norm": 0.0875392890457479,
+      "learning_rate": 7.780471617187056e-06,
+      "loss": 0.3375,
+      "step": 2103
+    },
+    {
+      "epoch": 1.228742427560032,
+      "grad_norm": 0.08192442544224701,
+      "learning_rate": 7.770530438538955e-06,
+      "loss": 0.2652,
+      "step": 2104
+    },
+    {
+      "epoch": 1.2293263265455077,
+      "grad_norm": 0.09569492854110273,
+      "learning_rate": 7.760591577914483e-06,
+      "loss": 0.3425,
+      "step": 2105
+    },
+    {
+      "epoch": 1.229910225530983,
+      "grad_norm": 0.09759206994871525,
+      "learning_rate": 7.750655045647268e-06,
+      "loss": 0.3106,
+      "step": 2106
+    },
+    {
+      "epoch": 1.2304941245164587,
+      "grad_norm": 0.10215416865492356,
+      "learning_rate": 7.740720852068524e-06,
+      "loss": 0.3694,
+      "step": 2107
+    },
+    {
+      "epoch": 1.231078023501934,
+      "grad_norm": 0.09044918736731351,
+      "learning_rate": 7.730789007507023e-06,
+      "loss": 0.3052,
+      "step": 2108
+    },
+    {
+      "epoch": 1.2316619224874097,
+      "grad_norm": 0.09648906954991572,
+      "learning_rate": 7.720859522289101e-06,
+      "loss": 0.3521,
+      "step": 2109
+    },
+    {
+      "epoch": 1.232245821472885,
+      "grad_norm": 0.09156966111680312,
+      "learning_rate": 7.710932406738643e-06,
+      "loss": 0.3111,
+      "step": 2110
+    },
+    {
+      "epoch": 1.2328297204583607,
+      "grad_norm": 0.08804907127190996,
+      "learning_rate": 7.701007671177066e-06,
+      "loss": 0.3109,
+      "step": 2111
+    },
+    {
+      "epoch": 1.2334136194438363,
+      "grad_norm": 0.0902333054552117,
+      "learning_rate": 7.691085325923317e-06,
+      "loss": 0.3402,
+      "step": 2112
+    },
+    {
+      "epoch": 1.2339975184293117,
+      "grad_norm": 0.08609652002605185,
+      "learning_rate": 7.68116538129385e-06,
+      "loss": 0.3154,
+      "step": 2113
+    },
+    {
+      "epoch": 1.2345814174147873,
+      "grad_norm": 0.09054624440504581,
+      "learning_rate": 7.671247847602628e-06,
+      "loss": 0.3238,
+      "step": 2114
+    },
+    {
+      "epoch": 1.2351653164002627,
+      "grad_norm": 0.08763567955584771,
+      "learning_rate": 7.661332735161111e-06,
+      "loss": 0.2911,
+      "step": 2115
+    },
+    {
+      "epoch": 1.2357492153857383,
+      "grad_norm": 0.09352958944925079,
+      "learning_rate": 7.651420054278234e-06,
+      "loss": 0.3317,
+      "step": 2116
+    },
+    {
+      "epoch": 1.2363331143712137,
+      "grad_norm": 0.08234809708541699,
+      "learning_rate": 7.641509815260412e-06,
+      "loss": 0.3142,
+      "step": 2117
+    },
+    {
+      "epoch": 1.2369170133566894,
+      "grad_norm": 0.09316405849329999,
+      "learning_rate": 7.631602028411512e-06,
+      "loss": 0.3631,
+      "step": 2118
+    },
+    {
+      "epoch": 1.2375009123421647,
+      "grad_norm": 0.08801808333169837,
+      "learning_rate": 7.621696704032857e-06,
+      "loss": 0.3183,
+      "step": 2119
+    },
+    {
+      "epoch": 1.2380848113276404,
+      "grad_norm": 0.082997131104761,
+      "learning_rate": 7.6117938524232105e-06,
+      "loss": 0.2669,
+      "step": 2120
+    },
+    {
+      "epoch": 1.2386687103131158,
+      "grad_norm": 0.08743670253483388,
+      "learning_rate": 7.601893483878761e-06,
+      "loss": 0.3247,
+      "step": 2121
+    },
+    {
+      "epoch": 1.2392526092985914,
+      "grad_norm": 0.0944002611638054,
+      "learning_rate": 7.591995608693118e-06,
+      "loss": 0.3259,
+      "step": 2122
+    },
+    {
+      "epoch": 1.2398365082840668,
+      "grad_norm": 0.09175840391356292,
+      "learning_rate": 7.5821002371573005e-06,
+      "loss": 0.3021,
+      "step": 2123
+    },
+    {
+      "epoch": 1.2404204072695424,
+      "grad_norm": 0.09100843909278548,
+      "learning_rate": 7.572207379559722e-06,
+      "loss": 0.353,
+      "step": 2124
+    },
+    {
+      "epoch": 1.2410043062550178,
+      "grad_norm": 0.08682926177985877,
+      "learning_rate": 7.562317046186182e-06,
+      "loss": 0.2936,
+      "step": 2125
+    },
+    {
+      "epoch": 1.2415882052404934,
+      "grad_norm": 0.08623572803821665,
+      "learning_rate": 7.552429247319854e-06,
+      "loss": 0.2906,
+      "step": 2126
+    },
+    {
+      "epoch": 1.2421721042259688,
+      "grad_norm": 0.08841725648588566,
+      "learning_rate": 7.542543993241278e-06,
+      "loss": 0.3278,
+      "step": 2127
+    },
+    {
+      "epoch": 1.2427560032114444,
+      "grad_norm": 0.08906706358654537,
+      "learning_rate": 7.53266129422835e-06,
+      "loss": 0.3035,
+      "step": 2128
+    },
+    {
+      "epoch": 1.24333990219692,
+      "grad_norm": 0.0889219029878735,
+      "learning_rate": 7.522781160556308e-06,
+      "loss": 0.3628,
+      "step": 2129
+    },
+    {
+      "epoch": 1.2439238011823954,
+      "grad_norm": 0.0827716486622492,
+      "learning_rate": 7.512903602497723e-06,
+      "loss": 0.269,
+      "step": 2130
+    },
+    {
+      "epoch": 1.244507700167871,
+      "grad_norm": 0.0908883574682588,
+      "learning_rate": 7.503028630322486e-06,
+      "loss": 0.3454,
+      "step": 2131
+    },
+    {
+      "epoch": 1.2450915991533464,
+      "grad_norm": 0.09016016461707722,
+      "learning_rate": 7.4931562542977994e-06,
+      "loss": 0.3507,
+      "step": 2132
+    },
+    {
+      "epoch": 1.245675498138822,
+      "grad_norm": 0.0938809689503389,
+      "learning_rate": 7.483286484688172e-06,
+      "loss": 0.3379,
+      "step": 2133
+    },
+    {
+      "epoch": 1.2462593971242975,
+      "grad_norm": 0.08299446642362061,
+      "learning_rate": 7.473419331755395e-06,
+      "loss": 0.2673,
+      "step": 2134
+    },
+    {
+      "epoch": 1.246843296109773,
+      "grad_norm": 0.08893820792149283,
+      "learning_rate": 7.463554805758546e-06,
+      "loss": 0.3343,
+      "step": 2135
+    },
+    {
+      "epoch": 1.2474271950952485,
+      "grad_norm": 0.08374215123508322,
+      "learning_rate": 7.453692916953965e-06,
+      "loss": 0.2898,
+      "step": 2136
+    },
+    {
+      "epoch": 1.248011094080724,
+      "grad_norm": 0.08740033284270947,
+      "learning_rate": 7.443833675595254e-06,
+      "loss": 0.282,
+      "step": 2137
+    },
+    {
+      "epoch": 1.2485949930661995,
+      "grad_norm": 0.08983472809871217,
+      "learning_rate": 7.433977091933262e-06,
+      "loss": 0.3263,
+      "step": 2138
+    },
+    {
+      "epoch": 1.249178892051675,
+      "grad_norm": 0.08723409592189417,
+      "learning_rate": 7.424123176216072e-06,
+      "loss": 0.3597,
+      "step": 2139
+    },
+    {
+      "epoch": 1.2497627910371505,
+      "grad_norm": 0.09062022031992072,
+      "learning_rate": 7.414271938689e-06,
+      "loss": 0.3288,
+      "step": 2140
+    },
+    {
+      "epoch": 1.250346690022626,
+      "grad_norm": 0.09414996872567236,
+      "learning_rate": 7.404423389594569e-06,
+      "loss": 0.3601,
+      "step": 2141
+    },
+    {
+      "epoch": 1.2509305890081017,
+      "grad_norm": 0.08835840380620273,
+      "learning_rate": 7.394577539172507e-06,
+      "loss": 0.2857,
+      "step": 2142
+    },
+    {
+      "epoch": 1.2515144879935771,
+      "grad_norm": 0.09322214868117433,
+      "learning_rate": 7.3847343976597454e-06,
+      "loss": 0.3299,
+      "step": 2143
+    },
+    {
+      "epoch": 1.2520983869790525,
+      "grad_norm": 0.09999874858809561,
+      "learning_rate": 7.374893975290391e-06,
+      "loss": 0.3516,
+      "step": 2144
+    },
+    {
+      "epoch": 1.2526822859645281,
+      "grad_norm": 0.09077789228321342,
+      "learning_rate": 7.3650562822957285e-06,
+      "loss": 0.2877,
+      "step": 2145
+    },
+    {
+      "epoch": 1.2532661849500037,
+      "grad_norm": 0.08976926338794472,
+      "learning_rate": 7.355221328904196e-06,
+      "loss": 0.3282,
+      "step": 2146
+    },
+    {
+      "epoch": 1.2538500839354791,
+      "grad_norm": 0.08735801570811194,
+      "learning_rate": 7.3453891253413935e-06,
+      "loss": 0.3199,
+      "step": 2147
+    },
+    {
+      "epoch": 1.2544339829209545,
+      "grad_norm": 0.09333409012221447,
+      "learning_rate": 7.335559681830058e-06,
+      "loss": 0.3487,
+      "step": 2148
+    },
+    {
+      "epoch": 1.2550178819064302,
+      "grad_norm": 0.08740722352807123,
+      "learning_rate": 7.325733008590053e-06,
+      "loss": 0.2942,
+      "step": 2149
+    },
+    {
+      "epoch": 1.2556017808919058,
+      "grad_norm": 0.08384971605506603,
+      "learning_rate": 7.315909115838367e-06,
+      "loss": 0.318,
+      "step": 2150
+    },
+    {
+      "epoch": 1.2561856798773812,
+      "grad_norm": 0.08110743025770902,
+      "learning_rate": 7.306088013789097e-06,
+      "loss": 0.2703,
+      "step": 2151
+    },
+    {
+      "epoch": 1.2567695788628568,
+      "grad_norm": 0.09698546468027436,
+      "learning_rate": 7.296269712653436e-06,
+      "loss": 0.3722,
+      "step": 2152
+    },
+    {
+      "epoch": 1.2573534778483322,
+      "grad_norm": 0.08362084154105635,
+      "learning_rate": 7.28645422263967e-06,
+      "loss": 0.2727,
+      "step": 2153
+    },
+    {
+      "epoch": 1.2579373768338078,
+      "grad_norm": 0.09033486363821534,
+      "learning_rate": 7.27664155395315e-06,
+      "loss": 0.3621,
+      "step": 2154
+    },
+    {
+      "epoch": 1.2585212758192832,
+      "grad_norm": 0.0909618839906228,
+      "learning_rate": 7.266831716796307e-06,
+      "loss": 0.3246,
+      "step": 2155
+    },
+    {
+      "epoch": 1.2591051748047588,
+      "grad_norm": 0.0820684392746737,
+      "learning_rate": 7.257024721368624e-06,
+      "loss": 0.2956,
+      "step": 2156
+    },
+    {
+      "epoch": 1.2596890737902342,
+      "grad_norm": 0.08272833286810921,
+      "learning_rate": 7.247220577866625e-06,
+      "loss": 0.2899,
+      "step": 2157
+    },
+    {
+      "epoch": 1.2602729727757098,
+      "grad_norm": 0.08927618848855626,
+      "learning_rate": 7.237419296483876e-06,
+      "loss": 0.2874,
+      "step": 2158
+    },
+    {
+      "epoch": 1.2608568717611854,
+      "grad_norm": 0.08699471176725997,
+      "learning_rate": 7.227620887410958e-06,
+      "loss": 0.3201,
+      "step": 2159
+    },
+    {
+      "epoch": 1.2614407707466608,
+      "grad_norm": 0.08809593155349488,
+      "learning_rate": 7.217825360835475e-06,
+      "loss": 0.3309,
+      "step": 2160
+    },
+    {
+      "epoch": 1.2620246697321362,
+      "grad_norm": 0.08992358581948587,
+      "learning_rate": 7.208032726942027e-06,
+      "loss": 0.3087,
+      "step": 2161
+    },
+    {
+      "epoch": 1.2626085687176118,
+      "grad_norm": 0.08444492384729725,
+      "learning_rate": 7.198242995912211e-06,
+      "loss": 0.2641,
+      "step": 2162
+    },
+    {
+      "epoch": 1.2631924677030875,
+      "grad_norm": 0.09476987044378767,
+      "learning_rate": 7.1884561779246055e-06,
+      "loss": 0.3504,
+      "step": 2163
+    },
+    {
+      "epoch": 1.2637763666885629,
+      "grad_norm": 0.08680771071849888,
+      "learning_rate": 7.178672283154756e-06,
+      "loss": 0.2949,
+      "step": 2164
+    },
+    {
+      "epoch": 1.2643602656740383,
+      "grad_norm": 0.09401994425338037,
+      "learning_rate": 7.168891321775172e-06,
+      "loss": 0.3947,
+      "step": 2165
+    },
+    {
+      "epoch": 1.2649441646595139,
+      "grad_norm": 0.08215430574640893,
+      "learning_rate": 7.159113303955314e-06,
+      "loss": 0.2854,
+      "step": 2166
+    },
+    {
+      "epoch": 1.2655280636449895,
+      "grad_norm": 0.08409263268739804,
+      "learning_rate": 7.149338239861579e-06,
+      "loss": 0.3006,
+      "step": 2167
+    },
+    {
+      "epoch": 1.2661119626304649,
+      "grad_norm": 0.08618188516028233,
+      "learning_rate": 7.139566139657298e-06,
+      "loss": 0.3179,
+      "step": 2168
+    },
+    {
+      "epoch": 1.2666958616159405,
+      "grad_norm": 0.08831967570068174,
+      "learning_rate": 7.129797013502713e-06,
+      "loss": 0.3254,
+      "step": 2169
+    },
+    {
+      "epoch": 1.267279760601416,
+      "grad_norm": 0.08807399735185314,
+      "learning_rate": 7.12003087155498e-06,
+      "loss": 0.3237,
+      "step": 2170
+    },
+    {
+      "epoch": 1.2678636595868915,
+      "grad_norm": 0.08868432635952946,
+      "learning_rate": 7.110267723968147e-06,
+      "loss": 0.3116,
+      "step": 2171
+    },
+    {
+      "epoch": 1.268447558572367,
+      "grad_norm": 0.08575571469017852,
+      "learning_rate": 7.100507580893156e-06,
+      "loss": 0.3352,
+      "step": 2172
+    },
+    {
+      "epoch": 1.2690314575578425,
+      "grad_norm": 0.07902707327027708,
+      "learning_rate": 7.09075045247782e-06,
+      "loss": 0.303,
+      "step": 2173
+    },
+    {
+      "epoch": 1.269615356543318,
+      "grad_norm": 0.0866798542763208,
+      "learning_rate": 7.080996348866817e-06,
+      "loss": 0.322,
+      "step": 2174
+    },
+    {
+      "epoch": 1.2701992555287935,
+      "grad_norm": 0.08241742470439148,
+      "learning_rate": 7.071245280201682e-06,
+      "loss": 0.279,
+      "step": 2175
+    },
+    {
+      "epoch": 1.2707831545142692,
+      "grad_norm": 0.09048470361011259,
+      "learning_rate": 7.061497256620793e-06,
+      "loss": 0.3182,
+      "step": 2176
+    },
+    {
+      "epoch": 1.2713670534997445,
+      "grad_norm": 0.08960739217963698,
+      "learning_rate": 7.051752288259366e-06,
+      "loss": 0.3057,
+      "step": 2177
+    },
+    {
+      "epoch": 1.27195095248522,
+      "grad_norm": 0.08743898676404914,
+      "learning_rate": 7.042010385249433e-06,
+      "loss": 0.3009,
+      "step": 2178
+    },
+    {
+      "epoch": 1.2725348514706956,
+      "grad_norm": 0.08955293473062877,
+      "learning_rate": 7.032271557719847e-06,
+      "loss": 0.3593,
+      "step": 2179
+    },
+    {
+      "epoch": 1.2731187504561712,
+      "grad_norm": 0.0864491154353338,
+      "learning_rate": 7.022535815796261e-06,
+      "loss": 0.3116,
+      "step": 2180
+    },
+    {
+      "epoch": 1.2737026494416466,
+      "grad_norm": 0.08925233499472403,
+      "learning_rate": 7.012803169601118e-06,
+      "loss": 0.307,
+      "step": 2181
+    },
+    {
+      "epoch": 1.2742865484271222,
+      "grad_norm": 0.09424753137969301,
+      "learning_rate": 7.003073629253638e-06,
+      "loss": 0.2828,
+      "step": 2182
+    },
+    {
+      "epoch": 1.2748704474125976,
+      "grad_norm": 0.08449638279597557,
+      "learning_rate": 6.9933472048698225e-06,
+      "loss": 0.288,
+      "step": 2183
+    },
+    {
+      "epoch": 1.2754543463980732,
+      "grad_norm": 0.0897382775051815,
+      "learning_rate": 6.983623906562422e-06,
+      "loss": 0.3266,
+      "step": 2184
+    },
+    {
+      "epoch": 1.2760382453835486,
+      "grad_norm": 0.08875604074341699,
+      "learning_rate": 6.973903744440949e-06,
+      "loss": 0.3092,
+      "step": 2185
+    },
+    {
+      "epoch": 1.2766221443690242,
+      "grad_norm": 0.0970598221853414,
+      "learning_rate": 6.964186728611644e-06,
+      "loss": 0.3327,
+      "step": 2186
+    },
+    {
+      "epoch": 1.2772060433544996,
+      "grad_norm": 0.09223837467956714,
+      "learning_rate": 6.954472869177479e-06,
+      "loss": 0.3184,
+      "step": 2187
+    },
+    {
+      "epoch": 1.2777899423399752,
+      "grad_norm": 0.08852883785955053,
+      "learning_rate": 6.944762176238149e-06,
+      "loss": 0.3526,
+      "step": 2188
+    },
+    {
+      "epoch": 1.2783738413254506,
+      "grad_norm": 0.09636270789025539,
+      "learning_rate": 6.935054659890053e-06,
+      "loss": 0.3644,
+      "step": 2189
+    },
+    {
+      "epoch": 1.2789577403109262,
+      "grad_norm": 0.09988413367869864,
+      "learning_rate": 6.9253503302262855e-06,
+      "loss": 0.3433,
+      "step": 2190
+    },
+    {
+      "epoch": 1.2795416392964016,
+      "grad_norm": 0.0872454129642266,
+      "learning_rate": 6.915649197336638e-06,
+      "loss": 0.3048,
+      "step": 2191
+    },
+    {
+      "epoch": 1.2801255382818773,
+      "grad_norm": 0.08466175639352268,
+      "learning_rate": 6.905951271307561e-06,
+      "loss": 0.3148,
+      "step": 2192
+    },
+    {
+      "epoch": 1.2807094372673529,
+      "grad_norm": 0.09431440614265761,
+      "learning_rate": 6.896256562222184e-06,
+      "loss": 0.3427,
+      "step": 2193
+    },
+    {
+      "epoch": 1.2812933362528283,
+      "grad_norm": 0.10056864802084728,
+      "learning_rate": 6.8865650801602855e-06,
+      "loss": 0.4092,
+      "step": 2194
+    },
+    {
+      "epoch": 1.2818772352383037,
+      "grad_norm": 0.0807965522528703,
+      "learning_rate": 6.8768768351982964e-06,
+      "loss": 0.2709,
+      "step": 2195
+    },
+    {
+      "epoch": 1.2824611342237793,
+      "grad_norm": 0.08592363514480962,
+      "learning_rate": 6.867191837409275e-06,
+      "loss": 0.288,
+      "step": 2196
+    },
+    {
+      "epoch": 1.283045033209255,
+      "grad_norm": 0.08886405122974861,
+      "learning_rate": 6.857510096862901e-06,
+      "loss": 0.3085,
+      "step": 2197
+    },
+    {
+      "epoch": 1.2836289321947303,
+      "grad_norm": 0.09467573843816428,
+      "learning_rate": 6.847831623625476e-06,
+      "loss": 0.3503,
+      "step": 2198
+    },
+    {
+      "epoch": 1.284212831180206,
+      "grad_norm": 0.08730354402134316,
+      "learning_rate": 6.8381564277598974e-06,
+      "loss": 0.2808,
+      "step": 2199
+    },
+    {
+      "epoch": 1.2847967301656813,
+      "grad_norm": 0.09038495868216823,
+      "learning_rate": 6.82848451932566e-06,
+      "loss": 0.3245,
+      "step": 2200
+    },
+    {
+      "epoch": 1.285380629151157,
+      "grad_norm": 0.1025227010932707,
+      "learning_rate": 6.81881590837884e-06,
+      "loss": 0.3848,
+      "step": 2201
+    },
+    {
+      "epoch": 1.2859645281366323,
+      "grad_norm": 0.0820787066831925,
+      "learning_rate": 6.809150604972079e-06,
+      "loss": 0.2918,
+      "step": 2202
+    },
+    {
+      "epoch": 1.286548427122108,
+      "grad_norm": 0.08071723233495022,
+      "learning_rate": 6.799488619154586e-06,
+      "loss": 0.2936,
+      "step": 2203
+    },
+    {
+      "epoch": 1.2871323261075833,
+      "grad_norm": 0.08836013494378199,
+      "learning_rate": 6.7898299609721186e-06,
+      "loss": 0.3342,
+      "step": 2204
+    },
+    {
+      "epoch": 1.287716225093059,
+      "grad_norm": 0.09356375553981922,
+      "learning_rate": 6.780174640466976e-06,
+      "loss": 0.3444,
+      "step": 2205
+    },
+    {
+      "epoch": 1.2883001240785343,
+      "grad_norm": 0.10150621071483237,
+      "learning_rate": 6.7705226676779855e-06,
+      "loss": 0.3453,
+      "step": 2206
+    },
+    {
+      "epoch": 1.28888402306401,
+      "grad_norm": 0.09346037530782342,
+      "learning_rate": 6.760874052640494e-06,
+      "loss": 0.3049,
+      "step": 2207
+    },
+    {
+      "epoch": 1.2894679220494853,
+      "grad_norm": 0.08155918133261261,
+      "learning_rate": 6.751228805386363e-06,
+      "loss": 0.3068,
+      "step": 2208
+    },
+    {
+      "epoch": 1.290051821034961,
+      "grad_norm": 0.0900765954577318,
+      "learning_rate": 6.741586935943937e-06,
+      "loss": 0.3192,
+      "step": 2209
+    },
+    {
+      "epoch": 1.2906357200204366,
+      "grad_norm": 0.09449874798759857,
+      "learning_rate": 6.731948454338064e-06,
+      "loss": 0.293,
+      "step": 2210
+    },
+    {
+      "epoch": 1.291219619005912,
+      "grad_norm": 0.09377292087853051,
+      "learning_rate": 6.7223133705900635e-06,
+      "loss": 0.3714,
+      "step": 2211
+    },
+    {
+      "epoch": 1.2918035179913874,
+      "grad_norm": 0.09579011671547145,
+      "learning_rate": 6.712681694717723e-06,
+      "loss": 0.3319,
+      "step": 2212
+    },
+    {
+      "epoch": 1.292387416976863,
+      "grad_norm": 0.09932022559574918,
+      "learning_rate": 6.7030534367352884e-06,
+      "loss": 0.3518,
+      "step": 2213
+    },
+    {
+      "epoch": 1.2929713159623386,
+      "grad_norm": 0.09563476665706167,
+      "learning_rate": 6.693428606653445e-06,
+      "loss": 0.353,
+      "step": 2214
+    },
+    {
+      "epoch": 1.293555214947814,
+      "grad_norm": 0.09357821548814335,
+      "learning_rate": 6.683807214479323e-06,
+      "loss": 0.3267,
+      "step": 2215
+    },
+    {
+      "epoch": 1.2941391139332896,
+      "grad_norm": 0.09046701048851333,
+      "learning_rate": 6.6741892702164735e-06,
+      "loss": 0.3452,
+      "step": 2216
+    },
+    {
+      "epoch": 1.294723012918765,
+      "grad_norm": 0.08447339465364394,
+      "learning_rate": 6.664574783864862e-06,
+      "loss": 0.2892,
+      "step": 2217
+    },
+    {
+      "epoch": 1.2953069119042406,
+      "grad_norm": 0.08968597679349956,
+      "learning_rate": 6.654963765420866e-06,
+      "loss": 0.3209,
+      "step": 2218
+    },
+    {
+      "epoch": 1.295890810889716,
+      "grad_norm": 0.07833632853580665,
+      "learning_rate": 6.645356224877242e-06,
+      "loss": 0.29,
+      "step": 2219
+    },
+    {
+      "epoch": 1.2964747098751916,
+      "grad_norm": 0.09220051359565674,
+      "learning_rate": 6.635752172223146e-06,
+      "loss": 0.3485,
+      "step": 2220
+    },
+    {
+      "epoch": 1.297058608860667,
+      "grad_norm": 0.08617728043072292,
+      "learning_rate": 6.626151617444103e-06,
+      "loss": 0.2992,
+      "step": 2221
+    },
+    {
+      "epoch": 1.2976425078461427,
+      "grad_norm": 0.08603360724088882,
+      "learning_rate": 6.6165545705219955e-06,
+      "loss": 0.288,
+      "step": 2222
+    },
+    {
+      "epoch": 1.298226406831618,
+      "grad_norm": 0.09352762665828722,
+      "learning_rate": 6.606961041435068e-06,
+      "loss": 0.3658,
+      "step": 2223
+    },
+    {
+      "epoch": 1.2988103058170937,
+      "grad_norm": 0.09054859432258221,
+      "learning_rate": 6.5973710401578985e-06,
+      "loss": 0.3278,
+      "step": 2224
+    },
+    {
+      "epoch": 1.299394204802569,
+      "grad_norm": 0.0906091442712072,
+      "learning_rate": 6.587784576661401e-06,
+      "loss": 0.2819,
+      "step": 2225
+    },
+    {
+      "epoch": 1.2999781037880447,
+      "grad_norm": 0.08525950299070298,
+      "learning_rate": 6.578201660912814e-06,
+      "loss": 0.2719,
+      "step": 2226
+    },
+    {
+      "epoch": 1.3005620027735203,
+      "grad_norm": 0.09502117233898026,
+      "learning_rate": 6.568622302875682e-06,
+      "loss": 0.349,
+      "step": 2227
+    },
+    {
+      "epoch": 1.3011459017589957,
+      "grad_norm": 0.09042401802294561,
+      "learning_rate": 6.559046512509859e-06,
+      "loss": 0.2973,
+      "step": 2228
+    },
+    {
+      "epoch": 1.301729800744471,
+      "grad_norm": 0.08733060343325119,
+      "learning_rate": 6.5494742997714765e-06,
+      "loss": 0.3216,
+      "step": 2229
+    },
+    {
+      "epoch": 1.3023136997299467,
+      "grad_norm": 0.08599452614624836,
+      "learning_rate": 6.539905674612956e-06,
+      "loss": 0.3449,
+      "step": 2230
+    },
+    {
+      "epoch": 1.3028975987154223,
+      "grad_norm": 0.09222126565501329,
+      "learning_rate": 6.530340646982987e-06,
+      "loss": 0.3544,
+      "step": 2231
+    },
+    {
+      "epoch": 1.3034814977008977,
+      "grad_norm": 0.0923115966005353,
+      "learning_rate": 6.520779226826517e-06,
+      "loss": 0.3043,
+      "step": 2232
+    },
+    {
+      "epoch": 1.3040653966863733,
+      "grad_norm": 0.08960963652296645,
+      "learning_rate": 6.511221424084748e-06,
+      "loss": 0.3172,
+      "step": 2233
+    },
+    {
+      "epoch": 1.3046492956718487,
+      "grad_norm": 0.09450916312634182,
+      "learning_rate": 6.501667248695107e-06,
+      "loss": 0.3071,
+      "step": 2234
+    },
+    {
+      "epoch": 1.3052331946573243,
+      "grad_norm": 0.08163969110816581,
+      "learning_rate": 6.4921167105912696e-06,
+      "loss": 0.2567,
+      "step": 2235
+    },
+    {
+      "epoch": 1.3058170936427997,
+      "grad_norm": 0.08934397482201767,
+      "learning_rate": 6.482569819703117e-06,
+      "loss": 0.2754,
+      "step": 2236
+    },
+    {
+      "epoch": 1.3064009926282754,
+      "grad_norm": 0.09403971367817346,
+      "learning_rate": 6.473026585956736e-06,
+      "loss": 0.3682,
+      "step": 2237
+    },
+    {
+      "epoch": 1.3069848916137508,
+      "grad_norm": 0.09806291074218748,
+      "learning_rate": 6.4634870192744205e-06,
+      "loss": 0.3274,
+      "step": 2238
+    },
+    {
+      "epoch": 1.3075687905992264,
+      "grad_norm": 0.08856185719227691,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 0.3083,
+      "step": 2239
+    },
+    {
+      "epoch": 1.3081526895847018,
+      "grad_norm": 0.09406449435848557,
+      "learning_rate": 6.44441892677206e-06,
+      "loss": 0.3505,
+      "step": 2240
+    },
+    {
+      "epoch": 1.3087365885701774,
+      "grad_norm": 0.09393679588121236,
+      "learning_rate": 6.434890420777491e-06,
+      "loss": 0.3111,
+      "step": 2241
+    },
+    {
+      "epoch": 1.3093204875556528,
+      "grad_norm": 0.08947926861239962,
+      "learning_rate": 6.42536562149791e-06,
+      "loss": 0.3196,
+      "step": 2242
+    },
+    {
+      "epoch": 1.3099043865411284,
+      "grad_norm": 0.09495481312221282,
+      "learning_rate": 6.41584453883644e-06,
+      "loss": 0.3079,
+      "step": 2243
+    },
+    {
+      "epoch": 1.310488285526604,
+      "grad_norm": 0.09100802159030147,
+      "learning_rate": 6.40632718269234e-06,
+      "loss": 0.3179,
+      "step": 2244
+    },
+    {
+      "epoch": 1.3110721845120794,
+      "grad_norm": 0.08967348883753556,
+      "learning_rate": 6.396813562960993e-06,
+      "loss": 0.3213,
+      "step": 2245
+    },
+    {
+      "epoch": 1.3116560834975548,
+      "grad_norm": 0.09529333320235595,
+      "learning_rate": 6.387303689533899e-06,
+      "loss": 0.285,
+      "step": 2246
+    },
+    {
+      "epoch": 1.3122399824830304,
+      "grad_norm": 0.08762255670635358,
+      "learning_rate": 6.377797572298661e-06,
+      "loss": 0.3528,
+      "step": 2247
+    },
+    {
+      "epoch": 1.312823881468506,
+      "grad_norm": 0.08910793500360832,
+      "learning_rate": 6.3682952211389735e-06,
+      "loss": 0.3301,
+      "step": 2248
+    },
+    {
+      "epoch": 1.3134077804539814,
+      "grad_norm": 0.08531659189279815,
+      "learning_rate": 6.358796645934624e-06,
+      "loss": 0.2894,
+      "step": 2249
+    },
+    {
+      "epoch": 1.313991679439457,
+      "grad_norm": 0.08639189138387236,
+      "learning_rate": 6.349301856561468e-06,
+      "loss": 0.3184,
+      "step": 2250
+    },
+    {
+      "epoch": 1.3145755784249324,
+      "grad_norm": 0.08832635017696683,
+      "learning_rate": 6.3398108628914264e-06,
+      "loss": 0.3229,
+      "step": 2251
+    },
+    {
+      "epoch": 1.315159477410408,
+      "grad_norm": 0.08685482458543095,
+      "learning_rate": 6.330323674792472e-06,
+      "loss": 0.3183,
+      "step": 2252
+    },
+    {
+      "epoch": 1.3157433763958835,
+      "grad_norm": 0.079854555064024,
+      "learning_rate": 6.320840302128619e-06,
+      "loss": 0.2991,
+      "step": 2253
+    },
+    {
+      "epoch": 1.316327275381359,
+      "grad_norm": 0.08220818916983169,
+      "learning_rate": 6.311360754759923e-06,
+      "loss": 0.3008,
+      "step": 2254
+    },
+    {
+      "epoch": 1.3169111743668345,
+      "grad_norm": 0.08686732726476201,
+      "learning_rate": 6.301885042542455e-06,
+      "loss": 0.3217,
+      "step": 2255
+    },
+    {
+      "epoch": 1.31749507335231,
+      "grad_norm": 0.09078138665349783,
+      "learning_rate": 6.292413175328302e-06,
+      "loss": 0.3288,
+      "step": 2256
+    },
+    {
+      "epoch": 1.3180789723377855,
+      "grad_norm": 0.09012347042593322,
+      "learning_rate": 6.282945162965548e-06,
+      "loss": 0.33,
+      "step": 2257
+    },
+    {
+      "epoch": 1.318662871323261,
+      "grad_norm": 0.08247622088935579,
+      "learning_rate": 6.273481015298275e-06,
+      "loss": 0.2914,
+      "step": 2258
+    },
+    {
+      "epoch": 1.3192467703087365,
+      "grad_norm": 0.09125482701573914,
+      "learning_rate": 6.264020742166543e-06,
+      "loss": 0.3035,
+      "step": 2259
+    },
+    {
+      "epoch": 1.319830669294212,
+      "grad_norm": 0.08744548149312226,
+      "learning_rate": 6.2545643534063894e-06,
+      "loss": 0.3055,
+      "step": 2260
+    },
+    {
+      "epoch": 1.3204145682796877,
+      "grad_norm": 0.08806689003088922,
+      "learning_rate": 6.245111858849808e-06,
+      "loss": 0.3096,
+      "step": 2261
+    },
+    {
+      "epoch": 1.3209984672651631,
+      "grad_norm": 0.0858006913761187,
+      "learning_rate": 6.235663268324735e-06,
+      "loss": 0.299,
+      "step": 2262
+    },
+    {
+      "epoch": 1.3215823662506385,
+      "grad_norm": 0.09340611054497164,
+      "learning_rate": 6.226218591655071e-06,
+      "loss": 0.3401,
+      "step": 2263
+    },
+    {
+      "epoch": 1.3221662652361141,
+      "grad_norm": 0.09372086309480247,
+      "learning_rate": 6.216777838660627e-06,
+      "loss": 0.3372,
+      "step": 2264
+    },
+    {
+      "epoch": 1.3227501642215898,
+      "grad_norm": 0.0901682737549222,
+      "learning_rate": 6.2073410191571395e-06,
+      "loss": 0.3186,
+      "step": 2265
+    },
+    {
+      "epoch": 1.3233340632070651,
+      "grad_norm": 0.08411000343559022,
+      "learning_rate": 6.1979081429562575e-06,
+      "loss": 0.294,
+      "step": 2266
+    },
+    {
+      "epoch": 1.3239179621925408,
+      "grad_norm": 0.08326427008425046,
+      "learning_rate": 6.188479219865529e-06,
+      "loss": 0.3111,
+      "step": 2267
+    },
+    {
+      "epoch": 1.3245018611780162,
+      "grad_norm": 0.09158941344378,
+      "learning_rate": 6.179054259688393e-06,
+      "loss": 0.2955,
+      "step": 2268
+    },
+    {
