diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9282 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.995459704880817,
+  "eval_steps": 500,
+  "global_step": 1320,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003026863412788498,
+      "grad_norm": 2.6606235888736935,
+      "learning_rate": 1.5151515151515152e-07,
+      "loss": 0.7542,
+      "step": 1
+    },
+    {
+      "epoch": 0.006053726825576996,
+      "grad_norm": 2.5839037503194775,
+      "learning_rate": 3.0303030303030305e-07,
+      "loss": 0.7427,
+      "step": 2
+    },
+    {
+      "epoch": 0.009080590238365494,
+      "grad_norm": 2.538792065209064,
+      "learning_rate": 4.5454545454545457e-07,
+      "loss": 0.7122,
+      "step": 3
+    },
+    {
+      "epoch": 0.012107453651153992,
+      "grad_norm": 2.7115171560089286,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 0.7928,
+      "step": 4
+    },
+    {
+      "epoch": 0.01513431706394249,
+      "grad_norm": 2.542982907435522,
+      "learning_rate": 7.575757575757576e-07,
+      "loss": 0.6912,
+      "step": 5
+    },
+    {
+      "epoch": 0.018161180476730987,
+      "grad_norm": 2.535131270912158,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.7308,
+      "step": 6
+    },
+    {
+      "epoch": 0.021188043889519486,
+      "grad_norm": 2.4786634639454177,
+      "learning_rate": 1.0606060606060608e-06,
+      "loss": 0.734,
+      "step": 7
+    },
+    {
+      "epoch": 0.024214907302307985,
+      "grad_norm": 2.2834574429347896,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 0.7021,
+      "step": 8
+    },
+    {
+      "epoch": 0.02724177071509648,
+      "grad_norm": 2.023610912927571,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 0.7152,
+      "step": 9
+    },
+    {
+      "epoch": 0.03026863412788498,
+      "grad_norm": 1.9388870257848552,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 0.7244,
+      "step": 10
+    },
+    {
+      "epoch": 0.03329549754067348,
+      "grad_norm": 1.8231701908739162,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.723,
+      "step": 11
+    },
+    {
+      "epoch": 0.036322360953461974,
+      "grad_norm": 1.2538936876411566,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.6874,
+      "step": 12
+    },
+    {
+      "epoch": 0.03934922436625047,
+      "grad_norm": 1.1878260184560339,
+      "learning_rate": 1.96969696969697e-06,
+      "loss": 0.6935,
+      "step": 13
+    },
+    {
+      "epoch": 0.04237608777903897,
+      "grad_norm": 1.075490067168476,
+      "learning_rate": 2.1212121212121216e-06,
+      "loss": 0.6601,
+      "step": 14
+    },
+    {
+      "epoch": 0.04540295119182747,
+      "grad_norm": 1.2245634128452174,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 0.6715,
+      "step": 15
+    },
+    {
+      "epoch": 0.04842981460461597,
+      "grad_norm": 1.389616062583451,
+      "learning_rate": 2.4242424242424244e-06,
+      "loss": 0.6176,
+      "step": 16
+    },
+    {
+      "epoch": 0.051456678017404466,
+      "grad_norm": 2.383243540383084,
+      "learning_rate": 2.575757575757576e-06,
+      "loss": 0.6237,
+      "step": 17
+    },
+    {
+      "epoch": 0.05448354143019296,
+      "grad_norm": 2.5463826407488863,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.6453,
+      "step": 18
+    },
+    {
+      "epoch": 0.057510404842981463,
+      "grad_norm": 2.174941527981311,
+      "learning_rate": 2.8787878787878793e-06,
+      "loss": 0.6325,
+      "step": 19
+    },
+    {
+      "epoch": 0.06053726825576996,
+      "grad_norm": 1.7496720593232848,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.6591,
+      "step": 20
+    },
+    {
+      "epoch": 0.06356413166855845,
+      "grad_norm": 1.2731775454551602,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 0.6136,
+      "step": 21
+    },
+    {
+      "epoch": 0.06659099508134696,
+      "grad_norm": 1.0250356902342999,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.5672,
+      "step": 22
+    },
+    {
+      "epoch": 0.06961785849413545,
+      "grad_norm": 1.0446542443194742,
+      "learning_rate": 3.4848484848484854e-06,
+      "loss": 0.5808,
+      "step": 23
+    },
+    {
+      "epoch": 0.07264472190692395,
+      "grad_norm": 1.1929304910101342,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.6272,
+      "step": 24
+    },
+    {
+      "epoch": 0.07567158531971245,
+      "grad_norm": 1.046668421491083,
+      "learning_rate": 3.7878787878787882e-06,
+      "loss": 0.5632,
+      "step": 25
+    },
+    {
+      "epoch": 0.07869844873250094,
+      "grad_norm": 1.0312675660867232,
+      "learning_rate": 3.93939393939394e-06,
+      "loss": 0.6227,
+      "step": 26
+    },
+    {
+      "epoch": 0.08172531214528944,
+      "grad_norm": 0.8221949255460383,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 0.5444,
+      "step": 27
+    },
+    {
+      "epoch": 0.08475217555807794,
+      "grad_norm": 0.8090471808121372,
+      "learning_rate": 4.242424242424243e-06,
+      "loss": 0.5242,
+      "step": 28
+    },
+    {
+      "epoch": 0.08777903897086645,
+      "grad_norm": 0.8266306035485538,
+      "learning_rate": 4.393939393939394e-06,
+      "loss": 0.5787,
+      "step": 29
+    },
+    {
+      "epoch": 0.09080590238365494,
+      "grad_norm": 0.7870753097865048,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.5659,
+      "step": 30
+    },
+    {
+      "epoch": 0.09383276579644344,
+      "grad_norm": 0.7171740857000073,
+      "learning_rate": 4.696969696969698e-06,
+      "loss": 0.5393,
+      "step": 31
+    },
+    {
+      "epoch": 0.09685962920923194,
+      "grad_norm": 0.709028343177713,
+      "learning_rate": 4.848484848484849e-06,
+      "loss": 0.5542,
+      "step": 32
+    },
+    {
+      "epoch": 0.09988649262202043,
+      "grad_norm": 0.7113678116591485,
+      "learning_rate": 5e-06,
+      "loss": 0.5495,
+      "step": 33
+    },
+    {
+      "epoch": 0.10291335603480893,
+      "grad_norm": 0.6976424936614798,
+      "learning_rate": 5.151515151515152e-06,
+      "loss": 0.5284,
+      "step": 34
+    },
+    {
+      "epoch": 0.10594021944759743,
+      "grad_norm": 0.7167523594774555,
+      "learning_rate": 5.303030303030303e-06,
+      "loss": 0.5689,
+      "step": 35
+    },
+    {
+      "epoch": 0.10896708286038592,
+      "grad_norm": 0.7010496641607645,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.5497,
+      "step": 36
+    },
+    {
+      "epoch": 0.11199394627317442,
+      "grad_norm": 0.689054511178451,
+      "learning_rate": 5.606060606060606e-06,
+      "loss": 0.553,
+      "step": 37
+    },
+    {
+      "epoch": 0.11502080968596293,
+      "grad_norm": 0.6652002574007199,
+      "learning_rate": 5.7575757575757586e-06,
+      "loss": 0.5246,
+      "step": 38
+    },
+    {
+      "epoch": 0.11804767309875142,
+      "grad_norm": 0.6649094767152581,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 0.5169,
+      "step": 39
+    },
+    {
+      "epoch": 0.12107453651153992,
+      "grad_norm": 0.6499112412939582,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.5003,
+      "step": 40
+    },
+    {
+      "epoch": 0.12410139992432842,
+      "grad_norm": 0.6685165805875313,
+      "learning_rate": 6.212121212121213e-06,
+      "loss": 0.5365,
+      "step": 41
+    },
+    {
+      "epoch": 0.1271282633371169,
+      "grad_norm": 0.6447179244866998,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.5272,
+      "step": 42
+    },
+    {
+      "epoch": 0.1301551267499054,
+      "grad_norm": 0.6311213402589272,
+      "learning_rate": 6.515151515151516e-06,
+      "loss": 0.5148,
+      "step": 43
+    },
+    {
+      "epoch": 0.13318199016269391,
+      "grad_norm": 0.6572907252913919,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.5024,
+      "step": 44
+    },
+    {
+      "epoch": 0.1362088535754824,
+      "grad_norm": 0.6944690534241844,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 0.5154,
+      "step": 45
+    },
+    {
+      "epoch": 0.1392357169882709,
+      "grad_norm": 2.0821292514118666,
+      "learning_rate": 6.969696969696971e-06,
+      "loss": 0.5398,
+      "step": 46
+    },
+    {
+      "epoch": 0.1422625804010594,
+      "grad_norm": 0.654173173145158,
+      "learning_rate": 7.121212121212122e-06,
+      "loss": 0.5293,
+      "step": 47
+    },
+    {
+      "epoch": 0.1452894438138479,
+      "grad_norm": 0.6443278098370018,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.4891,
+      "step": 48
+    },
+    {
+      "epoch": 0.14831630722663638,
+      "grad_norm": 0.6163289947589778,
+      "learning_rate": 7.424242424242425e-06,
+      "loss": 0.5456,
+      "step": 49
+    },
+    {
+      "epoch": 0.1513431706394249,
+      "grad_norm": 0.6293184142816125,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.507,
+      "step": 50
+    },
+    {
+      "epoch": 0.1543700340522134,
+      "grad_norm": 0.6108980595053909,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 0.4668,
+      "step": 51
+    },
+    {
+      "epoch": 0.15739689746500188,
+      "grad_norm": 0.650693680931454,
+      "learning_rate": 7.87878787878788e-06,
+      "loss": 0.5277,
+      "step": 52
+    },
+    {
+      "epoch": 0.1604237608777904,
+      "grad_norm": 0.6533375201792784,
+      "learning_rate": 8.03030303030303e-06,
+      "loss": 0.4782,
+      "step": 53
+    },
+    {
+      "epoch": 0.16345062429057888,
+      "grad_norm": 0.6013603945752811,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.4934,
+      "step": 54
+    },
+    {
+      "epoch": 0.16647748770336737,
+      "grad_norm": 0.6452710057086444,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.5203,
+      "step": 55
+    },
+    {
+      "epoch": 0.1695043511161559,
+      "grad_norm": 0.6731130320903949,
+      "learning_rate": 8.484848484848486e-06,
+      "loss": 0.4867,
+      "step": 56
+    },
+    {
+      "epoch": 0.17253121452894438,
+      "grad_norm": 0.6337296958994292,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 0.515,
+      "step": 57
+    },
+    {
+      "epoch": 0.1755580779417329,
+      "grad_norm": 0.6151239786380825,
+      "learning_rate": 8.787878787878788e-06,
+      "loss": 0.5148,
+      "step": 58
+    },
+    {
+      "epoch": 0.17858494135452138,
+      "grad_norm": 0.6166411355268289,
+      "learning_rate": 8.93939393939394e-06,
+      "loss": 0.5045,
+      "step": 59
+    },
+    {
+      "epoch": 0.18161180476730987,
+      "grad_norm": 0.6549047654923988,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.5363,
+      "step": 60
+    },
+    {
+      "epoch": 0.1846386681800984,
+      "grad_norm": 0.620456217872379,
+      "learning_rate": 9.242424242424244e-06,
+      "loss": 0.4807,
+      "step": 61
+    },
+    {
+      "epoch": 0.18766553159288688,
+      "grad_norm": 0.6107133997833717,
+      "learning_rate": 9.393939393939396e-06,
+      "loss": 0.5244,
+      "step": 62
+    },
+    {
+      "epoch": 0.19069239500567536,
+      "grad_norm": 0.6001484795084376,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 0.4877,
+      "step": 63
+    },
+    {
+      "epoch": 0.19371925841846388,
+      "grad_norm": 0.6334561016682858,
+      "learning_rate": 9.696969696969698e-06,
+      "loss": 0.5201,
+      "step": 64
+    },
+    {
+      "epoch": 0.19674612183125237,
+      "grad_norm": 0.6313491176735707,
+      "learning_rate": 9.84848484848485e-06,
+      "loss": 0.5353,
+      "step": 65
+    },
+    {
+      "epoch": 0.19977298524404086,
+      "grad_norm": 0.6281421539981944,
+      "learning_rate": 1e-05,
+      "loss": 0.5022,
+      "step": 66
+    },
+    {
+      "epoch": 0.20279984865682937,
+      "grad_norm": 0.6488947061768228,
+      "learning_rate": 1.0151515151515152e-05,
+      "loss": 0.5056,
+      "step": 67
+    },
+    {
+      "epoch": 0.20582671206961786,
+      "grad_norm": 0.6265731479985902,
+      "learning_rate": 1.0303030303030304e-05,
+      "loss": 0.4715,
+      "step": 68
+    },
+    {
+      "epoch": 0.20885357548240635,
+      "grad_norm": 0.6732834661626839,
+      "learning_rate": 1.0454545454545455e-05,
+      "loss": 0.4765,
+      "step": 69
+    },
+    {
+      "epoch": 0.21188043889519487,
+      "grad_norm": 0.6368639935663205,
+      "learning_rate": 1.0606060606060606e-05,
+      "loss": 0.5095,
+      "step": 70
+    },
+    {
+      "epoch": 0.21490730230798336,
+      "grad_norm": 0.6491897501445154,
+      "learning_rate": 1.0757575757575758e-05,
+      "loss": 0.4865,
+      "step": 71
+    },
+    {
+      "epoch": 0.21793416572077184,
+      "grad_norm": 0.6337243907517766,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 0.5088,
+      "step": 72
+    },
+    {
+      "epoch": 0.22096102913356036,
+      "grad_norm": 0.6430509617718421,
+      "learning_rate": 1.1060606060606061e-05,
+      "loss": 0.4968,
+      "step": 73
+    },
+    {
+      "epoch": 0.22398789254634885,
+      "grad_norm": 0.6750099206855208,
+      "learning_rate": 1.1212121212121212e-05,
+      "loss": 0.5181,
+      "step": 74
+    },
+    {
+      "epoch": 0.22701475595913734,
+      "grad_norm": 0.6622703450799444,
+      "learning_rate": 1.1363636363636366e-05,
+      "loss": 0.4902,
+      "step": 75
+    },
+    {
+      "epoch": 0.23004161937192585,
+      "grad_norm": 0.7066772099002263,
+      "learning_rate": 1.1515151515151517e-05,
+      "loss": 0.4934,
+      "step": 76
+    },
+    {
+      "epoch": 0.23306848278471434,
+      "grad_norm": 0.6020193501055748,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 0.4507,
+      "step": 77
+    },
+    {
+      "epoch": 0.23609534619750283,
+      "grad_norm": 0.7057620538054212,
+      "learning_rate": 1.181818181818182e-05,
+      "loss": 0.4698,
+      "step": 78
+    },
+    {
+      "epoch": 0.23912220961029135,
+      "grad_norm": 0.6284349041314211,
+      "learning_rate": 1.1969696969696971e-05,
+      "loss": 0.5077,
+      "step": 79
+    },
+    {
+      "epoch": 0.24214907302307984,
+      "grad_norm": 0.6395883872578291,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.4741,
+      "step": 80
+    },
+    {
+      "epoch": 0.24517593643586832,
+      "grad_norm": 0.662340257016944,
+      "learning_rate": 1.2272727272727274e-05,
+      "loss": 0.5136,
+      "step": 81
+    },
+    {
+      "epoch": 0.24820279984865684,
+      "grad_norm": 0.6652949592137235,
+      "learning_rate": 1.2424242424242425e-05,
+      "loss": 0.4794,
+      "step": 82
+    },
+    {
+      "epoch": 0.2512296632614453,
+      "grad_norm": 0.6289204818101842,
+      "learning_rate": 1.2575757575757576e-05,
+      "loss": 0.4736,
+      "step": 83
+    },
+    {
+      "epoch": 0.2542565266742338,
+      "grad_norm": 0.6975756512181838,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 0.4911,
+      "step": 84
+    },
+    {
+      "epoch": 0.25728339008702233,
+      "grad_norm": 0.6507951862093668,
+      "learning_rate": 1.287878787878788e-05,
+      "loss": 0.4752,
+      "step": 85
+    },
+    {
+      "epoch": 0.2603102534998108,
+      "grad_norm": 0.7058524333854576,
+      "learning_rate": 1.3030303030303032e-05,
+      "loss": 0.502,
+      "step": 86
+    },
+    {
+      "epoch": 0.2633371169125993,
+      "grad_norm": 0.6825600262163648,
+      "learning_rate": 1.3181818181818183e-05,
+      "loss": 0.4685,
+      "step": 87
+    },
+    {
+      "epoch": 0.26636398032538783,
+      "grad_norm": 0.6282924725201364,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.4696,
+      "step": 88
+    },
+    {
+      "epoch": 0.2693908437381763,
+      "grad_norm": 0.7217912272702046,
+      "learning_rate": 1.3484848484848486e-05,
+      "loss": 0.4805,
+      "step": 89
+    },
+    {
+      "epoch": 0.2724177071509648,
+      "grad_norm": 0.7136002686639734,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.4708,
+      "step": 90
+    },
+    {
+      "epoch": 0.2754445705637533,
+      "grad_norm": 0.6995423717696146,
+      "learning_rate": 1.378787878787879e-05,
+      "loss": 0.4774,
+      "step": 91
+    },
+    {
+      "epoch": 0.2784714339765418,
+      "grad_norm": 0.7473449328898674,
+      "learning_rate": 1.3939393939393942e-05,
+      "loss": 0.4626,
+      "step": 92
+    },
+    {
+      "epoch": 0.2814982973893303,
+      "grad_norm": 0.6483504620152055,
+      "learning_rate": 1.4090909090909092e-05,
+      "loss": 0.4755,
+      "step": 93
+    },
+    {
+      "epoch": 0.2845251608021188,
+      "grad_norm": 0.6975235832315637,
+      "learning_rate": 1.4242424242424245e-05,
+      "loss": 0.4547,
+      "step": 94
+    },
+    {
+      "epoch": 0.2875520242149073,
+      "grad_norm": 0.6966049725790425,
+      "learning_rate": 1.4393939393939396e-05,
+      "loss": 0.512,
+      "step": 95
+    },
+    {
+      "epoch": 0.2905788876276958,
+      "grad_norm": 0.6708628835116337,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 0.46,
+      "step": 96
+    },
+    {
+      "epoch": 0.2936057510404843,
+      "grad_norm": 0.7019632452427482,
+      "learning_rate": 1.4696969696969699e-05,
+      "loss": 0.515,
+      "step": 97
+    },
+    {
+      "epoch": 0.29663261445327277,
+      "grad_norm": 0.6579408524705841,
+      "learning_rate": 1.484848484848485e-05,
+      "loss": 0.5235,
+      "step": 98
+    },
+    {
+      "epoch": 0.2996594778660613,
+      "grad_norm": 0.7387270323911181,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.4887,
+      "step": 99
+    },
+    {
+      "epoch": 0.3026863412788498,
+      "grad_norm": 0.7408534215864312,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.4901,
+      "step": 100
+    },
+    {
+      "epoch": 0.30571320469163826,
+      "grad_norm": 0.7333427591372811,
+      "learning_rate": 1.5303030303030304e-05,
+      "loss": 0.5469,
+      "step": 101
+    },
+    {
+      "epoch": 0.3087400681044268,
+      "grad_norm": 0.8238078052850119,
+      "learning_rate": 1.5454545454545454e-05,
+      "loss": 0.4785,
+      "step": 102
+    },
+    {
+      "epoch": 0.3117669315172153,
+      "grad_norm": 0.6366634489094568,
+      "learning_rate": 1.5606060606060605e-05,
+      "loss": 0.4643,
+      "step": 103
+    },
+    {
+      "epoch": 0.31479379493000376,
+      "grad_norm": 0.812031906728069,
+      "learning_rate": 1.575757575757576e-05,
+      "loss": 0.4514,
+      "step": 104
+    },
+    {
+      "epoch": 0.3178206583427923,
+      "grad_norm": 0.6140216101397699,
+      "learning_rate": 1.590909090909091e-05,
+      "loss": 0.4701,
+      "step": 105
+    },
+    {
+      "epoch": 0.3208475217555808,
+      "grad_norm": 0.7441546346678057,
+      "learning_rate": 1.606060606060606e-05,
+      "loss": 0.4865,
+      "step": 106
+    },
+    {
+      "epoch": 0.32387438516836925,
+      "grad_norm": 0.6532723220933617,
+      "learning_rate": 1.6212121212121212e-05,
+      "loss": 0.501,
+      "step": 107
+    },
+    {
+      "epoch": 0.32690124858115777,
+      "grad_norm": 0.6874345837715945,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 0.4832,
+      "step": 108
+    },
+    {
+      "epoch": 0.3299281119939463,
+      "grad_norm": 0.6713124602281403,
+      "learning_rate": 1.6515151515151517e-05,
+      "loss": 0.4965,
+      "step": 109
+    },
+    {
+      "epoch": 0.33295497540673474,
+      "grad_norm": 0.7288733236572627,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.4853,
+      "step": 110
+    },
+    {
+      "epoch": 0.33598183881952326,
+      "grad_norm": 0.620498067103792,
+      "learning_rate": 1.681818181818182e-05,
+      "loss": 0.4718,
+      "step": 111
+    },
+    {
+      "epoch": 0.3390087022323118,
+      "grad_norm": 0.7373087910954631,
+      "learning_rate": 1.6969696969696972e-05,
+      "loss": 0.4565,
+      "step": 112
+    },
+    {
+      "epoch": 0.3420355656451003,
+      "grad_norm": 0.76102953307224,
+      "learning_rate": 1.7121212121212123e-05,
+      "loss": 0.4504,
+      "step": 113
+    },
+    {
+      "epoch": 0.34506242905788875,
+      "grad_norm": 0.644249717295401,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 0.5006,
+      "step": 114
+    },
+    {
+      "epoch": 0.34808929247067727,
+      "grad_norm": 0.6869402969563104,
+      "learning_rate": 1.7424242424242425e-05,
+      "loss": 0.4721,
+      "step": 115
+    },
+    {
+      "epoch": 0.3511161558834658,
+      "grad_norm": 0.6806378537764589,
+      "learning_rate": 1.7575757575757576e-05,
+      "loss": 0.4514,
+      "step": 116
+    },
+    {
+      "epoch": 0.35414301929625425,
+      "grad_norm": 0.7784088663994104,
+      "learning_rate": 1.772727272727273e-05,
+      "loss": 0.4453,
+      "step": 117
+    },
+    {
+      "epoch": 0.35716988270904276,
+      "grad_norm": 0.6304667330410376,
+      "learning_rate": 1.787878787878788e-05,
+      "loss": 0.4775,
+      "step": 118
+    },
+    {
+      "epoch": 0.3601967461218313,
+      "grad_norm": 0.7370651006381368,
+      "learning_rate": 1.803030303030303e-05,
+      "loss": 0.4667,
+      "step": 119
+    },
+    {
+      "epoch": 0.36322360953461974,
+      "grad_norm": 0.6485556701840409,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.4712,
+      "step": 120
+    },
+    {
+      "epoch": 0.36625047294740826,
+      "grad_norm": 0.6385859741542873,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.481,
+      "step": 121
+    },
+    {
+      "epoch": 0.3692773363601968,
+      "grad_norm": 0.7097042166108211,
+      "learning_rate": 1.8484848484848487e-05,
+      "loss": 0.4716,
+      "step": 122
+    },
+    {
+      "epoch": 0.37230419977298523,
+      "grad_norm": 0.6629971905848386,
+      "learning_rate": 1.8636363636363638e-05,
+      "loss": 0.4785,
+      "step": 123
+    },
+    {
+      "epoch": 0.37533106318577375,
+      "grad_norm": 0.6555134637210983,
+      "learning_rate": 1.8787878787878792e-05,
+      "loss": 0.5161,
+      "step": 124
+    },
+    {
+      "epoch": 0.37835792659856227,
+      "grad_norm": 0.6476284373047034,
+      "learning_rate": 1.8939393939393943e-05,
+      "loss": 0.4658,
+      "step": 125
+    },
+    {
+      "epoch": 0.3813847900113507,
+      "grad_norm": 0.6679665980147139,