+      "epoch": 1.3250857601634918,
+      "grad_norm": 0.08592292077584306,
+      "learning_rate": 6.169633272224167e-06,
+      "loss": 0.2849,
+      "step": 2269
+    },
+    {
+      "epoch": 1.3256696591489672,
+      "grad_norm": 0.08256587523186022,
+      "learning_rate": 6.160216267268037e-06,
+      "loss": 0.2941,
+      "step": 2270
+    },
+    {
+      "epoch": 1.3262535581344428,
+      "grad_norm": 0.08750258565526414,
+      "learning_rate": 6.1508032546110485e-06,
+      "loss": 0.327,
+      "step": 2271
+    },
+    {
+      "epoch": 1.3268374571199182,
+      "grad_norm": 0.09191854081229216,
+      "learning_rate": 6.1413942440400994e-06,
+      "loss": 0.3339,
+      "step": 2272
+    },
+    {
+      "epoch": 1.3274213561053938,
+      "grad_norm": 0.09473183817076396,
+      "learning_rate": 6.1319892453379235e-06,
+      "loss": 0.3432,
+      "step": 2273
+    },
+    {
+      "epoch": 1.3280052550908692,
+      "grad_norm": 0.09470052506274296,
+      "learning_rate": 6.122588268283085e-06,
+      "loss": 0.3565,
+      "step": 2274
+    },
+    {
+      "epoch": 1.3285891540763448,
+      "grad_norm": 0.07994432668654308,
+      "learning_rate": 6.113191322649964e-06,
+      "loss": 0.2624,
+      "step": 2275
+    },
+    {
+      "epoch": 1.3291730530618202,
+      "grad_norm": 0.09448709705368633,
+      "learning_rate": 6.10379841820875e-06,
+      "loss": 0.3401,
+      "step": 2276
+    },
+    {
+      "epoch": 1.3297569520472958,
+      "grad_norm": 0.08945302369365801,
+      "learning_rate": 6.094409564725435e-06,
+      "loss": 0.3183,
+      "step": 2277
+    },
+    {
+      "epoch": 1.3303408510327714,
+      "grad_norm": 0.0851894560623977,
+      "learning_rate": 6.085024771961792e-06,
+      "loss": 0.2984,
+      "step": 2278
+    },
+    {
+      "epoch": 1.3309247500182468,
+      "grad_norm": 0.08912861735934718,
+      "learning_rate": 6.07564404967538e-06,
+      "loss": 0.3,
+      "step": 2279
+    },
+    {
+      "epoch": 1.3315086490037222,
+      "grad_norm": 0.08807983439949119,
+      "learning_rate": 6.06626740761952e-06,
+      "loss": 0.3158,
+      "step": 2280
+    },
+    {
+      "epoch": 1.3320925479891979,
+      "grad_norm": 0.08895552192616234,
+      "learning_rate": 6.056894855543289e-06,
+      "loss": 0.3134,
+      "step": 2281
+    },
+    {
+      "epoch": 1.3326764469746735,
+      "grad_norm": 0.09098271721887957,
+      "learning_rate": 6.047526403191517e-06,
+      "loss": 0.3412,
+      "step": 2282
+    },
+    {
+      "epoch": 1.3332603459601489,
+      "grad_norm": 0.09542398683776629,
+      "learning_rate": 6.038162060304771e-06,
+      "loss": 0.342,
+      "step": 2283
+    },
+    {
+      "epoch": 1.3338442449456245,
+      "grad_norm": 0.08186472746147123,
+      "learning_rate": 6.028801836619345e-06,
+      "loss": 0.3128,
+      "step": 2284
+    },
+    {
+      "epoch": 1.3344281439310999,
+      "grad_norm": 0.0861327665716666,
+      "learning_rate": 6.019445741867245e-06,
+      "loss": 0.3314,
+      "step": 2285
+    },
+    {
+      "epoch": 1.3350120429165755,
+      "grad_norm": 0.0988009145871849,
+      "learning_rate": 6.010093785776188e-06,
+      "loss": 0.3224,
+      "step": 2286
+    },
+    {
+      "epoch": 1.3355959419020509,
+      "grad_norm": 0.08449443845814798,
+      "learning_rate": 6.0007459780695885e-06,
+      "loss": 0.2934,
+      "step": 2287
+    },
+    {
+      "epoch": 1.3361798408875265,
+      "grad_norm": 0.0871146044831975,
+      "learning_rate": 5.991402328466549e-06,
+      "loss": 0.2908,
+      "step": 2288
+    },
+    {
+      "epoch": 1.336763739873002,
+      "grad_norm": 0.09105223324610011,
+      "learning_rate": 5.982062846681848e-06,
+      "loss": 0.3466,
+      "step": 2289
+    },
+    {
+      "epoch": 1.3373476388584775,
+      "grad_norm": 0.08831579617214036,
+      "learning_rate": 5.97272754242592e-06,
+      "loss": 0.2966,
+      "step": 2290
+    },
+    {
+      "epoch": 1.3379315378439531,
+      "grad_norm": 0.08883583273864665,
+      "learning_rate": 5.963396425404877e-06,
+      "loss": 0.334,
+      "step": 2291
+    },
+    {
+      "epoch": 1.3385154368294285,
+      "grad_norm": 0.08393589751040886,
+      "learning_rate": 5.954069505320466e-06,
+      "loss": 0.2969,
+      "step": 2292
+    },
+    {
+      "epoch": 1.339099335814904,
+      "grad_norm": 0.08263403306246712,
+      "learning_rate": 5.944746791870062e-06,
+      "loss": 0.2866,
+      "step": 2293
+    },
+    {
+      "epoch": 1.3396832348003795,
+      "grad_norm": 0.09384827013225032,
+      "learning_rate": 5.935428294746679e-06,
+      "loss": 0.3477,
+      "step": 2294
+    },
+    {
+      "epoch": 1.3402671337858552,
+      "grad_norm": 0.0892650553575312,
+      "learning_rate": 5.926114023638944e-06,
+      "loss": 0.3269,
+      "step": 2295
+    },
+    {
+      "epoch": 1.3408510327713306,
+      "grad_norm": 0.08464776955505508,
+      "learning_rate": 5.916803988231087e-06,
+      "loss": 0.2848,
+      "step": 2296
+    },
+    {
+      "epoch": 1.341434931756806,
+      "grad_norm": 0.09326930970380357,
+      "learning_rate": 5.907498198202939e-06,
+      "loss": 0.351,
+      "step": 2297
+    },
+    {
+      "epoch": 1.3420188307422816,
+      "grad_norm": 0.08291422448306884,
+      "learning_rate": 5.898196663229912e-06,
+      "loss": 0.3147,
+      "step": 2298
+    },
+    {
+      "epoch": 1.3426027297277572,
+      "grad_norm": 0.09008940957847367,
+      "learning_rate": 5.888899392982994e-06,
+      "loss": 0.3373,
+      "step": 2299
+    },
+    {
+      "epoch": 1.3431866287132326,
+      "grad_norm": 0.08984609887158414,
+      "learning_rate": 5.879606397128743e-06,
+      "loss": 0.3151,
+      "step": 2300
+    },
+    {
+      "epoch": 1.3437705276987082,
+      "grad_norm": 0.08704510163485692,
+      "learning_rate": 5.8703176853292705e-06,
+      "loss": 0.3349,
+      "step": 2301
+    },
+    {
+      "epoch": 1.3443544266841836,
+      "grad_norm": 0.09160845052385129,
+      "learning_rate": 5.861033267242238e-06,
+      "loss": 0.3357,
+      "step": 2302
+    },
+    {
+      "epoch": 1.3449383256696592,
+      "grad_norm": 0.07994301659773378,
+      "learning_rate": 5.85175315252083e-06,
+      "loss": 0.2894,
+      "step": 2303
+    },
+    {
+      "epoch": 1.3455222246551346,
+      "grad_norm": 0.08993720895816876,
+      "learning_rate": 5.842477350813773e-06,
+      "loss": 0.3005,
+      "step": 2304
+    },
+    {
+      "epoch": 1.3461061236406102,
+      "grad_norm": 0.09321965660463899,
+      "learning_rate": 5.833205871765297e-06,
+      "loss": 0.3731,
+      "step": 2305
+    },
+    {
+      "epoch": 1.3466900226260856,
+      "grad_norm": 0.08391522548020883,
+      "learning_rate": 5.823938725015148e-06,
+      "loss": 0.2669,
+      "step": 2306
+    },
+    {
+      "epoch": 1.3472739216115612,
+      "grad_norm": 0.09634991839706769,
+      "learning_rate": 5.8146759201985525e-06,
+      "loss": 0.3685,
+      "step": 2307
+    },
+    {
+      "epoch": 1.3478578205970368,
+      "grad_norm": 0.10158352675617438,
+      "learning_rate": 5.8054174669462425e-06,
+      "loss": 0.3657,
+      "step": 2308
+    },
+    {
+      "epoch": 1.3484417195825122,
+      "grad_norm": 0.0965664648916933,
+      "learning_rate": 5.796163374884406e-06,
+      "loss": 0.3383,
+      "step": 2309
+    },
+    {
+      "epoch": 1.3490256185679876,
+      "grad_norm": 0.0843045258327265,
+      "learning_rate": 5.786913653634714e-06,
+      "loss": 0.2945,
+      "step": 2310
+    },
+    {
+      "epoch": 1.3496095175534633,
+      "grad_norm": 0.09918330059091131,
+      "learning_rate": 5.77766831281428e-06,
+      "loss": 0.3961,
+      "step": 2311
+    },
+    {
+      "epoch": 1.3501934165389389,
+      "grad_norm": 0.09142904338792841,
+      "learning_rate": 5.768427362035665e-06,
+      "loss": 0.3361,
+      "step": 2312
+    },
+    {
+      "epoch": 1.3507773155244143,
+      "grad_norm": 0.09112187632468056,
+      "learning_rate": 5.759190810906876e-06,
+      "loss": 0.2949,
+      "step": 2313
+    },
+    {
+      "epoch": 1.3513612145098897,
+      "grad_norm": 0.08198007777666597,
+      "learning_rate": 5.749958669031329e-06,
+      "loss": 0.2939,
+      "step": 2314
+    },
+    {
+      "epoch": 1.3519451134953653,
+      "grad_norm": 0.09760012007423624,
+      "learning_rate": 5.740730946007874e-06,
+      "loss": 0.3463,
+      "step": 2315
+    },
+    {
+      "epoch": 1.352529012480841,
+      "grad_norm": 0.09444991340613111,
+      "learning_rate": 5.7315076514307535e-06,
+      "loss": 0.3297,
+      "step": 2316
+    },
+    {
+      "epoch": 1.3531129114663163,
+      "grad_norm": 0.08992390151158787,
+      "learning_rate": 5.722288794889603e-06,
+      "loss": 0.312,
+      "step": 2317
+    },
+    {
+      "epoch": 1.353696810451792,
+      "grad_norm": 0.088057565090489,
+      "learning_rate": 5.713074385969457e-06,
+      "loss": 0.3052,
+      "step": 2318
+    },
+    {
+      "epoch": 1.3542807094372673,
+      "grad_norm": 0.08896206703165421,
+      "learning_rate": 5.703864434250721e-06,
+      "loss": 0.3135,
+      "step": 2319
+    },
+    {
+      "epoch": 1.354864608422743,
+      "grad_norm": 0.08939944209751066,
+      "learning_rate": 5.694658949309158e-06,
+      "loss": 0.2955,
+      "step": 2320
+    },
+    {
+      "epoch": 1.3554485074082183,
+      "grad_norm": 0.0921946861000385,
+      "learning_rate": 5.685457940715898e-06,
+      "loss": 0.311,
+      "step": 2321
+    },
+    {
+      "epoch": 1.356032406393694,
+      "grad_norm": 0.0869804895777747,
+      "learning_rate": 5.67626141803741e-06,
+      "loss": 0.3024,
+      "step": 2322
+    },
+    {
+      "epoch": 1.3566163053791693,
+      "grad_norm": 0.08934604297383186,
+      "learning_rate": 5.667069390835496e-06,
+      "loss": 0.31,
+      "step": 2323
+    },
+    {
+      "epoch": 1.357200204364645,
+      "grad_norm": 0.09418100206357892,
+      "learning_rate": 5.657881868667296e-06,
+      "loss": 0.3228,
+      "step": 2324
+    },
+    {
+      "epoch": 1.3577841033501206,
+      "grad_norm": 0.08335043706927758,
+      "learning_rate": 5.648698861085254e-06,
+      "loss": 0.3009,
+      "step": 2325
+    },
+    {
+      "epoch": 1.358368002335596,
+      "grad_norm": 0.08476923797625893,
+      "learning_rate": 5.639520377637127e-06,
+      "loss": 0.286,
+      "step": 2326
+    },
+    {
+      "epoch": 1.3589519013210714,
+      "grad_norm": 0.09081686268832614,
+      "learning_rate": 5.630346427865965e-06,
+      "loss": 0.3289,
+      "step": 2327
+    },
+    {
+      "epoch": 1.359535800306547,
+      "grad_norm": 0.09377687840948912,
+      "learning_rate": 5.621177021310101e-06,
+      "loss": 0.3338,
+      "step": 2328
+    },
+    {
+      "epoch": 1.3601196992920226,
+      "grad_norm": 0.09147409372002607,
+      "learning_rate": 5.612012167503157e-06,
+      "loss": 0.3385,
+      "step": 2329
+    },
+    {
+      "epoch": 1.360703598277498,
+      "grad_norm": 0.08811846013202936,
+      "learning_rate": 5.602851875974005e-06,
+      "loss": 0.2953,
+      "step": 2330
+    },
+    {
+      "epoch": 1.3612874972629734,
+      "grad_norm": 0.08620706418824692,
+      "learning_rate": 5.593696156246788e-06,
+      "loss": 0.2935,
+      "step": 2331
+    },
+    {
+      "epoch": 1.361871396248449,
+      "grad_norm": 0.09503769253323702,
+      "learning_rate": 5.584545017840886e-06,
+      "loss": 0.3647,
+      "step": 2332
+    },
+    {
+      "epoch": 1.3624552952339246,
+      "grad_norm": 0.10134151512108507,
+      "learning_rate": 5.575398470270913e-06,
+      "loss": 0.3765,
+      "step": 2333
+    },
+    {
+      "epoch": 1.3630391942194,
+      "grad_norm": 0.08808667527271069,
+      "learning_rate": 5.566256523046727e-06,
+      "loss": 0.2861,
+      "step": 2334
+    },
+    {
+      "epoch": 1.3636230932048756,
+      "grad_norm": 0.08788477268580215,
+      "learning_rate": 5.5571191856733795e-06,
+      "loss": 0.3124,
+      "step": 2335
+    },
+    {
+      "epoch": 1.364206992190351,
+      "grad_norm": 0.09223601921700202,
+      "learning_rate": 5.547986467651152e-06,
+      "loss": 0.3172,
+      "step": 2336
+    },
+    {
+      "epoch": 1.3647908911758266,
+      "grad_norm": 0.09448659312288109,
+      "learning_rate": 5.538858378475508e-06,
+      "loss": 0.3583,
+      "step": 2337
+    },
+    {
+      "epoch": 1.365374790161302,
+      "grad_norm": 0.09889184572441016,
+      "learning_rate": 5.529734927637096e-06,
+      "loss": 0.3105,
+      "step": 2338
+    },
+    {
+      "epoch": 1.3659586891467776,
+      "grad_norm": 0.09063715899077882,
+      "learning_rate": 5.520616124621759e-06,
+      "loss": 0.3062,
+      "step": 2339
+    },
+    {
+      "epoch": 1.366542588132253,
+      "grad_norm": 0.08397657868556636,
+      "learning_rate": 5.511501978910488e-06,
+      "loss": 0.3,
+      "step": 2340
+    },
+    {
+      "epoch": 1.3671264871177287,
+      "grad_norm": 0.08689802084891038,
+      "learning_rate": 5.50239249997945e-06,
+      "loss": 0.3166,
+      "step": 2341
+    },
+    {
+      "epoch": 1.3677103861032043,
+      "grad_norm": 0.09093708434083195,
+      "learning_rate": 5.493287697299943e-06,
+      "loss": 0.3385,
+      "step": 2342
+    },
+    {
+      "epoch": 1.3682942850886797,
+      "grad_norm": 0.08651886418436673,
+      "learning_rate": 5.484187580338409e-06,
+      "loss": 0.2812,
+      "step": 2343
+    },
+    {
+      "epoch": 1.368878184074155,
+      "grad_norm": 0.0884936596486878,
+      "learning_rate": 5.475092158556429e-06,
+      "loss": 0.3101,
+      "step": 2344
+    },
+    {
+      "epoch": 1.3694620830596307,
+      "grad_norm": 0.09885106418766211,
+      "learning_rate": 5.4660014414106825e-06,
+      "loss": 0.3751,
+      "step": 2345
+    },
+    {
+      "epoch": 1.3700459820451063,
+      "grad_norm": 0.07832408661953612,
+      "learning_rate": 5.4569154383529736e-06,
+      "loss": 0.2909,
+      "step": 2346
+    },
+    {
+      "epoch": 1.3706298810305817,
+      "grad_norm": 0.08164708132460531,
+      "learning_rate": 5.447834158830202e-06,
+      "loss": 0.2831,
+      "step": 2347
+    },
+    {
+      "epoch": 1.371213780016057,
+      "grad_norm": 0.09024437095238565,
+      "learning_rate": 5.438757612284348e-06,
+      "loss": 0.3212,
+      "step": 2348
+    },
+    {
+      "epoch": 1.3717976790015327,
+      "grad_norm": 0.09152393945408889,
+      "learning_rate": 5.429685808152483e-06,
+      "loss": 0.325,
+      "step": 2349
+    },
+    {
+      "epoch": 1.3723815779870083,
+      "grad_norm": 0.09116266710600476,
+      "learning_rate": 5.420618755866736e-06,
+      "loss": 0.3423,
+      "step": 2350
+    },
+    {
+      "epoch": 1.3729654769724837,
+      "grad_norm": 0.0873800050731155,
+      "learning_rate": 5.411556464854301e-06,
+      "loss": 0.3094,
+      "step": 2351
+    },
+    {
+      "epoch": 1.3735493759579593,
+      "grad_norm": 0.094640977361212,
+      "learning_rate": 5.4024989445374245e-06,
+      "loss": 0.3184,
+      "step": 2352
+    },
+    {
+      "epoch": 1.3741332749434347,
+      "grad_norm": 0.09205122515573358,
+      "learning_rate": 5.393446204333386e-06,
+      "loss": 0.3121,
+      "step": 2353
+    },
+    {
+      "epoch": 1.3747171739289104,
+      "grad_norm": 0.08428365850110131,
+      "learning_rate": 5.384398253654504e-06,
+      "loss": 0.2976,
+      "step": 2354
+    },
+    {
+      "epoch": 1.3753010729143857,
+      "grad_norm": 0.09528420177286316,
+      "learning_rate": 5.3753551019081065e-06,
+      "loss": 0.34,
+      "step": 2355
+    },
+    {
+      "epoch": 1.3758849718998614,
+      "grad_norm": 0.08196645026248169,
+      "learning_rate": 5.366316758496537e-06,
+      "loss": 0.2962,
+      "step": 2356
+    },
+    {
+      "epoch": 1.3764688708853368,
+      "grad_norm": 0.08542941558535018,
+      "learning_rate": 5.357283232817147e-06,
+      "loss": 0.3012,
+      "step": 2357
+    },
+    {
+      "epoch": 1.3770527698708124,
+      "grad_norm": 0.09145770306495496,
+      "learning_rate": 5.348254534262262e-06,
+      "loss": 0.3408,
+      "step": 2358
+    },
+    {
+      "epoch": 1.377636668856288,
+      "grad_norm": 0.10152316542812455,
+      "learning_rate": 5.339230672219209e-06,
+      "loss": 0.3659,
+      "step": 2359
+    },
+    {
+      "epoch": 1.3782205678417634,
+      "grad_norm": 0.09176312043853951,
+      "learning_rate": 5.330211656070269e-06,
+      "loss": 0.3269,
+      "step": 2360
+    },
+    {
+      "epoch": 1.3788044668272388,
+      "grad_norm": 0.09643971695506928,
+      "learning_rate": 5.3211974951926906e-06,
+      "loss": 0.3521,
+      "step": 2361
+    },
+    {
+      "epoch": 1.3793883658127144,
+      "grad_norm": 0.09072196009029657,
+      "learning_rate": 5.312188198958681e-06,
+      "loss": 0.3014,
+      "step": 2362
+    },
+    {
+      "epoch": 1.37997226479819,
+      "grad_norm": 0.08996563862970773,
+      "learning_rate": 5.303183776735379e-06,
+      "loss": 0.2809,
+      "step": 2363
+    },
+    {
+      "epoch": 1.3805561637836654,
+      "grad_norm": 0.09166663024029108,
+      "learning_rate": 5.294184237884865e-06,
+      "loss": 0.3082,
+      "step": 2364
+    },
+    {
+      "epoch": 1.3811400627691408,
+      "grad_norm": 0.08554336365786482,
+      "learning_rate": 5.2851895917641345e-06,
+      "loss": 0.3217,
+      "step": 2365
+    },
+    {
+      "epoch": 1.3817239617546164,
+      "grad_norm": 0.08364144652054088,
+      "learning_rate": 5.276199847725098e-06,
+      "loss": 0.2915,
+      "step": 2366
+    },
+    {
+      "epoch": 1.382307860740092,
+      "grad_norm": 0.09018588814408537,
+      "learning_rate": 5.267215015114574e-06,
+      "loss": 0.3064,
+      "step": 2367
+    },
+    {
+      "epoch": 1.3828917597255674,
+      "grad_norm": 0.08180785670684027,
+      "learning_rate": 5.258235103274265e-06,
+      "loss": 0.2917,
+      "step": 2368
+    },
+    {
+      "epoch": 1.383475658711043,
+      "grad_norm": 0.0862732597210528,
+      "learning_rate": 5.249260121540772e-06,
+      "loss": 0.3089,
+      "step": 2369
+    },
+    {
+      "epoch": 1.3840595576965184,
+      "grad_norm": 0.08703733050081265,
+      "learning_rate": 5.240290079245555e-06,
+      "loss": 0.2892,
+      "step": 2370
+    },
+    {
+      "epoch": 1.384643456681994,
+      "grad_norm": 0.0877943041583372,
+      "learning_rate": 5.231324985714942e-06,
+      "loss": 0.3066,
+      "step": 2371
+    },
+    {
+      "epoch": 1.3852273556674695,
+      "grad_norm": 0.10048405188867884,
+      "learning_rate": 5.222364850270125e-06,
+      "loss": 0.3385,
+      "step": 2372
+    },
+    {
+      "epoch": 1.385811254652945,
+      "grad_norm": 0.08433687099354982,
+      "learning_rate": 5.213409682227129e-06,
+      "loss": 0.3193,
+      "step": 2373
+    },
+    {
+      "epoch": 1.3863951536384205,
+      "grad_norm": 0.0893095191787227,
+      "learning_rate": 5.204459490896818e-06,
+      "loss": 0.3303,
+      "step": 2374
+    },
+    {
+      "epoch": 1.386979052623896,
+      "grad_norm": 0.08450218996079915,
+      "learning_rate": 5.195514285584893e-06,
+      "loss": 0.3007,
+      "step": 2375
+    },
+    {
+      "epoch": 1.3875629516093717,
+      "grad_norm": 0.09182029930894198,
+      "learning_rate": 5.1865740755918496e-06,
+      "loss": 0.3189,
+      "step": 2376
+    },
+    {
+      "epoch": 1.388146850594847,
+      "grad_norm": 0.09603245870496026,
+      "learning_rate": 5.177638870213008e-06,
+      "loss": 0.3731,
+      "step": 2377
+    },
+    {
+      "epoch": 1.3887307495803225,
+      "grad_norm": 0.0829478772437656,
+      "learning_rate": 5.1687086787384786e-06,
+      "loss": 0.3114,
+      "step": 2378
+    },
+    {
+      "epoch": 1.3893146485657981,
+      "grad_norm": 0.0874721626034095,
+      "learning_rate": 5.1597835104531514e-06,
+      "loss": 0.3142,
+      "step": 2379
+    },
+    {
+      "epoch": 1.3898985475512737,
+      "grad_norm": 0.0771748578080161,
+      "learning_rate": 5.1508633746367075e-06,
+      "loss": 0.2657,
+      "step": 2380
+    },
+    {
+      "epoch": 1.3904824465367491,
+      "grad_norm": 0.08294597758235477,
+      "learning_rate": 5.141948280563582e-06,
+      "loss": 0.2962,
+      "step": 2381
+    },
+    {
+      "epoch": 1.3910663455222245,
+      "grad_norm": 0.08537210014134117,
+      "learning_rate": 5.133038237502983e-06,
+      "loss": 0.2954,
+      "step": 2382
+    },
+    {
+      "epoch": 1.3916502445077001,
+      "grad_norm": 0.08577972042054197,
+      "learning_rate": 5.1241332547188535e-06,
+      "loss": 0.3177,
+      "step": 2383
+    },
+    {
+      "epoch": 1.3922341434931758,
+      "grad_norm": 0.08624466523236124,
+      "learning_rate": 5.1152333414698774e-06,
+      "loss": 0.3236,
+      "step": 2384
+    },
+    {
+      "epoch": 1.3928180424786512,
+      "grad_norm": 0.09353283334846067,
+      "learning_rate": 5.106338507009478e-06,
+      "loss": 0.3221,
+      "step": 2385
+    },
+    {
+      "epoch": 1.3934019414641268,
+      "grad_norm": 0.09571475344725501,
+      "learning_rate": 5.097448760585784e-06,
+      "loss": 0.3221,
+      "step": 2386
+    },
+    {
+      "epoch": 1.3939858404496022,
+      "grad_norm": 0.08748919193647021,
+      "learning_rate": 5.088564111441645e-06,
+      "loss": 0.3202,
+      "step": 2387
+    },
+    {
+      "epoch": 1.3945697394350778,
+      "grad_norm": 0.08785411498501466,
+      "learning_rate": 5.079684568814607e-06,
+      "loss": 0.3203,
+      "step": 2388
+    },
+    {
+      "epoch": 1.3951536384205532,
+      "grad_norm": 0.085007277028097,
+      "learning_rate": 5.070810141936901e-06,
+      "loss": 0.2812,
+      "step": 2389
+    },
+    {
+      "epoch": 1.3957375374060288,
+      "grad_norm": 0.08782301937527626,
+      "learning_rate": 5.06194084003545e-06,
+      "loss": 0.3074,
+      "step": 2390
+    },
+    {
+      "epoch": 1.3963214363915042,
+      "grad_norm": 0.08577821061519456,
+      "learning_rate": 5.053076672331837e-06,
+      "loss": 0.3116,
+      "step": 2391
+    },
+    {
+      "epoch": 1.3969053353769798,
+      "grad_norm": 0.08719319621376763,
+      "learning_rate": 5.04421764804232e-06,
+      "loss": 0.2893,
+      "step": 2392
+    },
+    {
+      "epoch": 1.3974892343624554,
+      "grad_norm": 0.09409745826865988,
+      "learning_rate": 5.035363776377797e-06,
+      "loss": 0.284,
+      "step": 2393
+    },
+    {
+      "epoch": 1.3980731333479308,
+      "grad_norm": 0.09491367832335258,
+      "learning_rate": 5.026515066543813e-06,
+      "loss": 0.3567,
+      "step": 2394
+    },
+    {
+      "epoch": 1.3986570323334062,
+      "grad_norm": 0.08699483500295582,
+      "learning_rate": 5.017671527740551e-06,
+      "loss": 0.2984,
+      "step": 2395
+    },
+    {
+      "epoch": 1.3992409313188818,
+      "grad_norm": 0.08534261926447664,
+      "learning_rate": 5.008833169162805e-06,
+      "loss": 0.2905,
+      "step": 2396
+    },
+    {
+      "epoch": 1.3998248303043574,
+      "grad_norm": 0.08896682234944585,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.3109,
+      "step": 2397
+    },
+    {
+      "epoch": 1.4004087292898328,
+      "grad_norm": 0.09542845472048445,
+      "learning_rate": 4.991172029436157e-06,
+      "loss": 0.3127,
+      "step": 2398
+    },
+    {
+      "epoch": 1.4009926282753082,
+      "grad_norm": 0.08267081227693292,
+      "learning_rate": 4.982349266649884e-06,
+      "loss": 0.2984,
+      "step": 2399
+    },
+    {
+      "epoch": 1.4015765272607839,
+      "grad_norm": 0.07555139697184274,
+      "learning_rate": 4.97353172081439e-06,
+      "loss": 0.2562,
+      "step": 2400
+    },
+    {
+      "epoch": 1.4021604262462595,
+      "grad_norm": 0.08762790621849294,
+      "learning_rate": 4.964719401097444e-06,
+      "loss": 0.3321,
+      "step": 2401
+    },
+    {
+      "epoch": 1.4027443252317349,
+      "grad_norm": 0.07992063126116558,
+      "learning_rate": 4.9559123166613935e-06,
+      "loss": 0.2853,
+      "step": 2402
+    },
+    {
+      "epoch": 1.4033282242172105,
+      "grad_norm": 0.09566578912037214,
+      "learning_rate": 4.947110476663143e-06,
+      "loss": 0.332,
+      "step": 2403
+    },
+    {
+      "epoch": 1.4039121232026859,
+      "grad_norm": 0.08774570136250857,
+      "learning_rate": 4.93831389025413e-06,
+      "loss": 0.3162,
+      "step": 2404
+    },
+    {
+      "epoch": 1.4044960221881615,
+      "grad_norm": 0.0862545039885994,
+      "learning_rate": 4.9295225665803504e-06,
+      "loss": 0.2613,
+      "step": 2405
+    },
+    {
+      "epoch": 1.405079921173637,
+      "grad_norm": 0.09097106258557923,
+      "learning_rate": 4.9207365147823114e-06,
+      "loss": 0.3394,
+      "step": 2406
+    },
+    {
+      "epoch": 1.4056638201591125,
+      "grad_norm": 0.09287318621150203,
+      "learning_rate": 4.911955743995042e-06,
+      "loss": 0.3543,
+      "step": 2407
+    },
+    {
+      "epoch": 1.406247719144588,
+      "grad_norm": 0.08189020334822691,
+      "learning_rate": 4.90318026334809e-06,
+      "loss": 0.2916,
+      "step": 2408
+    },
+    {
+      "epoch": 1.4068316181300635,
+      "grad_norm": 0.08878781225031367,
+      "learning_rate": 4.894410081965489e-06,
+      "loss": 0.3321,
+      "step": 2409
+    },
+    {
+      "epoch": 1.4074155171155391,
+      "grad_norm": 0.082758651822754,
+      "learning_rate": 4.885645208965779e-06,
+      "loss": 0.2702,
+      "step": 2410
+    },
+    {
+      "epoch": 1.4079994161010145,
+      "grad_norm": 0.09404912360995255,
+      "learning_rate": 4.876885653461967e-06,
+      "loss": 0.3657,
+      "step": 2411
+    },
+    {
+      "epoch": 1.40858331508649,
+      "grad_norm": 0.08608491471312288,
+      "learning_rate": 4.868131424561532e-06,
+      "loss": 0.2857,
+      "step": 2412
+    },
+    {
+      "epoch": 1.4091672140719655,
+      "grad_norm": 0.08666305586963424,
+      "learning_rate": 4.859382531366428e-06,
+      "loss": 0.3049,
+      "step": 2413
+    },
+    {
+      "epoch": 1.4097511130574412,
+      "grad_norm": 0.0882071734113627,
+      "learning_rate": 4.850638982973043e-06,
+      "loss": 0.3226,
+      "step": 2414
+    },
+    {
+      "epoch": 1.4103350120429166,
+      "grad_norm": 0.08999970022914001,
+      "learning_rate": 4.841900788472227e-06,
+      "loss": 0.2961,
+      "step": 2415
+    },
+    {
+      "epoch": 1.410918911028392,
+      "grad_norm": 0.09166370951676155,
+      "learning_rate": 4.833167956949249e-06,
+      "loss": 0.3046,
+      "step": 2416
+    },
+    {
+      "epoch": 1.4115028100138676,
+      "grad_norm": 0.08465015329242134,
+      "learning_rate": 4.824440497483802e-06,
+      "loss": 0.2776,
+      "step": 2417
+    },
+    {
+      "epoch": 1.4120867089993432,
+      "grad_norm": 0.09075217748032732,
+      "learning_rate": 4.815718419150007e-06,
+      "loss": 0.2953,
+      "step": 2418
+    },
+    {
+      "epoch": 1.4126706079848186,
+      "grad_norm": 0.09270352928642503,
+      "learning_rate": 4.807001731016374e-06,
+      "loss": 0.3487,
+      "step": 2419
+    },
+    {
+      "epoch": 1.4132545069702942,
+      "grad_norm": 0.0890897409402727,
+      "learning_rate": 4.7982904421458245e-06,
+      "loss": 0.3153,
+      "step": 2420
+    },
+    {
+      "epoch": 1.4138384059557696,
+      "grad_norm": 0.09086247470280864,
+      "learning_rate": 4.789584561595651e-06,
+      "loss": 0.3226,
+      "step": 2421
+    },
+    {
+      "epoch": 1.4144223049412452,
+      "grad_norm": 0.09642511015513601,
+      "learning_rate": 4.780884098417531e-06,
+      "loss": 0.3498,
+      "step": 2422
+    },
+    {
+      "epoch": 1.4150062039267206,
+      "grad_norm": 0.08838015093516857,
+      "learning_rate": 4.772189061657511e-06,
+      "loss": 0.2888,
+      "step": 2423
+    },
+    {
+      "epoch": 1.4155901029121962,
+      "grad_norm": 0.08774580734779357,
+      "learning_rate": 4.763499460355988e-06,
+      "loss": 0.2835,
+      "step": 2424
+    },
+    {
+      "epoch": 1.4161740018976716,
+      "grad_norm": 0.09392867797902595,
+      "learning_rate": 4.7548153035477185e-06,
+      "loss": 0.327,
+      "step": 2425
+    },
+    {
+      "epoch": 1.4167579008831472,
+      "grad_norm": 0.08513587394673863,
+      "learning_rate": 4.746136600261791e-06,
+      "loss": 0.2922,
+      "step": 2426
+    },
+    {
+      "epoch": 1.4173417998686229,
+      "grad_norm": 0.08598712982665328,
+      "learning_rate": 4.737463359521618e-06,
+      "loss": 0.3111,
+      "step": 2427
+    },
+    {
+      "epoch": 1.4179256988540982,
+      "grad_norm": 0.09614569543969825,
+      "learning_rate": 4.728795590344948e-06,
+      "loss": 0.3494,
+      "step": 2428
+    },
+    {
+      "epoch": 1.4185095978395736,
+      "grad_norm": 0.09363530354029877,
+      "learning_rate": 4.7201333017438266e-06,
+      "loss": 0.3317,
+      "step": 2429
+    },
+    {
+      "epoch": 1.4190934968250493,
+      "grad_norm": 0.08825998766715494,
+      "learning_rate": 4.711476502724609e-06,
+      "loss": 0.3199,
+      "step": 2430
+    },
+    {
+      "epoch": 1.4196773958105249,
+      "grad_norm": 0.08544593938086152,
+      "learning_rate": 4.702825202287944e-06,
+      "loss": 0.2963,
+      "step": 2431
+    },
+    {
+      "epoch": 1.4202612947960003,
+      "grad_norm": 0.09204331837830147,
+      "learning_rate": 4.694179409428752e-06,
+      "loss": 0.3226,
+      "step": 2432
+    },
+    {
+      "epoch": 1.4208451937814757,
+      "grad_norm": 0.07907389215844314,
+      "learning_rate": 4.685539133136244e-06,
+      "loss": 0.2663,
+      "step": 2433
+    },
+    {
+      "epoch": 1.4214290927669513,
+      "grad_norm": 0.09382573108501736,
+      "learning_rate": 4.6769043823938806e-06,
+      "loss": 0.3079,
+      "step": 2434
+    },
+    {
+      "epoch": 1.422012991752427,
+      "grad_norm": 0.0975248486171169,
+      "learning_rate": 4.668275166179383e-06,
+      "loss": 0.3508,
+      "step": 2435
+    },
+    {
+      "epoch": 1.4225968907379023,
+      "grad_norm": 0.0863274966040468,
+      "learning_rate": 4.659651493464721e-06,
+      "loss": 0.2917,
+      "step": 2436
+    },
+    {
+      "epoch": 1.423180789723378,
+      "grad_norm": 0.0921352606848789,
+      "learning_rate": 4.6510333732160915e-06,
+      "loss": 0.3197,
+      "step": 2437
+    },
+    {
+      "epoch": 1.4237646887088533,
+      "grad_norm": 0.08140522124410668,
+      "learning_rate": 4.642420814393934e-06,
+      "loss": 0.2944,
+      "step": 2438
+    },
+    {
+      "epoch": 1.424348587694329,
+      "grad_norm": 0.08946623920696974,
+      "learning_rate": 4.633813825952892e-06,
+      "loss": 0.2938,
+      "step": 2439
+    },
+    {
+      "epoch": 1.4249324866798043,
+      "grad_norm": 0.08626232010439691,
+      "learning_rate": 4.625212416841816e-06,
+      "loss": 0.3057,
+      "step": 2440
+    },
+    {
+      "epoch": 1.42551638566528,
+      "grad_norm": 0.08721477450878029,
+      "learning_rate": 4.616616596003772e-06,
+      "loss": 0.3459,
+      "step": 2441
+    },
+    {
+      "epoch": 1.4261002846507553,
+      "grad_norm": 0.09586819438924457,
+      "learning_rate": 4.6080263723759955e-06,
+      "loss": 0.3454,
+      "step": 2442
+    },
+    {
+      "epoch": 1.426684183636231,
+      "grad_norm": 0.10022023133650289,
+      "learning_rate": 4.599441754889919e-06,
+      "loss": 0.3872,
+      "step": 2443
+    },
+    {
+      "epoch": 1.4272680826217066,
+      "grad_norm": 0.0951556647976026,
+      "learning_rate": 4.590862752471138e-06,
+      "loss": 0.3435,
+      "step": 2444
+    },
+    {
+      "epoch": 1.427851981607182,
+      "grad_norm": 0.09013735167006554,
+      "learning_rate": 4.582289374039405e-06,
+      "loss": 0.2984,
+      "step": 2445
+    },
+    {
+      "epoch": 1.4284358805926574,
+      "grad_norm": 0.09192561567421383,
+      "learning_rate": 4.573721628508638e-06,
+      "loss": 0.3211,
+      "step": 2446
+    },
+    {
+      "epoch": 1.429019779578133,
+      "grad_norm": 0.08658841492618176,
+      "learning_rate": 4.565159524786888e-06,
+      "loss": 0.2843,
+      "step": 2447
+    },
+    {
+      "epoch": 1.4296036785636086,
+      "grad_norm": 0.08465490994194322,
+      "learning_rate": 4.556603071776347e-06,
+      "loss": 0.2609,
+      "step": 2448
+    },
+    {
+      "epoch": 1.430187577549084,
+      "grad_norm": 0.09346263044513814,
+      "learning_rate": 4.548052278373327e-06,
+      "loss": 0.2985,
+      "step": 2449
+    },
+    {
+      "epoch": 1.4307714765345594,
+      "grad_norm": 0.09273803515427384,
+      "learning_rate": 4.539507153468254e-06,
+      "loss": 0.3277,
+      "step": 2450
+    },
+    {
+      "epoch": 1.431355375520035,
+      "grad_norm": 0.09549070998421809,
+      "learning_rate": 4.530967705945668e-06,
+      "loss": 0.3502,
+      "step": 2451
+    },
+    {
+      "epoch": 1.4319392745055106,
+      "grad_norm": 0.09501560892946584,
+      "learning_rate": 4.522433944684197e-06,
+      "loss": 0.3388,
+      "step": 2452
+    },
+    {
+      "epoch": 1.432523173490986,
+      "grad_norm": 0.07837538135781677,
+      "learning_rate": 4.513905878556568e-06,
+      "loss": 0.2731,
+      "step": 2453
+    },
+    {
+      "epoch": 1.4331070724764616,
+      "grad_norm": 0.09005603605395537,
+      "learning_rate": 4.505383516429577e-06,
+      "loss": 0.3295,
+      "step": 2454
+    },
+    {
+      "epoch": 1.433690971461937,
+      "grad_norm": 0.08573250184570594,
+      "learning_rate": 4.496866867164087e-06,
+      "loss": 0.308,
+      "step": 2455
+    },
+    {
+      "epoch": 1.4342748704474126,
+      "grad_norm": 0.08149194420795365,
+      "learning_rate": 4.488355939615035e-06,
+      "loss": 0.289,
+      "step": 2456
+    },
+    {
+      "epoch": 1.434858769432888,
+      "grad_norm": 0.08809247107130126,
+      "learning_rate": 4.479850742631396e-06,
+      "loss": 0.3247,
+      "step": 2457
+    },
+    {
+      "epoch": 1.4354426684183637,
+      "grad_norm": 0.08286467904174094,
+      "learning_rate": 4.471351285056192e-06,
+      "loss": 0.3205,
+      "step": 2458
+    },
+    {
+      "epoch": 1.436026567403839,
+      "grad_norm": 0.08539404201916423,
+      "learning_rate": 4.462857575726482e-06,
+      "loss": 0.2813,
+      "step": 2459
+    },
+    {
+      "epoch": 1.4366104663893147,
+      "grad_norm": 0.0867626928741596,
+      "learning_rate": 4.454369623473337e-06,
+      "loss": 0.3171,
+      "step": 2460
+    },
+    {
+      "epoch": 1.4371943653747903,
+      "grad_norm": 0.08793526882415201,
+      "learning_rate": 4.445887437121855e-06,
+      "loss": 0.3077,
+      "step": 2461
+    },
+    {
+      "epoch": 1.4377782643602657,
+      "grad_norm": 0.08385443334026675,
+      "learning_rate": 4.437411025491131e-06,
+      "loss": 0.2824,
+      "step": 2462
+    },
+    {
+      "epoch": 1.438362163345741,
+      "grad_norm": 0.0902830362514876,
+      "learning_rate": 4.428940397394253e-06,
+      "loss": 0.3203,
+      "step": 2463
+    },
+    {
+      "epoch": 1.4389460623312167,
+      "grad_norm": 0.0900230659293957,
+      "learning_rate": 4.420475561638309e-06,
+      "loss": 0.3116,
+      "step": 2464
+    },
+    {
+      "epoch": 1.4395299613166923,
+      "grad_norm": 0.09086123510874876,
+      "learning_rate": 4.412016527024348e-06,
+      "loss": 0.2883,
+      "step": 2465
+    },
+    {
+      "epoch": 1.4401138603021677,
+      "grad_norm": 0.08856685186558118,
+      "learning_rate": 4.4035633023474035e-06,
+      "loss": 0.3056,
+      "step": 2466
+    },
+    {
+      "epoch": 1.4406977592876433,
+      "grad_norm": 0.09319706644645542,
+      "learning_rate": 4.395115896396457e-06,
+      "loss": 0.3283,
+      "step": 2467
+    },
+    {
+      "epoch": 1.4412816582731187,
+      "grad_norm": 0.08655798014194706,
+      "learning_rate": 4.386674317954439e-06,
+      "loss": 0.2805,
+      "step": 2468
+    },
+    {
+      "epoch": 1.4418655572585943,
+      "grad_norm": 0.0957281520050778,
+      "learning_rate": 4.378238575798233e-06,
+      "loss": 0.3286,
+      "step": 2469
+    },
+    {
+      "epoch": 1.4424494562440697,
+      "grad_norm": 0.09174174516430338,
+      "learning_rate": 4.3698086786986425e-06,
+      "loss": 0.3133,
+      "step": 2470
+    },
+    {
+      "epoch": 1.4430333552295453,
+      "grad_norm": 0.087874782509742,
+      "learning_rate": 4.3613846354204025e-06,
+      "loss": 0.301,
+      "step": 2471
+    },
+    {
+      "epoch": 1.4436172542150207,
+      "grad_norm": 0.0891886133675583,
+      "learning_rate": 4.352966454722155e-06,
+      "loss": 0.3226,
+      "step": 2472
+    },
+    {
+      "epoch": 1.4442011532004964,
+      "grad_norm": 0.09246839920441298,
+      "learning_rate": 4.344554145356447e-06,
+      "loss": 0.3327,
+      "step": 2473
+    },
+    {
+      "epoch": 1.4447850521859718,
+      "grad_norm": 0.08521094566517058,
+      "learning_rate": 4.336147716069727e-06,
+      "loss": 0.2907,
+      "step": 2474
+    },
+    {
+      "epoch": 1.4453689511714474,
+      "grad_norm": 0.08996415330368689,
+      "learning_rate": 4.327747175602321e-06,
+      "loss": 0.3529,
+      "step": 2475
+    },
+    {
+      "epoch": 1.4459528501569228,
+      "grad_norm": 0.09170278116841499,
+      "learning_rate": 4.319352532688444e-06,
+      "loss": 0.3308,
+      "step": 2476
+    },
+    {
+      "epoch": 1.4465367491423984,
+      "grad_norm": 0.08784407763378779,
+      "learning_rate": 4.310963796056168e-06,
+      "loss": 0.2927,
+      "step": 2477
+    },
+    {
+      "epoch": 1.447120648127874,
+      "grad_norm": 0.08426729745354707,
+      "learning_rate": 4.302580974427426e-06,
+      "loss": 0.2855,
+      "step": 2478
+    },
+    {
+      "epoch": 1.4477045471133494,
+      "grad_norm": 0.09017139402593817,
+      "learning_rate": 4.29420407651801e-06,
+      "loss": 0.3339,
+      "step": 2479
+    },
+    {
+      "epoch": 1.4482884460988248,
+      "grad_norm": 0.09570502138010142,
+      "learning_rate": 4.2858331110375406e-06,
+      "loss": 0.3337,
+      "step": 2480
+    },
+    {
+      "epoch": 1.4488723450843004,
+      "grad_norm": 0.08761709335228135,
+      "learning_rate": 4.277468086689481e-06,
+      "loss": 0.3242,
+      "step": 2481
+    },
+    {
+      "epoch": 1.449456244069776,
+      "grad_norm": 0.08952369898267282,
+      "learning_rate": 4.269109012171112e-06,
+      "loss": 0.3215,
+      "step": 2482
+    },
+    {
+      "epoch": 1.4500401430552514,
+      "grad_norm": 0.09692146205841282,
+      "learning_rate": 4.260755896173523e-06,
+      "loss": 0.349,
+      "step": 2483
+    },
+    {
+      "epoch": 1.450624042040727,
+      "grad_norm": 0.09227519712941402,
+      "learning_rate": 4.252408747381622e-06,
+      "loss": 0.316,
+      "step": 2484
+    },
+    {
+      "epoch": 1.4512079410262024,
+      "grad_norm": 0.08826145566539562,
+      "learning_rate": 4.244067574474098e-06,
+      "loss": 0.3059,
+      "step": 2485
+    },
+    {
+      "epoch": 1.451791840011678,
+      "grad_norm": 0.09194191139863438,
+      "learning_rate": 4.235732386123437e-06,
+      "loss": 0.3416,
+      "step": 2486
+    },
+    {
+      "epoch": 1.4523757389971534,
+      "grad_norm": 0.08608168029458155,
+      "learning_rate": 4.227403190995901e-06,
+      "loss": 0.3003,
+      "step": 2487
+    },
+    {
+      "epoch": 1.452959637982629,
+      "grad_norm": 0.0959907646309117,
+      "learning_rate": 4.219079997751515e-06,
+      "loss": 0.3489,
+      "step": 2488
+    },
+    {
+      "epoch": 1.4535435369681045,
+      "grad_norm": 0.09464982585528525,
+      "learning_rate": 4.210762815044073e-06,
+      "loss": 0.3455,
+      "step": 2489
+    },
+    {
+      "epoch": 1.45412743595358,
+      "grad_norm": 0.08218023649052496,
+      "learning_rate": 4.20245165152111e-06,
+      "loss": 0.2836,
+      "step": 2490
+    },
+    {
+      "epoch": 1.4547113349390555,
+      "grad_norm": 0.0879463226813803,
+      "learning_rate": 4.194146515823906e-06,
+      "loss": 0.3077,
+      "step": 2491
+    },
+    {
+      "epoch": 1.455295233924531,
+      "grad_norm": 0.10010851134303636,
+      "learning_rate": 4.185847416587481e-06,
+      "loss": 0.3313,
+      "step": 2492
+    },
+    {
+      "epoch": 1.4558791329100065,
+      "grad_norm": 0.09235132961683659,
+      "learning_rate": 4.177554362440565e-06,
+      "loss": 0.3136,
+      "step": 2493
+    },
+    {
+      "epoch": 1.456463031895482,
+      "grad_norm": 0.08961660670287878,
+      "learning_rate": 4.169267362005619e-06,
+      "loss": 0.3028,
+      "step": 2494
+    },
+    {
+      "epoch": 1.4570469308809577,
+      "grad_norm": 0.09266136047414188,
+      "learning_rate": 4.160986423898798e-06,
+      "loss": 0.334,
+      "step": 2495
+    },
+    {
+      "epoch": 1.457630829866433,
+      "grad_norm": 0.0821071262326694,
+      "learning_rate": 4.15271155672995e-06,
+      "loss": 0.2923,
+      "step": 2496
+    },
+    {
+      "epoch": 1.4582147288519085,
+      "grad_norm": 0.09444673141507093,
+      "learning_rate": 4.14444276910263e-06,
+      "loss": 0.337,
+      "step": 2497
+    },
+    {
+      "epoch": 1.4587986278373841,
+      "grad_norm": 0.08975942151674045,
+      "learning_rate": 4.1361800696140505e-06,
+      "loss": 0.3001,
+      "step": 2498
+    },
+    {
+      "epoch": 1.4593825268228597,
+      "grad_norm": 0.09215004508796767,
+      "learning_rate": 4.127923466855111e-06,
+      "loss": 0.3095,
+      "step": 2499
+    },
+    {
+      "epoch": 1.4599664258083351,
+      "grad_norm": 0.08708867568318807,
+      "learning_rate": 4.119672969410362e-06,
+      "loss": 0.308,
+      "step": 2500
+    },
+    {
+      "epoch": 1.4605503247938108,
+      "grad_norm": 0.09326820332236949,
+      "learning_rate": 4.111428585858005e-06,
+      "loss": 0.3422,
+      "step": 2501
+    },
+    {
+      "epoch": 1.4611342237792861,
+      "grad_norm": 0.08289959794029202,
+      "learning_rate": 4.103190324769895e-06,
+      "loss": 0.2926,
+      "step": 2502
+    },
+    {
+      "epoch": 1.4617181227647618,
+      "grad_norm": 0.0834547058990696,
+      "learning_rate": 4.0949581947115106e-06,
+      "loss": 0.2835,
+      "step": 2503
+    },
+    {
+      "epoch": 1.4623020217502372,
+      "grad_norm": 0.08700797377548518,
+      "learning_rate": 4.086732204241964e-06,
+      "loss": 0.3013,
+      "step": 2504
+    },
+    {
+      "epoch": 1.4628859207357128,
+      "grad_norm": 0.09353897806755798,
+      "learning_rate": 4.07851236191398e-06,
+      "loss": 0.358,
+      "step": 2505
+    },
+    {
+      "epoch": 1.4634698197211882,
+      "grad_norm": 0.08453195798529367,
+      "learning_rate": 4.070298676273886e-06,
+      "loss": 0.2849,
+      "step": 2506
+    },
+    {
+      "epoch": 1.4640537187066638,
+      "grad_norm": 0.08699951849009825,
+      "learning_rate": 4.06209115586162e-06,
+      "loss": 0.3116,
+      "step": 2507
+    },
+    {
+      "epoch": 1.4646376176921392,
+      "grad_norm": 0.08702980350606691,
+      "learning_rate": 4.053889809210698e-06,
+      "loss": 0.3199,
+      "step": 2508
+    },
+    {
+      "epoch": 1.4652215166776148,
+      "grad_norm": 0.08654692030685524,
+      "learning_rate": 4.045694644848228e-06,
+      "loss": 0.308,
+      "step": 2509
+    },
+    {
+      "epoch": 1.4658054156630902,
+      "grad_norm": 0.08812140903377792,
+      "learning_rate": 4.037505671294883e-06,
+      "loss": 0.354,
+      "step": 2510
+    },
+    {
+      "epoch": 1.4663893146485658,
+      "grad_norm": 0.09341399788960499,
+      "learning_rate": 4.0293228970648955e-06,
+      "loss": 0.3465,
+      "step": 2511
+    },
+    {
+      "epoch": 1.4669732136340414,
+      "grad_norm": 0.09574093623005273,
+      "learning_rate": 4.021146330666065e-06,
+      "loss": 0.3282,
+      "step": 2512
+    },
+    {
+      "epoch": 1.4675571126195168,
+      "grad_norm": 0.09192262891393646,
+      "learning_rate": 4.012975980599724e-06,
+      "loss": 0.2745,
+      "step": 2513
+    },
+    {
+      "epoch": 1.4681410116049922,
+      "grad_norm": 0.09555074610348502,
+      "learning_rate": 4.0048118553607485e-06,
+      "loss": 0.3517,
+      "step": 2514
+    },
+    {
+      "epoch": 1.4687249105904678,
+      "grad_norm": 0.09453473794408544,
+      "learning_rate": 3.996653963437546e-06,
+      "loss": 0.3117,
+      "step": 2515
+    },
+    {
+      "epoch": 1.4693088095759435,
+      "grad_norm": 0.09111297853180635,
+      "learning_rate": 3.98850231331203e-06,
+      "loss": 0.3313,
+      "step": 2516
+    },
+    {
+      "epoch": 1.4698927085614188,
+      "grad_norm": 0.0915481812020442,
+      "learning_rate": 3.980356913459642e-06,
+      "loss": 0.3254,
+      "step": 2517
+    },
+    {
+      "epoch": 1.4704766075468945,
+      "grad_norm": 0.085935465765466,
+      "learning_rate": 3.972217772349309e-06,
+      "loss": 0.2881,
+      "step": 2518
+    },
+    {
+      "epoch": 1.4710605065323699,
+      "grad_norm": 0.08834085941059203,
+      "learning_rate": 3.9640848984434556e-06,
+      "loss": 0.3164,
+      "step": 2519
+    },
+    {
+      "epoch": 1.4716444055178455,
+      "grad_norm": 0.08690860511106675,
+      "learning_rate": 3.955958300197998e-06,
+      "loss": 0.3025,
+      "step": 2520
+    },
+    {
+      "epoch": 1.4722283045033209,
+      "grad_norm": 0.08148579331762949,
+      "learning_rate": 3.947837986062314e-06,
+      "loss": 0.2704,
+      "step": 2521
+    },
+    {
+      "epoch": 1.4728122034887965,
+      "grad_norm": 0.08387071056018273,
+      "learning_rate": 3.939723964479262e-06,
+      "loss": 0.295,
+      "step": 2522
+    },
+    {
+      "epoch": 1.4733961024742719,
+      "grad_norm": 0.08963981597529543,
+      "learning_rate": 3.931616243885148e-06,
+      "loss": 0.3272,
+      "step": 2523
+    },
+    {
+      "epoch": 1.4739800014597475,
+      "grad_norm": 0.08356233305497361,
+      "learning_rate": 3.923514832709725e-06,
+      "loss": 0.3111,
+      "step": 2524
+    },
+    {
+      "epoch": 1.474563900445223,
+      "grad_norm": 0.08596163294581784,
+      "learning_rate": 3.915419739376198e-06,
+      "loss": 0.3006,
+      "step": 2525
+    },
+    {
+      "epoch": 1.4751477994306985,
+      "grad_norm": 0.08804074162552146,
+      "learning_rate": 3.90733097230119e-06,
+      "loss": 0.2985,
+      "step": 2526
+    },
+    {
+      "epoch": 1.475731698416174,
+      "grad_norm": 0.0926115632772681,
+      "learning_rate": 3.899248539894756e-06,
+      "loss": 0.3014,
+      "step": 2527
+    },
+    {
+      "epoch": 1.4763155974016495,
+      "grad_norm": 0.0823610476263303,
+      "learning_rate": 3.891172450560362e-06,
+      "loss": 0.2724,
+      "step": 2528
+    },
+    {
+      "epoch": 1.4768994963871251,
+      "grad_norm": 0.09358401388599763,
+      "learning_rate": 3.883102712694871e-06,
+      "loss": 0.3373,
+      "step": 2529
+    },
+    {
+      "epoch": 1.4774833953726005,
+      "grad_norm": 0.09842095017714672,
+      "learning_rate": 3.875039334688556e-06,
+      "loss": 0.3286,
+      "step": 2530
+    },
+    {
+      "epoch": 1.478067294358076,
+      "grad_norm": 0.09159978300811596,
+      "learning_rate": 3.866982324925066e-06,
+      "loss": 0.3216,
+      "step": 2531
+    },
+    {
+      "epoch": 1.4786511933435516,
+      "grad_norm": 0.09143136021217294,
+      "learning_rate": 3.858931691781439e-06,
+      "loss": 0.3052,
+      "step": 2532
+    },
+    {
+      "epoch": 1.4792350923290272,
+      "grad_norm": 0.09286169910051506,
+      "learning_rate": 3.850887443628075e-06,
+      "loss": 0.3081,
+      "step": 2533
+    },
+    {
+      "epoch": 1.4798189913145026,
+      "grad_norm": 0.09104920587627532,
+      "learning_rate": 3.842849588828733e-06,
+      "loss": 0.3216,
+      "step": 2534
+    },
+    {
+      "epoch": 1.4804028902999782,
+      "grad_norm": 0.08598900659331898,
+      "learning_rate": 3.834818135740539e-06,
+      "loss": 0.2934,
+      "step": 2535
+    },
+    {
+      "epoch": 1.4809867892854536,
+      "grad_norm": 0.0858075751705307,
+      "learning_rate": 3.826793092713944e-06,
+      "loss": 0.3107,
+      "step": 2536
+    },
+    {
+      "epoch": 1.4815706882709292,
+      "grad_norm": 0.08846780349649745,
+      "learning_rate": 3.818774468092754e-06,
+      "loss": 0.3118,
+      "step": 2537
+    },
+    {
+      "epoch": 1.4821545872564046,
+      "grad_norm": 0.09183453806695745,
+      "learning_rate": 3.8107622702140856e-06,
+      "loss": 0.3421,
+      "step": 2538
+    },
+    {
+      "epoch": 1.4827384862418802,
+      "grad_norm": 0.08944407371905683,
+      "learning_rate": 3.802756507408377e-06,
+      "loss": 0.3556,
+      "step": 2539
+    },
+    {
+      "epoch": 1.4833223852273556,
+      "grad_norm": 0.08135966896893304,
+      "learning_rate": 3.794757187999386e-06,
+      "loss": 0.3142,
+      "step": 2540
+    },
+    {
+      "epoch": 1.4839062842128312,
+      "grad_norm": 0.0864453566387568,
+      "learning_rate": 3.7867643203041548e-06,
+      "loss": 0.3282,
+      "step": 2541
+    },
+    {
+      "epoch": 1.4844901831983066,
+      "grad_norm": 0.087031915097657,
+      "learning_rate": 3.7787779126330314e-06,
+      "loss": 0.3012,
+      "step": 2542
+    },
+    {
+      "epoch": 1.4850740821837822,
+      "grad_norm": 0.08715968884730894,
+      "learning_rate": 3.770797973289644e-06,
+      "loss": 0.3295,
+      "step": 2543
+    },
+    {
+      "epoch": 1.4856579811692576,
+      "grad_norm": 0.08197184994910715,
+      "learning_rate": 3.762824510570887e-06,
+      "loss": 0.2656,
+      "step": 2544
+    },
+    {
+      "epoch": 1.4862418801547332,
+      "grad_norm": 0.0919329697580388,
+      "learning_rate": 3.7548575327669345e-06,
+      "loss": 0.309,
+      "step": 2545
+    },
+    {
+      "epoch": 1.4868257791402089,
+      "grad_norm": 0.082394001534849,
+      "learning_rate": 3.7468970481612077e-06,
+      "loss": 0.2752,
+      "step": 2546
+    },
+    {
+      "epoch": 1.4874096781256843,
+      "grad_norm": 0.08370668747793851,
+      "learning_rate": 3.738943065030376e-06,
+      "loss": 0.2961,
+      "step": 2547
+    },
+    {
+      "epoch": 1.4879935771111596,
+      "grad_norm": 0.07957603365034668,
+      "learning_rate": 3.7309955916443597e-06,
+      "loss": 0.2674,
+      "step": 2548
+    },
+    {
+      "epoch": 1.4885774760966353,
+      "grad_norm": 0.08246809874533066,
+      "learning_rate": 3.723054636266299e-06,
+      "loss": 0.3126,
+      "step": 2549
+    },
+    {
+      "epoch": 1.4891613750821109,
+      "grad_norm": 0.08718669196366648,
+      "learning_rate": 3.715120207152567e-06,
+      "loss": 0.3173,
+      "step": 2550
+    },
+    {
+      "epoch": 1.4897452740675863,
+      "grad_norm": 0.08689759381949419,
+      "learning_rate": 3.7071923125527444e-06,
+      "loss": 0.3086,
+      "step": 2551
+    },
+    {
+      "epoch": 1.490329173053062,
+      "grad_norm": 0.09389728537694979,
+      "learning_rate": 3.6992709607096167e-06,
+      "loss": 0.3244,
+      "step": 2552
+    },
+    {
+      "epoch": 1.4909130720385373,
+      "grad_norm": 0.08613427950092989,
+      "learning_rate": 3.6913561598591775e-06,
+      "loss": 0.2945,
+      "step": 2553
+    },
+    {
+      "epoch": 1.491496971024013,
+      "grad_norm": 0.08529020215127872,
+      "learning_rate": 3.683447918230594e-06,
+      "loss": 0.3411,
+      "step": 2554
+    },
+    {
+      "epoch": 1.4920808700094883,
+      "grad_norm": 0.1027036660729519,
+      "learning_rate": 3.6755462440462288e-06,
+      "loss": 0.3751,
+      "step": 2555
+    },
+    {
+      "epoch": 1.492664768994964,
+      "grad_norm": 0.08834290547249343,
+      "learning_rate": 3.6676511455216056e-06,
+      "loss": 0.3017,
+      "step": 2556
+    },
+    {
+      "epoch": 1.4932486679804393,
+      "grad_norm": 0.0846440868371805,
+      "learning_rate": 3.659762630865411e-06,
+      "loss": 0.3056,
+      "step": 2557
+    },
+    {
+      "epoch": 1.493832566965915,
+      "grad_norm": 0.08777381834680356,
+      "learning_rate": 3.651880708279497e-06,
+      "loss": 0.3086,
+      "step": 2558
+    },
+    {
+      "epoch": 1.4944164659513903,
+      "grad_norm": 0.08739889061831228,
+      "learning_rate": 3.6440053859588478e-06,
+      "loss": 0.287,
+      "step": 2559
+    },
+    {
+      "epoch": 1.495000364936866,
+      "grad_norm": 0.08638891838508245,
+      "learning_rate": 3.636136672091598e-06,
+      "loss": 0.3143,
+      "step": 2560
+    },
+    {
+      "epoch": 1.4955842639223413,
+      "grad_norm": 0.08007564456312086,
+      "learning_rate": 3.628274574859002e-06,
+      "loss": 0.2941,
+      "step": 2561
+    },
+    {
+      "epoch": 1.496168162907817,
+      "grad_norm": 0.08633312279033968,
+      "learning_rate": 3.6204191024354352e-06,
+      "loss": 0.3187,
+      "step": 2562
+    },
+    {
+      "epoch": 1.4967520618932926,
+      "grad_norm": 0.09373092378957466,
+      "learning_rate": 3.612570262988393e-06,
+      "loss": 0.3079,
+      "step": 2563
+    },
+    {
+      "epoch": 1.497335960878768,
+      "grad_norm": 0.08035826546848758,
+      "learning_rate": 3.604728064678464e-06,
+      "loss": 0.2506,
+      "step": 2564
+    },
+    {
+      "epoch": 1.4979198598642434,
+      "grad_norm": 0.0952779286144571,
+      "learning_rate": 3.5968925156593426e-06,
+      "loss": 0.3257,
+      "step": 2565
+    },
+    {
+      "epoch": 1.498503758849719,
+      "grad_norm": 0.08780523201890598,
+      "learning_rate": 3.589063624077802e-06,
+      "loss": 0.325,
+      "step": 2566
+    },
+    {
+      "epoch": 1.4990876578351946,
+      "grad_norm": 0.09102900967251021,
+      "learning_rate": 3.5812413980736916e-06,
+      "loss": 0.3239,
+      "step": 2567
+    },
+    {
+      "epoch": 1.49967155682067,
+      "grad_norm": 0.09645319185255825,
+      "learning_rate": 3.5734258457799407e-06,
+      "loss": 0.317,
+      "step": 2568
+    },
+    {
+      "epoch": 1.5002554558061454,
+      "grad_norm": 0.09728138018797342,
+      "learning_rate": 3.5656169753225278e-06,
+      "loss": 0.3436,
+      "step": 2569
+    },
+    {
+      "epoch": 1.500839354791621,
+      "grad_norm": 0.08432992169526402,
+      "learning_rate": 3.5578147948204934e-06,
+      "loss": 0.2931,
+      "step": 2570
+    },
+    {
+      "epoch": 1.5014232537770966,
+      "grad_norm": 0.08427283193044549,
+      "learning_rate": 3.5500193123859227e-06,
+      "loss": 0.2991,
+      "step": 2571
+    },
+    {
+      "epoch": 1.5020071527625722,
+      "grad_norm": 0.08516205272169775,
+      "learning_rate": 3.542230536123925e-06,
+      "loss": 0.2804,
+      "step": 2572
+    },
+    {
+      "epoch": 1.5025910517480476,
+      "grad_norm": 0.08711687393383395,
+      "learning_rate": 3.5344484741326533e-06,
+      "loss": 0.3178,
+      "step": 2573
+    },
+    {
+      "epoch": 1.503174950733523,
+      "grad_norm": 0.09019648000036438,
+      "learning_rate": 3.526673134503267e-06,
+      "loss": 0.3104,
+      "step": 2574
+    },
+    {
+      "epoch": 1.5037588497189986,
+      "grad_norm": 0.09111892814962166,
+      "learning_rate": 3.5189045253199384e-06,
+      "loss": 0.3397,
+      "step": 2575
+    },
+    {
+      "epoch": 1.5043427487044743,
+      "grad_norm": 0.09497949977015242,
+      "learning_rate": 3.51114265465985e-06,
+      "loss": 0.3235,
+      "step": 2576
+    },
+    {
+      "epoch": 1.5049266476899497,
+      "grad_norm": 0.09510012063995095,
+      "learning_rate": 3.5033875305931662e-06,
+      "loss": 0.2968,
+      "step": 2577
+    },
+    {
+      "epoch": 1.505510546675425,
+      "grad_norm": 0.09354129049682741,
+      "learning_rate": 3.4956391611830486e-06,
+      "loss": 0.3337,
+      "step": 2578
+    },
+    {
+      "epoch": 1.5060944456609007,
+      "grad_norm": 0.08617321825149186,
+      "learning_rate": 3.4878975544856285e-06,
+      "loss": 0.2979,
+      "step": 2579
+    },
+    {
+      "epoch": 1.5066783446463763,
+      "grad_norm": 0.08513666832112077,
+      "learning_rate": 3.4801627185500033e-06,
+      "loss": 0.2859,
+      "step": 2580
+    },
+    {
+      "epoch": 1.5072622436318517,
+      "grad_norm": 0.07998316309489245,
+      "learning_rate": 3.4724346614182427e-06,
+      "loss": 0.2882,
+      "step": 2581
+    },
+    {
+      "epoch": 1.507846142617327,
+      "grad_norm": 0.08720438796369215,
+      "learning_rate": 3.4647133911253516e-06,
+      "loss": 0.2862,
+      "step": 2582
+    },
+    {
+      "epoch": 1.5084300416028027,
+      "grad_norm": 0.08582321851668794,
+      "learning_rate": 3.4569989156992965e-06,
+      "loss": 0.3183,
+      "step": 2583
+    },
+    {
+      "epoch": 1.5090139405882783,
+      "grad_norm": 0.0894397431538933,
+      "learning_rate": 3.449291243160966e-06,
+      "loss": 0.2947,
+      "step": 2584
+    },
+    {
+      "epoch": 1.5095978395737537,
+      "grad_norm": 0.09886279619188078,
+      "learning_rate": 3.4415903815241757e-06,
+      "loss": 0.297,
+      "step": 2585
+    },
+    {
+      "epoch": 1.510181738559229,
+      "grad_norm": 0.08162536132423849,
+      "learning_rate": 3.4338963387956726e-06,
+      "loss": 0.3016,
+      "step": 2586
+    },
+    {
+      "epoch": 1.5107656375447047,
+      "grad_norm": 0.08217689376017993,
+      "learning_rate": 3.4262091229750973e-06,
+      "loss": 0.284,
+      "step": 2587
+    },
+    {
+      "epoch": 1.5113495365301803,
+      "grad_norm": 0.08697801738625241,
+      "learning_rate": 3.418528742055006e-06,
+      "loss": 0.3309,
+      "step": 2588
+    },
+    {
+      "epoch": 1.511933435515656,
+      "grad_norm": 0.08833226088593424,
+      "learning_rate": 3.4108552040208408e-06,
+      "loss": 0.3372,
+      "step": 2589
+    },
+    {
+      "epoch": 1.5125173345011313,
+      "grad_norm": 0.09153555701698159,
+      "learning_rate": 3.403188516850927e-06,
+      "loss": 0.3063,
+      "step": 2590
+    },
+    {
+      "epoch": 1.5131012334866067,
+      "grad_norm": 0.08701016215704771,
+      "learning_rate": 3.3955286885164786e-06,
+      "loss": 0.3091,
+      "step": 2591
+    },
+    {
+      "epoch": 1.5136851324720824,
+      "grad_norm": 0.08942316194979019,
+      "learning_rate": 3.387875726981563e-06,
+      "loss": 0.3136,
+      "step": 2592
+    },
+    {
+      "epoch": 1.514269031457558,
+      "grad_norm": 0.08685296712477052,
+      "learning_rate": 3.3802296402031234e-06,
+      "loss": 0.3019,
+      "step": 2593
+    },
+    {
+      "epoch": 1.5148529304430334,
+      "grad_norm": 0.08885651692227328,
+      "learning_rate": 3.3725904361309426e-06,
+      "loss": 0.293,
+      "step": 2594
+    },
+    {
+      "epoch": 1.5154368294285088,
+      "grad_norm": 0.0824016008849094,
+      "learning_rate": 3.36495812270765e-06,
+      "loss": 0.2541,
+      "step": 2595
+    },
+    {
+      "epoch": 1.5160207284139844,
+      "grad_norm": 0.08932890817357096,
+      "learning_rate": 3.35733270786872e-06,
+      "loss": 0.3116,
+      "step": 2596
+    },
+    {
+      "epoch": 1.51660462739946,
+      "grad_norm": 0.08482491120408663,
+      "learning_rate": 3.3497141995424397e-06,
+      "loss": 0.3292,
+      "step": 2597
+    },
+    {
+      "epoch": 1.5171885263849354,
+      "grad_norm": 0.08119165014150981,
+      "learning_rate": 3.3421026056499273e-06,
+      "loss": 0.277,
+      "step": 2598
+    },
+    {
+      "epoch": 1.5177724253704108,
+      "grad_norm": 0.0868736270830272,
+      "learning_rate": 3.3344979341051108e-06,
+      "loss": 0.3243,
+      "step": 2599
+    },
+    {
+      "epoch": 1.5183563243558864,
+      "grad_norm": 0.09033672235249522,