+      "learning_rate": 1.9090909090909094e-05,
+      "loss": 0.4571,
+      "step": 126
+    },
+    {
+      "epoch": 0.38441165342413924,
+      "grad_norm": 0.6322618485695937,
+      "learning_rate": 1.9242424242424244e-05,
+      "loss": 0.4657,
+      "step": 127
+    },
+    {
+      "epoch": 0.38743851683692776,
+      "grad_norm": 0.664795025319193,
+      "learning_rate": 1.9393939393939395e-05,
+      "loss": 0.4613,
+      "step": 128
+    },
+    {
+      "epoch": 0.3904653802497162,
+      "grad_norm": 0.6387866647495783,
+      "learning_rate": 1.9545454545454546e-05,
+      "loss": 0.4875,
+      "step": 129
+    },
+    {
+      "epoch": 0.39349224366250474,
+      "grad_norm": 0.6594468734623755,
+      "learning_rate": 1.96969696969697e-05,
+      "loss": 0.465,
+      "step": 130
+    },
+    {
+      "epoch": 0.39651910707529325,
+      "grad_norm": 0.6540871209001573,
+      "learning_rate": 1.984848484848485e-05,
+      "loss": 0.4881,
+      "step": 131
+    },
+    {
+      "epoch": 0.3995459704880817,
+      "grad_norm": 0.6158180696861089,
+      "learning_rate": 2e-05,
+      "loss": 0.4868,
+      "step": 132
+    },
+    {
+      "epoch": 0.40257283390087023,
+      "grad_norm": 0.6231530945175385,
+      "learning_rate": 1.9999965034751806e-05,
+      "loss": 0.4474,
+      "step": 133
+    },
+    {
+      "epoch": 0.40559969731365875,
+      "grad_norm": 0.6115961601328362,
+      "learning_rate": 1.9999860139251737e-05,
+      "loss": 0.4686,
+      "step": 134
+    },
+    {
+      "epoch": 0.4086265607264472,
+      "grad_norm": 0.5865383585327205,
+      "learning_rate": 1.9999685314233333e-05,
+      "loss": 0.4497,
+      "step": 135
+    },
+    {
+      "epoch": 0.4116534241392357,
+      "grad_norm": 0.6484637197502119,
+      "learning_rate": 1.9999440560919153e-05,
+      "loss": 0.4755,
+      "step": 136
+    },
+    {
+      "epoch": 0.41468028755202424,
+      "grad_norm": 0.6454398224400972,
+      "learning_rate": 1.9999125881020766e-05,
+      "loss": 0.4605,
+      "step": 137
+    },
+    {
+      "epoch": 0.4177071509648127,
+      "grad_norm": 0.6160056103447329,
+      "learning_rate": 1.9998741276738753e-05,
+      "loss": 0.4937,
+      "step": 138
+    },
+    {
+      "epoch": 0.4207340143776012,
+      "grad_norm": 0.6345924962959403,
+      "learning_rate": 1.9998286750762662e-05,
+      "loss": 0.511,
+      "step": 139
+    },
+    {
+      "epoch": 0.42376087779038973,
+      "grad_norm": 0.6764586642825581,
+      "learning_rate": 1.999776230627102e-05,
+      "loss": 0.4674,
+      "step": 140
+    },
+    {
+      "epoch": 0.4267877412031782,
+      "grad_norm": 0.6030835337989207,
+      "learning_rate": 1.9997167946931293e-05,
+      "loss": 0.4347,
+      "step": 141
+    },
+    {
+      "epoch": 0.4298146046159667,
+      "grad_norm": 0.6491168779385929,
+      "learning_rate": 1.9996503676899863e-05,
+      "loss": 0.4467,
+      "step": 142
+    },
+    {
+      "epoch": 0.4328414680287552,
+      "grad_norm": 0.5963441023505116,
+      "learning_rate": 1.9995769500822007e-05,
+      "loss": 0.4493,
+      "step": 143
+    },
+    {
+      "epoch": 0.4358683314415437,
+      "grad_norm": 0.6469840567377747,
+      "learning_rate": 1.9994965423831853e-05,
+      "loss": 0.4884,
+      "step": 144
+    },
+    {
+      "epoch": 0.4388951948543322,
+      "grad_norm": 0.5998640796027529,
+      "learning_rate": 1.999409145155235e-05,
+      "loss": 0.4509,
+      "step": 145
+    },
+    {
+      "epoch": 0.4419220582671207,
+      "grad_norm": 0.6113480788929524,
+      "learning_rate": 1.9993147590095232e-05,
+      "loss": 0.4524,
+      "step": 146
+    },
+    {
+      "epoch": 0.4449489216799092,
+      "grad_norm": 0.6940496896831555,
+      "learning_rate": 1.999213384606097e-05,
+      "loss": 0.4893,
+      "step": 147
+    },
+    {
+      "epoch": 0.4479757850926977,
+      "grad_norm": 0.6788833825248451,
+      "learning_rate": 1.999105022653872e-05,
+      "loss": 0.4502,
+      "step": 148
+    },
+    {
+      "epoch": 0.4510026485054862,
+      "grad_norm": 0.6383713935922738,
+      "learning_rate": 1.998989673910629e-05,
+      "loss": 0.4673,
+      "step": 149
+    },
+    {
+      "epoch": 0.4540295119182747,
+      "grad_norm": 0.698929579953909,
+      "learning_rate": 1.9988673391830082e-05,
+      "loss": 0.5125,
+      "step": 150
+    },
+    {
+      "epoch": 0.4570563753310632,
+      "grad_norm": 0.5962795793605712,
+      "learning_rate": 1.9987380193265014e-05,
+      "loss": 0.457,
+      "step": 151
+    },
+    {
+      "epoch": 0.4600832387438517,
+      "grad_norm": 0.7144462815328115,
+      "learning_rate": 1.9986017152454497e-05,
+      "loss": 0.4675,
+      "step": 152
+    },
+    {
+      "epoch": 0.46311010215664017,
+      "grad_norm": 0.6206531294656699,
+      "learning_rate": 1.9984584278930333e-05,
+      "loss": 0.4523,
+      "step": 153
+    },
+    {
+      "epoch": 0.4661369655694287,
+      "grad_norm": 0.657930813368399,
+      "learning_rate": 1.9983081582712684e-05,
+      "loss": 0.4161,
+      "step": 154
+    },
+    {
+      "epoch": 0.4691638289822172,
+      "grad_norm": 0.6292499158579902,
+      "learning_rate": 1.998150907430998e-05,
+      "loss": 0.4465,
+      "step": 155
+    },
+    {
+      "epoch": 0.47219069239500566,
+      "grad_norm": 0.706157954430773,
+      "learning_rate": 1.9979866764718846e-05,
+      "loss": 0.4766,
+      "step": 156
+    },
+    {
+      "epoch": 0.4752175558077942,
+      "grad_norm": 0.6904245881262766,
+      "learning_rate": 1.9978154665424037e-05,
+      "loss": 0.472,
+      "step": 157
+    },
+    {
+      "epoch": 0.4782444192205827,
+      "grad_norm": 0.6511174460017086,
+      "learning_rate": 1.997637278839835e-05,
+      "loss": 0.4847,
+      "step": 158
+    },
+    {
+      "epoch": 0.48127128263337116,
+      "grad_norm": 0.6441313711204578,
+      "learning_rate": 1.9974521146102535e-05,
+      "loss": 0.4472,
+      "step": 159
+    },
+    {
+      "epoch": 0.48429814604615967,
+      "grad_norm": 0.7307622354204739,
+      "learning_rate": 1.9972599751485225e-05,
+      "loss": 0.4704,
+      "step": 160
+    },
+    {
+      "epoch": 0.4873250094589482,
+      "grad_norm": 0.6293074204572366,
+      "learning_rate": 1.9970608617982824e-05,
+      "loss": 0.473,
+      "step": 161
+    },
+    {
+      "epoch": 0.49035187287173665,
+      "grad_norm": 0.783057097813894,
+      "learning_rate": 1.9968547759519426e-05,
+      "loss": 0.463,
+      "step": 162
+    },
+    {
+      "epoch": 0.49337873628452517,
+      "grad_norm": 0.6361817312995955,
+      "learning_rate": 1.9966417190506718e-05,
+      "loss": 0.4773,
+      "step": 163
+    },
+    {
+      "epoch": 0.4964055996973137,
+      "grad_norm": 0.8833002388363584,
+      "learning_rate": 1.9964216925843876e-05,
+      "loss": 0.4477,
+      "step": 164
+    },
+    {
+      "epoch": 0.49943246311010214,
+      "grad_norm": 0.7113014547078169,
+      "learning_rate": 1.9961946980917457e-05,
+      "loss": 0.4347,
+      "step": 165
+    },
+    {
+      "epoch": 0.5024593265228906,
+      "grad_norm": 0.7536370851341563,
+      "learning_rate": 1.9959607371601303e-05,
+      "loss": 0.4621,
+      "step": 166
+    },
+    {
+      "epoch": 0.5054861899356792,
+      "grad_norm": 0.7543931147947287,
+      "learning_rate": 1.9957198114256414e-05,
+      "loss": 0.4617,
+      "step": 167
+    },
+    {
+      "epoch": 0.5085130533484676,
+      "grad_norm": 0.6998599711689649,
+      "learning_rate": 1.9954719225730847e-05,
+      "loss": 0.4665,
+      "step": 168
+    },
+    {
+      "epoch": 0.5115399167612561,
+      "grad_norm": 0.7315397821949086,
+      "learning_rate": 1.9952170723359594e-05,
+      "loss": 0.4723,
+      "step": 169
+    },
+    {
+      "epoch": 0.5145667801740447,
+      "grad_norm": 0.6578390785098822,
+      "learning_rate": 1.994955262496446e-05,
+      "loss": 0.4549,
+      "step": 170
+    },
+    {
+      "epoch": 0.5175936435868331,
+      "grad_norm": 0.6742675430284679,
+      "learning_rate": 1.9946864948853936e-05,
+      "loss": 0.4637,
+      "step": 171
+    },
+    {
+      "epoch": 0.5206205069996216,
+      "grad_norm": 0.6791468806842567,
+      "learning_rate": 1.9944107713823068e-05,
+      "loss": 0.4462,
+      "step": 172
+    },
+    {
+      "epoch": 0.5236473704124102,
+      "grad_norm": 0.640202378386533,
+      "learning_rate": 1.9941280939153347e-05,
+      "loss": 0.499,
+      "step": 173
+    },
+    {
+      "epoch": 0.5266742338251986,
+      "grad_norm": 0.674330286006925,
+      "learning_rate": 1.9938384644612542e-05,
+      "loss": 0.4986,
+      "step": 174
+    },
+    {
+      "epoch": 0.5297010972379871,
+      "grad_norm": 0.6296118652982142,
+      "learning_rate": 1.993541885045459e-05,
+      "loss": 0.4472,
+      "step": 175
+    },
+    {
+      "epoch": 0.5327279606507757,
+      "grad_norm": 0.7083916537208047,
+      "learning_rate": 1.9932383577419432e-05,
+      "loss": 0.4533,
+      "step": 176
+    },
+    {
+      "epoch": 0.5357548240635641,
+      "grad_norm": 0.5739267977505365,
+      "learning_rate": 1.9929278846732883e-05,
+      "loss": 0.4387,
+      "step": 177
+    },
+    {
+      "epoch": 0.5387816874763526,
+      "grad_norm": 0.6672862398343221,
+      "learning_rate": 1.9926104680106484e-05,
+      "loss": 0.4739,
+      "step": 178
+    },
+    {
+      "epoch": 0.5418085508891411,
+      "grad_norm": 0.5786685744355408,
+      "learning_rate": 1.9922861099737335e-05,
+      "loss": 0.4431,
+      "step": 179
+    },
+    {
+      "epoch": 0.5448354143019296,
+      "grad_norm": 0.5970062034803544,
+      "learning_rate": 1.9919548128307954e-05,
+      "loss": 0.4425,
+      "step": 180
+    },
+    {
+      "epoch": 0.5478622777147181,
+      "grad_norm": 0.6385209445100716,
+      "learning_rate": 1.991616578898612e-05,
+      "loss": 0.4562,
+      "step": 181
+    },
+    {
+      "epoch": 0.5508891411275066,
+      "grad_norm": 0.6219220609600679,
+      "learning_rate": 1.9912714105424694e-05,
+      "loss": 0.4431,
+      "step": 182
+    },
+    {
+      "epoch": 0.5539160045402951,
+      "grad_norm": 0.656180609302135,
+      "learning_rate": 1.990919310176147e-05,
+      "loss": 0.465,
+      "step": 183
+    },
+    {
+      "epoch": 0.5569428679530836,
+      "grad_norm": 0.6330992050510046,
+      "learning_rate": 1.990560280261901e-05,
+      "loss": 0.4642,
+      "step": 184
+    },
+    {
+      "epoch": 0.5599697313658721,
+      "grad_norm": 0.6589569269871743,
+      "learning_rate": 1.9901943233104444e-05,
+      "loss": 0.4648,
+      "step": 185
+    },
+    {
+      "epoch": 0.5629965947786606,
+      "grad_norm": 0.5992638500172491,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.4232,
+      "step": 186
+    },
+    {
+      "epoch": 0.5660234581914491,
+      "grad_norm": 0.6496383819751609,
+      "learning_rate": 1.9894416385809444e-05,
+      "loss": 0.5094,
+      "step": 187
+    },
+    {
+      "epoch": 0.5690503216042376,