+      "learning_rate": 3.3269001928147103e-06,
+      "loss": 0.3016,
+      "step": 2600
+    },
+    {
+      "epoch": 1.518940223341362,
+      "grad_norm": 0.09944004646395133,
+      "learning_rate": 3.3193093896782546e-06,
+      "loss": 0.3519,
+      "step": 2601
+    },
+    {
+      "epoch": 1.5195241223268374,
+      "grad_norm": 0.09597794239635642,
+      "learning_rate": 3.311725532588049e-06,
+      "loss": 0.3187,
+      "step": 2602
+    },
+    {
+      "epoch": 1.5201080213123128,
+      "grad_norm": 0.0890008325539451,
+      "learning_rate": 3.3041486294291767e-06,
+      "loss": 0.2949,
+      "step": 2603
+    },
+    {
+      "epoch": 1.5206919202977884,
+      "grad_norm": 0.09454414583281463,
+      "learning_rate": 3.2965786880795005e-06,
+      "loss": 0.3619,
+      "step": 2604
+    },
+    {
+      "epoch": 1.521275819283264,
+      "grad_norm": 0.08636795796983165,
+      "learning_rate": 3.2890157164096315e-06,
+      "loss": 0.2883,
+      "step": 2605
+    },
+    {
+      "epoch": 1.5218597182687397,
+      "grad_norm": 0.09003312444819137,
+      "learning_rate": 3.2814597222829468e-06,
+      "loss": 0.3229,
+      "step": 2606
+    },
+    {
+      "epoch": 1.522443617254215,
+      "grad_norm": 0.0854418480269585,
+      "learning_rate": 3.2739107135555603e-06,
+      "loss": 0.3161,
+      "step": 2607
+    },
+    {
+      "epoch": 1.5230275162396905,
+      "grad_norm": 0.08841945236143223,
+      "learning_rate": 3.266368698076323e-06,
+      "loss": 0.3158,
+      "step": 2608
+    },
+    {
+      "epoch": 1.523611415225166,
+      "grad_norm": 0.08059841810495402,
+      "learning_rate": 3.258833683686824e-06,
+      "loss": 0.2673,
+      "step": 2609
+    },
+    {
+      "epoch": 1.5241953142106417,
+      "grad_norm": 0.08668302153005344,
+      "learning_rate": 3.251305678221359e-06,
+      "loss": 0.3137,
+      "step": 2610
+    },
+    {
+      "epoch": 1.524779213196117,
+      "grad_norm": 0.09260615073993698,
+      "learning_rate": 3.2437846895069535e-06,
+      "loss": 0.3278,
+      "step": 2611
+    },
+    {
+      "epoch": 1.5253631121815925,
+      "grad_norm": 0.08205763982597179,
+      "learning_rate": 3.236270725363323e-06,
+      "loss": 0.2898,
+      "step": 2612
+    },
+    {
+      "epoch": 1.525947011167068,
+      "grad_norm": 0.08638284077619501,
+      "learning_rate": 3.2287637936028814e-06,
+      "loss": 0.3314,
+      "step": 2613
+    },
+    {
+      "epoch": 1.5265309101525437,
+      "grad_norm": 0.08420395409632624,
+      "learning_rate": 3.2212639020307423e-06,
+      "loss": 0.2812,
+      "step": 2614
+    },
+    {
+      "epoch": 1.5271148091380191,
+      "grad_norm": 0.09587121444326648,
+      "learning_rate": 3.2137710584446837e-06,
+      "loss": 0.3294,
+      "step": 2615
+    },
+    {
+      "epoch": 1.5276987081234945,
+      "grad_norm": 0.09785324881432501,
+      "learning_rate": 3.20628527063517e-06,
+      "loss": 0.3227,
+      "step": 2616
+    },
+    {
+      "epoch": 1.5282826071089701,
+      "grad_norm": 0.08488280863699148,
+      "learning_rate": 3.1988065463853204e-06,
+      "loss": 0.302,
+      "step": 2617
+    },
+    {
+      "epoch": 1.5288665060944457,
+      "grad_norm": 0.08681742054960384,
+      "learning_rate": 3.1913348934709076e-06,
+      "loss": 0.3161,
+      "step": 2618
+    },
+    {
+      "epoch": 1.5294504050799211,
+      "grad_norm": 0.0851285568551277,
+      "learning_rate": 3.183870319660365e-06,
+      "loss": 0.3123,
+      "step": 2619
+    },
+    {
+      "epoch": 1.5300343040653965,
+      "grad_norm": 0.0852526574679343,
+      "learning_rate": 3.1764128327147515e-06,
+      "loss": 0.286,
+      "step": 2620
+    },
+    {
+      "epoch": 1.5306182030508722,
+      "grad_norm": 0.09291643393499747,
+      "learning_rate": 3.1689624403877685e-06,
+      "loss": 0.3256,
+      "step": 2621
+    },
+    {
+      "epoch": 1.5312021020363478,
+      "grad_norm": 0.09157277831990508,
+      "learning_rate": 3.161519150425735e-06,
+      "loss": 0.2874,
+      "step": 2622
+    },
+    {
+      "epoch": 1.5317860010218234,
+      "grad_norm": 0.08940371873080839,
+      "learning_rate": 3.1540829705675835e-06,
+      "loss": 0.3077,
+      "step": 2623
+    },
+    {
+      "epoch": 1.5323699000072988,
+      "grad_norm": 0.08728878185257997,
+      "learning_rate": 3.1466539085448624e-06,
+      "loss": 0.2869,
+      "step": 2624
+    },
+    {
+      "epoch": 1.5329537989927742,
+      "grad_norm": 0.09316312257007421,
+      "learning_rate": 3.139231972081709e-06,
+      "loss": 0.3298,
+      "step": 2625
+    },
+    {
+      "epoch": 1.5335376979782498,
+      "grad_norm": 0.08439190437917227,
+      "learning_rate": 3.1318171688948618e-06,
+      "loss": 0.311,
+      "step": 2626
+    },
+    {
+      "epoch": 1.5341215969637254,
+      "grad_norm": 0.08619739543320243,
+      "learning_rate": 3.1244095066936396e-06,
+      "loss": 0.2968,
+      "step": 2627
+    },
+    {
+      "epoch": 1.5347054959492008,
+      "grad_norm": 0.09031836659741173,
+      "learning_rate": 3.1170089931799296e-06,
+      "loss": 0.3067,
+      "step": 2628
+    },
+    {
+      "epoch": 1.5352893949346762,
+      "grad_norm": 0.09147912579465926,
+      "learning_rate": 3.1096156360482e-06,
+      "loss": 0.3066,
+      "step": 2629
+    },
+    {
+      "epoch": 1.5358732939201518,
+      "grad_norm": 0.09239847993874097,
+      "learning_rate": 3.102229442985466e-06,
+      "loss": 0.2842,
+      "step": 2630
+    },
+    {
+      "epoch": 1.5364571929056274,
+      "grad_norm": 0.0932862490367047,
+      "learning_rate": 3.094850421671295e-06,
+      "loss": 0.3193,
+      "step": 2631
+    },
+    {
+      "epoch": 1.5370410918911028,
+      "grad_norm": 0.0802383691153122,
+      "learning_rate": 3.0874785797778096e-06,
+      "loss": 0.2713,
+      "step": 2632
+    },
+    {
+      "epoch": 1.5376249908765782,
+      "grad_norm": 0.0946901563961915,
+      "learning_rate": 3.080113924969652e-06,
+      "loss": 0.3588,
+      "step": 2633
+    },
+    {
+      "epoch": 1.5382088898620538,
+      "grad_norm": 0.09047466756448537,
+      "learning_rate": 3.0727564649040066e-06,
+      "loss": 0.3214,
+      "step": 2634
+    },
+    {
+      "epoch": 1.5387927888475295,
+      "grad_norm": 0.08869462488459272,
+      "learning_rate": 3.0654062072305667e-06,
+      "loss": 0.3281,
+      "step": 2635
+    },
+    {
+      "epoch": 1.5393766878330049,
+      "grad_norm": 0.09153942732800854,
+      "learning_rate": 3.0580631595915368e-06,
+      "loss": 0.2945,
+      "step": 2636
+    },
+    {
+      "epoch": 1.5399605868184802,
+      "grad_norm": 0.10299553091052183,
+      "learning_rate": 3.050727329621637e-06,
+      "loss": 0.3749,
+      "step": 2637
+    },
+    {
+      "epoch": 1.5405444858039559,
+      "grad_norm": 0.08315789724672386,
+      "learning_rate": 3.043398724948068e-06,
+      "loss": 0.277,
+      "step": 2638
+    },
+    {
+      "epoch": 1.5411283847894315,
+      "grad_norm": 0.09154821697236457,
+      "learning_rate": 3.03607735319053e-06,
+      "loss": 0.3227,
+      "step": 2639
+    },
+    {
+      "epoch": 1.541712283774907,
+      "grad_norm": 0.09442003790507683,
+      "learning_rate": 3.028763221961196e-06,
+      "loss": 0.3264,
+      "step": 2640
+    },
+    {
+      "epoch": 1.5422961827603825,
+      "grad_norm": 0.09246789405883815,
+      "learning_rate": 3.02145633886471e-06,
+      "loss": 0.3177,
+      "step": 2641
+    },
+    {
+      "epoch": 1.542880081745858,
+      "grad_norm": 0.08642002676090976,
+      "learning_rate": 3.0141567114981897e-06,
+      "loss": 0.2907,
+      "step": 2642
+    },
+    {
+      "epoch": 1.5434639807313335,
+      "grad_norm": 0.08836257297793011,
+      "learning_rate": 3.006864347451195e-06,
+      "loss": 0.2962,
+      "step": 2643
+    },
+    {
+      "epoch": 1.5440478797168091,
+      "grad_norm": 0.09291633328775067,
+      "learning_rate": 2.999579254305748e-06,
+      "loss": 0.3212,
+      "step": 2644
+    },
+    {
+      "epoch": 1.5446317787022845,
+      "grad_norm": 0.09011500748885783,
+      "learning_rate": 2.992301439636299e-06,
+      "loss": 0.3215,
+      "step": 2645
+    },
+    {
+      "epoch": 1.54521567768776,
+      "grad_norm": 0.08633034570673986,
+      "learning_rate": 2.9850309110097364e-06,
+      "loss": 0.2964,
+      "step": 2646
+    },
+    {
+      "epoch": 1.5457995766732355,
+      "grad_norm": 0.09348633585019633,
+      "learning_rate": 2.977767675985377e-06,
+      "loss": 0.3514,
+      "step": 2647
+    },
+    {
+      "epoch": 1.5463834756587111,
+      "grad_norm": 0.09262124931307275,
+      "learning_rate": 2.970511742114943e-06,
+      "loss": 0.3744,
+      "step": 2648
+    },
+    {
+      "epoch": 1.5469673746441865,
+      "grad_norm": 0.07889584587092038,
+      "learning_rate": 2.963263116942581e-06,
+      "loss": 0.2881,
+      "step": 2649
+    },
+    {
+      "epoch": 1.547551273629662,
+      "grad_norm": 0.09101433026563387,
+      "learning_rate": 2.9560218080048243e-06,
+      "loss": 0.3095,
+      "step": 2650
+    },
+    {
+      "epoch": 1.5481351726151376,
+      "grad_norm": 0.0946802736960726,
+      "learning_rate": 2.9487878228306044e-06,
+      "loss": 0.3038,
+      "step": 2651
+    },
+    {
+      "epoch": 1.5487190716006132,
+      "grad_norm": 0.09843400435634764,
+      "learning_rate": 2.9415611689412426e-06,
+      "loss": 0.3143,
+      "step": 2652
+    },
+    {
+      "epoch": 1.5493029705860886,
+      "grad_norm": 0.09142875415468128,
+      "learning_rate": 2.9343418538504297e-06,
+      "loss": 0.3292,
+      "step": 2653
+    },
+    {
+      "epoch": 1.549886869571564,
+      "grad_norm": 0.08697338615257248,
+      "learning_rate": 2.9271298850642337e-06,
+      "loss": 0.2679,
+      "step": 2654
+    },
+    {
+      "epoch": 1.5504707685570396,
+      "grad_norm": 0.08213806150610116,
+      "learning_rate": 2.9199252700810833e-06,
+      "loss": 0.2746,
+      "step": 2655
+    },
+    {
+      "epoch": 1.5510546675425152,
+      "grad_norm": 0.08583937128074208,
+      "learning_rate": 2.912728016391753e-06,
+      "loss": 0.323,
+      "step": 2656
+    },
+    {
+      "epoch": 1.5516385665279908,
+      "grad_norm": 0.08695464857967662,
+      "learning_rate": 2.905538131479376e-06,
+      "loss": 0.2822,
+      "step": 2657
+    },
+    {
+      "epoch": 1.5522224655134662,
+      "grad_norm": 0.08971552092254298,
+      "learning_rate": 2.8983556228194165e-06,
+      "loss": 0.2985,
+      "step": 2658
+    },
+    {
+      "epoch": 1.5528063644989416,
+      "grad_norm": 0.0952358224924773,
+      "learning_rate": 2.8911804978796664e-06,
+      "loss": 0.3492,
+      "step": 2659
+    },
+    {
+      "epoch": 1.5533902634844172,
+      "grad_norm": 0.08555028019688338,
+      "learning_rate": 2.884012764120252e-06,
+      "loss": 0.3112,
+      "step": 2660
+    },
+    {
+      "epoch": 1.5539741624698928,
+      "grad_norm": 0.09044783672651677,
+      "learning_rate": 2.8768524289936007e-06,
+      "loss": 0.3176,
+      "step": 2661
+    },
+    {
+      "epoch": 1.5545580614553682,
+      "grad_norm": 0.09292301654711868,
+      "learning_rate": 2.8696994999444614e-06,
+      "loss": 0.3414,
+      "step": 2662
+    },
+    {
+      "epoch": 1.5551419604408436,
+      "grad_norm": 0.0934974704673205,
+      "learning_rate": 2.8625539844098736e-06,
+      "loss": 0.2956,
+      "step": 2663
+    },
+    {
+      "epoch": 1.5557258594263192,
+      "grad_norm": 0.09232821875741438,
+      "learning_rate": 2.8554158898191674e-06,
+      "loss": 0.308,
+      "step": 2664
+    },
+    {
+      "epoch": 1.5563097584117949,
+      "grad_norm": 0.08639882552804652,
+      "learning_rate": 2.8482852235939672e-06,
+      "loss": 0.3154,
+      "step": 2665
+    },
+    {
+      "epoch": 1.5568936573972703,
+      "grad_norm": 0.09106787089257534,
+      "learning_rate": 2.8411619931481627e-06,
+      "loss": 0.2869,
+      "step": 2666
+    },
+    {
+      "epoch": 1.5574775563827457,
+      "grad_norm": 0.09022645860786016,
+      "learning_rate": 2.8340462058879214e-06,
+      "loss": 0.3322,
+      "step": 2667
+    },
+    {
+      "epoch": 1.5580614553682213,
+      "grad_norm": 0.09635330942764082,
+      "learning_rate": 2.8269378692116676e-06,
+      "loss": 0.3223,
+      "step": 2668
+    },
+    {
+      "epoch": 1.5586453543536969,
+      "grad_norm": 0.08718498522948262,
+      "learning_rate": 2.8198369905100754e-06,
+      "loss": 0.3067,
+      "step": 2669
+    },
+    {
+      "epoch": 1.5592292533391723,
+      "grad_norm": 0.08762630565891273,
+      "learning_rate": 2.812743577166075e-06,
+      "loss": 0.3164,
+      "step": 2670
+    },
+    {
+      "epoch": 1.5598131523246477,
+      "grad_norm": 0.08713602733188572,
+      "learning_rate": 2.8056576365548216e-06,
+      "loss": 0.2797,
+      "step": 2671
+    },
+    {
+      "epoch": 1.5603970513101233,
+      "grad_norm": 0.08945157378832412,
+      "learning_rate": 2.7985791760437163e-06,
+      "loss": 0.3021,
+      "step": 2672
+    },
+    {
+      "epoch": 1.560980950295599,
+      "grad_norm": 0.09585268184892844,
+      "learning_rate": 2.79150820299237e-06,
+      "loss": 0.3119,
+      "step": 2673
+    },
+    {
+      "epoch": 1.5615648492810745,
+      "grad_norm": 0.09658673329891977,
+      "learning_rate": 2.784444724752611e-06,
+      "loss": 0.3361,
+      "step": 2674
+    },
+    {
+      "epoch": 1.56214874826655,
+      "grad_norm": 0.09035076582430443,
+      "learning_rate": 2.7773887486684815e-06,
+      "loss": 0.333,
+      "step": 2675
+    },
+    {
+      "epoch": 1.5627326472520253,
+      "grad_norm": 0.09235752881616759,
+      "learning_rate": 2.770340282076216e-06,
+      "loss": 0.3368,
+      "step": 2676
+    },
+    {
+      "epoch": 1.563316546237501,
+      "grad_norm": 0.0874545511605651,
+      "learning_rate": 2.76329933230425e-06,
+      "loss": 0.287,
+      "step": 2677
+    },
+    {
+      "epoch": 1.5639004452229766,
+      "grad_norm": 0.0844439618806292,
+      "learning_rate": 2.7562659066731947e-06,
+      "loss": 0.2764,
+      "step": 2678
+    },
+    {
+      "epoch": 1.564484344208452,
+      "grad_norm": 0.08974563057015433,
+      "learning_rate": 2.7492400124958397e-06,
+      "loss": 0.3294,
+      "step": 2679
+    },
+    {
+      "epoch": 1.5650682431939273,
+      "grad_norm": 0.08425795265746196,
+      "learning_rate": 2.742221657077151e-06,
+      "loss": 0.2954,
+      "step": 2680
+    },
+    {
+      "epoch": 1.565652142179403,
+      "grad_norm": 0.08840087881502673,
+      "learning_rate": 2.735210847714247e-06,
+      "loss": 0.3141,
+      "step": 2681
+    },
+    {
+      "epoch": 1.5662360411648786,
+      "grad_norm": 0.08560886993535796,
+      "learning_rate": 2.7282075916964077e-06,
+      "loss": 0.3242,
+      "step": 2682
+    },
+    {
+      "epoch": 1.566819940150354,
+      "grad_norm": 0.08642529204832461,
+      "learning_rate": 2.721211896305059e-06,
+      "loss": 0.3061,
+      "step": 2683
+    },
+    {
+      "epoch": 1.5674038391358294,
+      "grad_norm": 0.08177485377064425,
+      "learning_rate": 2.7142237688137594e-06,
+      "loss": 0.2948,
+      "step": 2684
+    },
+    {
+      "epoch": 1.567987738121305,
+      "grad_norm": 0.08811436945738257,
+      "learning_rate": 2.707243216488208e-06,
+      "loss": 0.3437,
+      "step": 2685
+    },
+    {
+      "epoch": 1.5685716371067806,
+      "grad_norm": 0.08952419293557991,
+      "learning_rate": 2.7002702465862206e-06,
+      "loss": 0.2908,
+      "step": 2686
+    },
+    {
+      "epoch": 1.569155536092256,
+      "grad_norm": 0.09494806104995257,
+      "learning_rate": 2.6933048663577297e-06,
+      "loss": 0.2992,
+      "step": 2687
+    },
+    {
+      "epoch": 1.5697394350777314,
+      "grad_norm": 0.0845677584206805,
+      "learning_rate": 2.6863470830447837e-06,
+      "loss": 0.3136,
+      "step": 2688
+    },
+    {
+      "epoch": 1.570323334063207,
+      "grad_norm": 0.07765771231810609,
+      "learning_rate": 2.6793969038815224e-06,
+      "loss": 0.239,
+      "step": 2689
+    },
+    {
+      "epoch": 1.5709072330486826,
+      "grad_norm": 0.09209148571564438,
+      "learning_rate": 2.672454336094191e-06,
+      "loss": 0.3279,
+      "step": 2690
+    },
+    {
+      "epoch": 1.5714911320341582,
+      "grad_norm": 0.0908520187039667,
+      "learning_rate": 2.665519386901111e-06,
+      "loss": 0.3333,
+      "step": 2691
+    },
+    {
+      "epoch": 1.5720750310196336,
+      "grad_norm": 0.08341119611021972,
+      "learning_rate": 2.658592063512684e-06,
+      "loss": 0.3354,
+      "step": 2692
+    },
+    {
+      "epoch": 1.572658930005109,
+      "grad_norm": 0.09966068268863304,
+      "learning_rate": 2.6516723731313896e-06,
+      "loss": 0.3392,
+      "step": 2693
+    },
+    {
+      "epoch": 1.5732428289905847,
+      "grad_norm": 0.08720191107348217,
+      "learning_rate": 2.644760322951764e-06,
+      "loss": 0.2809,
+      "step": 2694
+    },
+    {
+      "epoch": 1.5738267279760603,
+      "grad_norm": 0.08899505152690246,
+      "learning_rate": 2.6378559201604047e-06,
+      "loss": 0.3419,
+      "step": 2695
+    },
+    {
+      "epoch": 1.5744106269615357,
+      "grad_norm": 0.08708563071812123,
+      "learning_rate": 2.6309591719359563e-06,
+      "loss": 0.3173,
+      "step": 2696
+    },
+    {
+      "epoch": 1.574994525947011,
+      "grad_norm": 0.09029581199349246,
+      "learning_rate": 2.6240700854490988e-06,
+      "loss": 0.3133,
+      "step": 2697
+    },
+    {
+      "epoch": 1.5755784249324867,
+      "grad_norm": 0.09080300286980403,
+      "learning_rate": 2.6171886678625593e-06,
+      "loss": 0.3171,
+      "step": 2698
+    },
+    {
+      "epoch": 1.5761623239179623,
+      "grad_norm": 0.08512867126994779,
+      "learning_rate": 2.6103149263310768e-06,
+      "loss": 0.3073,
+      "step": 2699
+    },
+    {
+      "epoch": 1.5767462229034377,
+      "grad_norm": 0.08193357723517525,
+      "learning_rate": 2.6034488680014236e-06,
+      "loss": 0.2774,
+      "step": 2700
+    },
+    {
+      "epoch": 1.577330121888913,
+      "grad_norm": 0.08610359016356871,
+      "learning_rate": 2.5965905000123736e-06,
+      "loss": 0.2882,
+      "step": 2701
+    },
+    {
+      "epoch": 1.5779140208743887,
+      "grad_norm": 0.09241811761268019,
+      "learning_rate": 2.5897398294947027e-06,
+      "loss": 0.3114,
+      "step": 2702
+    },
+    {
+      "epoch": 1.5784979198598643,
+      "grad_norm": 0.09269353421785698,
+      "learning_rate": 2.582896863571197e-06,
+      "loss": 0.3005,
+      "step": 2703
+    },
+    {
+      "epoch": 1.5790818188453397,
+      "grad_norm": 0.09086974730035968,
+      "learning_rate": 2.576061609356617e-06,
+      "loss": 0.3318,
+      "step": 2704
+    },
+    {
+      "epoch": 1.579665717830815,
+      "grad_norm": 0.07965816529814923,
+      "learning_rate": 2.569234073957717e-06,
+      "loss": 0.273,
+      "step": 2705
+    },
+    {
+      "epoch": 1.5802496168162907,
+      "grad_norm": 0.08524798289778211,
+      "learning_rate": 2.5624142644732177e-06,
+      "loss": 0.2918,
+      "step": 2706
+    },
+    {
+      "epoch": 1.5808335158017663,
+      "grad_norm": 0.09000775576636812,
+      "learning_rate": 2.5556021879938074e-06,
+      "loss": 0.3144,
+      "step": 2707
+    },
+    {
+      "epoch": 1.581417414787242,
+      "grad_norm": 0.08510243835639211,
+      "learning_rate": 2.5487978516021426e-06,
+      "loss": 0.3365,
+      "step": 2708
+    },
+    {
+      "epoch": 1.5820013137727174,
+      "grad_norm": 0.0893035531194083,
+      "learning_rate": 2.542001262372821e-06,
+      "loss": 0.3021,
+      "step": 2709
+    },
+    {
+      "epoch": 1.5825852127581927,
+      "grad_norm": 0.08665641690465571,
+      "learning_rate": 2.535212427372393e-06,
+      "loss": 0.2937,
+      "step": 2710
+    },
+    {
+      "epoch": 1.5831691117436684,
+      "grad_norm": 0.0982902671764726,
+      "learning_rate": 2.52843135365935e-06,
+      "loss": 0.3613,
+      "step": 2711
+    },
+    {
+      "epoch": 1.583753010729144,
+      "grad_norm": 0.08621243950107117,
+      "learning_rate": 2.5216580482840993e-06,
+      "loss": 0.2939,
+      "step": 2712
+    },
+    {
+      "epoch": 1.5843369097146194,
+      "grad_norm": 0.09008128586856522,
+      "learning_rate": 2.514892518288988e-06,
+      "loss": 0.311,
+      "step": 2713
+    },
+    {
+      "epoch": 1.5849208087000948,
+      "grad_norm": 0.09556105811123047,
+      "learning_rate": 2.50813477070827e-06,
+      "loss": 0.3261,
+      "step": 2714
+    },
+    {
+      "epoch": 1.5855047076855704,
+      "grad_norm": 0.0891816295673628,
+      "learning_rate": 2.501384812568104e-06,
+      "loss": 0.3046,
+      "step": 2715
+    },
+    {
+      "epoch": 1.586088606671046,
+      "grad_norm": 0.09225837410608656,
+      "learning_rate": 2.494642650886563e-06,
+      "loss": 0.3597,
+      "step": 2716
+    },
+    {
+      "epoch": 1.5866725056565214,
+      "grad_norm": 0.08572392475278785,
+      "learning_rate": 2.4879082926735974e-06,
+      "loss": 0.3017,
+      "step": 2717
+    },
+    {
+      "epoch": 1.5872564046419968,
+      "grad_norm": 0.08315285015950266,
+      "learning_rate": 2.4811817449310615e-06,
+      "loss": 0.2635,
+      "step": 2718
+    },
+    {
+      "epoch": 1.5878403036274724,
+      "grad_norm": 0.09094807653958127,
+      "learning_rate": 2.4744630146526762e-06,
+      "loss": 0.3246,
+      "step": 2719
+    },
+    {
+      "epoch": 1.588424202612948,
+      "grad_norm": 0.0861275798471693,
+      "learning_rate": 2.467752108824034e-06,
+      "loss": 0.2844,
+      "step": 2720
+    },
+    {
+      "epoch": 1.5890081015984234,
+      "grad_norm": 0.08701887864702935,
+      "learning_rate": 2.4610490344226034e-06,
+      "loss": 0.3191,
+      "step": 2721
+    },
+    {
+      "epoch": 1.5895920005838988,
+      "grad_norm": 0.08163393163744767,
+      "learning_rate": 2.454353798417698e-06,
+      "loss": 0.2802,
+      "step": 2722
+    },
+    {
+      "epoch": 1.5901758995693744,
+      "grad_norm": 0.09065144758615648,
+      "learning_rate": 2.4476664077704926e-06,
+      "loss": 0.343,
+      "step": 2723
+    },
+    {
+      "epoch": 1.59075979855485,
+      "grad_norm": 0.09061571321744781,
+      "learning_rate": 2.4409868694339965e-06,
+      "loss": 0.311,
+      "step": 2724
+    },
+    {
+      "epoch": 1.5913436975403257,
+      "grad_norm": 0.08744874136506844,
+      "learning_rate": 2.434315190353056e-06,
+      "loss": 0.3212,
+      "step": 2725
+    },
+    {
+      "epoch": 1.591927596525801,
+      "grad_norm": 0.09855482784529292,
+      "learning_rate": 2.427651377464353e-06,
+      "loss": 0.3338,
+      "step": 2726
+    },
+    {
+      "epoch": 1.5925114955112765,
+      "grad_norm": 0.08638158575807878,
+      "learning_rate": 2.4209954376963797e-06,
+      "loss": 0.2806,
+      "step": 2727
+    },
+    {
+      "epoch": 1.593095394496752,
+      "grad_norm": 0.08256881472116842,
+      "learning_rate": 2.4143473779694548e-06,
+      "loss": 0.2868,
+      "step": 2728
+    },
+    {
+      "epoch": 1.5936792934822277,
+      "grad_norm": 0.08977771489388456,
+      "learning_rate": 2.407707205195694e-06,
+      "loss": 0.317,
+      "step": 2729
+    },
+    {
+      "epoch": 1.594263192467703,
+      "grad_norm": 0.0875707366161125,
+      "learning_rate": 2.4010749262790136e-06,
+      "loss": 0.2937,
+      "step": 2730
+    },
+    {
+      "epoch": 1.5948470914531785,
+      "grad_norm": 0.08449905096150369,
+      "learning_rate": 2.3944505481151303e-06,
+      "loss": 0.3007,
+      "step": 2731
+    },
+    {
+      "epoch": 1.595430990438654,
+      "grad_norm": 0.0878929002911304,
+      "learning_rate": 2.387834077591538e-06,
+      "loss": 0.2878,
+      "step": 2732
+    },
+    {
+      "epoch": 1.5960148894241297,
+      "grad_norm": 0.09257535557016475,
+      "learning_rate": 2.3812255215875147e-06,
+      "loss": 0.3266,
+      "step": 2733
+    },
+    {
+      "epoch": 1.5965987884096051,
+      "grad_norm": 0.0893229346547595,
+      "learning_rate": 2.374624886974106e-06,
+      "loss": 0.3118,
+      "step": 2734
+    },
+    {
+      "epoch": 1.5971826873950805,
+      "grad_norm": 0.09387590022085404,
+      "learning_rate": 2.3680321806141182e-06,
+      "loss": 0.3125,
+      "step": 2735
+    },
+    {
+      "epoch": 1.5977665863805561,
+      "grad_norm": 0.09324719840266338,
+      "learning_rate": 2.3614474093621255e-06,
+      "loss": 0.3089,
+      "step": 2736
+    },
+    {
+      "epoch": 1.5983504853660317,
+      "grad_norm": 0.08855432382674419,
+      "learning_rate": 2.354870580064439e-06,
+      "loss": 0.289,
+      "step": 2737
+    },
+    {
+      "epoch": 1.5989343843515071,
+      "grad_norm": 0.09172204469569499,
+      "learning_rate": 2.34830169955912e-06,
+      "loss": 0.3334,
+      "step": 2738
+    },
+    {
+      "epoch": 1.5995182833369825,
+      "grad_norm": 0.08717567713643702,
+      "learning_rate": 2.341740774675968e-06,
+      "loss": 0.3057,
+      "step": 2739
+    },
+    {
+      "epoch": 1.6001021823224582,
+      "grad_norm": 0.08652915801742803,
+      "learning_rate": 2.335187812236499e-06,
+      "loss": 0.3104,
+      "step": 2740
+    },
+    {
+      "epoch": 1.6006860813079338,
+      "grad_norm": 0.09322153719740107,
+      "learning_rate": 2.3286428190539645e-06,
+      "loss": 0.3413,
+      "step": 2741
+    },
+    {
+      "epoch": 1.6012699802934094,
+      "grad_norm": 0.0856674554368565,
+      "learning_rate": 2.322105801933321e-06,
+      "loss": 0.3,
+      "step": 2742
+    },
+    {
+      "epoch": 1.6018538792788848,
+      "grad_norm": 0.09541520542074722,
+      "learning_rate": 2.3155767676712317e-06,
+      "loss": 0.3231,
+      "step": 2743
+    },
+    {
+      "epoch": 1.6024377782643602,
+      "grad_norm": 0.08295347393512019,
+      "learning_rate": 2.3090557230560673e-06,
+      "loss": 0.3022,
+      "step": 2744
+    },
+    {
+      "epoch": 1.6030216772498358,
+      "grad_norm": 0.09024262274305489,
+      "learning_rate": 2.3025426748678814e-06,
+      "loss": 0.3227,
+      "step": 2745
+    },
+    {
+      "epoch": 1.6036055762353114,
+      "grad_norm": 0.08700274609416091,
+      "learning_rate": 2.296037629878426e-06,
+      "loss": 0.2979,
+      "step": 2746
+    },
+    {
+      "epoch": 1.6041894752207868,
+      "grad_norm": 0.08851633366836065,
+      "learning_rate": 2.289540594851122e-06,
+      "loss": 0.3156,
+      "step": 2747
+    },
+    {
+      "epoch": 1.6047733742062622,
+      "grad_norm": 0.08189364944502212,
+      "learning_rate": 2.283051576541062e-06,
+      "loss": 0.2841,
+      "step": 2748
+    },
+    {
+      "epoch": 1.6053572731917378,
+      "grad_norm": 0.08296570991459747,
+      "learning_rate": 2.2765705816950124e-06,
+      "loss": 0.3029,
+      "step": 2749
+    },
+    {
+      "epoch": 1.6059411721772134,
+      "grad_norm": 0.0894619671438262,
+      "learning_rate": 2.2700976170513855e-06,
+      "loss": 0.32,
+      "step": 2750
+    },
+    {
+      "epoch": 1.6065250711626888,
+      "grad_norm": 0.07995017617560513,
+      "learning_rate": 2.263632689340257e-06,
+      "loss": 0.2627,
+      "step": 2751
+    },
+    {
+      "epoch": 1.6071089701481642,
+      "grad_norm": 0.08327144066744055,
+      "learning_rate": 2.257175805283338e-06,
+      "loss": 0.2984,
+      "step": 2752
+    },
+    {
+      "epoch": 1.6076928691336398,
+      "grad_norm": 0.08764391972478744,
+      "learning_rate": 2.250726971593976e-06,
+      "loss": 0.275,
+      "step": 2753
+    },
+    {
+      "epoch": 1.6082767681191155,
+      "grad_norm": 0.09479075442045409,
+      "learning_rate": 2.2442861949771554e-06,
+      "loss": 0.3535,
+      "step": 2754
+    },
+    {
+      "epoch": 1.6088606671045909,
+      "grad_norm": 0.08561785212401186,
+      "learning_rate": 2.237853482129475e-06,
+      "loss": 0.2891,
+      "step": 2755
+    },
+    {
+      "epoch": 1.6094445660900663,
+      "grad_norm": 0.08123943908863514,
+      "learning_rate": 2.231428839739157e-06,
+      "loss": 0.2766,
+      "step": 2756
+    },
+    {
+      "epoch": 1.6100284650755419,
+      "grad_norm": 0.08771232533442673,
+      "learning_rate": 2.225012274486028e-06,
+      "loss": 0.2993,
+      "step": 2757
+    },
+    {
+      "epoch": 1.6106123640610175,
+      "grad_norm": 0.0946760181825588,
+      "learning_rate": 2.218603793041516e-06,
+      "loss": 0.3121,
+      "step": 2758
+    },
+    {
+      "epoch": 1.611196263046493,
+      "grad_norm": 0.09000970906342437,
+      "learning_rate": 2.21220340206865e-06,
+      "loss": 0.2921,
+      "step": 2759
+    },
+    {
+      "epoch": 1.6117801620319685,
+      "grad_norm": 0.08563356971455253,
+      "learning_rate": 2.205811108222038e-06,
+      "loss": 0.3049,
+      "step": 2760
+    },
+    {
+      "epoch": 1.612364061017444,
+      "grad_norm": 0.08841076019432903,
+      "learning_rate": 2.19942691814788e-06,
+      "loss": 0.3147,
+      "step": 2761
+    },
+    {
+      "epoch": 1.6129479600029195,
+      "grad_norm": 0.08674310355781881,
+      "learning_rate": 2.193050838483942e-06,
+      "loss": 0.2891,
+      "step": 2762
+    },
+    {
+      "epoch": 1.6135318589883951,
+      "grad_norm": 0.08909771473191294,
+      "learning_rate": 2.186682875859557e-06,
+      "loss": 0.2911,
+      "step": 2763
+    },
+    {
+      "epoch": 1.6141157579738705,
+      "grad_norm": 0.09901891990099063,
+      "learning_rate": 2.1803230368956296e-06,
+      "loss": 0.3225,
+      "step": 2764
+    },
+    {
+      "epoch": 1.614699656959346,
+      "grad_norm": 0.08934040750096521,
+      "learning_rate": 2.1739713282046017e-06,
+      "loss": 0.3184,
+      "step": 2765
+    },
+    {
+      "epoch": 1.6152835559448215,
+      "grad_norm": 0.08920801576672022,
+      "learning_rate": 2.1676277563904747e-06,
+      "loss": 0.2763,
+      "step": 2766
+    },
+    {
+      "epoch": 1.6158674549302972,
+      "grad_norm": 0.09010040275065445,
+      "learning_rate": 2.1612923280487883e-06,
+      "loss": 0.3056,
+      "step": 2767
+    },
+    {
+      "epoch": 1.6164513539157725,
+      "grad_norm": 0.08640111667290108,
+      "learning_rate": 2.1549650497666096e-06,
+      "loss": 0.3196,
+      "step": 2768
+    },
+    {
+      "epoch": 1.617035252901248,
+      "grad_norm": 0.08253086657745881,
+      "learning_rate": 2.1486459281225337e-06,
+      "loss": 0.3224,
+      "step": 2769
+    },
+    {
+      "epoch": 1.6176191518867236,
+      "grad_norm": 0.08538805082715221,
+      "learning_rate": 2.14233496968668e-06,
+      "loss": 0.297,
+      "step": 2770
+    },
+    {
+      "epoch": 1.6182030508721992,
+      "grad_norm": 0.08726838171216537,
+      "learning_rate": 2.136032181020673e-06,
+      "loss": 0.3135,
+      "step": 2771
+    },
+    {
+      "epoch": 1.6187869498576746,
+      "grad_norm": 0.09165710199284505,
+      "learning_rate": 2.1297375686776522e-06,
+      "loss": 0.3246,
+      "step": 2772
+    },
+    {
+      "epoch": 1.6193708488431502,
+      "grad_norm": 0.08360147948018504,
+      "learning_rate": 2.1234511392022473e-06,
+      "loss": 0.276,
+      "step": 2773
+    },
+    {
+      "epoch": 1.6199547478286256,
+      "grad_norm": 0.08504294669014584,
+      "learning_rate": 2.1171728991305797e-06,
+      "loss": 0.286,
+      "step": 2774
+    },
+    {
+      "epoch": 1.6205386468141012,
+      "grad_norm": 0.09089154456258475,
+      "learning_rate": 2.110902854990268e-06,
+      "loss": 0.3247,
+      "step": 2775
+    },
+    {
+      "epoch": 1.6211225457995768,
+      "grad_norm": 0.09017537855064596,
+      "learning_rate": 2.1046410133003923e-06,
+      "loss": 0.2936,
+      "step": 2776
+    },
+    {
+      "epoch": 1.6217064447850522,
+      "grad_norm": 0.08407130918362843,
+      "learning_rate": 2.0983873805715216e-06,
+      "loss": 0.2866,
+      "step": 2777
+    },
+    {
+      "epoch": 1.6222903437705276,
+      "grad_norm": 0.09486269937410233,
+      "learning_rate": 2.0921419633056782e-06,
+      "loss": 0.3094,
+      "step": 2778
+    },
+    {
+      "epoch": 1.6228742427560032,
+      "grad_norm": 0.07942592083346355,
+      "learning_rate": 2.085904767996343e-06,
+      "loss": 0.2823,
+      "step": 2779
+    },
+    {
+      "epoch": 1.6234581417414788,
+      "grad_norm": 0.08439176569373646,
+      "learning_rate": 2.0796758011284567e-06,
+      "loss": 0.2962,
+      "step": 2780
+    },
+    {
+      "epoch": 1.6240420407269542,
+      "grad_norm": 0.08268741143145944,
+      "learning_rate": 2.0734550691783937e-06,
+      "loss": 0.29,
+      "step": 2781
+    },
+    {
+      "epoch": 1.6246259397124296,
+      "grad_norm": 0.08764906208941274,
+      "learning_rate": 2.0672425786139794e-06,
+      "loss": 0.3036,
+      "step": 2782
+    },
+    {
+      "epoch": 1.6252098386979053,
+      "grad_norm": 0.08783964427244415,
+      "learning_rate": 2.0610383358944584e-06,
+      "loss": 0.3314,
+      "step": 2783
+    },
+    {
+      "epoch": 1.6257937376833809,
+      "grad_norm": 0.08379328949873344,
+      "learning_rate": 2.0548423474705024e-06,
+      "loss": 0.3095,
+      "step": 2784
+    },
+    {
+      "epoch": 1.6263776366688563,
+      "grad_norm": 0.08126643030798154,
+      "learning_rate": 2.0486546197842096e-06,
+      "loss": 0.2777,
+      "step": 2785
+    },
+    {
+      "epoch": 1.6269615356543317,
+      "grad_norm": 0.09095367051705196,
+      "learning_rate": 2.0424751592690762e-06,
+      "loss": 0.3098,
+      "step": 2786
+    },
+    {
+      "epoch": 1.6275454346398073,
+      "grad_norm": 0.08482520104712579,
+      "learning_rate": 2.0363039723500155e-06,
+      "loss": 0.294,
+      "step": 2787
+    },
+    {
+      "epoch": 1.628129333625283,
+      "grad_norm": 0.08865619653285754,
+      "learning_rate": 2.0301410654433307e-06,
+      "loss": 0.3012,
+      "step": 2788
+    },
+    {
+      "epoch": 1.6287132326107583,
+      "grad_norm": 0.08481335898026474,
+      "learning_rate": 2.023986444956715e-06,
+      "loss": 0.3022,
+      "step": 2789
+    },
+    {
+      "epoch": 1.629297131596234,
+      "grad_norm": 0.0882374945787197,
+      "learning_rate": 2.017840117289254e-06,
+      "loss": 0.3164,
+      "step": 2790
+    },
+    {
+      "epoch": 1.6298810305817093,
+      "grad_norm": 0.08491814775448743,
+      "learning_rate": 2.0117020888313998e-06,
+      "loss": 0.2933,
+      "step": 2791
+    },
+    {
+      "epoch": 1.630464929567185,
+      "grad_norm": 0.09669196728742827,
+      "learning_rate": 2.0055723659649907e-06,
+      "loss": 0.3313,
+      "step": 2792
+    },
+    {
+      "epoch": 1.6310488285526605,
+      "grad_norm": 0.08938081610423863,
+      "learning_rate": 1.999450955063216e-06,
+      "loss": 0.2991,
+      "step": 2793
+    },
+    {
+      "epoch": 1.631632727538136,
+      "grad_norm": 0.09534126553002377,
+      "learning_rate": 1.9933378624906218e-06,
+      "loss": 0.357,
+      "step": 2794
+    },
+    {
+      "epoch": 1.6322166265236113,
+      "grad_norm": 0.08649948641001376,
+      "learning_rate": 1.9872330946031237e-06,
+      "loss": 0.3115,
+      "step": 2795
+    },
+    {
+      "epoch": 1.632800525509087,
+      "grad_norm": 0.08889270098109296,
+      "learning_rate": 1.981136657747963e-06,
+      "loss": 0.3149,
+      "step": 2796
+    },
+    {
+      "epoch": 1.6333844244945626,
+      "grad_norm": 0.08253933405792702,
+      "learning_rate": 1.9750485582637245e-06,
+      "loss": 0.2745,
+      "step": 2797
+    },
+    {
+      "epoch": 1.633968323480038,
+      "grad_norm": 0.08296681901183083,
+      "learning_rate": 1.9689688024803298e-06,
+      "loss": 0.2961,
+      "step": 2798
+    },
+    {
+      "epoch": 1.6345522224655133,
+      "grad_norm": 0.08485867991452802,
+      "learning_rate": 1.962897396719018e-06,
+      "loss": 0.307,
+      "step": 2799
+    },
+    {
+      "epoch": 1.635136121450989,
+      "grad_norm": 0.08914231559703806,
+      "learning_rate": 1.9568343472923524e-06,
+      "loss": 0.3177,
+      "step": 2800
+    },
+    {
+      "epoch": 1.6357200204364646,
+      "grad_norm": 0.09265859916170287,
+      "learning_rate": 1.950779660504204e-06,
+      "loss": 0.2954,
+      "step": 2801
+    },
+    {
+      "epoch": 1.63630391942194,
+      "grad_norm": 0.0922993584161075,
+      "learning_rate": 1.944733342649748e-06,
+      "loss": 0.3297,
+      "step": 2802
+    },
+    {
+      "epoch": 1.6368878184074154,
+      "grad_norm": 0.08538897867184332,
+      "learning_rate": 1.938695400015467e-06,
+      "loss": 0.3015,
+      "step": 2803
+    },
+    {
+      "epoch": 1.637471717392891,
+      "grad_norm": 0.08662306176050659,
+      "learning_rate": 1.932665838879123e-06,
+      "loss": 0.3168,
+      "step": 2804
+    },
+    {
+      "epoch": 1.6380556163783666,
+      "grad_norm": 0.08717383505087216,
+      "learning_rate": 1.926644665509775e-06,
+      "loss": 0.2977,
+      "step": 2805
+    },
+    {
+      "epoch": 1.638639515363842,
+      "grad_norm": 0.08976952396928667,
+      "learning_rate": 1.920631886167754e-06,
+      "loss": 0.3053,
+      "step": 2806
+    },
+    {
+      "epoch": 1.6392234143493176,
+      "grad_norm": 0.08862903587197135,
+      "learning_rate": 1.9146275071046626e-06,
+      "loss": 0.3071,
+      "step": 2807
+    },
+    {
+      "epoch": 1.639807313334793,
+      "grad_norm": 0.08989318090675634,
+      "learning_rate": 1.9086315345633786e-06,
+      "loss": 0.3136,
+      "step": 2808
+    },
+    {
+      "epoch": 1.6403912123202686,
+      "grad_norm": 0.08815179950159056,
+      "learning_rate": 1.9026439747780278e-06,
+      "loss": 0.3209,
+      "step": 2809
+    },
+    {
+      "epoch": 1.6409751113057442,
+      "grad_norm": 0.08566298569253793,
+      "learning_rate": 1.8966648339740002e-06,
+      "loss": 0.3077,
+      "step": 2810
+    },
+    {
+      "epoch": 1.6415590102912196,
+      "grad_norm": 0.09057698608179941,
+      "learning_rate": 1.8906941183679227e-06,
+      "loss": 0.3245,
+      "step": 2811
+    },
+    {
+      "epoch": 1.642142909276695,
+      "grad_norm": 0.08950538404923337,
+      "learning_rate": 1.8847318341676657e-06,
+      "loss": 0.2865,
+      "step": 2812
+    },
+    {
+      "epoch": 1.6427268082621707,
+      "grad_norm": 0.10044778778804353,
+      "learning_rate": 1.8787779875723389e-06,
+      "loss": 0.3654,
+      "step": 2813
+    },
+    {
+      "epoch": 1.6433107072476463,
+      "grad_norm": 0.09245404437711423,
+      "learning_rate": 1.8728325847722684e-06,
+      "loss": 0.3155,
+      "step": 2814
+    },
+    {
+      "epoch": 1.6438946062331217,
+      "grad_norm": 0.09355968940240383,
+      "learning_rate": 1.8668956319490128e-06,
+      "loss": 0.3379,
+      "step": 2815
+    },
+    {
+      "epoch": 1.644478505218597,
+      "grad_norm": 0.086379425575275,
+      "learning_rate": 1.8609671352753367e-06,
+      "loss": 0.31,
+      "step": 2816
+    },
+    {
+      "epoch": 1.6450624042040727,
+      "grad_norm": 0.08072548858336856,
+      "learning_rate": 1.8550471009152138e-06,
+      "loss": 0.2641,
+      "step": 2817
+    },
+    {
+      "epoch": 1.6456463031895483,
+      "grad_norm": 0.0952707200165274,
+      "learning_rate": 1.849135535023825e-06,
+      "loss": 0.3456,
+      "step": 2818
+    },
+    {
+      "epoch": 1.6462302021750237,
+      "grad_norm": 0.0951804174890332,
+      "learning_rate": 1.843232443747538e-06,
+      "loss": 0.3459,
+      "step": 2819
+    },
+    {
+      "epoch": 1.646814101160499,
+      "grad_norm": 0.08972407208962244,
+      "learning_rate": 1.8373378332239177e-06,
+      "loss": 0.3052,
+      "step": 2820
+    },
+    {
+      "epoch": 1.6473980001459747,
+      "grad_norm": 0.08537922256509099,
+      "learning_rate": 1.8314517095817052e-06,
+      "loss": 0.3136,
+      "step": 2821
+    },
+    {
+      "epoch": 1.6479818991314503,
+      "grad_norm": 0.09488967053391885,
+      "learning_rate": 1.8255740789408161e-06,
+      "loss": 0.3339,
+      "step": 2822
+    },
+    {
+      "epoch": 1.6485657981169257,
+      "grad_norm": 0.08723722346762633,
+      "learning_rate": 1.8197049474123475e-06,
+      "loss": 0.3352,
+      "step": 2823
+    },
+    {
+      "epoch": 1.6491496971024013,
+      "grad_norm": 0.09183094012000258,
+      "learning_rate": 1.8138443210985468e-06,
+      "loss": 0.3073,
+      "step": 2824
+    },
+    {
+      "epoch": 1.6497335960878767,
+      "grad_norm": 0.08736882199499582,
+      "learning_rate": 1.8079922060928223e-06,
+      "loss": 0.2749,
+      "step": 2825
+    },
+    {
+      "epoch": 1.6503174950733523,
+      "grad_norm": 0.08072166004255642,
+      "learning_rate": 1.8021486084797368e-06,
+      "loss": 0.2845,
+      "step": 2826
+    },
+    {
+      "epoch": 1.650901394058828,
+      "grad_norm": 0.08048323996193933,
+      "learning_rate": 1.7963135343349914e-06,
+      "loss": 0.2903,
+      "step": 2827
+    },
+    {
+      "epoch": 1.6514852930443034,
+      "grad_norm": 0.08197019336455492,
+      "learning_rate": 1.7904869897254308e-06,
+      "loss": 0.2863,
+      "step": 2828
+    },
+    {
+      "epoch": 1.6520691920297788,
+      "grad_norm": 0.09368709195871784,
+      "learning_rate": 1.7846689807090277e-06,
+      "loss": 0.3023,
+      "step": 2829
+    },
+    {
+      "epoch": 1.6526530910152544,
+      "grad_norm": 0.09022671760323137,
+      "learning_rate": 1.7788595133348796e-06,
+      "loss": 0.3328,
+      "step": 2830
+    },
+    {
+      "epoch": 1.65323699000073,
+      "grad_norm": 0.08372521170087517,
+      "learning_rate": 1.7730585936432077e-06,
+      "loss": 0.2927,
+      "step": 2831
+    },
+    {
+      "epoch": 1.6538208889862054,
+      "grad_norm": 0.08572204115942626,
+      "learning_rate": 1.7672662276653384e-06,
+      "loss": 0.3149,
+      "step": 2832
+    },
+    {
+      "epoch": 1.6544047879716808,
+      "grad_norm": 0.08448339093176169,
+      "learning_rate": 1.7614824214237158e-06,
+      "loss": 0.2942,
+      "step": 2833
+    },
+    {
+      "epoch": 1.6549886869571564,
+      "grad_norm": 0.0892875756212391,
+      "learning_rate": 1.7557071809318737e-06,
+      "loss": 0.3152,
+      "step": 2834
+    },
+    {
+      "epoch": 1.655572585942632,
+      "grad_norm": 0.0929354859365066,
+      "learning_rate": 1.7499405121944423e-06,
+      "loss": 0.3296,
+      "step": 2835
+    },
+    {
+      "epoch": 1.6561564849281074,
+      "grad_norm": 0.0866926171329897,
+      "learning_rate": 1.7441824212071455e-06,
+      "loss": 0.3047,
+      "step": 2836
+    },
+    {
+      "epoch": 1.6567403839135828,
+      "grad_norm": 0.08779710170951847,
+      "learning_rate": 1.73843291395678e-06,
+      "loss": 0.2817,
+      "step": 2837
+    },
+    {
+      "epoch": 1.6573242828990584,
+      "grad_norm": 0.08930025748118767,
+      "learning_rate": 1.7326919964212275e-06,
+      "loss": 0.3078,
+      "step": 2838
+    },
+    {
+      "epoch": 1.657908181884534,
+      "grad_norm": 0.09138928873300248,
+      "learning_rate": 1.7269596745694295e-06,
+      "loss": 0.3354,
+      "step": 2839
+    },
+    {
+      "epoch": 1.6584920808700094,
+      "grad_norm": 0.09365254920248223,
+      "learning_rate": 1.7212359543613943e-06,
+      "loss": 0.3241,
+      "step": 2840
+    },
+    {
+      "epoch": 1.659075979855485,
+      "grad_norm": 0.09791257858527587,
+      "learning_rate": 1.7155208417481906e-06,
+      "loss": 0.3332,
+      "step": 2841
+    },
+    {
+      "epoch": 1.6596598788409604,
+      "grad_norm": 0.09084775987588738,
+      "learning_rate": 1.7098143426719293e-06,
+      "loss": 0.3169,
+      "step": 2842
+    },
+    {
+      "epoch": 1.660243777826436,
+      "grad_norm": 0.08351386258445684,
+      "learning_rate": 1.7041164630657757e-06,
+      "loss": 0.3005,
+      "step": 2843
+    },
+    {
+      "epoch": 1.6608276768119117,
+      "grad_norm": 0.08976040255982419,
+      "learning_rate": 1.6984272088539256e-06,
+      "loss": 0.331,
+      "step": 2844
+    },
+    {
+      "epoch": 1.661411575797387,
+      "grad_norm": 0.0948756643638096,
+      "learning_rate": 1.6927465859516057e-06,
+      "loss": 0.3354,
+      "step": 2845
+    },
+    {
+      "epoch": 1.6619954747828625,
+      "grad_norm": 0.10384250157288312,
+      "learning_rate": 1.6870746002650784e-06,
+      "loss": 0.3448,
+      "step": 2846
+    },
+    {
+      "epoch": 1.662579373768338,
+      "grad_norm": 0.08376240050822872,
+      "learning_rate": 1.6814112576916142e-06,
+      "loss": 0.2646,
+      "step": 2847
+    },
+    {
+      "epoch": 1.6631632727538137,
+      "grad_norm": 0.0923496129707507,
+      "learning_rate": 1.6757565641195073e-06,
+      "loss": 0.3216,
+      "step": 2848
+    },
+    {
+      "epoch": 1.663747171739289,
+      "grad_norm": 0.09654573898926508,
+      "learning_rate": 1.6701105254280513e-06,
+      "loss": 0.3368,
+      "step": 2849
+    },
+    {
+      "epoch": 1.6643310707247645,
+      "grad_norm": 0.08682377983693461,
+      "learning_rate": 1.664473147487541e-06,
+      "loss": 0.2996,
+      "step": 2850
+    },
+    {
+      "epoch": 1.6649149697102401,
+      "grad_norm": 0.0868710907329911,
+      "learning_rate": 1.658844436159277e-06,
+      "loss": 0.3228,
+      "step": 2851
+    },
+    {
+      "epoch": 1.6654988686957157,
+      "grad_norm": 0.0892936885324465,
+      "learning_rate": 1.6532243972955397e-06,
+      "loss": 0.2665,
+      "step": 2852
+    },
+    {
+      "epoch": 1.6660827676811911,
+      "grad_norm": 0.0922961336517733,
+      "learning_rate": 1.6476130367395914e-06,
+      "loss": 0.288,
+      "step": 2853
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.08884003920583328,
+      "learning_rate": 1.64201036032568e-06,
+      "loss": 0.3028,
+      "step": 2854
+    },
+    {
+      "epoch": 1.6672505656521421,
+      "grad_norm": 0.0881281234037118,
+      "learning_rate": 1.6364163738790128e-06,
+      "loss": 0.2777,
+      "step": 2855
+    },
+    {
+      "epoch": 1.6678344646376178,
+      "grad_norm": 0.08738493726752065,
+      "learning_rate": 1.6308310832157737e-06,
+      "loss": 0.3286,
+      "step": 2856
+    },
+    {
+      "epoch": 1.6684183636230934,
+      "grad_norm": 0.08762330668794399,
+      "learning_rate": 1.6252544941430982e-06,
+      "loss": 0.2923,
+      "step": 2857
+    },
+    {
+      "epoch": 1.6690022626085688,
+      "grad_norm": 0.08681897374370598,
+      "learning_rate": 1.6196866124590737e-06,
+      "loss": 0.2784,
+      "step": 2858
+    },
+    {
+      "epoch": 1.6695861615940442,
+      "grad_norm": 0.08376334321988234,
+      "learning_rate": 1.614127443952741e-06,
+      "loss": 0.292,
+      "step": 2859
+    },
+    {
+      "epoch": 1.6701700605795198,
+      "grad_norm": 0.08349112856652988,
+      "learning_rate": 1.608576994404074e-06,
+      "loss": 0.2794,
+      "step": 2860
+    },
+    {
+      "epoch": 1.6707539595649954,
+      "grad_norm": 0.08357314837595409,
+      "learning_rate": 1.603035269583989e-06,
+      "loss": 0.2759,
+      "step": 2861
+    },
+    {
+      "epoch": 1.6713378585504708,
+      "grad_norm": 0.09666979023552605,
+      "learning_rate": 1.5975022752543247e-06,
+      "loss": 0.3678,
+      "step": 2862
+    },
+    {
+      "epoch": 1.6719217575359462,
+      "grad_norm": 0.08903627487344483,
+      "learning_rate": 1.5919780171678412e-06,
+      "loss": 0.2762,
+      "step": 2863
+    },
+    {
+      "epoch": 1.6725056565214218,
+      "grad_norm": 0.09147760055649448,
+      "learning_rate": 1.5864625010682266e-06,
+      "loss": 0.3066,
+      "step": 2864
+    },
+    {
+      "epoch": 1.6730895555068974,
+      "grad_norm": 0.08833116465368375,
+      "learning_rate": 1.580955732690065e-06,
+      "loss": 0.2888,
+      "step": 2865
+    },
+    {
+      "epoch": 1.6736734544923728,
+      "grad_norm": 0.08453873112662393,
+      "learning_rate": 1.5754577177588581e-06,
+      "loss": 0.2987,
+      "step": 2866
+    },
+    {
+      "epoch": 1.6742573534778482,
+      "grad_norm": 0.08639345145179927,
+      "learning_rate": 1.5699684619909983e-06,
+      "loss": 0.3312,
+      "step": 2867
+    },
+    {
+      "epoch": 1.6748412524633238,
+      "grad_norm": 0.08887739793635734,
+      "learning_rate": 1.5644879710937722e-06,
+      "loss": 0.2737,
+      "step": 2868
+    },
+    {
+      "epoch": 1.6754251514487994,
+      "grad_norm": 0.09928756054330166,
+      "learning_rate": 1.5590162507653573e-06,
+      "loss": 0.3387,
+      "step": 2869
+    },
+    {
+      "epoch": 1.6760090504342748,
+      "grad_norm": 0.09336357804983127,
+      "learning_rate": 1.5535533066948062e-06,
+      "loss": 0.3215,
+      "step": 2870
+    },
+    {
+      "epoch": 1.6765929494197502,
+      "grad_norm": 0.09941545192356908,
+      "learning_rate": 1.5480991445620541e-06,
+      "loss": 0.3285,
+      "step": 2871
+    },
+    {
+      "epoch": 1.6771768484052259,
+      "grad_norm": 0.0966645308613163,
+      "learning_rate": 1.5426537700378985e-06,
+      "loss": 0.3223,
+      "step": 2872
+    },
+    {
+      "epoch": 1.6777607473907015,
+      "grad_norm": 0.0845008654355572,
+      "learning_rate": 1.5372171887840026e-06,
+      "loss": 0.3061,
+      "step": 2873
+    },
+    {
+      "epoch": 1.678344646376177,
+      "grad_norm": 0.09191780791329321,
+      "learning_rate": 1.5317894064528905e-06,
+      "loss": 0.3524,
+      "step": 2874
+    },
+    {
+      "epoch": 1.6789285453616525,
+      "grad_norm": 0.08958894414390065,
+      "learning_rate": 1.5263704286879311e-06,
+      "loss": 0.3145,
+      "step": 2875
+    },
+    {
+      "epoch": 1.6795124443471279,
+      "grad_norm": 0.09548452136018556,
+      "learning_rate": 1.5209602611233465e-06,
+      "loss": 0.327,
+      "step": 2876
+    },
+    {
+      "epoch": 1.6800963433326035,
+      "grad_norm": 0.08971837209227904,
+      "learning_rate": 1.5155589093841939e-06,
+      "loss": 0.2981,
+      "step": 2877
+    },
+    {
+      "epoch": 1.680680242318079,
+      "grad_norm": 0.08475160818817833,
+      "learning_rate": 1.5101663790863597e-06,
+      "loss": 0.2804,
+      "step": 2878
+    },
+    {
+      "epoch": 1.6812641413035545,
+      "grad_norm": 0.08998292250181139,
+      "learning_rate": 1.5047826758365748e-06,
+      "loss": 0.3447,
+      "step": 2879
+    },
+    {
+      "epoch": 1.68184804028903,
+      "grad_norm": 0.07976314715580113,
+      "learning_rate": 1.4994078052323767e-06,
+      "loss": 0.2755,
+      "step": 2880
+    },
+    {
+      "epoch": 1.6824319392745055,
+      "grad_norm": 0.08338205964186912,
+      "learning_rate": 1.4940417728621236e-06,
+      "loss": 0.2769,
+      "step": 2881
+    },
+    {
+      "epoch": 1.6830158382599811,
+      "grad_norm": 0.08607938805092073,
+      "learning_rate": 1.488684584304988e-06,
+      "loss": 0.2485,
+      "step": 2882
+    },
+    {
+      "epoch": 1.6835997372454565,
+      "grad_norm": 0.08797983663562785,
+      "learning_rate": 1.483336245130942e-06,
+      "loss": 0.3149,
+      "step": 2883
+    },
+    {
+      "epoch": 1.684183636230932,
+      "grad_norm": 0.09000539071289705,
+      "learning_rate": 1.477996760900764e-06,
+      "loss": 0.2868,
+      "step": 2884
+    },
+    {
+      "epoch": 1.6847675352164075,
+      "grad_norm": 0.08975067954698918,
+      "learning_rate": 1.4726661371660189e-06,
+      "loss": 0.3119,
+      "step": 2885
+    },
+    {
+      "epoch": 1.6853514342018832,
+      "grad_norm": 0.0884792867427891,
+      "learning_rate": 1.467344379469059e-06,
+      "loss": 0.3127,
+      "step": 2886
+    },
+    {
+      "epoch": 1.6859353331873586,
+      "grad_norm": 0.08366530934067863,
+      "learning_rate": 1.4620314933430269e-06,
+      "loss": 0.2698,
+      "step": 2887
+    },
+    {
+      "epoch": 1.686519232172834,
+      "grad_norm": 0.09530951870647125,
+      "learning_rate": 1.4567274843118296e-06,
+      "loss": 0.3493,
+      "step": 2888
+    },
+    {
+      "epoch": 1.6871031311583096,
+      "grad_norm": 0.08787362607764188,
+      "learning_rate": 1.4514323578901545e-06,
+      "loss": 0.3198,
+      "step": 2889
+    },
+    {
+      "epoch": 1.6876870301437852,
+      "grad_norm": 0.08836763642023365,
+      "learning_rate": 1.4461461195834491e-06,
+      "loss": 0.2954,
+      "step": 2890
+    },
+    {
+      "epoch": 1.6882709291292608,
+      "grad_norm": 0.08114731400208024,
+      "learning_rate": 1.4408687748879157e-06,
+      "loss": 0.2939,
+      "step": 2891
+    },
+    {
+      "epoch": 1.6888548281147362,
+      "grad_norm": 0.0894385082161046,
+      "learning_rate": 1.4356003292905197e-06,
+      "loss": 0.3396,
+      "step": 2892
+    },
+    {
+      "epoch": 1.6894387271002116,
+      "grad_norm": 0.0952035833086242,
+      "learning_rate": 1.4303407882689635e-06,
+      "loss": 0.3248,
+      "step": 2893
+    },
+    {
+      "epoch": 1.6900226260856872,
+      "grad_norm": 0.08194099139845323,
+      "learning_rate": 1.4250901572917009e-06,
+      "loss": 0.2857,
+      "step": 2894
+    },
+    {
+      "epoch": 1.6906065250711628,
+      "grad_norm": 0.08588422204004244,
+      "learning_rate": 1.4198484418179137e-06,
+      "loss": 0.3167,
+      "step": 2895
+    },
+    {
+      "epoch": 1.6911904240566382,
+      "grad_norm": 0.08893944284006498,
+      "learning_rate": 1.4146156472975147e-06,
+      "loss": 0.295,
+      "step": 2896
+    },
+    {
+      "epoch": 1.6917743230421136,
+      "grad_norm": 0.08623066840932737,
+      "learning_rate": 1.4093917791711497e-06,
+      "loss": 0.2721,
+      "step": 2897
+    },
+    {
+      "epoch": 1.6923582220275892,
+      "grad_norm": 0.08128463202888,
+      "learning_rate": 1.404176842870173e-06,
+      "loss": 0.2772,
+      "step": 2898
+    },
+    {
+      "epoch": 1.6929421210130648,
+      "grad_norm": 0.09046477989172445,
+      "learning_rate": 1.3989708438166605e-06,
+      "loss": 0.3053,
+      "step": 2899
+    },
+    {
+      "epoch": 1.6935260199985402,
+      "grad_norm": 0.09372810069578606,
+      "learning_rate": 1.3937737874233913e-06,
+      "loss": 0.2976,
+      "step": 2900
+    },
+    {
+      "epoch": 1.6941099189840156,
+      "grad_norm": 0.09281435568862173,
+      "learning_rate": 1.3885856790938457e-06,
+      "loss": 0.3753,
+      "step": 2901
+    },
+    {
+      "epoch": 1.6946938179694913,
+      "grad_norm": 0.0930790136624927,
+      "learning_rate": 1.383406524222206e-06,
+      "loss": 0.308,
+      "step": 2902
+    },
+    {
+      "epoch": 1.6952777169549669,
+      "grad_norm": 0.09359052556208464,
+      "learning_rate": 1.3782363281933387e-06,
+      "loss": 0.3129,
+      "step": 2903
+    },
+    {
+      "epoch": 1.6958616159404423,
+      "grad_norm": 0.08791074108491799,
+      "learning_rate": 1.3730750963828033e-06,
+      "loss": 0.3049,
+      "step": 2904
+    },
+    {
+      "epoch": 1.6964455149259177,
+      "grad_norm": 0.08429400485989422,
+      "learning_rate": 1.3679228341568308e-06,
+      "loss": 0.283,
+      "step": 2905
+    },
+    {
+      "epoch": 1.6970294139113933,
+      "grad_norm": 0.09426381487417643,
+      "learning_rate": 1.362779546872327e-06,
+      "loss": 0.3254,
+      "step": 2906
+    },
+    {
+      "epoch": 1.697613312896869,
+      "grad_norm": 0.08629029874131403,
+      "learning_rate": 1.357645239876879e-06,
+      "loss": 0.2926,
+      "step": 2907
+    },
+    {
+      "epoch": 1.6981972118823445,
+      "grad_norm": 0.08963886144686847,
+      "learning_rate": 1.3525199185087223e-06,
+      "loss": 0.3019,
+      "step": 2908
+    },
+    {
+      "epoch": 1.69878111086782,
+      "grad_norm": 0.09064130395030626,
+      "learning_rate": 1.3474035880967529e-06,
+      "loss": 0.354,
+      "step": 2909
+    },
+    {
+      "epoch": 1.6993650098532953,
+      "grad_norm": 0.08998563090120822,
+      "learning_rate": 1.3422962539605245e-06,
+      "loss": 0.3109,
+      "step": 2910
+    },
+    {
+      "epoch": 1.699948908838771,
+      "grad_norm": 0.08631197222515942,
+      "learning_rate": 1.3371979214102293e-06,
+      "loss": 0.2981,
+      "step": 2911
+    },
+    {
+      "epoch": 1.7005328078242465,
+      "grad_norm": 0.08518164142101413,
+      "learning_rate": 1.3321085957467107e-06,
+      "loss": 0.2841,
+      "step": 2912
+    },
+    {
+      "epoch": 1.701116706809722,
+      "grad_norm": 0.08487449180308386,
+      "learning_rate": 1.3270282822614366e-06,
+      "loss": 0.3031,
+      "step": 2913
+    },
+    {
+      "epoch": 1.7017006057951973,
+      "grad_norm": 0.08420082385138022,
+      "learning_rate": 1.321956986236509e-06,
+      "loss": 0.2999,
+      "step": 2914
+    },
+    {
+      "epoch": 1.702284504780673,
+      "grad_norm": 0.09287436406827683,
+      "learning_rate": 1.3168947129446574e-06,
+      "loss": 0.3206,
+      "step": 2915
+    },
+    {
+      "epoch": 1.7028684037661486,
+      "grad_norm": 0.08299714222257058,
+      "learning_rate": 1.3118414676492252e-06,
+      "loss": 0.2613,
+      "step": 2916
+    },
+    {
+      "epoch": 1.703452302751624,
+      "grad_norm": 0.09409380875982824,
+      "learning_rate": 1.3067972556041753e-06,
+      "loss": 0.312,
+      "step": 2917
+    },
+    {
+      "epoch": 1.7040362017370994,
+      "grad_norm": 0.08275575502103474,
+      "learning_rate": 1.3017620820540721e-06,
+      "loss": 0.2945,
+      "step": 2918
+    },
+    {
+      "epoch": 1.704620100722575,
+      "grad_norm": 0.09732248129241561,
+      "learning_rate": 1.2967359522340828e-06,
+      "loss": 0.3615,
+      "step": 2919
+    },
+    {
+      "epoch": 1.7052039997080506,
+      "grad_norm": 0.08450870904941496,
+      "learning_rate": 1.2917188713699791e-06,
+      "loss": 0.2632,
+      "step": 2920
+    },
+    {
+      "epoch": 1.705787898693526,
+      "grad_norm": 0.08014888031969351,
+      "learning_rate": 1.286710844678114e-06,
+      "loss": 0.2476,
+      "step": 2921
+    },
+    {
+      "epoch": 1.7063717976790014,
+      "grad_norm": 0.0918413902228895,
+      "learning_rate": 1.2817118773654381e-06,
+      "loss": 0.3323,
+      "step": 2922
+    },
+    {
+      "epoch": 1.706955696664477,
+      "grad_norm": 0.0800922403715263,
+      "learning_rate": 1.2767219746294724e-06,
+      "loss": 0.2991,
+      "step": 2923
+    },
+    {
+      "epoch": 1.7075395956499526,
+      "grad_norm": 0.08087779951049365,
+      "learning_rate": 1.271741141658317e-06,
+      "loss": 0.2778,
+      "step": 2924
+    },
+    {
+      "epoch": 1.7081234946354282,
+      "grad_norm": 0.08934717397355721,
+      "learning_rate": 1.266769383630646e-06,
+      "loss": 0.3083,
+      "step": 2925
+    },
+    {
+      "epoch": 1.7087073936209036,
+      "grad_norm": 0.09169601591933019,
+      "learning_rate": 1.2618067057156901e-06,
+      "loss": 0.3361,
+      "step": 2926
+    },
+    {
+      "epoch": 1.709291292606379,
+      "grad_norm": 0.093760920547733,
+      "learning_rate": 1.2568531130732498e-06,
+      "loss": 0.3534,
+      "step": 2927
+    },
+    {
+      "epoch": 1.7098751915918546,
+      "grad_norm": 0.09091977499103673,
+      "learning_rate": 1.2519086108536683e-06,
+      "loss": 0.31,
+      "step": 2928
+    },
+    {
+      "epoch": 1.7104590905773303,
+      "grad_norm": 0.08804765649485029,
+      "learning_rate": 1.2469732041978422e-06,
+      "loss": 0.3221,
+      "step": 2929
+    },
+    {
+      "epoch": 1.7110429895628056,
+      "grad_norm": 0.08949593905988745,
+      "learning_rate": 1.2420468982372158e-06,
+      "loss": 0.294,