+      "grad_norm": 0.6526253638494249,
+      "learning_rate": 1.9890549160664633e-05,
+      "loss": 0.4469,
+      "step": 188
+    },
+    {
+      "epoch": 0.5720771850170261,
+      "grad_norm": 0.6032717734930514,
+      "learning_rate": 1.988661277041858e-05,
+      "loss": 0.4342,
+      "step": 189
+    },
+    {
+      "epoch": 0.5751040484298146,
+      "grad_norm": 0.6041851799025165,
+      "learning_rate": 1.9882607242598663e-05,
+      "loss": 0.4711,
+      "step": 190
+    },
+    {
+      "epoch": 0.5781309118426031,
+      "grad_norm": 0.7137570067221294,
+      "learning_rate": 1.9878532605215738e-05,
+      "loss": 0.4516,
+      "step": 191
+    },
+    {
+      "epoch": 0.5811577752553916,
+      "grad_norm": 0.5880607828478984,
+      "learning_rate": 1.9874388886763944e-05,
+      "loss": 0.4337,
+      "step": 192
+    },
+    {
+      "epoch": 0.58418463866818,
+      "grad_norm": 0.6142061761250415,
+      "learning_rate": 1.9870176116220512e-05,
+      "loss": 0.4245,
+      "step": 193
+    },
+    {
+      "epoch": 0.5872115020809686,
+      "grad_norm": 0.67729983234337,
+      "learning_rate": 1.9865894323045558e-05,
+      "loss": 0.4366,
+      "step": 194
+    },
+    {
+      "epoch": 0.5902383654937571,
+      "grad_norm": 0.5993654344238707,
+      "learning_rate": 1.986154353718187e-05,
+      "loss": 0.4598,
+      "step": 195
+    },
+    {
+      "epoch": 0.5932652289065455,
+      "grad_norm": 0.5877536590691649,
+      "learning_rate": 1.9857123789054707e-05,
+      "loss": 0.4751,
+      "step": 196
+    },
+    {
+      "epoch": 0.5962920923193341,
+      "grad_norm": 0.5692776422765056,
+      "learning_rate": 1.9852635109571592e-05,
+      "loss": 0.4806,
+      "step": 197
+    },
+    {
+      "epoch": 0.5993189557321226,
+      "grad_norm": 0.6304514710183153,
+      "learning_rate": 1.9848077530122083e-05,
+      "loss": 0.4259,
+      "step": 198
+    },
+    {
+      "epoch": 0.602345819144911,
+      "grad_norm": 0.6591983995304184,
+      "learning_rate": 1.9843451082577557e-05,
+      "loss": 0.5223,
+      "step": 199
+    },
+    {
+      "epoch": 0.6053726825576996,
+      "grad_norm": 0.6328874451897569,
+      "learning_rate": 1.9838755799290993e-05,
+      "loss": 0.4926,
+      "step": 200
+    },
+    {
+      "epoch": 0.6083995459704881,
+      "grad_norm": 0.6487617091235212,
+      "learning_rate": 1.9833991713096742e-05,
+      "loss": 0.4485,
+      "step": 201
+    },
+    {
+      "epoch": 0.6114264093832765,
+      "grad_norm": 0.6595813238708705,
+      "learning_rate": 1.9829158857310288e-05,
+      "loss": 0.4665,
+      "step": 202
+    },
+    {
+      "epoch": 0.6144532727960651,
+      "grad_norm": 0.6707469006156764,
+      "learning_rate": 1.982425726572804e-05,
+      "loss": 0.516,
+      "step": 203
+    },
+    {
+      "epoch": 0.6174801362088536,
+      "grad_norm": 0.6703832373223567,
+      "learning_rate": 1.9819286972627066e-05,
+      "loss": 0.4291,
+      "step": 204
+    },
+    {
+      "epoch": 0.620506999621642,
+      "grad_norm": 0.6394533335802106,
+      "learning_rate": 1.9814248012764877e-05,
+      "loss": 0.4577,
+      "step": 205
+    },
+    {
+      "epoch": 0.6235338630344306,
+      "grad_norm": 0.5904225685988153,
+      "learning_rate": 1.9809140421379168e-05,
+      "loss": 0.4232,
+      "step": 206
+    },
+    {
+      "epoch": 0.626560726447219,
+      "grad_norm": 0.6433281247150431,
+      "learning_rate": 1.980396423418757e-05,
+      "loss": 0.4439,
+      "step": 207
+    },
+    {
+      "epoch": 0.6295875898600075,
+      "grad_norm": 0.5987618096945992,
+      "learning_rate": 1.979871948738743e-05,
+      "loss": 0.4654,
+      "step": 208
+    },
+    {
+      "epoch": 0.6326144532727961,
+      "grad_norm": 0.6150673329731747,
+      "learning_rate": 1.9793406217655516e-05,
+      "loss": 0.5075,
+      "step": 209
+    },
+    {
+      "epoch": 0.6356413166855845,
+      "grad_norm": 0.5982041486937786,
+      "learning_rate": 1.978802446214779e-05,
+      "loss": 0.4835,
+      "step": 210
+    },
+    {
+      "epoch": 0.638668180098373,
+      "grad_norm": 0.6191373665347453,
+      "learning_rate": 1.978257425849913e-05,
+      "loss": 0.4679,
+      "step": 211
+    },
+    {
+      "epoch": 0.6416950435111616,
+      "grad_norm": 0.568028781728923,
+      "learning_rate": 1.9777055644823087e-05,
+      "loss": 0.4792,
+      "step": 212
+    },
+    {
+      "epoch": 0.64472190692395,
+      "grad_norm": 0.6131608954195588,
+      "learning_rate": 1.9771468659711595e-05,
+      "loss": 0.4734,
+      "step": 213
+    },
+    {
+      "epoch": 0.6477487703367385,
+      "grad_norm": 0.588277200088701,
+      "learning_rate": 1.9765813342234726e-05,
+      "loss": 0.4989,
+      "step": 214
+    },
+    {
+      "epoch": 0.6507756337495271,
+      "grad_norm": 0.6124184471301078,
+      "learning_rate": 1.9760089731940384e-05,
+      "loss": 0.4525,
+      "step": 215
+    },
+    {
+      "epoch": 0.6538024971623155,
+      "grad_norm": 0.6131265896141542,
+      "learning_rate": 1.9754297868854075e-05,
+      "loss": 0.465,
+      "step": 216
+    },
+    {
+      "epoch": 0.656829360575104,
+      "grad_norm": 0.7170213677810617,
+      "learning_rate": 1.974843779347857e-05,
+      "loss": 0.4691,
+      "step": 217
+    },
+    {
+      "epoch": 0.6598562239878926,
+      "grad_norm": 0.6054191129948248,
+      "learning_rate": 1.9742509546793673e-05,
+      "loss": 0.4604,
+      "step": 218
+    },
+    {
+      "epoch": 0.662883087400681,
+      "grad_norm": 0.6768850863039694,
+      "learning_rate": 1.973651317025591e-05,
+      "loss": 0.4865,
+      "step": 219
+    },
+    {
+      "epoch": 0.6659099508134695,
+      "grad_norm": 0.5635844565615987,
+      "learning_rate": 1.973044870579824e-05,
+      "loss": 0.4334,
+      "step": 220
+    },
+    {
+      "epoch": 0.6689368142262581,
+      "grad_norm": 0.7310902330388573,
+      "learning_rate": 1.972431619582976e-05,
+      "loss": 0.4614,
+      "step": 221
+    },
+    {
+      "epoch": 0.6719636776390465,
+      "grad_norm": 0.6269600763130015,
+      "learning_rate": 1.9718115683235418e-05,
+      "loss": 0.4447,
+      "step": 222
+    },
+    {
+      "epoch": 0.6749905410518351,
+      "grad_norm": 0.6551879503010097,
+      "learning_rate": 1.971184721137571e-05,
+      "loss": 0.4245,
+      "step": 223
+    },
+    {
+      "epoch": 0.6780174044646236,
+      "grad_norm": 0.6811895057260012,
+      "learning_rate": 1.970551082408636e-05,
+      "loss": 0.4543,
+      "step": 224
+    },
+    {
+      "epoch": 0.681044267877412,
+      "grad_norm": 0.6122928882645706,
+      "learning_rate": 1.969910656567805e-05,
+      "loss": 0.4486,
+      "step": 225
+    },
+    {
+      "epoch": 0.6840711312902006,
+      "grad_norm": 0.655844638282308,
+      "learning_rate": 1.969263448093608e-05,
+      "loss": 0.448,
+      "step": 226
+    },
+    {
+      "epoch": 0.687097994702989,
+      "grad_norm": 0.6269055246746411,
+      "learning_rate": 1.9686094615120047e-05,
+      "loss": 0.4448,
+      "step": 227
+    },
+    {
+      "epoch": 0.6901248581157775,
+      "grad_norm": 0.5947643513462555,
+      "learning_rate": 1.9679487013963566e-05,
+      "loss": 0.4336,
+      "step": 228
+    },
+    {
+      "epoch": 0.6931517215285661,
+      "grad_norm": 0.6667866847919166,
+      "learning_rate": 1.9672811723673914e-05,
+      "loss": 0.5033,
+      "step": 229
+    },
+    {
+      "epoch": 0.6961785849413545,
+      "grad_norm": 0.5878142270152522,
+      "learning_rate": 1.9666068790931733e-05,
+      "loss": 0.4186,
+      "step": 230
+    },
+    {
+      "epoch": 0.699205448354143,
+      "grad_norm": 0.6336711393616944,
+      "learning_rate": 1.9659258262890683e-05,
+      "loss": 0.4808,
+      "step": 231
+    },
+    {
+      "epoch": 0.7022323117669316,
+      "grad_norm": 0.5341579990519509,
+      "learning_rate": 1.9652380187177128e-05,
+      "loss": 0.4461,
+      "step": 232
+    },
+    {
+      "epoch": 0.70525917517972,
+      "grad_norm": 0.5600750792739216,
+      "learning_rate": 1.964543461188979e-05,
+      "loss": 0.4241,
+      "step": 233
+    },
+    {
+      "epoch": 0.7082860385925085,
+      "grad_norm": 0.6366466985248083,
+      "learning_rate": 1.9638421585599422e-05,
+      "loss": 0.4717,
+      "step": 234
+    },
+    {
+      "epoch": 0.7113129020052971,
+      "grad_norm": 0.5954329176440126,
+      "learning_rate": 1.9631341157348464e-05,
+      "loss": 0.4363,
+      "step": 235
+    },
+    {
+      "epoch": 0.7143397654180855,
+      "grad_norm": 0.5997407963882663,
+      "learning_rate": 1.9624193376650708e-05,
+      "loss": 0.4206,
+      "step": 236
+    },
+    {
+      "epoch": 0.717366628830874,
+      "grad_norm": 0.6463956766728491,
+      "learning_rate": 1.961697829349093e-05,
+      "loss": 0.4774,
+      "step": 237
+    },
+    {
+      "epoch": 0.7203934922436626,
+      "grad_norm": 0.6088305138952317,
+      "learning_rate": 1.960969595832457e-05,
+      "loss": 0.4414,
+      "step": 238
+    },
+    {
+      "epoch": 0.723420355656451,
+      "grad_norm": 0.5931332715801896,
+      "learning_rate": 1.9602346422077364e-05,
+      "loss": 0.4695,
+      "step": 239
+    },
+    {
+      "epoch": 0.7264472190692395,
+      "grad_norm": 0.6156895289478719,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.4266,
+      "step": 240
+    },
+    {
+      "epoch": 0.729474082482028,
+      "grad_norm": 0.5498440021911983,
+      "learning_rate": 1.958744595239266e-05,
+      "loss": 0.4386,
+      "step": 241
+    },
+    {
+      "epoch": 0.7325009458948165,
+      "grad_norm": 0.5788331654405979,
+      "learning_rate": 1.957989512315489e-05,
+      "loss": 0.455,
+      "step": 242
+    },
+    {
+      "epoch": 0.735527809307605,
+      "grad_norm": 0.5603044322692867,
+      "learning_rate": 1.957227730123499e-05,
+      "loss": 0.4381,
+      "step": 243
+    },
+    {
+      "epoch": 0.7385546727203935,
+      "grad_norm": 0.6033474859365513,
+      "learning_rate": 1.956459253990476e-05,
+      "loss": 0.4442,
+      "step": 244
+    },
+    {
+      "epoch": 0.741581536133182,
+      "grad_norm": 0.5545123893174032,
+      "learning_rate": 1.9556840892904127e-05,
+      "loss": 0.4255,
+      "step": 245
+    },
+    {
+      "epoch": 0.7446083995459705,
+      "grad_norm": 0.6461011215558895,
+      "learning_rate": 1.9549022414440738e-05,
+      "loss": 0.4606,
+      "step": 246
+    },
+    {
+      "epoch": 0.747635262958759,
+      "grad_norm": 0.6089401612595773,
+      "learning_rate": 1.9541137159189606e-05,
+      "loss": 0.4863,
+      "step": 247
+    },
+    {
+      "epoch": 0.7506621263715475,
+      "grad_norm": 0.6438368810493399,
+      "learning_rate": 1.9533185182292705e-05,
+      "loss": 0.5077,
+      "step": 248
+    },
+    {
+      "epoch": 0.753688989784336,
+      "grad_norm": 0.5531427296677005,
+      "learning_rate": 1.9525166539358608e-05,
+      "loss": 0.4476,
+      "step": 249
+    },
+    {
+      "epoch": 0.7567158531971245,