+      "step": 2930
+    },
+    {
+      "epoch": 1.711626888548281,
+      "grad_norm": 0.09302260615601768,
+      "learning_rate": 1.237129698093762e-06,
+      "loss": 0.3201,
+      "step": 2931
+    },
+    {
+      "epoch": 1.7122107875337567,
+      "grad_norm": 0.0867742741419955,
+      "learning_rate": 1.2322216088799955e-06,
+      "loss": 0.2999,
+      "step": 2932
+    },
+    {
+      "epoch": 1.7127946865192323,
+      "grad_norm": 0.08589304565259366,
+      "learning_rate": 1.227322635698952e-06,
+      "loss": 0.2656,
+      "step": 2933
+    },
+    {
+      "epoch": 1.7133785855047077,
+      "grad_norm": 0.09349885872023175,
+      "learning_rate": 1.2224327836441863e-06,
+      "loss": 0.3117,
+      "step": 2934
+    },
+    {
+      "epoch": 1.713962484490183,
+      "grad_norm": 0.08284293116426811,
+      "learning_rate": 1.2175520577997834e-06,
+      "loss": 0.3006,
+      "step": 2935
+    },
+    {
+      "epoch": 1.7145463834756587,
+      "grad_norm": 0.08525720243529546,
+      "learning_rate": 1.2126804632403255e-06,
+      "loss": 0.3163,
+      "step": 2936
+    },
+    {
+      "epoch": 1.7151302824611343,
+      "grad_norm": 0.08571461057980632,
+      "learning_rate": 1.207818005030904e-06,
+      "loss": 0.3285,
+      "step": 2937
+    },
+    {
+      "epoch": 1.7157141814466097,
+      "grad_norm": 0.08322278893397721,
+      "learning_rate": 1.2029646882271173e-06,
+      "loss": 0.2939,
+      "step": 2938
+    },
+    {
+      "epoch": 1.716298080432085,
+      "grad_norm": 0.0910402296543325,
+      "learning_rate": 1.1981205178750511e-06,
+      "loss": 0.3463,
+      "step": 2939
+    },
+    {
+      "epoch": 1.7168819794175607,
+      "grad_norm": 0.08463085509397124,
+      "learning_rate": 1.1932854990112896e-06,
+      "loss": 0.2824,
+      "step": 2940
+    },
+    {
+      "epoch": 1.7174658784030363,
+      "grad_norm": 0.09217865336649095,
+      "learning_rate": 1.1884596366628942e-06,
+      "loss": 0.2924,
+      "step": 2941
+    },
+    {
+      "epoch": 1.718049777388512,
+      "grad_norm": 0.08823846948512909,
+      "learning_rate": 1.1836429358474077e-06,
+      "loss": 0.3051,
+      "step": 2942
+    },
+    {
+      "epoch": 1.7186336763739873,
+      "grad_norm": 0.08653676032826817,
+      "learning_rate": 1.1788354015728543e-06,
+      "loss": 0.2882,
+      "step": 2943
+    },
+    {
+      "epoch": 1.7192175753594627,
+      "grad_norm": 0.0927899321084126,
+      "learning_rate": 1.1740370388377188e-06,
+      "loss": 0.3594,
+      "step": 2944
+    },
+    {
+      "epoch": 1.7198014743449384,
+      "grad_norm": 0.09144793946907379,
+      "learning_rate": 1.1692478526309558e-06,
+      "loss": 0.3191,
+      "step": 2945
+    },
+    {
+      "epoch": 1.720385373330414,
+      "grad_norm": 0.0912659548405323,
+      "learning_rate": 1.1644678479319772e-06,
+      "loss": 0.2973,
+      "step": 2946
+    },
+    {
+      "epoch": 1.7209692723158894,
+      "grad_norm": 0.08531685512742015,
+      "learning_rate": 1.1596970297106458e-06,
+      "loss": 0.2925,
+      "step": 2947
+    },
+    {
+      "epoch": 1.7215531713013648,
+      "grad_norm": 0.09129339370760457,
+      "learning_rate": 1.1549354029272786e-06,
+      "loss": 0.3122,
+      "step": 2948
+    },
+    {
+      "epoch": 1.7221370702868404,
+      "grad_norm": 0.09228778676660264,
+      "learning_rate": 1.1501829725326307e-06,
+      "loss": 0.3456,
+      "step": 2949
+    },
+    {
+      "epoch": 1.722720969272316,
+      "grad_norm": 0.09121826294390589,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 0.2825,
+      "step": 2950
+    },
+    {
+      "epoch": 1.7233048682577914,
+      "grad_norm": 0.08896707916494323,
+      "learning_rate": 1.1407057206647188e-06,
+      "loss": 0.3021,
+      "step": 2951
+    },
+    {
+      "epoch": 1.7238887672432668,
+      "grad_norm": 0.08765575015981457,
+      "learning_rate": 1.1359809090451357e-06,
+      "loss": 0.2776,
+      "step": 2952
+    },
+    {
+      "epoch": 1.7244726662287424,
+      "grad_norm": 0.08932623833044764,
+      "learning_rate": 1.131265313521639e-06,
+      "loss": 0.3128,
+      "step": 2953
+    },
+    {
+      "epoch": 1.725056565214218,
+      "grad_norm": 0.09597771963463289,
+      "learning_rate": 1.126558938997121e-06,
+      "loss": 0.3373,
+      "step": 2954
+    },
+    {
+      "epoch": 1.7256404641996934,
+      "grad_norm": 0.0848574028432411,
+      "learning_rate": 1.1218617903648966e-06,
+      "loss": 0.292,
+      "step": 2955
+    },
+    {
+      "epoch": 1.7262243631851688,
+      "grad_norm": 0.08901974907959101,
+      "learning_rate": 1.1171738725086833e-06,
+      "loss": 0.3237,
+      "step": 2956
+    },
+    {
+      "epoch": 1.7268082621706444,
+      "grad_norm": 0.0794011300009036,
+      "learning_rate": 1.1124951903025981e-06,
+      "loss": 0.2557,
+      "step": 2957
+    },
+    {
+      "epoch": 1.72739216115612,
+      "grad_norm": 0.08140148758853254,
+      "learning_rate": 1.1078257486111654e-06,
+      "loss": 0.2799,
+      "step": 2958
+    },
+    {
+      "epoch": 1.7279760601415957,
+      "grad_norm": 0.09528368593672694,
+      "learning_rate": 1.1031655522892915e-06,
+      "loss": 0.2975,
+      "step": 2959
+    },
+    {
+      "epoch": 1.728559959127071,
+      "grad_norm": 0.09088312927300063,
+      "learning_rate": 1.0985146061822794e-06,
+      "loss": 0.2886,
+      "step": 2960
+    },
+    {
+      "epoch": 1.7291438581125465,
+      "grad_norm": 0.0871655840365122,
+      "learning_rate": 1.0938729151258065e-06,
+      "loss": 0.3286,
+      "step": 2961
+    },
+    {
+      "epoch": 1.729727757098022,
+      "grad_norm": 0.09005837151377202,
+      "learning_rate": 1.0892404839459269e-06,
+      "loss": 0.29,
+      "step": 2962
+    },
+    {
+      "epoch": 1.7303116560834977,
+      "grad_norm": 0.08513541480633148,
+      "learning_rate": 1.0846173174590802e-06,
+      "loss": 0.2959,
+      "step": 2963
+    },
+    {
+      "epoch": 1.730895555068973,
+      "grad_norm": 0.09100994567321888,
+      "learning_rate": 1.0800034204720588e-06,
+      "loss": 0.3131,
+      "step": 2964
+    },
+    {
+      "epoch": 1.7314794540544485,
+      "grad_norm": 0.08783273970386206,
+      "learning_rate": 1.0753987977820214e-06,
+      "loss": 0.2904,
+      "step": 2965
+    },
+    {
+      "epoch": 1.732063353039924,
+      "grad_norm": 0.08575336500142525,
+      "learning_rate": 1.07080345417649e-06,
+      "loss": 0.303,
+      "step": 2966
+    },
+    {
+      "epoch": 1.7326472520253997,
+      "grad_norm": 0.08947798309822216,
+      "learning_rate": 1.0662173944333288e-06,
+      "loss": 0.3049,
+      "step": 2967
+    },
+    {
+      "epoch": 1.733231151010875,
+      "grad_norm": 0.09518363799786442,
+      "learning_rate": 1.06164062332076e-06,
+      "loss": 0.3448,
+      "step": 2968
+    },
+    {
+      "epoch": 1.7338150499963505,
+      "grad_norm": 0.07988594836759536,
+      "learning_rate": 1.0570731455973415e-06,
+      "loss": 0.2662,
+      "step": 2969
+    },
+    {
+      "epoch": 1.7343989489818261,
+      "grad_norm": 0.08507839909670867,
+      "learning_rate": 1.052514966011966e-06,
+      "loss": 0.3165,
+      "step": 2970
+    },
+    {
+      "epoch": 1.7349828479673017,
+      "grad_norm": 0.08573812437016654,
+      "learning_rate": 1.0479660893038702e-06,
+      "loss": 0.2796,
+      "step": 2971
+    },
+    {
+      "epoch": 1.7355667469527771,
+      "grad_norm": 0.08720279476479635,
+      "learning_rate": 1.043426520202605e-06,
+      "loss": 0.3198,
+      "step": 2972
+    },
+    {
+      "epoch": 1.7361506459382525,
+      "grad_norm": 0.08178240404533646,
+      "learning_rate": 1.0388962634280543e-06,
+      "loss": 0.2737,
+      "step": 2973
+    },
+    {
+      "epoch": 1.7367345449237281,
+      "grad_norm": 0.08545939826506499,
+      "learning_rate": 1.0343753236904152e-06,
+      "loss": 0.2805,
+      "step": 2974
+    },
+    {
+      "epoch": 1.7373184439092038,
+      "grad_norm": 0.08070206180094003,
+      "learning_rate": 1.029863705690195e-06,
+      "loss": 0.2776,
+      "step": 2975
+    },
+    {
+      "epoch": 1.7379023428946794,
+      "grad_norm": 0.09201632691915795,
+      "learning_rate": 1.0253614141182167e-06,
+      "loss": 0.2985,
+      "step": 2976
+    },
+    {
+      "epoch": 1.7384862418801548,
+      "grad_norm": 0.09154559924090393,
+      "learning_rate": 1.0208684536555968e-06,
+      "loss": 0.2801,
+      "step": 2977
+    },
+    {
+      "epoch": 1.7390701408656302,
+      "grad_norm": 0.09035573627714134,
+      "learning_rate": 1.016384828973761e-06,
+      "loss": 0.3037,
+      "step": 2978
+    },
+    {
+      "epoch": 1.7396540398511058,
+      "grad_norm": 0.09003217266711798,
+      "learning_rate": 1.0119105447344203e-06,
+      "loss": 0.3084,
+      "step": 2979
+    },
+    {
+      "epoch": 1.7402379388365814,
+      "grad_norm": 0.08614657345447459,
+      "learning_rate": 1.007445605589573e-06,
+      "loss": 0.3112,
+      "step": 2980
+    },
+    {
+      "epoch": 1.7408218378220568,
+      "grad_norm": 0.09141827315267644,
+      "learning_rate": 1.0029900161815109e-06,
+      "loss": 0.3223,
+      "step": 2981
+    },
+    {
+      "epoch": 1.7414057368075322,
+      "grad_norm": 0.09263664925502296,
+      "learning_rate": 9.985437811427934e-07,
+      "loss": 0.3252,
+      "step": 2982
+    },
+    {
+      "epoch": 1.7419896357930078,
+      "grad_norm": 0.09310455150675177,
+      "learning_rate": 9.941069050962626e-07,
+      "loss": 0.309,
+      "step": 2983
+    },
+    {
+      "epoch": 1.7425735347784834,
+      "grad_norm": 0.08944734483288722,
+      "learning_rate": 9.896793926550252e-07,
+      "loss": 0.2894,
+      "step": 2984
+    },
+    {
+      "epoch": 1.7431574337639588,
+      "grad_norm": 0.08893793595293298,
+      "learning_rate": 9.8526124842245e-07,
+      "loss": 0.2922,
+      "step": 2985
+    },
+    {
+      "epoch": 1.7437413327494342,
+      "grad_norm": 0.08370831485933028,
+      "learning_rate": 9.808524769921756e-07,
+      "loss": 0.2726,
+      "step": 2986
+    },
+    {
+      "epoch": 1.7443252317349098,
+      "grad_norm": 0.09232717238662276,
+      "learning_rate": 9.764530829480822e-07,
+      "loss": 0.333,
+      "step": 2987
+    },
+    {
+      "epoch": 1.7449091307203854,
+      "grad_norm": 0.08402127403193392,
+      "learning_rate": 9.720630708643131e-07,
+      "loss": 0.2819,
+      "step": 2988
+    },
+    {
+      "epoch": 1.7454930297058608,
+      "grad_norm": 0.08942254687070669,
+      "learning_rate": 9.67682445305248e-07,
+      "loss": 0.3068,
+      "step": 2989
+    },
+    {
+      "epoch": 1.7460769286913362,
+      "grad_norm": 0.09102163708579147,
+      "learning_rate": 9.63311210825505e-07,
+      "loss": 0.3229,
+      "step": 2990
+    },
+    {
+      "epoch": 1.7466608276768119,
+      "grad_norm": 0.08726218260009509,
+      "learning_rate": 9.589493719699517e-07,
+      "loss": 0.3108,
+      "step": 2991
+    },
+    {
+      "epoch": 1.7472447266622875,
+      "grad_norm": 0.0930691327134997,
+      "learning_rate": 9.545969332736748e-07,
+      "loss": 0.3444,
+      "step": 2992
+    },
+    {
+      "epoch": 1.747828625647763,
+      "grad_norm": 0.0843230810014101,
+      "learning_rate": 9.502538992619892e-07,
+      "loss": 0.2566,
+      "step": 2993
+    },
+    {
+      "epoch": 1.7484125246332385,
+      "grad_norm": 0.08720604144464333,
+      "learning_rate": 9.459202744504359e-07,
+      "loss": 0.275,
+      "step": 2994
+    },
+    {
+      "epoch": 1.7489964236187139,
+      "grad_norm": 0.08056572574632331,
+      "learning_rate": 9.415960633447674e-07,
+      "loss": 0.2753,
+      "step": 2995
+    },
+    {
+      "epoch": 1.7495803226041895,
+      "grad_norm": 0.08140521278998084,
+      "learning_rate": 9.372812704409551e-07,
+      "loss": 0.2798,
+      "step": 2996
+    },
+    {
+      "epoch": 1.7501642215896651,
+      "grad_norm": 0.09400537521999199,
+      "learning_rate": 9.329759002251726e-07,
+      "loss": 0.3363,
+      "step": 2997
+    },
+    {
+      "epoch": 1.7507481205751405,
+      "grad_norm": 0.08075907383941486,
+      "learning_rate": 9.286799571737981e-07,
+      "loss": 0.2724,
+      "step": 2998
+    },
+    {
+      "epoch": 1.751332019560616,
+      "grad_norm": 0.08894513985235591,
+      "learning_rate": 9.243934457534098e-07,
+      "loss": 0.3105,
+      "step": 2999
+    },
+    {
+      "epoch": 1.7519159185460915,
+      "grad_norm": 0.08895736832294276,
+      "learning_rate": 9.201163704207771e-07,
+      "loss": 0.3048,
+      "step": 3000
+    },
+    {
+      "epoch": 1.7524998175315671,
+      "grad_norm": 0.089945688501075,
+      "learning_rate": 9.158487356228618e-07,
+      "loss": 0.2858,
+      "step": 3001
+    },
+    {
+      "epoch": 1.7530837165170425,
+      "grad_norm": 0.08853961342988431,
+      "learning_rate": 9.115905457968077e-07,
+      "loss": 0.2988,
+      "step": 3002
+    },
+    {
+      "epoch": 1.753667615502518,
+      "grad_norm": 0.08679984258769172,
+      "learning_rate": 9.073418053699368e-07,
+      "loss": 0.3112,
+      "step": 3003
+    },
+    {
+      "epoch": 1.7542515144879935,
+      "grad_norm": 0.08321185195950458,
+      "learning_rate": 9.031025187597519e-07,
+      "loss": 0.2872,
+      "step": 3004
+    },
+    {
+      "epoch": 1.7548354134734692,
+      "grad_norm": 0.08273327779028662,
+      "learning_rate": 8.988726903739197e-07,
+      "loss": 0.2701,
+      "step": 3005
+    },
+    {
+      "epoch": 1.7554193124589446,
+      "grad_norm": 0.084587328537699,
+      "learning_rate": 8.946523246102811e-07,
+      "loss": 0.29,
+      "step": 3006
+    },
+    {
+      "epoch": 1.75600321144442,
+      "grad_norm": 0.08920350939356088,
+      "learning_rate": 8.904414258568306e-07,
+      "loss": 0.3035,
+      "step": 3007
+    },
+    {
+      "epoch": 1.7565871104298956,
+      "grad_norm": 0.08316462072618035,
+      "learning_rate": 8.862399984917214e-07,
+      "loss": 0.2697,
+      "step": 3008
+    },
+    {
+      "epoch": 1.7571710094153712,
+      "grad_norm": 0.09287148592333239,
+      "learning_rate": 8.820480468832649e-07,
+      "loss": 0.3046,
+      "step": 3009
+    },
+    {
+      "epoch": 1.7577549084008468,
+      "grad_norm": 0.09785541207527797,
+      "learning_rate": 8.778655753899124e-07,
+      "loss": 0.3507,
+      "step": 3010
+    },
+    {
+      "epoch": 1.7583388073863222,
+      "grad_norm": 0.08281289719288021,
+      "learning_rate": 8.736925883602665e-07,
+      "loss": 0.2875,
+      "step": 3011
+    },
+    {
+      "epoch": 1.7589227063717976,
+      "grad_norm": 0.08527031624134249,
+      "learning_rate": 8.695290901330611e-07,
+      "loss": 0.3135,
+      "step": 3012
+    },
+    {
+      "epoch": 1.7595066053572732,
+      "grad_norm": 0.08586627646617032,
+      "learning_rate": 8.653750850371667e-07,
+      "loss": 0.2897,
+      "step": 3013
+    },
+    {
+      "epoch": 1.7600905043427488,
+      "grad_norm": 0.08713251653109393,
+      "learning_rate": 8.612305773915886e-07,
+      "loss": 0.3113,
+      "step": 3014
+    },
+    {
+      "epoch": 1.7606744033282242,
+      "grad_norm": 0.08968130364123778,
+      "learning_rate": 8.570955715054496e-07,
+      "loss": 0.3217,
+      "step": 3015
+    },
+    {
+      "epoch": 1.7612583023136996,
+      "grad_norm": 0.07921512383835833,
+      "learning_rate": 8.529700716780009e-07,
+      "loss": 0.3008,
+      "step": 3016
+    },
+    {
+      "epoch": 1.7618422012991752,
+      "grad_norm": 0.08789602661917328,
+      "learning_rate": 8.488540821986035e-07,
+      "loss": 0.3031,
+      "step": 3017
+    },
+    {
+      "epoch": 1.7624261002846509,
+      "grad_norm": 0.08785979067445888,
+      "learning_rate": 8.447476073467309e-07,
+      "loss": 0.3011,
+      "step": 3018
+    },
+    {
+      "epoch": 1.7630099992701262,
+      "grad_norm": 0.0904028495708559,
+      "learning_rate": 8.406506513919721e-07,
+      "loss": 0.2985,
+      "step": 3019
+    },
+    {
+      "epoch": 1.7635938982556016,
+      "grad_norm": 0.09688839879863544,
+      "learning_rate": 8.365632185940109e-07,
+      "loss": 0.3198,
+      "step": 3020
+    },
+    {
+      "epoch": 1.7641777972410773,
+      "grad_norm": 0.0826394342404232,
+      "learning_rate": 8.3248531320263e-07,
+      "loss": 0.2964,
+      "step": 3021
+    },
+    {
+      "epoch": 1.7647616962265529,
+      "grad_norm": 0.0894665329391872,
+      "learning_rate": 8.284169394577124e-07,
+      "loss": 0.2971,
+      "step": 3022
+    },
+    {
+      "epoch": 1.7653455952120283,
+      "grad_norm": 0.09185663656417631,
+      "learning_rate": 8.243581015892221e-07,
+      "loss": 0.3054,
+      "step": 3023
+    },
+    {
+      "epoch": 1.7659294941975037,
+      "grad_norm": 0.08518931752896243,
+      "learning_rate": 8.203088038172169e-07,
+      "loss": 0.2894,
+      "step": 3024
+    },
+    {
+      "epoch": 1.7665133931829793,
+      "grad_norm": 0.08322365199611717,
+      "learning_rate": 8.1626905035183e-07,
+      "loss": 0.2737,
+      "step": 3025
+    },
+    {
+      "epoch": 1.767097292168455,
+      "grad_norm": 0.08397715672037832,
+      "learning_rate": 8.122388453932728e-07,
+      "loss": 0.2843,
+      "step": 3026
+    },
+    {
+      "epoch": 1.7676811911539305,
+      "grad_norm": 0.08020847188537908,
+      "learning_rate": 8.082181931318311e-07,
+      "loss": 0.2728,
+      "step": 3027
+    },
+    {
+      "epoch": 1.768265090139406,
+      "grad_norm": 0.09412085440828606,
+      "learning_rate": 8.042070977478533e-07,
+      "loss": 0.294,
+      "step": 3028
+    },
+    {
+      "epoch": 1.7688489891248813,
+      "grad_norm": 0.09055258375077366,
+      "learning_rate": 8.002055634117578e-07,
+      "loss": 0.3052,
+      "step": 3029
+    },
+    {
+      "epoch": 1.769432888110357,
+      "grad_norm": 0.09057140596498112,
+      "learning_rate": 7.962135942840188e-07,
+      "loss": 0.3166,
+      "step": 3030
+    },
+    {
+      "epoch": 1.7700167870958325,
+      "grad_norm": 0.08811092750182023,
+      "learning_rate": 7.922311945151629e-07,
+      "loss": 0.3239,
+      "step": 3031
+    },
+    {
+      "epoch": 1.770600686081308,
+      "grad_norm": 0.09089354143068934,
+      "learning_rate": 7.882583682457734e-07,
+      "loss": 0.3453,
+      "step": 3032
+    },
+    {
+      "epoch": 1.7711845850667833,
+      "grad_norm": 0.08544428242945963,
+      "learning_rate": 7.84295119606473e-07,
+      "loss": 0.3135,
+      "step": 3033
+    },
+    {
+      "epoch": 1.771768484052259,
+      "grad_norm": 0.08561187754657379,
+      "learning_rate": 7.803414527179343e-07,
+      "loss": 0.3024,
+      "step": 3034
+    },
+    {
+      "epoch": 1.7723523830377346,
+      "grad_norm": 0.08088193947497127,
+      "learning_rate": 7.76397371690859e-07,
+      "loss": 0.2522,
+      "step": 3035
+    },
+    {
+      "epoch": 1.77293628202321,
+      "grad_norm": 0.08903871118395863,
+      "learning_rate": 7.72462880625986e-07,
+      "loss": 0.3169,
+      "step": 3036
+    },
+    {
+      "epoch": 1.7735201810086854,
+      "grad_norm": 0.08696975717596996,
+      "learning_rate": 7.685379836140872e-07,
+      "loss": 0.2901,
+      "step": 3037
+    },
+    {
+      "epoch": 1.774104079994161,
+      "grad_norm": 0.08118460480807976,
+      "learning_rate": 7.646226847359506e-07,
+      "loss": 0.2934,
+      "step": 3038
+    },
+    {
+      "epoch": 1.7746879789796366,
+      "grad_norm": 0.09264316210882517,
+      "learning_rate": 7.607169880623955e-07,
+      "loss": 0.3355,
+      "step": 3039
+    },
+    {
+      "epoch": 1.775271877965112,
+      "grad_norm": 0.08641335090159838,
+      "learning_rate": 7.568208976542491e-07,
+      "loss": 0.2946,
+      "step": 3040
+    },
+    {
+      "epoch": 1.7758557769505874,
+      "grad_norm": 0.08855129138066256,
+      "learning_rate": 7.529344175623521e-07,
+      "loss": 0.3113,
+      "step": 3041
+    },
+    {
+      "epoch": 1.776439675936063,
+      "grad_norm": 0.08193132434183854,
+      "learning_rate": 7.490575518275589e-07,
+      "loss": 0.2947,
+      "step": 3042
+    },
+    {
+      "epoch": 1.7770235749215386,
+      "grad_norm": 0.0853040641670257,
+      "learning_rate": 7.451903044807185e-07,
+      "loss": 0.2927,
+      "step": 3043
+    },
+    {
+      "epoch": 1.7776074739070142,
+      "grad_norm": 0.08895835922442463,
+      "learning_rate": 7.4133267954269e-07,
+      "loss": 0.2783,
+      "step": 3044
+    },
+    {
+      "epoch": 1.7781913728924896,
+      "grad_norm": 0.08927622906273894,
+      "learning_rate": 7.374846810243197e-07,
+      "loss": 0.3103,
+      "step": 3045
+    },
+    {
+      "epoch": 1.778775271877965,
+      "grad_norm": 0.09355958800983855,
+      "learning_rate": 7.336463129264437e-07,
+      "loss": 0.3347,
+      "step": 3046
+    },
+    {
+      "epoch": 1.7793591708634406,
+      "grad_norm": 0.08438872192014728,
+      "learning_rate": 7.298175792398976e-07,
+      "loss": 0.3064,
+      "step": 3047
+    },
+    {
+      "epoch": 1.7799430698489163,
+      "grad_norm": 0.08799185304588986,
+      "learning_rate": 7.25998483945487e-07,
+      "loss": 0.3301,
+      "step": 3048
+    },
+    {
+      "epoch": 1.7805269688343917,
+      "grad_norm": 0.08245352372786574,
+      "learning_rate": 7.22189031013999e-07,
+      "loss": 0.2859,
+      "step": 3049
+    },
+    {
+      "epoch": 1.781110867819867,
+      "grad_norm": 0.08197379308859429,
+      "learning_rate": 7.183892244062018e-07,
+      "loss": 0.2895,
+      "step": 3050
+    },
+    {
+      "epoch": 1.7816947668053427,
+      "grad_norm": 0.08771680638924784,
+      "learning_rate": 7.145990680728243e-07,
+      "loss": 0.297,
+      "step": 3051
+    },
+    {
+      "epoch": 1.7822786657908183,
+      "grad_norm": 0.07914840139776184,
+      "learning_rate": 7.10818565954573e-07,
+      "loss": 0.2702,
+      "step": 3052
+    },
+    {
+      "epoch": 1.7828625647762937,
+      "grad_norm": 0.09177798517369501,
+      "learning_rate": 7.07047721982107e-07,
+      "loss": 0.3279,
+      "step": 3053
+    },
+    {
+      "epoch": 1.783446463761769,
+      "grad_norm": 0.0879605596520252,
+      "learning_rate": 7.032865400760469e-07,
+      "loss": 0.3185,
+      "step": 3054
+    },
+    {
+      "epoch": 1.7840303627472447,
+      "grad_norm": 0.09484221308636585,
+      "learning_rate": 6.995350241469701e-07,
+      "loss": 0.3166,
+      "step": 3055
+    },
+    {
+      "epoch": 1.7846142617327203,
+      "grad_norm": 0.0804545110608819,
+      "learning_rate": 6.957931780954008e-07,
+      "loss": 0.284,
+      "step": 3056
+    },
+    {
+      "epoch": 1.7851981607181957,
+      "grad_norm": 0.08569995718272268,
+      "learning_rate": 6.920610058118105e-07,
+      "loss": 0.302,
+      "step": 3057
+    },
+    {
+      "epoch": 1.785782059703671,
+      "grad_norm": 0.0921317212054899,
+      "learning_rate": 6.883385111766139e-07,
+      "loss": 0.3177,
+      "step": 3058
+    },
+    {
+      "epoch": 1.7863659586891467,
+      "grad_norm": 0.08386527675179974,
+      "learning_rate": 6.846256980601596e-07,
+      "loss": 0.272,
+      "step": 3059
+    },
+    {
+      "epoch": 1.7869498576746223,
+      "grad_norm": 0.09088420058979575,
+      "learning_rate": 6.809225703227352e-07,
+      "loss": 0.2953,
+      "step": 3060
+    },
+    {
+      "epoch": 1.787533756660098,
+      "grad_norm": 0.08574771953516644,
+      "learning_rate": 6.772291318145541e-07,
+      "loss": 0.3006,
+      "step": 3061
+    },
+    {
+      "epoch": 1.7881176556455733,
+      "grad_norm": 0.08330097041251203,
+      "learning_rate": 6.735453863757602e-07,
+      "loss": 0.2526,
+      "step": 3062
+    },
+    {
+      "epoch": 1.7887015546310487,
+      "grad_norm": 0.09230866675138164,
+      "learning_rate": 6.698713378364142e-07,
+      "loss": 0.3252,
+      "step": 3063
+    },
+    {
+      "epoch": 1.7892854536165244,
+      "grad_norm": 0.08717974784159616,
+      "learning_rate": 6.662069900164969e-07,
+      "loss": 0.2746,
+      "step": 3064
+    },
+    {
+      "epoch": 1.789869352602,
+      "grad_norm": 0.0861227996512415,
+      "learning_rate": 6.625523467259043e-07,
+      "loss": 0.3308,
+      "step": 3065
+    },
+    {
+      "epoch": 1.7904532515874754,
+      "grad_norm": 0.08696621440537845,
+      "learning_rate": 6.589074117644411e-07,
+      "loss": 0.3051,
+      "step": 3066
+    },
+    {
+      "epoch": 1.7910371505729508,
+      "grad_norm": 0.08737047338750509,
+      "learning_rate": 6.552721889218194e-07,
+      "loss": 0.3019,
+      "step": 3067
+    },
+    {
+      "epoch": 1.7916210495584264,
+      "grad_norm": 0.0920790388231457,
+      "learning_rate": 6.516466819776502e-07,
+      "loss": 0.3356,
+      "step": 3068
+    },
+    {
+      "epoch": 1.792204948543902,
+      "grad_norm": 0.08919713513180166,
+      "learning_rate": 6.480308947014458e-07,
+      "loss": 0.315,
+      "step": 3069
+    },
+    {
+      "epoch": 1.7927888475293774,
+      "grad_norm": 0.09424576053389117,
+      "learning_rate": 6.444248308526125e-07,
+      "loss": 0.3717,
+      "step": 3070
+    },
+    {
+      "epoch": 1.7933727465148528,
+      "grad_norm": 0.0838416641966189,
+      "learning_rate": 6.408284941804444e-07,
+      "loss": 0.2793,
+      "step": 3071
+    },
+    {
+      "epoch": 1.7939566455003284,
+      "grad_norm": 0.0890216012583102,
+      "learning_rate": 6.372418884241271e-07,
+      "loss": 0.2948,
+      "step": 3072
+    },
+    {
+      "epoch": 1.794540544485804,
+      "grad_norm": 0.08909156969852251,
+      "learning_rate": 6.336650173127224e-07,
+      "loss": 0.3092,
+      "step": 3073
+    },
+    {
+      "epoch": 1.7951244434712794,
+      "grad_norm": 0.08700031289443887,
+      "learning_rate": 6.300978845651728e-07,
+      "loss": 0.3124,
+      "step": 3074
+    },
+    {
+      "epoch": 1.795708342456755,
+      "grad_norm": 0.08913364760921895,
+      "learning_rate": 6.26540493890303e-07,
+      "loss": 0.3218,
+      "step": 3075
+    },
+    {
+      "epoch": 1.7962922414422304,
+      "grad_norm": 0.08299333108288351,
+      "learning_rate": 6.229928489867987e-07,
+      "loss": 0.2984,
+      "step": 3076
+    },
+    {
+      "epoch": 1.796876140427706,
+      "grad_norm": 0.08371709473441251,
+      "learning_rate": 6.194549535432137e-07,
+      "loss": 0.2894,
+      "step": 3077
+    },
+    {
+      "epoch": 1.7974600394131817,
+      "grad_norm": 0.08657602870342991,
+      "learning_rate": 6.159268112379734e-07,
+      "loss": 0.2879,
+      "step": 3078
+    },
+    {
+      "epoch": 1.798043938398657,
+      "grad_norm": 0.08519140451128597,
+      "learning_rate": 6.124084257393525e-07,
+      "loss": 0.2913,
+      "step": 3079
+    },
+    {
+      "epoch": 1.7986278373841325,
+      "grad_norm": 0.08300224946686031,
+      "learning_rate": 6.088998007054903e-07,
+      "loss": 0.2996,
+      "step": 3080
+    },
+    {
+      "epoch": 1.799211736369608,
+      "grad_norm": 0.0932104456101071,
+      "learning_rate": 6.054009397843708e-07,
+      "loss": 0.3144,
+      "step": 3081
+    },
+    {
+      "epoch": 1.7997956353550837,
+      "grad_norm": 0.08559696164193707,
+      "learning_rate": 6.019118466138285e-07,
+      "loss": 0.2975,
+      "step": 3082
+    },
+    {
+      "epoch": 1.800379534340559,
+      "grad_norm": 0.08944302518031633,
+      "learning_rate": 5.98432524821545e-07,
+      "loss": 0.3045,
+      "step": 3083
+    },
+    {
+      "epoch": 1.8009634333260345,
+      "grad_norm": 0.08866393050792092,
+      "learning_rate": 5.949629780250376e-07,
+      "loss": 0.3221,
+      "step": 3084
+    },
+    {
+      "epoch": 1.80154733231151,
+      "grad_norm": 0.08766295952999233,
+      "learning_rate": 5.915032098316653e-07,
+      "loss": 0.305,
+      "step": 3085
+    },
+    {
+      "epoch": 1.8021312312969857,
+      "grad_norm": 0.08604490112310588,
+      "learning_rate": 5.880532238386161e-07,
+      "loss": 0.2883,
+      "step": 3086
+    },
+    {
+      "epoch": 1.802715130282461,
+      "grad_norm": 0.08960677044850361,
+      "learning_rate": 5.846130236329073e-07,
+      "loss": 0.3264,
+      "step": 3087
+    },
+    {
+      "epoch": 1.8032990292679365,
+      "grad_norm": 0.08360548375281042,
+      "learning_rate": 5.811826127913855e-07,
+      "loss": 0.2405,
+      "step": 3088
+    },
+    {
+      "epoch": 1.8038829282534121,
+      "grad_norm": 0.08203459351847663,
+      "learning_rate": 5.777619948807156e-07,
+      "loss": 0.284,
+      "step": 3089
+    },
+    {
+      "epoch": 1.8044668272388877,
+      "grad_norm": 0.08540071730036647,
+      "learning_rate": 5.743511734573837e-07,
+      "loss": 0.2904,
+      "step": 3090
+    },
+    {
+      "epoch": 1.8050507262243631,
+      "grad_norm": 0.09413825040019591,
+      "learning_rate": 5.709501520676853e-07,
+      "loss": 0.3261,
+      "step": 3091
+    },
+    {
+      "epoch": 1.8056346252098388,
+      "grad_norm": 0.08461267678204475,
+      "learning_rate": 5.675589342477305e-07,
+      "loss": 0.2471,
+      "step": 3092
+    },
+    {
+      "epoch": 1.8062185241953141,
+      "grad_norm": 0.08538893829406426,
+      "learning_rate": 5.641775235234381e-07,
+      "loss": 0.2967,
+      "step": 3093
+    },
+    {
+      "epoch": 1.8068024231807898,
+      "grad_norm": 0.0870621971034043,
+      "learning_rate": 5.608059234105234e-07,
+      "loss": 0.3441,
+      "step": 3094
+    },
+    {
+      "epoch": 1.8073863221662654,
+      "grad_norm": 0.09150970041698957,
+      "learning_rate": 5.5744413741451e-07,
+      "loss": 0.2865,
+      "step": 3095
+    },
+    {
+      "epoch": 1.8079702211517408,
+      "grad_norm": 0.09250988493366538,
+      "learning_rate": 5.540921690307111e-07,
+      "loss": 0.2921,
+      "step": 3096
+    },
+    {
+      "epoch": 1.8085541201372162,
+      "grad_norm": 0.09617099812727396,
+      "learning_rate": 5.507500217442341e-07,
+      "loss": 0.3318,
+      "step": 3097
+    },
+    {
+      "epoch": 1.8091380191226918,
+      "grad_norm": 0.08444627618579727,
+      "learning_rate": 5.474176990299773e-07,
+      "loss": 0.2886,
+      "step": 3098
+    },
+    {
+      "epoch": 1.8097219181081674,
+      "grad_norm": 0.08626503451288457,
+      "learning_rate": 5.440952043526215e-07,
+      "loss": 0.2833,
+      "step": 3099
+    },
+    {
+      "epoch": 1.8103058170936428,
+      "grad_norm": 0.09436229327893464,
+      "learning_rate": 5.407825411666312e-07,
+      "loss": 0.3297,
+      "step": 3100
+    },
+    {
+      "epoch": 1.8108897160791182,
+      "grad_norm": 0.08462486673426424,
+      "learning_rate": 5.374797129162468e-07,
+      "loss": 0.2787,
+      "step": 3101
+    },
+    {
+      "epoch": 1.8114736150645938,
+      "grad_norm": 0.0911560611708199,
+      "learning_rate": 5.341867230354824e-07,
+      "loss": 0.3261,
+      "step": 3102
+    },
+    {
+      "epoch": 1.8120575140500694,
+      "grad_norm": 0.09624376387767465,
+      "learning_rate": 5.309035749481295e-07,
+      "loss": 0.3042,
+      "step": 3103
+    },
+    {
+      "epoch": 1.8126414130355448,
+      "grad_norm": 0.09303145455429745,
+      "learning_rate": 5.276302720677395e-07,
+      "loss": 0.3141,
+      "step": 3104
+    },
+    {
+      "epoch": 1.8132253120210202,
+      "grad_norm": 0.08827962396356338,
+      "learning_rate": 5.243668177976291e-07,
+      "loss": 0.3035,
+      "step": 3105
+    },
+    {
+      "epoch": 1.8138092110064958,
+      "grad_norm": 0.08801293271583843,
+      "learning_rate": 5.211132155308785e-07,
+      "loss": 0.3195,
+      "step": 3106
+    },
+    {
+      "epoch": 1.8143931099919715,
+      "grad_norm": 0.08987372710837427,
+      "learning_rate": 5.178694686503205e-07,
+      "loss": 0.3107,
+      "step": 3107
+    },
+    {
+      "epoch": 1.8149770089774468,
+      "grad_norm": 0.08289489158136615,
+      "learning_rate": 5.146355805285452e-07,
+      "loss": 0.271,
+      "step": 3108
+    },
+    {
+      "epoch": 1.8155609079629225,
+      "grad_norm": 0.09480964276496395,
+      "learning_rate": 5.114115545278875e-07,
+      "loss": 0.3531,
+      "step": 3109
+    },
+    {
+      "epoch": 1.8161448069483979,
+      "grad_norm": 0.08618311950672712,
+      "learning_rate": 5.081973940004315e-07,
+      "loss": 0.2806,
+      "step": 3110
+    },
+    {
+      "epoch": 1.8167287059338735,
+      "grad_norm": 0.09239201627310456,
+      "learning_rate": 5.049931022880061e-07,
+      "loss": 0.3193,
+      "step": 3111
+    },
+    {
+      "epoch": 1.817312604919349,
+      "grad_norm": 0.09238081844257634,
+      "learning_rate": 5.017986827221733e-07,
+      "loss": 0.3457,
+      "step": 3112
+    },
+    {
+      "epoch": 1.8178965039048245,
+      "grad_norm": 0.09332885302694656,
+      "learning_rate": 4.986141386242371e-07,
+      "loss": 0.3319,
+      "step": 3113
+    },
+    {
+      "epoch": 1.8184804028902999,
+      "grad_norm": 0.08190144433254112,
+      "learning_rate": 4.954394733052293e-07,
+      "loss": 0.2834,
+      "step": 3114
+    },
+    {
+      "epoch": 1.8190643018757755,
+      "grad_norm": 0.08642558502640947,
+      "learning_rate": 4.922746900659125e-07,
+      "loss": 0.296,
+      "step": 3115
+    },
+    {
+      "epoch": 1.8196482008612511,
+      "grad_norm": 0.09202681201698579,
+      "learning_rate": 4.89119792196776e-07,
+      "loss": 0.3347,
+      "step": 3116
+    },
+    {
+      "epoch": 1.8202320998467265,
+      "grad_norm": 0.0833363476143065,
+      "learning_rate": 4.85974782978027e-07,
+      "loss": 0.3053,
+      "step": 3117
+    },
+    {
+      "epoch": 1.820815998832202,
+      "grad_norm": 0.08670604171879,
+      "learning_rate": 4.828396656795964e-07,
+      "loss": 0.2989,
+      "step": 3118
+    },
+    {
+      "epoch": 1.8213998978176775,
+      "grad_norm": 0.08762148872884952,
+      "learning_rate": 4.797144435611256e-07,
+      "loss": 0.3041,
+      "step": 3119
+    },
+    {
+      "epoch": 1.8219837968031531,
+      "grad_norm": 0.07929997473403812,
+      "learning_rate": 4.76599119871971e-07,
+      "loss": 0.2651,
+      "step": 3120
+    },
+    {
+      "epoch": 1.8225676957886285,
+      "grad_norm": 0.08377224940605213,
+      "learning_rate": 4.734936978511961e-07,
+      "loss": 0.2867,
+      "step": 3121
+    },
+    {
+      "epoch": 1.823151594774104,
+      "grad_norm": 0.08931276031631125,
+      "learning_rate": 4.7039818072756927e-07,
+      "loss": 0.2888,
+      "step": 3122
+    },
+    {
+      "epoch": 1.8237354937595796,
+      "grad_norm": 0.08704691695753657,
+      "learning_rate": 4.6731257171956256e-07,
+      "loss": 0.2817,
+      "step": 3123
+    },
+    {
+      "epoch": 1.8243193927450552,
+      "grad_norm": 0.09434333182504516,
+      "learning_rate": 4.642368740353431e-07,
+      "loss": 0.3297,
+      "step": 3124
+    },
+    {
+      "epoch": 1.8249032917305306,
+      "grad_norm": 0.09045381442652044,
+      "learning_rate": 4.61171090872774e-07,
+      "loss": 0.3152,
+      "step": 3125
+    },
+    {
+      "epoch": 1.8254871907160062,
+      "grad_norm": 0.07929089782676045,
+      "learning_rate": 4.581152254194121e-07,
+      "loss": 0.2686,
+      "step": 3126
+    },
+    {
+      "epoch": 1.8260710897014816,
+      "grad_norm": 0.08096124335118454,
+      "learning_rate": 4.5506928085250033e-07,
+      "loss": 0.2657,
+      "step": 3127
+    },
+    {
+      "epoch": 1.8266549886869572,
+      "grad_norm": 0.08177468966583025,
+      "learning_rate": 4.520332603389699e-07,
+      "loss": 0.2738,
+      "step": 3128
+    },
+    {
+      "epoch": 1.8272388876724328,
+      "grad_norm": 0.08936189404425426,
+      "learning_rate": 4.490071670354279e-07,
+      "loss": 0.3047,
+      "step": 3129
+    },
+    {
+      "epoch": 1.8278227866579082,
+      "grad_norm": 0.08825550331204482,
+      "learning_rate": 4.459910040881632e-07,
+      "loss": 0.3067,
+      "step": 3130
+    },
+    {
+      "epoch": 1.8284066856433836,
+      "grad_norm": 0.08811022706392743,
+      "learning_rate": 4.42984774633145e-07,
+      "loss": 0.3024,
+      "step": 3131
+    },
+    {
+      "epoch": 1.8289905846288592,
+      "grad_norm": 0.09472587031780937,
+      "learning_rate": 4.399884817960065e-07,
+      "loss": 0.3479,
+      "step": 3132
+    },
+    {
+      "epoch": 1.8295744836143348,
+      "grad_norm": 0.08959127579833713,
+      "learning_rate": 4.3700212869205117e-07,
+      "loss": 0.2945,
+      "step": 3133
+    },
+    {
+      "epoch": 1.8301583825998102,
+      "grad_norm": 0.09111172436104784,
+      "learning_rate": 4.34025718426252e-07,
+      "loss": 0.3135,
+      "step": 3134
+    },
+    {
+      "epoch": 1.8307422815852856,
+      "grad_norm": 0.09076980814635999,
+      "learning_rate": 4.310592540932401e-07,
+      "loss": 0.3237,
+      "step": 3135
+    },
+    {
+      "epoch": 1.8313261805707612,
+      "grad_norm": 0.09243097367030356,
+      "learning_rate": 4.2810273877730843e-07,
+      "loss": 0.3174,
+      "step": 3136
+    },
+    {
+      "epoch": 1.8319100795562369,
+      "grad_norm": 0.08835534189345208,
+      "learning_rate": 4.251561755524036e-07,
+      "loss": 0.293,
+      "step": 3137
+    },
+    {
+      "epoch": 1.8324939785417123,
+      "grad_norm": 0.08283881712655825,
+      "learning_rate": 4.222195674821239e-07,
+      "loss": 0.2815,
+      "step": 3138
+    },
+    {
+      "epoch": 1.8330778775271876,
+      "grad_norm": 0.07823255923527181,
+      "learning_rate": 4.192929176197236e-07,
+      "loss": 0.2814,
+      "step": 3139
+    },
+    {
+      "epoch": 1.8336617765126633,
+      "grad_norm": 0.09550047102597076,
+      "learning_rate": 4.1637622900809304e-07,
+      "loss": 0.3307,
+      "step": 3140
+    },
+    {
+      "epoch": 1.8342456754981389,
+      "grad_norm": 0.08442208448847098,
+      "learning_rate": 4.1346950467977545e-07,
+      "loss": 0.2904,
+      "step": 3141
+    },
+    {
+      "epoch": 1.8348295744836145,
+      "grad_norm": 0.08421308420888718,
+      "learning_rate": 4.1057274765694765e-07,
+      "loss": 0.2949,
+      "step": 3142
+    },
+    {
+      "epoch": 1.83541347346909,
+      "grad_norm": 0.0812597514232386,
+      "learning_rate": 4.0768596095142497e-07,
+      "loss": 0.272,
+      "step": 3143
+    },
+    {
+      "epoch": 1.8359973724545653,
+      "grad_norm": 0.08623785130910966,
+      "learning_rate": 4.048091475646576e-07,
+      "loss": 0.2951,
+      "step": 3144
+    },
+    {
+      "epoch": 1.836581271440041,
+      "grad_norm": 0.08997205042584555,
+      "learning_rate": 4.0194231048772514e-07,
+      "loss": 0.3116,
+      "step": 3145
+    },
+    {
+      "epoch": 1.8371651704255165,
+      "grad_norm": 0.084514263239577,
+      "learning_rate": 3.9908545270133436e-07,
+      "loss": 0.3128,
+      "step": 3146
+    },
+    {
+      "epoch": 1.837749069410992,
+      "grad_norm": 0.0915860638174057,
+      "learning_rate": 3.9623857717581813e-07,
+      "loss": 0.3074,
+      "step": 3147
+    },
+    {
+      "epoch": 1.8383329683964673,
+      "grad_norm": 0.08691073399383939,
+      "learning_rate": 3.934016868711266e-07,
+      "loss": 0.2887,
+      "step": 3148
+    },
+    {
+      "epoch": 1.838916867381943,
+      "grad_norm": 0.08355923300895465,
+      "learning_rate": 3.905747847368335e-07,
+      "loss": 0.3005,
+      "step": 3149
+    },
+    {
+      "epoch": 1.8395007663674185,
+      "grad_norm": 0.08825087270283859,
+      "learning_rate": 3.8775787371212346e-07,
+      "loss": 0.3137,
+      "step": 3150
+    },
+    {
+      "epoch": 1.840084665352894,
+      "grad_norm": 0.07874972837968709,
+      "learning_rate": 3.8495095672579584e-07,
+      "loss": 0.2716,
+      "step": 3151
+    },
+    {
+      "epoch": 1.8406685643383693,
+      "grad_norm": 0.09159646925794108,
+      "learning_rate": 3.8215403669625726e-07,
+      "loss": 0.3291,
+      "step": 3152
+    },
+    {
+      "epoch": 1.841252463323845,
+      "grad_norm": 0.10013682658527887,
+      "learning_rate": 3.793671165315194e-07,
+      "loss": 0.3015,
+      "step": 3153
+    },
+    {
+      "epoch": 1.8418363623093206,
+      "grad_norm": 0.08317894091138682,
+      "learning_rate": 3.765901991291998e-07,
+      "loss": 0.2941,
+      "step": 3154
+    },
+    {
+      "epoch": 1.842420261294796,
+      "grad_norm": 0.08967320183593953,
+      "learning_rate": 3.738232873765146e-07,
+      "loss": 0.3262,
+      "step": 3155
+    },
+    {
+      "epoch": 1.8430041602802714,
+      "grad_norm": 0.09155990726650498,
+      "learning_rate": 3.7106638415027594e-07,
+      "loss": 0.3182,
+      "step": 3156
+    },
+    {
+      "epoch": 1.843588059265747,
+      "grad_norm": 0.09385465837638611,
+      "learning_rate": 3.6831949231689203e-07,
+      "loss": 0.3459,
+      "step": 3157
+    },
+    {
+      "epoch": 1.8441719582512226,
+      "grad_norm": 0.09514724522367692,
+      "learning_rate": 3.65582614732356e-07,
+      "loss": 0.2884,
+      "step": 3158
+    },
+    {
+      "epoch": 1.8447558572366982,
+      "grad_norm": 0.08789971778831376,
+      "learning_rate": 3.628557542422606e-07,
+      "loss": 0.3037,
+      "step": 3159
+    },
+    {
+      "epoch": 1.8453397562221736,
+      "grad_norm": 0.08866112343668321,
+      "learning_rate": 3.6013891368177345e-07,
+      "loss": 0.2997,
+      "step": 3160
+    },
+    {
+      "epoch": 1.845923655207649,
+      "grad_norm": 0.09663836939156575,
+      "learning_rate": 3.574320958756461e-07,
+      "loss": 0.372,
+      "step": 3161
+    },
+    {
+      "epoch": 1.8465075541931246,
+      "grad_norm": 0.08971130807907668,
+      "learning_rate": 3.547353036382117e-07,
+      "loss": 0.3239,
+      "step": 3162
+    },
+    {
+      "epoch": 1.8470914531786002,
+      "grad_norm": 0.08757983654603695,
+      "learning_rate": 3.520485397733786e-07,
+      "loss": 0.2828,
+      "step": 3163
+    },
+    {
+      "epoch": 1.8476753521640756,
+      "grad_norm": 0.07568235233813884,
+      "learning_rate": 3.493718070746299e-07,
+      "loss": 0.2494,
+      "step": 3164
+    },
+    {
+      "epoch": 1.848259251149551,
+      "grad_norm": 0.08553360407094963,
+      "learning_rate": 3.467051083250161e-07,
+      "loss": 0.2876,
+      "step": 3165
+    },
+    {
+      "epoch": 1.8488431501350266,
+      "grad_norm": 0.08137325819620245,
+      "learning_rate": 3.440484462971549e-07,
+      "loss": 0.2815,
+      "step": 3166
+    },
+    {
+      "epoch": 1.8494270491205023,
+      "grad_norm": 0.08400540575868719,
+      "learning_rate": 3.414018237532335e-07,
+      "loss": 0.2918,
+      "step": 3167
+    },
+    {
+      "epoch": 1.8500109481059777,
+      "grad_norm": 0.08602669614112628,
+      "learning_rate": 3.3876524344499507e-07,
+      "loss": 0.3221,
+      "step": 3168
+    },
+    {
+      "epoch": 1.850594847091453,
+      "grad_norm": 0.08781031212883572,
+      "learning_rate": 3.3613870811374574e-07,
+      "loss": 0.3099,
+      "step": 3169
+    },
+    {
+      "epoch": 1.8511787460769287,
+      "grad_norm": 0.08306844298678354,
+      "learning_rate": 3.335222204903477e-07,
+      "loss": 0.3071,
+      "step": 3170
+    },
+    {
+      "epoch": 1.8517626450624043,
+      "grad_norm": 0.0854203297156832,
+      "learning_rate": 3.3091578329521147e-07,
+      "loss": 0.2871,
+      "step": 3171
+    },
+    {
+      "epoch": 1.8523465440478797,
+      "grad_norm": 0.08746449656902805,
+      "learning_rate": 3.283193992383049e-07,
+      "loss": 0.299,
+      "step": 3172
+    },
+    {
+      "epoch": 1.852930443033355,
+      "grad_norm": 0.08647666810832813,
+      "learning_rate": 3.2573307101913756e-07,
+      "loss": 0.292,
+      "step": 3173
+    },
+    {
+      "epoch": 1.8535143420188307,
+      "grad_norm": 0.0810278035339806,
+      "learning_rate": 3.231568013267672e-07,
+      "loss": 0.2572,
+      "step": 3174
+    },
+    {
+      "epoch": 1.8540982410043063,
+      "grad_norm": 0.0980012386328427,
+      "learning_rate": 3.205905928397923e-07,
+      "loss": 0.3378,
+      "step": 3175
+    },
+    {
+      "epoch": 1.854682139989782,
+      "grad_norm": 0.09653973467408311,
+      "learning_rate": 3.180344482263487e-07,
+      "loss": 0.3143,
+      "step": 3176
+    },
+    {
+      "epoch": 1.8552660389752573,
+      "grad_norm": 0.08385027433240437,
+      "learning_rate": 3.154883701441136e-07,
+      "loss": 0.272,
+      "step": 3177
+    },
+    {
+      "epoch": 1.8558499379607327,
+      "grad_norm": 0.08491885265963801,
+      "learning_rate": 3.129523612402918e-07,
+      "loss": 0.2752,
+      "step": 3178
+    },
+    {
+      "epoch": 1.8564338369462083,
+      "grad_norm": 0.08914676798011169,
+      "learning_rate": 3.1042642415162526e-07,
+      "loss": 0.3136,
+      "step": 3179
+    },
+    {
+      "epoch": 1.857017735931684,
+      "grad_norm": 0.09066710783232959,
+      "learning_rate": 3.079105615043787e-07,
+      "loss": 0.3118,
+      "step": 3180
+    },
+    {
+      "epoch": 1.8576016349171594,
+      "grad_norm": 0.0924604570837776,
+      "learning_rate": 3.0540477591434415e-07,
+      "loss": 0.3164,
+      "step": 3181
+    },
+    {
+      "epoch": 1.8581855339026347,
+      "grad_norm": 0.08868827412766157,
+      "learning_rate": 3.0290906998683755e-07,
+      "loss": 0.3012,
+      "step": 3182
+    },
+    {
+      "epoch": 1.8587694328881104,
+      "grad_norm": 0.08432166751587561,
+      "learning_rate": 3.0042344631669217e-07,
+      "loss": 0.2777,
+      "step": 3183
+    },
+    {
+      "epoch": 1.859353331873586,
+      "grad_norm": 0.09089092735538036,
+      "learning_rate": 2.9794790748826184e-07,
+      "loss": 0.3441,
+      "step": 3184
+    },
+    {
+      "epoch": 1.8599372308590614,
+      "grad_norm": 0.0833233822416718,
+      "learning_rate": 2.9548245607541326e-07,
+      "loss": 0.2735,
+      "step": 3185
+    },
+    {
+      "epoch": 1.8605211298445368,
+      "grad_norm": 0.08788146647829476,
+      "learning_rate": 2.9302709464152144e-07,
+      "loss": 0.2907,
+      "step": 3186
+    },
+    {
+      "epoch": 1.8611050288300124,
+      "grad_norm": 0.08750652520369745,
+      "learning_rate": 2.905818257394799e-07,
+      "loss": 0.3099,
+      "step": 3187
+    },
+    {
+      "epoch": 1.861688927815488,
+      "grad_norm": 0.07885783096976096,
+      "learning_rate": 2.881466519116793e-07,
+      "loss": 0.2818,
+      "step": 3188
+    },
+    {
+      "epoch": 1.8622728268009634,
+      "grad_norm": 0.08138112610132912,
+      "learning_rate": 2.857215756900189e-07,
+      "loss": 0.2906,
+      "step": 3189
+    },
+    {
+      "epoch": 1.8628567257864388,
+      "grad_norm": 0.08248114065113761,
+      "learning_rate": 2.8330659959589944e-07,
+      "loss": 0.2864,
+      "step": 3190
+    },
+    {
+      "epoch": 1.8634406247719144,
+      "grad_norm": 0.10013729848792587,
+      "learning_rate": 2.8090172614021804e-07,
+      "loss": 0.3166,
+      "step": 3191
+    },
+    {
+      "epoch": 1.86402452375739,
+      "grad_norm": 0.08778018259817816,
+      "learning_rate": 2.7850695782337124e-07,
+      "loss": 0.2584,
+      "step": 3192
+    },
+    {
+      "epoch": 1.8646084227428656,
+      "grad_norm": 0.09002304373718127,
+      "learning_rate": 2.761222971352451e-07,
+      "loss": 0.3043,
+      "step": 3193
+    },
+    {
+      "epoch": 1.865192321728341,
+      "grad_norm": 0.0930182053026977,
+      "learning_rate": 2.737477465552174e-07,
+      "loss": 0.3417,
+      "step": 3194
+    },
+    {
+      "epoch": 1.8657762207138164,
+      "grad_norm": 0.09634843250783759,
+      "learning_rate": 2.713833085521589e-07,
+      "loss": 0.3771,
+      "step": 3195
+    },
+    {
+      "epoch": 1.866360119699292,
+      "grad_norm": 0.07978999680208976,
+      "learning_rate": 2.690289855844186e-07,
+      "loss": 0.2777,
+      "step": 3196
+    },
+    {
+      "epoch": 1.8669440186847677,
+      "grad_norm": 0.08501056356556504,
+      "learning_rate": 2.666847800998362e-07,
+      "loss": 0.2715,
+      "step": 3197
+    },
+    {
+      "epoch": 1.867527917670243,
+      "grad_norm": 0.08415051694809264,
+      "learning_rate": 2.643506945357277e-07,
+      "loss": 0.3138,
+      "step": 3198
+    },
+    {
+      "epoch": 1.8681118166557185,
+      "grad_norm": 0.08667727196625621,
+      "learning_rate": 2.620267313188862e-07,
+      "loss": 0.2859,
+      "step": 3199
+    },
+    {
+      "epoch": 1.868695715641194,
+      "grad_norm": 0.09265032796899918,
+      "learning_rate": 2.5971289286558455e-07,
+      "loss": 0.3082,
+      "step": 3200
+    },
+    {
+      "epoch": 1.8692796146266697,
+      "grad_norm": 0.08568016543693206,
+      "learning_rate": 2.574091815815649e-07,
+      "loss": 0.2691,
+      "step": 3201
+    },
+    {
+      "epoch": 1.869863513612145,
+      "grad_norm": 0.08315796371653175,
+      "learning_rate": 2.5511559986204247e-07,
+      "loss": 0.2805,
+      "step": 3202
+    },
+    {
+      "epoch": 1.8704474125976205,
+      "grad_norm": 0.09394575929229129,
+      "learning_rate": 2.528321500916986e-07,
+      "loss": 0.3513,
+      "step": 3203
+    },
+    {
+      "epoch": 1.871031311583096,
+      "grad_norm": 0.0832153258472712,
+      "learning_rate": 2.505588346446808e-07,
+      "loss": 0.2758,
+      "step": 3204
+    },
+    {
+      "epoch": 1.8716152105685717,
+      "grad_norm": 0.08930126192828468,
+      "learning_rate": 2.482956558846017e-07,
+      "loss": 0.3182,
+      "step": 3205
+    },
+    {
+      "epoch": 1.8721991095540471,
+      "grad_norm": 0.08118777950327088,
+      "learning_rate": 2.460426161645324e-07,
+      "loss": 0.2614,
+      "step": 3206
+    },
+    {
+      "epoch": 1.8727830085395225,
+      "grad_norm": 0.09059797285382805,
+      "learning_rate": 2.437997178270035e-07,
+      "loss": 0.2972,
+      "step": 3207
+    },
+    {
+      "epoch": 1.8733669075249981,
+      "grad_norm": 0.08960921700831702,
+      "learning_rate": 2.4156696320399963e-07,
+      "loss": 0.2684,
+      "step": 3208
+    },
+    {
+      "epoch": 1.8739508065104737,
+      "grad_norm": 0.0966821886052414,
+      "learning_rate": 2.3934435461695936e-07,
+      "loss": 0.334,
+      "step": 3209
+    },
+    {
+      "epoch": 1.8745347054959494,
+      "grad_norm": 0.09395171964838248,
+      "learning_rate": 2.371318943767753e-07,
+      "loss": 0.3367,
+      "step": 3210
+    },
+    {
+      "epoch": 1.8751186044814248,
+      "grad_norm": 0.08787326693568194,
+      "learning_rate": 2.3492958478378402e-07,
+      "loss": 0.3008,
+      "step": 3211
+    },
+    {
+      "epoch": 1.8757025034669002,
+      "grad_norm": 0.09208125803804201,
+      "learning_rate": 2.3273742812777166e-07,
+      "loss": 0.326,
+      "step": 3212
+    },
+    {
+      "epoch": 1.8762864024523758,
+      "grad_norm": 0.09342802887635838,
+      "learning_rate": 2.3055542668796617e-07,
+      "loss": 0.3231,
+      "step": 3213
+    },
+    {
+      "epoch": 1.8768703014378514,
+      "grad_norm": 0.08486934855773866,
+      "learning_rate": 2.2838358273303717e-07,
+      "loss": 0.2829,
+      "step": 3214
+    },
+    {
+      "epoch": 1.8774542004233268,
+      "grad_norm": 0.09045340626851998,
+      "learning_rate": 2.26221898521094e-07,
+      "loss": 0.3011,
+      "step": 3215
+    },
+    {
+      "epoch": 1.8780380994088022,
+      "grad_norm": 0.0936406163064616,
+      "learning_rate": 2.2407037629968431e-07,
+      "loss": 0.3186,
+      "step": 3216
+    },
+    {
+      "epoch": 1.8786219983942778,
+      "grad_norm": 0.09178412076908621,
+      "learning_rate": 2.219290183057865e-07,
+      "loss": 0.355,
+      "step": 3217
+    },
+    {
+      "epoch": 1.8792058973797534,
+      "grad_norm": 0.08962622058153048,
+      "learning_rate": 2.1979782676581408e-07,
+      "loss": 0.296,
+      "step": 3218
+    },
+    {
+      "epoch": 1.8797897963652288,
+      "grad_norm": 0.08332687568912045,
+      "learning_rate": 2.1767680389560785e-07,
+      "loss": 0.3059,
+      "step": 3219
+    },
+    {
+      "epoch": 1.8803736953507042,
+      "grad_norm": 0.08814828921848539,
+      "learning_rate": 2.1556595190043718e-07,
+      "loss": 0.3114,
+      "step": 3220
+    },
+    {
+      "epoch": 1.8809575943361798,
+      "grad_norm": 0.08425676273658249,
+      "learning_rate": 2.1346527297499752e-07,
+      "loss": 0.3074,
+      "step": 3221
+    },
+    {
+      "epoch": 1.8815414933216554,
+      "grad_norm": 0.08434718710835229,
+      "learning_rate": 2.1137476930340628e-07,
+      "loss": 0.3077,
+      "step": 3222
+    },
+    {
+      "epoch": 1.8821253923071308,
+      "grad_norm": 0.08432170986832635,
+      "learning_rate": 2.0929444305920142e-07,
+      "loss": 0.2882,
+      "step": 3223
+    },
+    {
+      "epoch": 1.8827092912926062,
+      "grad_norm": 0.08637993322234078,
+      "learning_rate": 2.0722429640533948e-07,
+      "loss": 0.2948,
+      "step": 3224
+    },
+    {
+      "epoch": 1.8832931902780818,
+      "grad_norm": 0.09489175110302694,
+      "learning_rate": 2.051643314941909e-07,
+      "loss": 0.2993,
+      "step": 3225
+    },
+    {
+      "epoch": 1.8838770892635575,
+      "grad_norm": 0.08891721015551639,
+      "learning_rate": 2.0311455046754581e-07,
+      "loss": 0.2989,
+      "step": 3226
+    },
+    {
+      "epoch": 1.884460988249033,
+      "grad_norm": 0.09167106312929948,
+      "learning_rate": 2.0107495545659829e-07,
+      "loss": 0.302,
+      "step": 3227
+    },
+    {
+      "epoch": 1.8850448872345085,
+      "grad_norm": 0.08699042205183016,
+      "learning_rate": 1.990455485819587e-07,
+      "loss": 0.3256,
+      "step": 3228
+    },
+    {
+      "epoch": 1.8856287862199839,
+      "grad_norm": 0.09084983713310536,
+      "learning_rate": 1.9702633195363918e-07,
+      "loss": 0.2953,
+      "step": 3229
+    },
+    {
+      "epoch": 1.8862126852054595,
+      "grad_norm": 0.09383385568633446,
+      "learning_rate": 1.9501730767106043e-07,
+      "loss": 0.3062,
+      "step": 3230
+    },
+    {
+      "epoch": 1.886796584190935,
+      "grad_norm": 0.0837234249802778,
+      "learning_rate": 1.9301847782304484e-07,
+      "loss": 0.2986,
+      "step": 3231
+    },
+    {
+      "epoch": 1.8873804831764105,
+      "grad_norm": 0.08857392991618783,
+      "learning_rate": 1.9102984448781337e-07,
+      "loss": 0.3176,
+      "step": 3232
+    },
+    {
+      "epoch": 1.887964382161886,
+      "grad_norm": 0.08656967672259219,
+      "learning_rate": 1.8905140973299096e-07,
+      "loss": 0.3184,
+      "step": 3233
+    },
+    {
+      "epoch": 1.8885482811473615,
+      "grad_norm": 0.08976990713311463,
+      "learning_rate": 1.870831756155933e-07,
+      "loss": 0.3281,
+      "step": 3234
+    },
+    {
+      "epoch": 1.8891321801328371,
+      "grad_norm": 0.09174098245470733,
+      "learning_rate": 1.851251441820323e-07,
+      "loss": 0.2601,
+      "step": 3235
+    },
+    {
+      "epoch": 1.8897160791183125,
+      "grad_norm": 0.0877419662733352,
+      "learning_rate": 1.8317731746811285e-07,
+      "loss": 0.2943,
+      "step": 3236
+    },
+    {
+      "epoch": 1.890299978103788,
+      "grad_norm": 0.08792933605122286,
+      "learning_rate": 1.8123969749902714e-07,
+      "loss": 0.3242,
+      "step": 3237
+    },
+    {
+      "epoch": 1.8908838770892635,
+      "grad_norm": 0.09458027744263292,
+      "learning_rate": 1.7931228628935926e-07,
+      "loss": 0.3343,
+      "step": 3238
+    },
+    {
+      "epoch": 1.8914677760747391,
+      "grad_norm": 0.09107194378306366,
+      "learning_rate": 1.773950858430762e-07,
+      "loss": 0.2962,
+      "step": 3239
+    },
+    {
+      "epoch": 1.8920516750602145,
+      "grad_norm": 0.08413784115411445,
+      "learning_rate": 1.7548809815352785e-07,
+      "loss": 0.3015,
+      "step": 3240
+    },
+    {
+      "epoch": 1.89263557404569,
+      "grad_norm": 0.0911617935886624,
+      "learning_rate": 1.7359132520344823e-07,
+      "loss": 0.3254,
+      "step": 3241
+    },
+    {
+      "epoch": 1.8932194730311656,
+      "grad_norm": 0.09079456462290109,
+      "learning_rate": 1.717047689649487e-07,
+      "loss": 0.2945,
+      "step": 3242
+    },
+    {
+      "epoch": 1.8938033720166412,
+      "grad_norm": 0.08966260488997663,
+      "learning_rate": 1.6982843139952022e-07,
+      "loss": 0.2768,
+      "step": 3243
+    },
+    {
+      "epoch": 1.8943872710021168,
+      "grad_norm": 0.08943950036891886,
+      "learning_rate": 1.6796231445802892e-07,
+      "loss": 0.3035,
+      "step": 3244
+    },
+    {
+      "epoch": 1.8949711699875922,
+      "grad_norm": 0.0810749491430996,
+      "learning_rate": 1.6610642008071166e-07,
+      "loss": 0.2635,
+      "step": 3245
+    },
+    {
+      "epoch": 1.8955550689730676,
+      "grad_norm": 0.08932620447075977,
+      "learning_rate": 1.6426075019717935e-07,
+      "loss": 0.321,
+      "step": 3246
+    },
+    {
+      "epoch": 1.8961389679585432,
+      "grad_norm": 0.08750921257942008,
+      "learning_rate": 1.6242530672641143e-07,
+      "loss": 0.265,
+      "step": 3247
+    },
+    {
+      "epoch": 1.8967228669440188,
+      "grad_norm": 0.08875912637695582,
+      "learning_rate": 1.6060009157675472e-07,
+      "loss": 0.3051,
+      "step": 3248
+    },
+    {
+      "epoch": 1.8973067659294942,
+      "grad_norm": 0.08614639821782472,
+      "learning_rate": 1.5878510664592116e-07,
+      "loss": 0.2705,
+      "step": 3249
+    },
+    {
+      "epoch": 1.8978906649149696,
+      "grad_norm": 0.0879684254585697,
+      "learning_rate": 1.5698035382098687e-07,
+      "loss": 0.2962,
+      "step": 3250
+    },
+    {
+      "epoch": 1.8984745639004452,
+      "grad_norm": 0.08828834194867435,
+      "learning_rate": 1.551858349783908e-07,
+      "loss": 0.2988,
+      "step": 3251
+    },
+    {
+      "epoch": 1.8990584628859208,
+      "grad_norm": 0.08519332805888366,
+      "learning_rate": 1.5340155198392716e-07,
+      "loss": 0.2984,
+      "step": 3252
+    },
+    {
+      "epoch": 1.8996423618713962,
+      "grad_norm": 0.08511332372469033,
+      "learning_rate": 1.5162750669274973e-07,
+      "loss": 0.2927,
+      "step": 3253
+    },
+    {
+      "epoch": 1.9002262608568716,
+      "grad_norm": 0.08771274300049751,
+      "learning_rate": 1.4986370094937197e-07,
+      "loss": 0.3081,
+      "step": 3254
+    },
+    {
+      "epoch": 1.9008101598423472,
+      "grad_norm": 0.08273804207822551,
+      "learning_rate": 1.4811013658765471e-07,
+      "loss": 0.2846,
+      "step": 3255
+    },
+    {
+      "epoch": 1.9013940588278229,
+      "grad_norm": 0.08324330320189118,
+      "learning_rate": 1.46366815430814e-07,
+      "loss": 0.281,
+      "step": 3256
+    },
+    {
+      "epoch": 1.9019779578132983,
+      "grad_norm": 0.0953682159842167,
+      "learning_rate": 1.4463373929141766e-07,
+      "loss": 0.3435,
+      "step": 3257
+    },
+    {
+      "epoch": 1.9025618567987737,
+      "grad_norm": 0.08604112753118452,
+      "learning_rate": 1.4291090997137547e-07,
+      "loss": 0.2663,
+      "step": 3258
+    },
+    {
+      "epoch": 1.9031457557842493,
+      "grad_norm": 0.0851146798833154,
+      "learning_rate": 1.411983292619501e-07,
+      "loss": 0.2565,
+      "step": 3259
+    },
+    {
+      "epoch": 1.9037296547697249,
+      "grad_norm": 0.08033382758542612,
+      "learning_rate": 1.3949599894374276e-07,
+      "loss": 0.2675,
+      "step": 3260
+    },
+    {
+      "epoch": 1.9043135537552005,
+      "grad_norm": 0.08985508538737608,
+      "learning_rate": 1.3780392078670436e-07,
+      "loss": 0.3115,
+      "step": 3261