+      "grad_norm": 0.6233426376238553,
+      "learning_rate": 1.9517081286462082e-05,
+      "loss": 0.4354,
+      "step": 250
+    },
+    {
+      "epoch": 0.759742716609913,
+      "grad_norm": 0.6142404925721421,
+      "learning_rate": 1.9508929480143704e-05,
+      "loss": 0.4314,
+      "step": 251
+    },
+    {
+      "epoch": 0.7627695800227015,
+      "grad_norm": 0.665401320826178,
+      "learning_rate": 1.9500711177409456e-05,
+      "loss": 0.4353,
+      "step": 252
+    },
+    {
+      "epoch": 0.76579644343549,
+      "grad_norm": 0.6120835470962498,
+      "learning_rate": 1.949242643573034e-05,
+      "loss": 0.4756,
+      "step": 253
+    },
+    {
+      "epoch": 0.7688233068482785,
+      "grad_norm": 0.7094920407141364,
+      "learning_rate": 1.9484075313041968e-05,
+      "loss": 0.4684,
+      "step": 254
+    },
+    {
+      "epoch": 0.771850170261067,
+      "grad_norm": 0.6001165448646515,
+      "learning_rate": 1.9475657867744153e-05,
+      "loss": 0.4279,
+      "step": 255
+    },
+    {
+      "epoch": 0.7748770336738555,
+      "grad_norm": 0.6083344128435288,
+      "learning_rate": 1.9467174158700507e-05,
+      "loss": 0.4171,
+      "step": 256
+    },
+    {
+      "epoch": 0.777903897086644,
+      "grad_norm": 0.6942479425067493,
+      "learning_rate": 1.9458624245238027e-05,
+      "loss": 0.4395,
+      "step": 257
+    },
+    {
+      "epoch": 0.7809307604994324,
+      "grad_norm": 0.5622414945747122,
+      "learning_rate": 1.9450008187146685e-05,
+      "loss": 0.461,
+      "step": 258
+    },
+    {
+      "epoch": 0.783957623912221,
+      "grad_norm": 0.6541709481904641,
+      "learning_rate": 1.9441326044679007e-05,
+      "loss": 0.4496,
+      "step": 259
+    },
+    {
+      "epoch": 0.7869844873250095,
+      "grad_norm": 0.629784744214303,
+      "learning_rate": 1.9432577878549635e-05,
+      "loss": 0.4589,
+      "step": 260
+    },
+    {
+      "epoch": 0.7900113507377979,
+      "grad_norm": 0.6543850330522473,
+      "learning_rate": 1.9423763749934942e-05,
+      "loss": 0.4554,
+      "step": 261
+    },
+    {
+      "epoch": 0.7930382141505865,
+      "grad_norm": 0.627984507397876,
+      "learning_rate": 1.9414883720472557e-05,
+      "loss": 0.4668,
+      "step": 262
+    },
+    {
+      "epoch": 0.796065077563375,
+      "grad_norm": 0.5992067102669638,
+      "learning_rate": 1.940593785226097e-05,
+      "loss": 0.4712,
+      "step": 263
+    },
+    {
+      "epoch": 0.7990919409761634,
+      "grad_norm": 0.6230369697839225,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.4375,
+      "step": 264
+    },
+    {
+      "epoch": 0.802118804388952,
+      "grad_norm": 0.6012172803933326,
+      "learning_rate": 1.9387848850285773e-05,
+      "loss": 0.4621,
+      "step": 265
+    },
+    {
+      "epoch": 0.8051456678017405,
+      "grad_norm": 0.59112158493013,
+      "learning_rate": 1.937870584301945e-05,
+      "loss": 0.4644,
+      "step": 266
+    },
+    {
+      "epoch": 0.8081725312145289,
+      "grad_norm": 0.6395024270471195,
+      "learning_rate": 1.936949724999762e-05,
+      "loss": 0.4395,
+      "step": 267
+    },
+    {
+      "epoch": 0.8111993946273175,
+      "grad_norm": 0.5888271780616533,
+      "learning_rate": 1.9360223135616423e-05,
+      "loss": 0.4506,
+      "step": 268
+    },
+    {
+      "epoch": 0.814226258040106,
+      "grad_norm": 0.6305434640303268,
+      "learning_rate": 1.9350883564730213e-05,
+      "loss": 0.4451,
+      "step": 269
+    },
+    {
+      "epoch": 0.8172531214528944,
+      "grad_norm": 0.6498899485625302,
+      "learning_rate": 1.9341478602651068e-05,
+      "loss": 0.4215,
+      "step": 270
+    },
+    {
+      "epoch": 0.820279984865683,
+      "grad_norm": 0.640374031704103,
+      "learning_rate": 1.9332008315148355e-05,
+      "loss": 0.4436,
+      "step": 271
+    },
+    {
+      "epoch": 0.8233068482784714,
+      "grad_norm": 0.6118286196425239,
+      "learning_rate": 1.932247276844826e-05,
+      "loss": 0.4254,
+      "step": 272
+    },
+    {
+      "epoch": 0.8263337116912599,
+      "grad_norm": 0.5860437980388866,
+      "learning_rate": 1.931287202923334e-05,
+      "loss": 0.447,
+      "step": 273
+    },
+    {
+      "epoch": 0.8293605751040485,
+      "grad_norm": 0.6333290481445468,
+      "learning_rate": 1.9303206164642037e-05,
+      "loss": 0.4361,
+      "step": 274
+    },
+    {
+      "epoch": 0.8323874385168369,
+      "grad_norm": 0.6040888543372642,
+      "learning_rate": 1.9293475242268224e-05,
+      "loss": 0.4738,
+      "step": 275
+    },
+    {
+      "epoch": 0.8354143019296254,
+      "grad_norm": 0.5520452343326311,
+      "learning_rate": 1.9283679330160726e-05,
+      "loss": 0.4377,
+      "step": 276
+    },
+    {
+      "epoch": 0.838441165342414,
+      "grad_norm": 0.5667290122721785,
+      "learning_rate": 1.9273818496822843e-05,
+      "loss": 0.4548,
+      "step": 277
+    },
+    {
+      "epoch": 0.8414680287552024,
+      "grad_norm": 0.5667411734714476,
+      "learning_rate": 1.9263892811211865e-05,
+      "loss": 0.4445,
+      "step": 278
+    },
+    {
+      "epoch": 0.8444948921679909,
+      "grad_norm": 0.6141906929623148,
+      "learning_rate": 1.9253902342738612e-05,
+      "loss": 0.4436,
+      "step": 279
+    },
+    {
+      "epoch": 0.8475217555807795,
+      "grad_norm": 0.5884252757702806,
+      "learning_rate": 1.9243847161266924e-05,
+      "loss": 0.4852,
+      "step": 280
+    },
+    {
+      "epoch": 0.8505486189935679,
+      "grad_norm": 0.6099603912522954,
+      "learning_rate": 1.9233727337113176e-05,
+      "loss": 0.4324,
+      "step": 281
+    },
+    {
+      "epoch": 0.8535754824063564,
+      "grad_norm": 0.5897819370932302,
+      "learning_rate": 1.9223542941045817e-05,
+      "loss": 0.4468,
+      "step": 282
+    },
+    {
+      "epoch": 0.856602345819145,
+      "grad_norm": 0.6119594306169949,
+      "learning_rate": 1.9213294044284822e-05,
+      "loss": 0.4554,
+      "step": 283
+    },
+    {
+      "epoch": 0.8596292092319334,
+      "grad_norm": 0.628966980633244,
+      "learning_rate": 1.920298071850123e-05,
+      "loss": 0.4898,
+      "step": 284
+    },
+    {
+      "epoch": 0.8626560726447219,
+      "grad_norm": 0.6016445644573405,
+      "learning_rate": 1.9192603035816657e-05,
+      "loss": 0.4561,
+      "step": 285
+    },
+    {
+      "epoch": 0.8656829360575105,
+      "grad_norm": 0.5910267553637264,
+      "learning_rate": 1.9182161068802742e-05,
+      "loss": 0.4135,
+      "step": 286
+    },
+    {
+      "epoch": 0.8687097994702989,
+      "grad_norm": 2.2562101038886833,
+      "learning_rate": 1.917165489048068e-05,
+      "loss": 0.4378,
+      "step": 287
+    },
+    {
+      "epoch": 0.8717366628830874,
+      "grad_norm": 3.915774754246098,
+      "learning_rate": 1.9161084574320696e-05,
+      "loss": 0.4793,
+      "step": 288
+    },
+    {
+      "epoch": 0.874763526295876,
+      "grad_norm": 12.42765768289753,
+      "learning_rate": 1.915045019424154e-05,
+      "loss": 0.4896,
+      "step": 289
+    },
+    {
+      "epoch": 0.8777903897086644,
+      "grad_norm": 2.102654805617899,
+      "learning_rate": 1.913975182460996e-05,
+      "loss": 0.4806,
+      "step": 290
+    },
+    {
+      "epoch": 0.8808172531214529,
+      "grad_norm": 1.9180125280711635,
+      "learning_rate": 1.9128989540240178e-05,
+      "loss": 0.4536,
+      "step": 291
+    },
+    {
+      "epoch": 0.8838441165342414,
+      "grad_norm": 3.024053877402373,
+      "learning_rate": 1.9118163416393392e-05,
+      "loss": 0.4786,
+      "step": 292
+    },
+    {
+      "epoch": 0.8868709799470299,
+      "grad_norm": 0.9025683149862834,
+      "learning_rate": 1.910727352877722e-05,
+      "loss": 0.4503,
+      "step": 293
+    },
+    {
+      "epoch": 0.8898978433598184,
+      "grad_norm": 0.9184323174205071,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.4386,
+      "step": 294
+    },
+    {
+      "epoch": 0.8929247067726069,
+      "grad_norm": 1.8645802635358135,
+      "learning_rate": 1.9085302767296183e-05,
+      "loss": 0.478,
+      "step": 295
+    },
+    {
+      "epoch": 0.8959515701853954,
+      "grad_norm": 0.6917218083242985,
+      "learning_rate": 1.9074222047073945e-05,
+      "loss": 0.4397,
+      "step": 296
+    },
+    {
+      "epoch": 0.8989784335981839,
+      "grad_norm": 0.6704591599921869,
+      "learning_rate": 1.9063077870366504e-05,
+      "loss": 0.4413,
+      "step": 297
+    },
+    {
+      "epoch": 0.9020052970109724,
+      "grad_norm": 0.7025186137447532,
+      "learning_rate": 1.9051870315105626e-05,
+      "loss": 0.4282,
+      "step": 298
+    },
+    {
+      "epoch": 0.9050321604237609,
+      "grad_norm": 0.6213166662119899,
+      "learning_rate": 1.9040599459666314e-05,
+      "loss": 0.4086,
+      "step": 299
+    },
+    {
+      "epoch": 0.9080590238365494,
+      "grad_norm": 0.5729906114498917,
+      "learning_rate": 1.9029265382866216e-05,
+      "loss": 0.4253,
+      "step": 300
+    },
+    {
+      "epoch": 0.9110858872493379,
+      "grad_norm": 0.5902417602489661,
+      "learning_rate": 1.901786816396509e-05,
+      "loss": 0.4392,
+      "step": 301
+    },
+    {
+      "epoch": 0.9141127506621264,
+      "grad_norm": 0.6609431058820643,
+      "learning_rate": 1.9006407882664256e-05,
+      "loss": 0.4395,
+      "step": 302
+    },
+    {
+      "epoch": 0.9171396140749148,
+      "grad_norm": 0.5711764470953805,
+      "learning_rate": 1.8994884619106034e-05,
+      "loss": 0.4372,
+      "step": 303
+    },
+    {
+      "epoch": 0.9201664774877034,
+      "grad_norm": 0.6515175506072429,
+      "learning_rate": 1.8983298453873172e-05,
+      "loss": 0.4225,
+      "step": 304
+    },
+    {
+      "epoch": 0.9231933409004919,
+      "grad_norm": 0.5598608937336701,
+      "learning_rate": 1.8971649467988303e-05,
+      "loss": 0.4297,
+      "step": 305
+    },
+    {
+      "epoch": 0.9262202043132803,
+      "grad_norm": 0.6125311656842891,
+      "learning_rate": 1.895993774291336e-05,
+      "loss": 0.4295,
+      "step": 306
+    },
+    {
+      "epoch": 0.9292470677260689,
+      "grad_norm": 0.6190585129861108,
+      "learning_rate": 1.894816336054902e-05,
+      "loss": 0.4682,
+      "step": 307
+    },
+    {
+      "epoch": 0.9322739311388574,
+      "grad_norm": 0.6000135242301603,
+      "learning_rate": 1.8936326403234125e-05,
+      "loss": 0.4318,
+      "step": 308
+    },
+    {
+      "epoch": 0.9353007945516458,
+      "grad_norm": 0.5734596359927081,
+      "learning_rate": 1.89244269537451e-05,
+      "loss": 0.4401,
+      "step": 309
+    },
+    {
+      "epoch": 0.9383276579644344,
+      "grad_norm": 0.5769938462905732,
+      "learning_rate": 1.891246509529539e-05,
+      "loss": 0.4206,
+      "step": 310
+    },
+    {
+      "epoch": 0.9413545213772229,
+      "grad_norm": 0.5552022969206694,
+      "learning_rate": 1.8900440911534864e-05,
+      "loss": 0.4266,
+      "step": 311
+    },
+    {
+      "epoch": 0.9443813847900113,
+      "grad_norm": 0.5587774158032164,
+      "learning_rate": 1.8888354486549238e-05,
+      "loss": 0.4436,
+      "step": 312
+    },
+    {
+      "epoch": 0.9474082482027999,
+      "grad_norm": 0.5164937863909608,
+      "learning_rate": 1.8876205904859474e-05,
+      "loss": 0.4163,
+      "step": 313
+    },
+    {
+      "epoch": 0.9504351116155884,
+      "grad_norm": 0.5781871387848945,
+      "learning_rate": 1.886399525142122e-05,
+      "loss": 0.4359,
+      "step": 314
+    },
+    {
+      "epoch": 0.9534619750283768,
+      "grad_norm": 0.5904372663736467,
+      "learning_rate": 1.8851722611624166e-05,
+      "loss": 0.4337,
+      "step": 315
+    },
+    {
+      "epoch": 0.9564888384411654,
+      "grad_norm": 0.5239073597982347,
+      "learning_rate": 1.8839388071291506e-05,
+      "loss": 0.4324,
+      "step": 316
+    },
+    {
+      "epoch": 0.9595157018539539,
+      "grad_norm": 0.5641488660039699,
+      "learning_rate": 1.8826991716679283e-05,
+      "loss": 0.4731,
+      "step": 317
+    },
+    {
+      "epoch": 0.9625425652667423,
+      "grad_norm": 0.5637805932939142,
+      "learning_rate": 1.881453363447582e-05,
+      "loss": 0.4556,
+      "step": 318
+    },
+    {
+      "epoch": 0.9655694286795309,
+      "grad_norm": 0.5553627246803207,
+      "learning_rate": 1.880201391180111e-05,
+      "loss": 0.43,
+      "step": 319
+    },
+    {
+      "epoch": 0.9685962920923193,
+      "grad_norm": 0.5725803563111447,
+      "learning_rate": 1.8789432636206197e-05,
+      "loss": 0.443,
+      "step": 320
+    },
+    {
+      "epoch": 0.9716231555051078,
+      "grad_norm": 0.5559115512270395,
+      "learning_rate": 1.8776789895672557e-05,
+      "loss": 0.4425,
+      "step": 321
+    },
+    {
+      "epoch": 0.9746500189178964,
+      "grad_norm": 0.5538490941014681,
+      "learning_rate": 1.8764085778611507e-05,
+      "loss": 0.442,
+      "step": 322
+    },
+    {
+      "epoch": 0.9776768823306848,
+      "grad_norm": 0.5841673268563595,
+      "learning_rate": 1.875132037386357e-05,
+      "loss": 0.4308,
+      "step": 323
+    },
+    {
+      "epoch": 0.9807037457434733,
+      "grad_norm": 0.5341527953806847,
+      "learning_rate": 1.873849377069785e-05,
+      "loss": 0.4363,
+      "step": 324
+    },
+    {
+      "epoch": 0.9837306091562619,
+      "grad_norm": 0.627999348860883,
+      "learning_rate": 1.8725606058811424e-05,
+      "loss": 0.474,
+      "step": 325
+    },
+    {
+      "epoch": 0.9867574725690503,
+      "grad_norm": 0.5632706540055739,
+      "learning_rate": 1.87126573283287e-05,
+      "loss": 0.4456,
+      "step": 326
+    },
+    {
+      "epoch": 0.9897843359818388,
+      "grad_norm": 0.5686686012576577,
+      "learning_rate": 1.869964766980079e-05,
+      "loss": 0.413,
+      "step": 327
+    },
+    {
+      "epoch": 0.9928111993946274,
+      "grad_norm": 0.5450099662265838,
+      "learning_rate": 1.8686577174204887e-05,
+      "loss": 0.4619,
+      "step": 328
+    },
+    {
+      "epoch": 0.9958380628074158,
+      "grad_norm": 0.5647164069185754,
+      "learning_rate": 1.8673445932943612e-05,
+      "loss": 0.4274,
+      "step": 329
+    },
+    {
+      "epoch": 0.9988649262202043,
+      "grad_norm": 0.5874338649495462,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.4519,
+      "step": 330
+    },
+    {
+      "epoch": 1.0018917896329929,
+      "grad_norm": 0.6049925134989617,
+      "learning_rate": 1.864700158115879e-05,
+      "loss": 0.3815,
+      "step": 331
+    },
+    {
+      "epoch": 1.0049186530457812,
+      "grad_norm": 0.6739135034230719,
+      "learning_rate": 1.863368865556191e-05,
+      "loss": 0.3727,
+      "step": 332
+    },
+    {
+      "epoch": 1.0079455164585698,
+      "grad_norm": 0.6190322483104117,
+      "learning_rate": 1.8620315354151695e-05,
+      "loss": 0.3643,
+      "step": 333
+    },
+    {
+      "epoch": 1.0109723798713584,
+      "grad_norm": 0.7332598959692239,
+      "learning_rate": 1.8606881770448305e-05,
+      "loss": 0.3863,
+      "step": 334
+    },
+    {
+      "epoch": 1.0139992432841467,
+      "grad_norm": 0.7364849222703389,
+      "learning_rate": 1.8593387998393457e-05,
+      "loss": 0.3885,
+      "step": 335
+    },
+    {
+      "epoch": 1.0170261066969353,
+      "grad_norm": 0.7149178622824649,
+      "learning_rate": 1.8579834132349773e-05,
+      "loss": 0.359,
+      "step": 336
+    },
+    {
+      "epoch": 1.0200529701097238,
+      "grad_norm": 0.8094059062600422,
+      "learning_rate": 1.8566220267100107e-05,
+      "loss": 0.3769,
+      "step": 337
+    },
+    {
+      "epoch": 1.0230798335225122,
+      "grad_norm": 0.650420416087458,
+      "learning_rate": 1.8552546497846893e-05,
+      "loss": 0.3765,
+      "step": 338
+    },
+    {
+      "epoch": 1.0261066969353008,
+      "grad_norm": 0.7112597409598952,
+      "learning_rate": 1.8538812920211484e-05,
+      "loss": 0.3499,
+      "step": 339
+    },
+    {
+      "epoch": 1.0291335603480893,
+      "grad_norm": 0.5868999988443542,
+      "learning_rate": 1.8525019630233463e-05,
+      "loss": 0.3413,
+      "step": 340
+    },
+    {
+      "epoch": 1.0321604237608777,
+      "grad_norm": 0.7182905859785886,
+      "learning_rate": 1.8511166724369997e-05,
+      "loss": 0.3438,
+      "step": 341
+    },
+    {
+      "epoch": 1.0351872871736663,
+      "grad_norm": 0.588164391183548,
+      "learning_rate": 1.8497254299495147e-05,
+      "loss": 0.3579,
+      "step": 342
+    },
+    {
+      "epoch": 1.0382141505864548,
+      "grad_norm": 0.5994632927208723,
+      "learning_rate": 1.8483282452899183e-05,
+      "loss": 0.3598,
+      "step": 343
+    },
+    {
+      "epoch": 1.0412410139992432,
+      "grad_norm": 0.5962827035742624,
+      "learning_rate": 1.8469251282287925e-05,
+      "loss": 0.336,
+      "step": 344
+    },
+    {
+      "epoch": 1.0442678774120318,
+      "grad_norm": 0.6700459284924142,
+      "learning_rate": 1.8455160885782045e-05,
+      "loss": 0.3364,
+      "step": 345
+    },
+    {
+      "epoch": 1.0472947408248203,
+      "grad_norm": 0.5833987927122629,
+      "learning_rate": 1.8441011361916387e-05,
+      "loss": 0.3441,
+      "step": 346
+    },
+    {
+      "epoch": 1.0503216042376087,
+      "grad_norm": 0.6632977637188374,
+      "learning_rate": 1.842680280963927e-05,
+      "loss": 0.368,
+      "step": 347
+    },
+    {
+      "epoch": 1.0533484676503972,
+      "grad_norm": 0.5781665006826253,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.3608,
+      "step": 348
+    },
+    {
+      "epoch": 1.0563753310631858,
+      "grad_norm": 0.5557225712059319,
+      "learning_rate": 1.8398209017707214e-05,
+      "loss": 0.3666,
+      "step": 349
+    },
+    {
+      "epoch": 1.0594021944759742,
+      "grad_norm": 0.5654439640099864,
+      "learning_rate": 1.8383823978010077e-05,
+      "loss": 0.3551,
+      "step": 350
+    },
+    {
+      "epoch": 1.0624290578887627,
+      "grad_norm": 0.5704866046263746,
+      "learning_rate": 1.83693803098157e-05,
+      "loss": 0.3456,
+      "step": 351
+    },
+    {
+      "epoch": 1.0654559213015513,
+      "grad_norm": 0.5889292171226277,
+      "learning_rate": 1.8354878114129368e-05,
+      "loss": 0.3742,
+      "step": 352
+    },
+    {
+      "epoch": 1.0684827847143397,
+      "grad_norm": 0.5684059106350144,
+      "learning_rate": 1.8340317492365654e-05,
+      "loss": 0.3713,
+      "step": 353
+    },
+    {
+      "epoch": 1.0715096481271282,
+      "grad_norm": 0.5707686797944012,
+      "learning_rate": 1.8325698546347714e-05,
+      "loss": 0.3471,
+      "step": 354
+    },
+    {
+      "epoch": 1.0745365115399168,
+      "grad_norm": 0.5722684940871366,
+      "learning_rate": 1.8311021378306563e-05,
+      "loss": 0.3654,
+      "step": 355
+    },
+    {
+      "epoch": 1.0775633749527052,
+      "grad_norm": 0.6078309352773806,
+      "learning_rate": 1.8296286090880362e-05,
+      "loss": 0.374,
+      "step": 356
+    },
+    {
+      "epoch": 1.0805902383654937,
+      "grad_norm": 0.5473554484573657,
+      "learning_rate": 1.8281492787113707e-05,
+      "loss": 0.3466,
+      "step": 357
+    },
+    {
+      "epoch": 1.0836171017782823,
+      "grad_norm": 0.5565867448567463,
+      "learning_rate": 1.8266641570456915e-05,
+      "loss": 0.3605,
+      "step": 358
+    },
+    {
+      "epoch": 1.0866439651910706,
+      "grad_norm": 0.5257208229693764,
+      "learning_rate": 1.8251732544765267e-05,
+      "loss": 0.3427,
+      "step": 359
+    },
+    {
+      "epoch": 1.0896708286038592,
+      "grad_norm": 0.5708413246498837,
+      "learning_rate": 1.8236765814298328e-05,
+      "loss": 0.3448,
+      "step": 360
+    },
+    {
+      "epoch": 1.0926976920166478,
+      "grad_norm": 0.5585777940028106,
+      "learning_rate": 1.822174148371919e-05,
+      "loss": 0.3656,
+      "step": 361
+    },
+    {
+      "epoch": 1.0957245554294364,
+      "grad_norm": 0.5554431717522996,
+      "learning_rate": 1.820665965809373e-05,
+      "loss": 0.3715,
+      "step": 362
+    },
+    {
+      "epoch": 1.0987514188422247,
+      "grad_norm": 0.5626458110501423,
+      "learning_rate": 1.819152044288992e-05,
+      "loss": 0.3504,
+      "step": 363
+    },
+    {
+      "epoch": 1.1017782822550133,
+      "grad_norm": 0.5457115805883949,
+      "learning_rate": 1.8176323943977034e-05,
+      "loss": 0.3436,
+      "step": 364
+    },
+    {
+      "epoch": 1.1048051456678016,
+      "grad_norm": 0.6027824486301956,
+      "learning_rate": 1.8161070267624938e-05,
+      "loss": 0.3691,
+      "step": 365
+    },
+    {
+      "epoch": 1.1078320090805902,
+      "grad_norm": 0.5488305621537294,
+      "learning_rate": 1.814575952050336e-05,
+      "loss": 0.3589,
+      "step": 366
+    },
+    {
+      "epoch": 1.1108588724933788,
+      "grad_norm": 0.637759255703908,
+      "learning_rate": 1.8130391809681105e-05,
+      "loss": 0.373,
+      "step": 367
+    },
+    {
+      "epoch": 1.1138857359061674,
+      "grad_norm": 0.5802318872040859,
+      "learning_rate": 1.8114967242625342e-05,
+      "loss": 0.3428,
+      "step": 368
+    },
+    {
+      "epoch": 1.1169125993189557,
+      "grad_norm": 0.6094030302304407,
+      "learning_rate": 1.809948592720084e-05,
+      "loss": 0.3285,
+      "step": 369
+    },
+    {
+      "epoch": 1.1199394627317443,
+      "grad_norm": 0.5897315049855668,
+      "learning_rate": 1.808394797166919e-05,
+      "loss": 0.3358,
+      "step": 370
+    },
+    {
+      "epoch": 1.1229663261445326,
+      "grad_norm": 0.6014694139962191,
+      "learning_rate": 1.8068353484688098e-05,
+      "loss": 0.3606,
+      "step": 371
+    },
+    {
+      "epoch": 1.1259931895573212,
+      "grad_norm": 0.6055585760090961,
+      "learning_rate": 1.8052702575310588e-05,
+      "loss": 0.3456,
+      "step": 372
+    },
+    {
+      "epoch": 1.1290200529701098,
+      "grad_norm": 0.5976899637331201,
+      "learning_rate": 1.803699535298424e-05,
+      "loss": 0.3447,
+      "step": 373
+    },
+    {
+      "epoch": 1.1320469163828983,