+    },
+    {
+      "epoch": 1.904897452740676,
+      "grad_norm": 0.08396430745585091,
+      "learning_rate": 1.3612209655011866e-07,
+      "loss": 0.2944,
+      "step": 3262
+    },
+    {
+      "epoch": 1.9054813517261513,
+      "grad_norm": 0.0867258398254808,
+      "learning_rate": 1.3445052798261137e-07,
+      "loss": 0.3139,
+      "step": 3263
+    },
+    {
+      "epoch": 1.906065250711627,
+      "grad_norm": 0.09233210610326666,
+      "learning_rate": 1.3278921682214784e-07,
+      "loss": 0.3245,
+      "step": 3264
+    },
+    {
+      "epoch": 1.9066491496971025,
+      "grad_norm": 0.08621102322533856,
+      "learning_rate": 1.3113816479602304e-07,
+      "loss": 0.3095,
+      "step": 3265
+    },
+    {
+      "epoch": 1.907233048682578,
+      "grad_norm": 0.0799357482093539,
+      "learning_rate": 1.2949737362087156e-07,
+      "loss": 0.2624,
+      "step": 3266
+    },
+    {
+      "epoch": 1.9078169476680533,
+      "grad_norm": 0.0902174151582053,
+      "learning_rate": 1.2786684500265546e-07,
+      "loss": 0.3257,
+      "step": 3267
+    },
+    {
+      "epoch": 1.908400846653529,
+      "grad_norm": 0.08688867735884609,
+      "learning_rate": 1.262465806366664e-07,
+      "loss": 0.2847,
+      "step": 3268
+    },
+    {
+      "epoch": 1.9089847456390046,
+      "grad_norm": 0.09055661264358911,
+      "learning_rate": 1.2463658220752683e-07,
+      "loss": 0.3163,
+      "step": 3269
+    },
+    {
+      "epoch": 1.90956864462448,
+      "grad_norm": 0.08253503170116067,
+      "learning_rate": 1.230368513891822e-07,
+      "loss": 0.3061,
+      "step": 3270
+    },
+    {
+      "epoch": 1.9101525436099553,
+      "grad_norm": 0.0885613762724995,
+      "learning_rate": 1.2144738984490533e-07,
+      "loss": 0.2883,
+      "step": 3271
+    },
+    {
+      "epoch": 1.910736442595431,
+      "grad_norm": 0.08588209112933573,
+      "learning_rate": 1.1986819922729209e-07,
+      "loss": 0.2887,
+      "step": 3272
+    },
+    {
+      "epoch": 1.9113203415809066,
+      "grad_norm": 0.08259863477309273,
+      "learning_rate": 1.1829928117825685e-07,
+      "loss": 0.2948,
+      "step": 3273
+    },
+    {
+      "epoch": 1.911904240566382,
+      "grad_norm": 0.08781060833998246,
+      "learning_rate": 1.1674063732903473e-07,
+      "loss": 0.313,
+      "step": 3274
+    },
+    {
+      "epoch": 1.9124881395518574,
+      "grad_norm": 0.07800928560429705,
+      "learning_rate": 1.1519226930017946e-07,
+      "loss": 0.2471,
+      "step": 3275
+    },
+    {
+      "epoch": 1.913072038537333,
+      "grad_norm": 0.09438116030362541,
+      "learning_rate": 1.1365417870155881e-07,
+      "loss": 0.3579,
+      "step": 3276
+    },
+    {
+      "epoch": 1.9136559375228086,
+      "grad_norm": 0.0888049023557599,
+      "learning_rate": 1.1212636713235581e-07,
+      "loss": 0.3115,
+      "step": 3277
+    },
+    {
+      "epoch": 1.9142398365082842,
+      "grad_norm": 0.08749539087352674,
+      "learning_rate": 1.1060883618106754e-07,
+      "loss": 0.2872,
+      "step": 3278
+    },
+    {
+      "epoch": 1.9148237354937596,
+      "grad_norm": 0.08449612607203431,
+      "learning_rate": 1.0910158742550081e-07,
+      "loss": 0.2699,
+      "step": 3279
+    },
+    {
+      "epoch": 1.915407634479235,
+      "grad_norm": 0.08373605300564951,
+      "learning_rate": 1.0760462243277204e-07,
+      "loss": 0.2941,
+      "step": 3280
+    },
+    {
+      "epoch": 1.9159915334647106,
+      "grad_norm": 0.08543920487635473,
+      "learning_rate": 1.0611794275930398e-07,
+      "loss": 0.2828,
+      "step": 3281
+    },
+    {
+      "epoch": 1.9165754324501862,
+      "grad_norm": 0.0883860155897194,
+      "learning_rate": 1.0464154995082909e-07,
+      "loss": 0.3001,
+      "step": 3282
+    },
+    {
+      "epoch": 1.9171593314356616,
+      "grad_norm": 0.08644895600908652,
+      "learning_rate": 1.0317544554238058e-07,
+      "loss": 0.2775,
+      "step": 3283
+    },
+    {
+      "epoch": 1.917743230421137,
+      "grad_norm": 0.08924774098696027,
+      "learning_rate": 1.0171963105829686e-07,
+      "loss": 0.3156,
+      "step": 3284
+    },
+    {
+      "epoch": 1.9183271294066127,
+      "grad_norm": 0.08889517922219568,
+      "learning_rate": 1.0027410801221604e-07,
+      "loss": 0.3026,
+      "step": 3285
+    },
+    {
+      "epoch": 1.9189110283920883,
+      "grad_norm": 0.0889077945705487,
+      "learning_rate": 9.883887790707814e-08,
+      "loss": 0.3106,
+      "step": 3286
+    },
+    {
+      "epoch": 1.9194949273775637,
+      "grad_norm": 0.09210317923853588,
+      "learning_rate": 9.741394223512057e-08,
+      "loss": 0.3296,
+      "step": 3287
+    },
+    {
+      "epoch": 1.920078826363039,
+      "grad_norm": 0.09453765765660951,
+      "learning_rate": 9.599930247787604e-08,
+      "loss": 0.305,
+      "step": 3288
+    },
+    {
+      "epoch": 1.9206627253485147,
+      "grad_norm": 0.08873585035845115,
+      "learning_rate": 9.459496010617464e-08,
+      "loss": 0.3234,
+      "step": 3289
+    },
+    {
+      "epoch": 1.9212466243339903,
+      "grad_norm": 0.08756919273356488,
+      "learning_rate": 9.320091658013841e-08,
+      "loss": 0.2981,
+      "step": 3290
+    },
+    {
+      "epoch": 1.9218305233194657,
+      "grad_norm": 0.08310739555407189,
+      "learning_rate": 9.181717334918127e-08,
+      "loss": 0.2858,
+      "step": 3291
+    },
+    {
+      "epoch": 1.922414422304941,
+      "grad_norm": 0.09673147595351422,
+      "learning_rate": 9.044373185200906e-08,
+      "loss": 0.336,
+      "step": 3292
+    },
+    {
+      "epoch": 1.9229983212904167,
+      "grad_norm": 0.08324882756681469,
+      "learning_rate": 8.908059351661725e-08,
+      "loss": 0.2583,
+      "step": 3293
+    },
+    {
+      "epoch": 1.9235822202758923,
+      "grad_norm": 0.08840073143828481,
+      "learning_rate": 8.772775976028547e-08,
+      "loss": 0.2925,
+      "step": 3294
+    },
+    {
+      "epoch": 1.924166119261368,
+      "grad_norm": 0.08169262018525472,
+      "learning_rate": 8.638523198958415e-08,
+      "loss": 0.3132,
+      "step": 3295
+    },
+    {
+      "epoch": 1.9247500182468433,
+      "grad_norm": 0.08563499806240714,
+      "learning_rate": 8.505301160036339e-08,
+      "loss": 0.2747,
+      "step": 3296
+    },
+    {
+      "epoch": 1.9253339172323187,
+      "grad_norm": 0.09364279015240973,
+      "learning_rate": 8.373109997776185e-08,
+      "loss": 0.2786,
+      "step": 3297
+    },
+    {
+      "epoch": 1.9259178162177943,
+      "grad_norm": 0.08811137414677286,
+      "learning_rate": 8.241949849619457e-08,
+      "loss": 0.2993,
+      "step": 3298
+    },
+    {
+      "epoch": 1.92650171520327,
+      "grad_norm": 0.09650951937213867,
+      "learning_rate": 8.11182085193607e-08,
+      "loss": 0.3514,
+      "step": 3299
+    },
+    {
+      "epoch": 1.9270856141887454,
+      "grad_norm": 0.08503289378115884,
+      "learning_rate": 7.982723140023906e-08,
+      "loss": 0.2723,
+      "step": 3300
+    },
+    {
+      "epoch": 1.9276695131742208,
+      "grad_norm": 0.08461056135829302,
+      "learning_rate": 7.854656848108044e-08,
+      "loss": 0.308,
+      "step": 3301
+    },
+    {
+      "epoch": 1.9282534121596964,
+      "grad_norm": 0.08576853459491707,
+      "learning_rate": 7.727622109341859e-08,
+      "loss": 0.2886,
+      "step": 3302
+    },
+    {
+      "epoch": 1.928837311145172,
+      "grad_norm": 0.08865278880622722,
+      "learning_rate": 7.601619055805697e-08,
+      "loss": 0.315,
+      "step": 3303
+    },
+    {
+      "epoch": 1.9294212101306474,
+      "grad_norm": 0.09539147794132567,
+      "learning_rate": 7.476647818507542e-08,
+      "loss": 0.3568,
+      "step": 3304
+    },
+    {
+      "epoch": 1.9300051091161228,
+      "grad_norm": 0.07761163491199877,
+      "learning_rate": 7.352708527382346e-08,
+      "loss": 0.2632,
+      "step": 3305
+    },
+    {
+      "epoch": 1.9305890081015984,
+      "grad_norm": 0.08533237598639366,
+      "learning_rate": 7.229801311292361e-08,
+      "loss": 0.3125,
+      "step": 3306
+    },
+    {
+      "epoch": 1.931172907087074,
+      "grad_norm": 0.08847538834680153,
+      "learning_rate": 7.10792629802659e-08,
+      "loss": 0.287,
+      "step": 3307
+    },
+    {
+      "epoch": 1.9317568060725494,
+      "grad_norm": 0.09537042694152782,
+      "learning_rate": 6.987083614300893e-08,
+      "loss": 0.3387,
+      "step": 3308
+    },
+    {
+      "epoch": 1.9323407050580248,
+      "grad_norm": 0.08632859807068002,
+      "learning_rate": 6.867273385757767e-08,
+      "loss": 0.2974,
+      "step": 3309
+    },
+    {
+      "epoch": 1.9329246040435004,
+      "grad_norm": 0.09619285487797331,
+      "learning_rate": 6.748495736966454e-08,
+      "loss": 0.3422,
+      "step": 3310
+    },
+    {
+      "epoch": 1.933508503028976,
+      "grad_norm": 0.09141794237098383,
+      "learning_rate": 6.630750791422169e-08,
+      "loss": 0.3164,
+      "step": 3311
+    },
+    {
+      "epoch": 1.9340924020144517,
+      "grad_norm": 0.09084630148325167,
+      "learning_rate": 6.514038671546874e-08,
+      "loss": 0.2986,
+      "step": 3312
+    },
+    {
+      "epoch": 1.934676300999927,
+      "grad_norm": 0.0907609353472204,
+      "learning_rate": 6.398359498688278e-08,
+      "loss": 0.3196,
+      "step": 3313
+    },
+    {
+      "epoch": 1.9352601999854024,
+      "grad_norm": 0.07503258219047194,
+      "learning_rate": 6.283713393120505e-08,
+      "loss": 0.2637,
+      "step": 3314
+    },
+    {
+      "epoch": 1.935844098970878,
+      "grad_norm": 0.08659922375515655,
+      "learning_rate": 6.170100474043206e-08,
+      "loss": 0.3077,
+      "step": 3315
+    },
+    {
+      "epoch": 1.9364279979563537,
+      "grad_norm": 0.09162438207726716,
+      "learning_rate": 6.05752085958211e-08,
+      "loss": 0.3356,
+      "step": 3316
+    },
+    {
+      "epoch": 1.937011896941829,
+      "grad_norm": 0.0826535538666985,
+      "learning_rate": 5.945974666788479e-08,
+      "loss": 0.2826,
+      "step": 3317
+    },
+    {
+      "epoch": 1.9375957959273045,
+      "grad_norm": 0.08112764255175492,
+      "learning_rate": 5.835462011638982e-08,
+      "loss": 0.2888,
+      "step": 3318
+    },
+    {
+      "epoch": 1.93817969491278,
+      "grad_norm": 0.08473385645129632,
+      "learning_rate": 5.725983009035818e-08,
+      "loss": 0.3272,
+      "step": 3319
+    },
+    {
+      "epoch": 1.9387635938982557,
+      "grad_norm": 0.08263858758524921,
+      "learning_rate": 5.617537772806603e-08,
+      "loss": 0.2778,
+      "step": 3320
+    },
+    {
+      "epoch": 1.939347492883731,
+      "grad_norm": 0.08714126481647054,
+      "learning_rate": 5.5101264157039203e-08,
+      "loss": 0.29,
+      "step": 3321
+    },
+    {
+      "epoch": 1.9399313918692065,
+      "grad_norm": 0.08612721229120282,
+      "learning_rate": 5.403749049405438e-08,
+      "loss": 0.3274,
+      "step": 3322
+    },
+    {
+      "epoch": 1.940515290854682,
+      "grad_norm": 0.08179451740463552,
+      "learning_rate": 5.298405784513905e-08,
+      "loss": 0.2616,
+      "step": 3323
+    },
+    {
+      "epoch": 1.9410991898401577,
+      "grad_norm": 0.08895224797358554,
+      "learning_rate": 5.194096730556708e-08,
+      "loss": 0.3375,
+      "step": 3324
+    },
+    {
+      "epoch": 1.9416830888256331,
+      "grad_norm": 0.0806249885976326,
+      "learning_rate": 5.090821995986095e-08,
+      "loss": 0.2847,
+      "step": 3325
+    },
+    {
+      "epoch": 1.9422669878111085,
+      "grad_norm": 0.08725342168358602,
+      "learning_rate": 4.9885816881787287e-08,
+      "loss": 0.2801,
+      "step": 3326
+    },
+    {
+      "epoch": 1.9428508867965841,
+      "grad_norm": 0.08397668838363491,
+      "learning_rate": 4.887375913436132e-08,
+      "loss": 0.3061,
+      "step": 3327
+    },
+    {
+      "epoch": 1.9434347857820597,
+      "grad_norm": 0.08311493018246177,
+      "learning_rate": 4.787204776983689e-08,
+      "loss": 0.2852,
+      "step": 3328
+    },
+    {
+      "epoch": 1.9440186847675354,
+      "grad_norm": 0.08132389367275121,
+      "learning_rate": 4.6880683829715335e-08,
+      "loss": 0.2782,
+      "step": 3329
+    },
+    {
+      "epoch": 1.9446025837530108,
+      "grad_norm": 0.08133958791623556,
+      "learning_rate": 4.589966834473547e-08,
+      "loss": 0.289,
+      "step": 3330
+    },
+    {
+      "epoch": 1.9451864827384862,
+      "grad_norm": 0.09206433800152225,
+      "learning_rate": 4.492900233488029e-08,
+      "loss": 0.2817,
+      "step": 3331
+    },
+    {
+      "epoch": 1.9457703817239618,
+      "grad_norm": 0.09414037671837533,
+      "learning_rate": 4.3968686809369165e-08,
+      "loss": 0.2989,
+      "step": 3332
+    },
+    {
+      "epoch": 1.9463542807094374,
+      "grad_norm": 0.08296730047872414,
+      "learning_rate": 4.3018722766661194e-08,
+      "loss": 0.2916,
+      "step": 3333
+    },
+    {
+      "epoch": 1.9469381796949128,
+      "grad_norm": 0.0880310852879677,
+      "learning_rate": 4.207911119445296e-08,
+      "loss": 0.3074,
+      "step": 3334
+    },
+    {
+      "epoch": 1.9475220786803882,
+      "grad_norm": 0.0959745051458115,
+      "learning_rate": 4.114985306967745e-08,
+      "loss": 0.2968,
+      "step": 3335
+    },
+    {
+      "epoch": 1.9481059776658638,
+      "grad_norm": 0.08273103877297201,
+      "learning_rate": 4.0230949358502915e-08,
+      "loss": 0.2589,
+      "step": 3336
+    },
+    {
+      "epoch": 1.9486898766513394,
+      "grad_norm": 0.09296593816380899,
+      "learning_rate": 3.932240101633178e-08,
+      "loss": 0.3107,
+      "step": 3337
+    },
+    {
+      "epoch": 1.9492737756368148,
+      "grad_norm": 0.08136191865828954,
+      "learning_rate": 3.8424208987798415e-08,
+      "loss": 0.2487,
+      "step": 3338
+    },
+    {
+      "epoch": 1.9498576746222902,
+      "grad_norm": 0.09058544275619738,
+      "learning_rate": 3.7536374206772475e-08,
+      "loss": 0.3109,
+      "step": 3339
+    },
+    {
+      "epoch": 1.9504415736077658,
+      "grad_norm": 0.07694637322003581,
+      "learning_rate": 3.665889759635222e-08,
+      "loss": 0.2761,
+      "step": 3340
+    },
+    {
+      "epoch": 1.9510254725932414,
+      "grad_norm": 0.07982952965228436,
+      "learning_rate": 3.579178006886896e-08,
+      "loss": 0.2866,
+      "step": 3341
+    },
+    {
+      "epoch": 1.9516093715787168,
+      "grad_norm": 0.09260072866321262,
+      "learning_rate": 3.4935022525880434e-08,
+      "loss": 0.3574,
+      "step": 3342
+    },
+    {
+      "epoch": 1.9521932705641922,
+      "grad_norm": 0.08237078614942933,
+      "learning_rate": 3.4088625858174075e-08,
+      "loss": 0.2739,
+      "step": 3343
+    },
+    {
+      "epoch": 1.9527771695496678,
+      "grad_norm": 0.08504126457069763,
+      "learning_rate": 3.3252590945767047e-08,
+      "loss": 0.2649,
+      "step": 3344
+    },
+    {
+      "epoch": 1.9533610685351435,
+      "grad_norm": 0.08397940540913501,
+      "learning_rate": 3.242691865790071e-08,
+      "loss": 0.312,
+      "step": 3345
+    },
+    {
+      "epoch": 1.953944967520619,
+      "grad_norm": 0.0937908923265636,
+      "learning_rate": 3.161160985304168e-08,
+      "loss": 0.3142,
+      "step": 3346
+    },
+    {
+      "epoch": 1.9545288665060945,
+      "grad_norm": 0.08645231463187328,
+      "learning_rate": 3.0806665378884106e-08,
+      "loss": 0.2974,
+      "step": 3347
+    },
+    {
+      "epoch": 1.9551127654915699,
+      "grad_norm": 0.0930396155565426,
+      "learning_rate": 3.001208607234407e-08,
+      "loss": 0.3053,
+      "step": 3348
+    },
+    {
+      "epoch": 1.9556966644770455,
+      "grad_norm": 0.08984322208002168,
+      "learning_rate": 2.922787275956074e-08,
+      "loss": 0.2911,
+      "step": 3349
+    },
+    {
+      "epoch": 1.956280563462521,
+      "grad_norm": 0.08848884936006156,
+      "learning_rate": 2.845402625589855e-08,
+      "loss": 0.3095,
+      "step": 3350
+    },
+    {
+      "epoch": 1.9568644624479965,
+      "grad_norm": 0.08878169921273348,
+      "learning_rate": 2.7690547365938348e-08,
+      "loss": 0.2801,
+      "step": 3351
+    },
+    {
+      "epoch": 1.957448361433472,
+      "grad_norm": 0.08444589281407235,
+      "learning_rate": 2.693743688348627e-08,
+      "loss": 0.3048,
+      "step": 3352
+    },
+    {
+      "epoch": 1.9580322604189475,
+      "grad_norm": 0.08365770785547919,
+      "learning_rate": 2.6194695591563733e-08,
+      "loss": 0.2731,
+      "step": 3353
+    },
+    {
+      "epoch": 1.9586161594044231,
+      "grad_norm": 0.08859156119827412,
+      "learning_rate": 2.546232426241635e-08,
+      "loss": 0.3129,
+      "step": 3354
+    },
+    {
+      "epoch": 1.9592000583898985,
+      "grad_norm": 0.0830060806441253,
+      "learning_rate": 2.4740323657503895e-08,
+      "loss": 0.2902,
+      "step": 3355
+    },
+    {
+      "epoch": 1.959783957375374,
+      "grad_norm": 0.0894978516565684,
+      "learning_rate": 2.4028694527503673e-08,
+      "loss": 0.2761,
+      "step": 3356
+    },
+    {
+      "epoch": 1.9603678563608495,
+      "grad_norm": 0.09439828917867563,
+      "learning_rate": 2.33274376123116e-08,
+      "loss": 0.35,
+      "step": 3357
+    },
+    {
+      "epoch": 1.9609517553463252,
+      "grad_norm": 0.08267329975177361,
+      "learning_rate": 2.2636553641040003e-08,
+      "loss": 0.3061,
+      "step": 3358
+    },
+    {
+      "epoch": 1.9615356543318005,
+      "grad_norm": 0.08930647174651028,
+      "learning_rate": 2.1956043332010957e-08,
+      "loss": 0.3234,
+      "step": 3359
+    },
+    {
+      "epoch": 1.9621195533172762,
+      "grad_norm": 0.08520907708355664,
+      "learning_rate": 2.1285907392767367e-08,
+      "loss": 0.2686,
+      "step": 3360
+    },
+    {
+      "epoch": 1.9627034523027516,
+      "grad_norm": 0.09453666948066386,
+      "learning_rate": 2.0626146520061897e-08,
+      "loss": 0.311,
+      "step": 3361
+    },
+    {
+      "epoch": 1.9632873512882272,
+      "grad_norm": 0.08767455534350097,
+      "learning_rate": 1.997676139986138e-08,
+      "loss": 0.3059,
+      "step": 3362
+    },
+    {
+      "epoch": 1.9638712502737028,
+      "grad_norm": 0.08686753460244756,
+      "learning_rate": 1.9337752707343504e-08,
+      "loss": 0.3191,
+      "step": 3363
+    },
+    {
+      "epoch": 1.9644551492591782,
+      "grad_norm": 0.08564873656027894,
+      "learning_rate": 1.8709121106899043e-08,
+      "loss": 0.2879,
+      "step": 3364
+    },
+    {
+      "epoch": 1.9650390482446536,
+      "grad_norm": 0.09425614638184439,
+      "learning_rate": 1.8090867252127387e-08,
+      "loss": 0.3066,
+      "step": 3365
+    },
+    {
+      "epoch": 1.9656229472301292,
+      "grad_norm": 0.08789698222748536,
+      "learning_rate": 1.748299178584101e-08,
+      "loss": 0.2837,
+      "step": 3366
+    },
+    {
+      "epoch": 1.9662068462156048,
+      "grad_norm": 0.09041254127113027,
+      "learning_rate": 1.68854953400599e-08,
+      "loss": 0.3112,
+      "step": 3367
+    },
+    {
+      "epoch": 1.9667907452010802,
+      "grad_norm": 0.0919964493149233,
+      "learning_rate": 1.6298378536012682e-08,
+      "loss": 0.3131,
+      "step": 3368
+    },
+    {
+      "epoch": 1.9673746441865556,
+      "grad_norm": 0.09061762810698683,
+      "learning_rate": 1.5721641984135505e-08,
+      "loss": 0.2989,
+      "step": 3369
+    },
+    {
+      "epoch": 1.9679585431720312,
+      "grad_norm": 0.08885739500219834,
+      "learning_rate": 1.5155286284073146e-08,
+      "loss": 0.2931,
+      "step": 3370
+    },
+    {
+      "epoch": 1.9685424421575068,
+      "grad_norm": 0.08829728675580047,
+      "learning_rate": 1.4599312024676792e-08,
+      "loss": 0.3196,
+      "step": 3371
+    },
+    {
+      "epoch": 1.9691263411429822,
+      "grad_norm": 0.0879976583844318,
+      "learning_rate": 1.405371978400516e-08,
+      "loss": 0.3191,
+      "step": 3372
+    },
+    {
+      "epoch": 1.9697102401284576,
+      "grad_norm": 0.0845225197448845,
+      "learning_rate": 1.351851012931893e-08,
+      "loss": 0.3084,
+      "step": 3373
+    },
+    {
+      "epoch": 1.9702941391139333,
+      "grad_norm": 0.09535173903567114,
+      "learning_rate": 1.2993683617088526e-08,
+      "loss": 0.2948,
+      "step": 3374
+    },
+    {
+      "epoch": 1.9708780380994089,
+      "grad_norm": 0.09404301255052072,
+      "learning_rate": 1.2479240792985237e-08,
+      "loss": 0.3177,
+      "step": 3375
+    },
+    {
+      "epoch": 1.9714619370848843,
+      "grad_norm": 0.22470472185817955,
+      "learning_rate": 1.1975182191885648e-08,
+      "loss": 0.3162,
+      "step": 3376
+    },
+    {
+      "epoch": 1.9720458360703599,
+      "grad_norm": 0.08564007484358507,
+      "learning_rate": 1.1481508337869429e-08,
+      "loss": 0.291,
+      "step": 3377
+    },
+    {
+      "epoch": 1.9726297350558353,
+      "grad_norm": 0.09645757791018185,
+      "learning_rate": 1.099821974421933e-08,
+      "loss": 0.3548,
+      "step": 3378
+    },
+    {
+      "epoch": 1.973213634041311,
+      "grad_norm": 0.08678124234324852,
+      "learning_rate": 1.0525316913420069e-08,
+      "loss": 0.2948,
+      "step": 3379
+    },
+    {
+      "epoch": 1.9737975330267865,
+      "grad_norm": 0.0896991287527311,
+      "learning_rate": 1.006280033715723e-08,
+      "loss": 0.313,
+      "step": 3380
+    },
+    {
+      "epoch": 1.974381432012262,
+      "grad_norm": 0.08701107505291639,
+      "learning_rate": 9.610670496319475e-09,
+      "loss": 0.3041,
+      "step": 3381
+    },
+    {
+      "epoch": 1.9749653309977373,
+      "grad_norm": 0.09028490035369065,
+      "learning_rate": 9.168927860994104e-09,
+      "loss": 0.2995,
+      "step": 3382
+    },
+    {
+      "epoch": 1.975549229983213,
+      "grad_norm": 0.09104900055373935,
+      "learning_rate": 8.737572890470391e-09,
+      "loss": 0.3101,
+      "step": 3383
+    },
+    {
+      "epoch": 1.9761331289686885,
+      "grad_norm": 0.08686736438325564,
+      "learning_rate": 8.316606033237362e-09,
+      "loss": 0.3048,
+      "step": 3384
+    },
+    {
+      "epoch": 1.976717027954164,
+      "grad_norm": 0.08840548190337588,
+      "learning_rate": 7.906027726981568e-09,
+      "loss": 0.3108,
+      "step": 3385
+    },
+    {
+      "epoch": 1.9773009269396393,
+      "grad_norm": 0.08926798479268085,
+      "learning_rate": 7.505838398589316e-09,
+      "loss": 0.312,
+      "step": 3386
+    },
+    {
+      "epoch": 1.977884825925115,
+      "grad_norm": 0.09001788956239173,
+      "learning_rate": 7.1160384641455475e-09,
+      "loss": 0.2844,
+      "step": 3387
+    },
+    {
+      "epoch": 1.9784687249105906,
+      "grad_norm": 0.09452923785089407,
+      "learning_rate": 6.736628328933847e-09,
+      "loss": 0.3454,
+      "step": 3388
+    },
+    {
+      "epoch": 1.979052623896066,
+      "grad_norm": 0.08582916109981235,
+      "learning_rate": 6.367608387433111e-09,
+      "loss": 0.307,
+      "step": 3389
+    },
+    {
+      "epoch": 1.9796365228815413,
+      "grad_norm": 0.09260456514767397,
+      "learning_rate": 6.008979023320871e-09,
+      "loss": 0.3304,
+      "step": 3390
+    },
+    {
+      "epoch": 1.980220421867017,
+      "grad_norm": 0.08586680764241691,
+      "learning_rate": 5.660740609472193e-09,
+      "loss": 0.2851,
+      "step": 3391
+    },
+    {
+      "epoch": 1.9808043208524926,
+      "grad_norm": 0.11215718400014124,
+      "learning_rate": 5.322893507956339e-09,
+      "loss": 0.405,
+      "step": 3392
+    },
+    {
+      "epoch": 1.981388219837968,
+      "grad_norm": 0.08258366551295686,
+      "learning_rate": 4.995438070041214e-09,
+      "loss": 0.3266,
+      "step": 3393
+    },
+    {
+      "epoch": 1.9819721188234436,
+      "grad_norm": 0.09343253408958818,
+      "learning_rate": 4.6783746361867e-09,
+      "loss": 0.345,
+      "step": 3394
+    },
+    {
+      "epoch": 1.982556017808919,
+      "grad_norm": 0.0909096520499017,
+      "learning_rate": 4.3717035360502094e-09,
+      "loss": 0.3349,
+      "step": 3395
+    },
+    {
+      "epoch": 1.9831399167943946,
+      "grad_norm": 0.09122709999757472,
+      "learning_rate": 4.075425088485574e-09,
+      "loss": 0.326,
+      "step": 3396
+    },
+    {
+      "epoch": 1.9837238157798702,
+      "grad_norm": 0.07958173922292036,
+      "learning_rate": 3.7895396015374955e-09,
+      "loss": 0.2752,
+      "step": 3397
+    },
+    {
+      "epoch": 1.9843077147653456,
+      "grad_norm": 0.08624117898251214,
+      "learning_rate": 3.514047372448204e-09,
+      "loss": 0.2848,
+      "step": 3398
+    },
+    {
+      "epoch": 1.984891613750821,
+      "grad_norm": 0.08999212926837558,
+      "learning_rate": 3.248948687650799e-09,
+      "loss": 0.3277,
+      "step": 3399
+    },
+    {
+      "epoch": 1.9854755127362966,
+      "grad_norm": 0.08259564950939226,
+      "learning_rate": 2.9942438227748004e-09,
+      "loss": 0.2781,
+      "step": 3400
+    },
+    {
+      "epoch": 1.9860594117217722,
+      "grad_norm": 0.0903701390679294,
+      "learning_rate": 2.749933042641706e-09,
+      "loss": 0.2865,
+      "step": 3401
+    },
+    {
+      "epoch": 1.9866433107072476,
+      "grad_norm": 0.08664047399989915,
+      "learning_rate": 2.5160166012661024e-09,
+      "loss": 0.2916,
+      "step": 3402
+    },
+    {
+      "epoch": 1.987227209692723,
+      "grad_norm": 0.08748391045458065,
+      "learning_rate": 2.2924947418556666e-09,
+      "loss": 0.2992,
+      "step": 3403
+    },
+    {
+      "epoch": 1.9878111086781987,
+      "grad_norm": 0.08890827386420758,
+      "learning_rate": 2.079367696810053e-09,
+      "loss": 0.3057,
+      "step": 3404
+    },
+    {
+      "epoch": 1.9883950076636743,
+      "grad_norm": 0.09034869082218763,
+      "learning_rate": 1.876635687722006e-09,
+      "loss": 0.3049,
+      "step": 3405
+    },
+    {
+      "epoch": 1.9889789066491497,
+      "grad_norm": 0.08374105709499309,
+      "learning_rate": 1.684298925377359e-09,
+      "loss": 0.3065,
+      "step": 3406
+    },
+    {
+      "epoch": 1.989562805634625,
+      "grad_norm": 0.09195756130161492,
+      "learning_rate": 1.502357609749483e-09,
+      "loss": 0.3375,
+      "step": 3407
+    },
+    {
+      "epoch": 1.9901467046201007,
+      "grad_norm": 0.08805907951583584,
+      "learning_rate": 1.3308119300092793e-09,
+      "loss": 0.3473,
+      "step": 3408
+    },
+    {
+      "epoch": 1.9907306036055763,
+      "grad_norm": 0.09684439445666133,
+      "learning_rate": 1.1696620645140765e-09,
+      "loss": 0.3569,
+      "step": 3409
+    },
+    {
+      "epoch": 1.9913145025910517,
+      "grad_norm": 0.09277529313333982,
+      "learning_rate": 1.0189081808154033e-09,
+      "loss": 0.3599,
+      "step": 3410
+    },
+    {
+      "epoch": 1.9918984015765273,
+      "grad_norm": 0.09060612540551786,
+      "learning_rate": 8.785504356556562e-10,
+      "loss": 0.3009,
+      "step": 3411
+    },
+    {
+      "epoch": 1.9924823005620027,
+      "grad_norm": 0.08585812303107306,
+      "learning_rate": 7.485889749658803e-10,
+      "loss": 0.2628,
+      "step": 3412
+    },
+    {
+      "epoch": 1.9930661995474783,
+      "grad_norm": 0.09198639265234065,
+      "learning_rate": 6.290239338724302e-10,
+      "loss": 0.3087,
+      "step": 3413
+    },
+    {
+      "epoch": 1.993650098532954,
+      "grad_norm": 0.08193437983571096,
+      "learning_rate": 5.198554366858676e-10,
+      "loss": 0.2718,
+      "step": 3414
+    },
+    {
+      "epoch": 1.9942339975184293,
+      "grad_norm": 0.0904653246578968,
+      "learning_rate": 4.210835969142846e-10,
+      "loss": 0.3157,
+      "step": 3415
+    },
+    {
+      "epoch": 1.9948178965039047,
+      "grad_norm": 0.09141711530442853,
+      "learning_rate": 3.3270851724998e-10,
+      "loss": 0.3283,
+      "step": 3416
+    },
+    {
+      "epoch": 1.9954017954893803,
+      "grad_norm": 0.08939369201154362,
+      "learning_rate": 2.5473028957945234e-10,
+      "loss": 0.3319,
+      "step": 3417
+    },
+    {
+      "epoch": 1.995985694474856,
+      "grad_norm": 0.08516520399958326,
+      "learning_rate": 1.8714899497895845e-10,
+      "loss": 0.3246,
+      "step": 3418
+    },
+    {
+      "epoch": 1.9965695934603314,
+      "grad_norm": 0.08153202536443965,
+      "learning_rate": 1.2996470371229307e-10,
+      "loss": 0.2814,
+      "step": 3419
+    },
+    {
+      "epoch": 1.9971534924458068,
+      "grad_norm": 0.08929057643181486,
+      "learning_rate": 8.317747523745035e-11,
+      "loss": 0.303,
+      "step": 3420
+    },
+    {
+      "epoch": 1.9977373914312824,
+      "grad_norm": 0.08436051736800419,
+      "learning_rate": 4.678735819774183e-11,
+      "loss": 0.3066,
+      "step": 3421
+    },
+    {
+      "epoch": 1.998321290416758,
+      "grad_norm": 0.08819324455261725,
+      "learning_rate": 2.0794390429568212e-11,
+      "loss": 0.3217,
+      "step": 3422
+    },
+    {
+      "epoch": 1.9989051894022334,
+      "grad_norm": 0.08692607513678971,
+      "learning_rate": 5.198598959088586e-12,
+      "loss": 0.3155,
+      "step": 3423
+    },
+    {
+      "epoch": 1.9994890883877088,
+      "grad_norm": 0.08520074907130494,
+      "learning_rate": 0.0,
+      "loss": 0.3148,
+      "step": 3424
+    },
+    {
+      "epoch": 1.9994890883877088,
+      "step": 3424,
+      "total_flos": 314202706821120.0,
+      "train_loss": 0.38291217326665317,
+      "train_runtime": 11922.7119,
+      "train_samples_per_second": 18.385,
+      "train_steps_per_second": 0.287
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3424,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 314202706821120.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}