+      "grad_norm": 0.5954207174024013,
+      "learning_rate": 1.802123192755044e-05,
+      "loss": 0.347,
+      "step": 374
+    },
+    {
+      "epoch": 1.1350737797956867,
+      "grad_norm": 0.5849302903556542,
+      "learning_rate": 1.8005412409243604e-05,
+      "loss": 0.3436,
+      "step": 375
+    },
+    {
+      "epoch": 1.1381006432084753,
+      "grad_norm": 0.6078302103550508,
+      "learning_rate": 1.7989536908690413e-05,
+      "loss": 0.369,
+      "step": 376
+    },
+    {
+      "epoch": 1.1411275066212636,
+      "grad_norm": 0.5947317241727288,
+      "learning_rate": 1.797360553690903e-05,
+      "loss": 0.3514,
+      "step": 377
+    },
+    {
+      "epoch": 1.1441543700340522,
+      "grad_norm": 0.5929849176209275,
+      "learning_rate": 1.7957618405308323e-05,
+      "loss": 0.3472,
+      "step": 378
+    },
+    {
+      "epoch": 1.1471812334468408,
+      "grad_norm": 0.6124138664534822,
+      "learning_rate": 1.7941575625687103e-05,
+      "loss": 0.3715,
+      "step": 379
+    },
+    {
+      "epoch": 1.1502080968596293,
+      "grad_norm": 0.5898497394389981,
+      "learning_rate": 1.792547731023332e-05,
+      "loss": 0.3595,
+      "step": 380
+    },
+    {
+      "epoch": 1.1532349602724177,
+      "grad_norm": 0.627285753508268,
+      "learning_rate": 1.7909323571523295e-05,
+      "loss": 0.3704,
+      "step": 381
+    },
+    {
+      "epoch": 1.1562618236852062,
+      "grad_norm": 0.6007725551018329,
+      "learning_rate": 1.789311452252092e-05,
+      "loss": 0.3498,
+      "step": 382
+    },
+    {
+      "epoch": 1.1592886870979946,
+      "grad_norm": 0.5852737087821599,
+      "learning_rate": 1.787685027657689e-05,
+      "loss": 0.3512,
+      "step": 383
+    },
+    {
+      "epoch": 1.1623155505107832,
+      "grad_norm": 0.5973663653046741,
+      "learning_rate": 1.7860530947427878e-05,
+      "loss": 0.3384,
+      "step": 384
+    },
+    {
+      "epoch": 1.1653424139235717,
+      "grad_norm": 0.6085075097881701,
+      "learning_rate": 1.784415664919576e-05,
+      "loss": 0.3809,
+      "step": 385
+    },
+    {
+      "epoch": 1.1683692773363603,
+      "grad_norm": 0.5613700359874143,
+      "learning_rate": 1.782772749638682e-05,
+      "loss": 0.3529,
+      "step": 386
+    },
+    {
+      "epoch": 1.1713961407491487,
+      "grad_norm": 0.6018742375235979,
+      "learning_rate": 1.7811243603890934e-05,
+      "loss": 0.3592,
+      "step": 387
+    },
+    {
+      "epoch": 1.1744230041619372,
+      "grad_norm": 0.5661324420520297,
+      "learning_rate": 1.779470508698079e-05,
+      "loss": 0.3629,
+      "step": 388
+    },
+    {
+      "epoch": 1.1774498675747256,
+      "grad_norm": 0.5934780512667347,
+      "learning_rate": 1.777811206131105e-05,
+      "loss": 0.3562,
+      "step": 389
+    },
+    {
+      "epoch": 1.1804767309875142,
+      "grad_norm": 0.5773257794207289,
+      "learning_rate": 1.776146464291757e-05,
+      "loss": 0.3411,
+      "step": 390
+    },
+    {
+      "epoch": 1.1835035944003027,
+      "grad_norm": 0.5993796588010996,
+      "learning_rate": 1.7744762948216568e-05,
+      "loss": 0.3539,
+      "step": 391
+    },
+    {
+      "epoch": 1.1865304578130913,
+      "grad_norm": 0.5790413221556278,
+      "learning_rate": 1.772800709400383e-05,
+      "loss": 0.3576,
+      "step": 392
+    },
+    {
+      "epoch": 1.1895573212258796,
+      "grad_norm": 0.5897972514828568,
+      "learning_rate": 1.771119719745388e-05,
+      "loss": 0.3546,
+      "step": 393
+    },
+    {
+      "epoch": 1.1925841846386682,
+      "grad_norm": 0.6171075943715888,
+      "learning_rate": 1.7694333376119144e-05,
+      "loss": 0.3816,
+      "step": 394
+    },
+    {
+      "epoch": 1.1956110480514566,
+      "grad_norm": 0.6848231833308182,
+      "learning_rate": 1.7677415747929173e-05,
+      "loss": 0.3684,
+      "step": 395
+    },
+    {
+      "epoch": 1.1986379114642451,
+      "grad_norm": 0.5859142847874673,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 0.3466,
+      "step": 396
+    },
+    {
+      "epoch": 1.2016647748770337,
+      "grad_norm": 0.6862639105339676,
+      "learning_rate": 1.7643419544582225e-05,
+      "loss": 0.3389,
+      "step": 397
+    },
+    {
+      "epoch": 1.2046916382898223,
+      "grad_norm": 0.5383989750060779,
+      "learning_rate": 1.762634120716238e-05,
+      "loss": 0.3563,
+      "step": 398
+    },
+    {
+      "epoch": 1.2077185017026106,
+      "grad_norm": 0.5860901630501475,
+      "learning_rate": 1.7609209538359917e-05,
+      "loss": 0.3638,
+      "step": 399
+    },
+    {
+      "epoch": 1.2107453651153992,
+      "grad_norm": 0.6179314317022487,
+      "learning_rate": 1.7592024657977432e-05,
+      "loss": 0.3329,
+      "step": 400
+    },
+    {
+      "epoch": 1.2137722285281876,
+      "grad_norm": 0.6043016907959041,
+      "learning_rate": 1.757478668618966e-05,
+      "loss": 0.415,
+      "step": 401
+    },
+    {
+      "epoch": 1.2167990919409761,
+      "grad_norm": 0.5813598088603198,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.3453,
+      "step": 402
+    },
+    {
+      "epoch": 1.2198259553537647,
+      "grad_norm": 0.6063437490363506,
+      "learning_rate": 1.7540151950952632e-05,
+      "loss": 0.3517,
+      "step": 403
+    },
+    {
+      "epoch": 1.2228528187665533,
+      "grad_norm": 0.5574293715557551,
+      "learning_rate": 1.75227554297058e-05,
+      "loss": 0.3878,
+      "step": 404
+    },
+    {
+      "epoch": 1.2258796821793416,
+      "grad_norm": 0.5876182742881972,
+      "learning_rate": 1.7505306301456823e-05,
+      "loss": 0.338,
+      "step": 405
+    },
+    {
+      "epoch": 1.2289065455921302,
+      "grad_norm": 0.5580036915187141,
+      "learning_rate": 1.7487804688228327e-05,
+      "loss": 0.3566,
+      "step": 406
+    },
+    {
+      "epoch": 1.2319334090049185,
+      "grad_norm": 0.5479301641322466,
+      "learning_rate": 1.7470250712409963e-05,
+      "loss": 0.3737,
+      "step": 407
+    },
+    {
+      "epoch": 1.2349602724177071,
+      "grad_norm": 0.637863488732352,
+      "learning_rate": 1.745264449675755e-05,
+      "loss": 0.3687,
+      "step": 408
+    },
+    {
+      "epoch": 1.2379871358304957,
+      "grad_norm": 0.5641204771234941,
+      "learning_rate": 1.743498616439223e-05,
+      "loss": 0.3405,
+      "step": 409
+    },
+    {
+      "epoch": 1.2410139992432843,
+      "grad_norm": 0.6033319464460466,
+      "learning_rate": 1.7417275838799596e-05,
+      "loss": 0.399,
+      "step": 410
+    },
+    {
+      "epoch": 1.2440408626560726,
+      "grad_norm": 0.6121574204398148,
+      "learning_rate": 1.739951364382884e-05,
+      "loss": 0.3577,
+      "step": 411
+    },
+    {
+      "epoch": 1.2470677260688612,
+      "grad_norm": 0.5999547165909839,
+      "learning_rate": 1.7381699703691866e-05,
+      "loss": 0.3663,
+      "step": 412
+    },
+    {
+      "epoch": 1.2500945894816495,
+      "grad_norm": 0.5511920141804303,
+      "learning_rate": 1.736383414296245e-05,
+      "loss": 0.3522,
+      "step": 413
+    },
+    {
+      "epoch": 1.253121452894438,
+      "grad_norm": 0.5870166340473345,
+      "learning_rate": 1.734591708657533e-05,
+      "loss": 0.3289,
+      "step": 414
+    },
+    {
+      "epoch": 1.2561483163072267,
+      "grad_norm": 0.58744505002037,
+      "learning_rate": 1.732794865982539e-05,
+      "loss": 0.3462,
+      "step": 415
+    },
+    {
+      "epoch": 1.2591751797200152,
+      "grad_norm": 0.5505997027750461,
+      "learning_rate": 1.730992898836672e-05,
+      "loss": 0.3224,
+      "step": 416
+    },
+    {
+      "epoch": 1.2622020431328036,
+      "grad_norm": 0.6195038911774278,
+      "learning_rate": 1.7291858198211772e-05,
+      "loss": 0.3379,
+      "step": 417
+    },
+    {
+      "epoch": 1.2652289065455922,
+      "grad_norm": 0.5824121792370334,
+      "learning_rate": 1.7273736415730488e-05,
+      "loss": 0.3586,
+      "step": 418
+    },
+    {
+      "epoch": 1.2682557699583805,
+      "grad_norm": 0.5705350849324659,
+      "learning_rate": 1.725556376764939e-05,
+      "loss": 0.3556,
+      "step": 419
+    },
+    {
+      "epoch": 1.271282633371169,
+      "grad_norm": 0.5915195806371565,
+      "learning_rate": 1.72373403810507e-05,
+      "loss": 0.3362,
+      "step": 420
+    },
+    {
+      "epoch": 1.2743094967839577,
+      "grad_norm": 0.5683755067310357,
+      "learning_rate": 1.721906638337148e-05,
+      "loss": 0.3328,
+      "step": 421
+    },
+    {
+      "epoch": 1.2773363601967462,
+      "grad_norm": 0.6195098889825341,
+      "learning_rate": 1.720074190240269e-05,
+      "loss": 0.3857,
+      "step": 422
+    },
+    {
+      "epoch": 1.2803632236095346,
+      "grad_norm": 0.5783839808430629,
+      "learning_rate": 1.7182367066288344e-05,
+      "loss": 0.3683,
+      "step": 423
+    },
+    {
+      "epoch": 1.2833900870223232,
+      "grad_norm": 0.618952833082765,
+      "learning_rate": 1.7163942003524574e-05,
+      "loss": 0.3566,
+      "step": 424
+    },
+    {
+      "epoch": 1.2864169504351115,
+      "grad_norm": 0.5535144927238279,
+      "learning_rate": 1.7145466842958764e-05,
+      "loss": 0.3553,
+      "step": 425
+    },
+    {
+      "epoch": 1.2894438138479,
+      "grad_norm": 0.607161908184491,
+      "learning_rate": 1.7126941713788633e-05,
+      "loss": 0.3685,
+      "step": 426
+    },
+    {
+      "epoch": 1.2924706772606886,
+      "grad_norm": 0.563370385137249,
+      "learning_rate": 1.7108366745561318e-05,
+      "loss": 0.3823,
+      "step": 427
+    },
+    {
+      "epoch": 1.2954975406734772,
+      "grad_norm": 0.5396798431456107,
+      "learning_rate": 1.70897420681725e-05,
+      "loss": 0.3561,
+      "step": 428
+    },
+    {
+      "epoch": 1.2985244040862656,
+      "grad_norm": 0.5454612530772457,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.35,
+      "step": 429
+    },
+    {
+      "epoch": 1.3015512674990541,
+      "grad_norm": 0.5377641159931936,
+      "learning_rate": 1.7052344107230244e-05,
+      "loss": 0.3306,
+      "step": 430
+    },
+    {
+      "epoch": 1.3045781309118425,
+      "grad_norm": 0.5630281580985383,
+      "learning_rate": 1.7033571085202597e-05,
+      "loss": 0.3702,
+      "step": 431
+    },
+    {
+      "epoch": 1.307604994324631,
+      "grad_norm": 0.5536501233890527,
+      "learning_rate": 1.7014748877063212e-05,
+      "loss": 0.3632,
+      "step": 432
+    },
+    {
+      "epoch": 1.3106318577374196,
+      "grad_norm": 0.5406979164533523,
+      "learning_rate": 1.699587761443673e-05,
+      "loss": 0.3683,
+      "step": 433
+    },
+    {
+      "epoch": 1.3136587211502082,
+      "grad_norm": 0.5544798357187173,
+      "learning_rate": 1.697695742929082e-05,
+      "loss": 0.3605,
+      "step": 434
+    },
+    {
+      "epoch": 1.3166855845629966,
+      "grad_norm": 0.5957958959218278,
+      "learning_rate": 1.6957988453935276e-05,
+      "loss": 0.3704,
+      "step": 435
+    },
+    {
+      "epoch": 1.3197124479757851,
+      "grad_norm": 0.5323299139101086,
+      "learning_rate": 1.693897082102109e-05,
+      "loss": 0.3706,
+      "step": 436
+    },
+    {
+      "epoch": 1.3227393113885735,
+      "grad_norm": 0.6085121378715731,
+      "learning_rate": 1.691990466353951e-05,
+      "loss": 0.3635,
+      "step": 437
+    },
+    {
+      "epoch": 1.325766174801362,
+      "grad_norm": 0.5959040345205575,
+      "learning_rate": 1.6900790114821122e-05,
+      "loss": 0.364,
+      "step": 438
+    },
+    {
+      "epoch": 1.3287930382141506,
+      "grad_norm": 0.5763811387387792,
+      "learning_rate": 1.688162730853491e-05,
+      "loss": 0.3634,
+      "step": 439
+    },
+    {
+      "epoch": 1.3318199016269392,
+      "grad_norm": 0.5532804033861637,
+      "learning_rate": 1.686241637868734e-05,
+      "loss": 0.3557,
+      "step": 440
+    },
+    {
+      "epoch": 1.3348467650397275,
+      "grad_norm": 0.5789035843759028,
+      "learning_rate": 1.6843157459621386e-05,
+      "loss": 0.3828,
+      "step": 441
+    },
+    {
+      "epoch": 1.3378736284525161,
+      "grad_norm": 0.549978824284729,
+      "learning_rate": 1.682385068601563e-05,
+      "loss": 0.3534,
+      "step": 442
+    },
+    {
+      "epoch": 1.3409004918653045,
+      "grad_norm": 0.5645222264148867,
+      "learning_rate": 1.68044961928833e-05,
+      "loss": 0.3468,
+      "step": 443
+    },
+    {
+      "epoch": 1.343927355278093,
+      "grad_norm": 0.6397706004062406,
+      "learning_rate": 1.6785094115571323e-05,
+      "loss": 0.3327,
+      "step": 444
+    },
+    {
+      "epoch": 1.3469542186908816,
+      "grad_norm": 0.5899152834199866,
+      "learning_rate": 1.6765644589759396e-05,
+      "loss": 0.3605,
+      "step": 445
+    },
+    {
+      "epoch": 1.3499810821036702,
+      "grad_norm": 0.5606890851946956,
+      "learning_rate": 1.674614775145901e-05,
+      "loss": 0.3595,
+      "step": 446
+    },
+    {
+      "epoch": 1.3530079455164585,
+      "grad_norm": 0.6103645345267558,
+      "learning_rate": 1.6726603737012527e-05,
+      "loss": 0.3828,
+      "step": 447
+    },
+    {
+      "epoch": 1.356034808929247,
+      "grad_norm": 0.5710541731700772,
+      "learning_rate": 1.670701268309221e-05,
+      "loss": 0.3688,
+      "step": 448
+    },
+    {
+      "epoch": 1.3590616723420355,
+      "grad_norm": 0.5638083232799351,
+      "learning_rate": 1.6687374726699274e-05,
+      "loss": 0.3715,
+      "step": 449
+    },
+    {
+      "epoch": 1.362088535754824,
+      "grad_norm": 0.5911467794323303,
+      "learning_rate": 1.666769000516292e-05,
+      "loss": 0.3674,
+      "step": 450
+    },
+    {
+      "epoch": 1.3651153991676126,
+      "grad_norm": 0.537162897039177,
+      "learning_rate": 1.6647958656139377e-05,
+      "loss": 0.3396,
+      "step": 451
+    },
+    {
+      "epoch": 1.3681422625804012,
+      "grad_norm": 0.5714722170551004,
+      "learning_rate": 1.6628180817610963e-05,
+      "loss": 0.3894,
+      "step": 452
+    },
+    {
+      "epoch": 1.3711691259931895,
+      "grad_norm": 0.6217322112592903,
+      "learning_rate": 1.660835662788507e-05,
+      "loss": 0.3692,
+      "step": 453
+    },
+    {
+      "epoch": 1.374195989405978,
+      "grad_norm": 0.5808420556146904,
+      "learning_rate": 1.658848622559325e-05,
+      "loss": 0.3834,
+      "step": 454
+    },
+    {
+      "epoch": 1.3772228528187664,
+      "grad_norm": 0.606602074453483,
+      "learning_rate": 1.6568569749690208e-05,
+      "loss": 0.3435,
+      "step": 455
+    },
+    {
+      "epoch": 1.380249716231555,
+      "grad_norm": 0.589823827669581,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.3672,
+      "step": 456
+    },
+    {
+      "epoch": 1.3832765796443436,
+      "grad_norm": 0.6266803721395025,
+      "learning_rate": 1.6528599134479305e-05,
+      "loss": 0.3427,
+      "step": 457
+    },
+    {
+      "epoch": 1.3863034430571322,
+      "grad_norm": 0.6039021182974211,
+      "learning_rate": 1.6508545274687936e-05,
+      "loss": 0.3832,
+      "step": 458
+    },
+    {
+      "epoch": 1.3893303064699205,
+      "grad_norm": 0.5481956682591254,
+      "learning_rate": 1.6488445900316388e-05,
+      "loss": 0.3475,
+      "step": 459
+    },
+    {
+      "epoch": 1.392357169882709,
+      "grad_norm": 0.5531149384890495,
+      "learning_rate": 1.6468301151920576e-05,
+      "loss": 0.3469,
+      "step": 460
+    },
+    {
+      "epoch": 1.3953840332954974,
+      "grad_norm": 0.5557054164574347,
+      "learning_rate": 1.644811117037373e-05,
+      "loss": 0.342,
+      "step": 461
+    },
+    {
+      "epoch": 1.398410896708286,
+      "grad_norm": 0.5928789728431401,
+      "learning_rate": 1.6427876096865394e-05,
+      "loss": 0.3887,
+      "step": 462
+    },
+    {
+      "epoch": 1.4014377601210746,
+      "grad_norm": 0.5413822670695531,
+      "learning_rate": 1.6407596072900444e-05,
+      "loss": 0.3426,
+      "step": 463
+    },
+    {
+      "epoch": 1.4044646235338631,
+      "grad_norm": 0.5948440448293565,
+      "learning_rate": 1.6387271240298082e-05,
+      "loss": 0.3655,
+      "step": 464
+    },
+    {
+      "epoch": 1.4074914869466515,
+      "grad_norm": 0.5662199416711721,
+      "learning_rate": 1.6366901741190885e-05,
+      "loss": 0.3923,
+      "step": 465
+    },
+    {
+      "epoch": 1.41051835035944,
+      "grad_norm": 0.5619953351532011,
+      "learning_rate": 1.6346487718023762e-05,
+      "loss": 0.3721,
+      "step": 466
+    },
+    {
+      "epoch": 1.4135452137722284,
+      "grad_norm": 0.5504888996943389,
+      "learning_rate": 1.6326029313552998e-05,
+      "loss": 0.3525,
+      "step": 467
+    },
+    {
+      "epoch": 1.416572077185017,
+      "grad_norm": 0.5740471858650602,
+      "learning_rate": 1.6305526670845225e-05,
+      "loss": 0.3676,
+      "step": 468
+    },
+    {
+      "epoch": 1.4195989405978056,
+      "grad_norm": 0.538488978010878,
+      "learning_rate": 1.6284979933276446e-05,
+      "loss": 0.3458,
+      "step": 469
+    },
+    {
+      "epoch": 1.4226258040105941,
+      "grad_norm": 0.5890875439941626,
+      "learning_rate": 1.6264389244531015e-05,
+      "loss": 0.3742,
+      "step": 470
+    },
+    {
+      "epoch": 1.4256526674233825,
+      "grad_norm": 0.6058199656890567,
+      "learning_rate": 1.6243754748600637e-05,
+      "loss": 0.3626,
+      "step": 471
+    },
+    {
+      "epoch": 1.428679530836171,
+      "grad_norm": 0.6000704541005174,
+      "learning_rate": 1.6223076589783368e-05,
+      "loss": 0.3602,
+      "step": 472
+    },
+    {
+      "epoch": 1.4317063942489594,
+      "grad_norm": 0.5695744765958708,
+      "learning_rate": 1.6202354912682602e-05,
+      "loss": 0.3722,
+      "step": 473
+    },
+    {
+      "epoch": 1.434733257661748,
+      "grad_norm": 0.6046473083778348,
+      "learning_rate": 1.6181589862206053e-05,
+      "loss": 0.3636,
+      "step": 474
+    },
+    {
+      "epoch": 1.4377601210745365,
+      "grad_norm": 0.5931096747575126,
+      "learning_rate": 1.616078158356475e-05,
+      "loss": 0.3706,
+      "step": 475
+    },
+    {
+      "epoch": 1.4407869844873251,
+      "grad_norm": 0.5810745029046898,
+      "learning_rate": 1.613993022227202e-05,
+      "loss": 0.3755,
+      "step": 476
+    },
+    {
+      "epoch": 1.4438138479001135,
+      "grad_norm": 0.5602620894587699,
+      "learning_rate": 1.6119035924142468e-05,
+      "loss": 0.384,
+      "step": 477
+    },
+    {
+      "epoch": 1.446840711312902,
+      "grad_norm": 0.5777953481104446,
+      "learning_rate": 1.6098098835290955e-05,
+      "loss": 0.3462,
+      "step": 478
+    },
+    {
+      "epoch": 1.4498675747256904,
+      "grad_norm": 0.5691638509904715,
+      "learning_rate": 1.6077119102131585e-05,
+      "loss": 0.3513,
+      "step": 479
+    },
+    {
+      "epoch": 1.452894438138479,
+      "grad_norm": 0.5470755102693828,
+      "learning_rate": 1.6056096871376667e-05,
+      "loss": 0.3346,
+      "step": 480
+    },
+    {
+      "epoch": 1.4559213015512675,
+      "grad_norm": 0.5414941063765051,
+      "learning_rate": 1.6035032290035715e-05,
+      "loss": 0.359,
+      "step": 481
+    },
+    {
+      "epoch": 1.458948164964056,
+      "grad_norm": 0.5311187278030568,
+      "learning_rate": 1.6013925505414386e-05,
+      "loss": 0.3339,
+      "step": 482
+    },
+    {
+      "epoch": 1.4619750283768445,
+      "grad_norm": 0.5543647164520338,
+      "learning_rate": 1.599277666511347e-05,
+      "loss": 0.3599,
+      "step": 483
+    },
+    {
+      "epoch": 1.465001891789633,
+      "grad_norm": 0.5680413514093865,
+      "learning_rate": 1.5971585917027864e-05,
+      "loss": 0.353,
+      "step": 484
+    },
+    {
+      "epoch": 1.4680287552024214,
+      "grad_norm": 0.5868824158963895,
+      "learning_rate": 1.5950353409345515e-05,
+      "loss": 0.3406,
+      "step": 485
+    },
+    {
+      "epoch": 1.47105561861521,
+      "grad_norm": 0.5428480968972231,
+      "learning_rate": 1.5929079290546408e-05,
+      "loss": 0.3312,
+      "step": 486
+    },
+    {
+      "epoch": 1.4740824820279985,
+      "grad_norm": 0.5313860654068874,
+      "learning_rate": 1.5907763709401507e-05,
+      "loss": 0.3244,
+      "step": 487
+    },
+    {
+      "epoch": 1.477109345440787,
+      "grad_norm": 0.5955614485616447,
+      "learning_rate": 1.5886406814971728e-05,
+      "loss": 0.3381,
+      "step": 488
+    },
+    {
+      "epoch": 1.4801362088535754,
+      "grad_norm": 0.5401200213586426,
+      "learning_rate": 1.5865008756606905e-05,
+      "loss": 0.356,
+      "step": 489
+    },
+    {
+      "epoch": 1.483163072266364,
+      "grad_norm": 0.5830180639918157,
+      "learning_rate": 1.584356968394471e-05,
+      "loss": 0.338,
+      "step": 490
+    },
+    {
+      "epoch": 1.4861899356791524,
+      "grad_norm": 0.546318980031365,
+      "learning_rate": 1.5822089746909647e-05,
+      "loss": 0.3575,
+      "step": 491
+    },
+    {
+      "epoch": 1.489216799091941,
+      "grad_norm": 0.5875210869251928,
+      "learning_rate": 1.5800569095711983e-05,
+      "loss": 0.3681,
+      "step": 492
+    },
+    {
+      "epoch": 1.4922436625047295,
+      "grad_norm": 0.5779525594026657,
+      "learning_rate": 1.57790078808467e-05,
+      "loss": 0.3608,
+      "step": 493
+    },
+    {
+      "epoch": 1.495270525917518,
+      "grad_norm": 0.5194134247082568,
+      "learning_rate": 1.575740625309244e-05,
+      "loss": 0.3248,
+      "step": 494
+    },
+    {
+      "epoch": 1.4982973893303064,
+      "grad_norm": 0.5328950931599159,
+      "learning_rate": 1.573576436351046e-05,
+      "loss": 0.3206,
+      "step": 495
+    },
+    {
+      "epoch": 1.501324252743095,
+      "grad_norm": 0.5783712714939293,
+      "learning_rate": 1.5714082363443576e-05,
+      "loss": 0.3599,
+      "step": 496
+    },
+    {
+      "epoch": 1.5043511161558833,
+      "grad_norm": 0.5676712793133903,
+      "learning_rate": 1.5692360404515078e-05,
+      "loss": 0.3715,
+      "step": 497
+    },
+    {
+      "epoch": 1.507377979568672,
+      "grad_norm": 0.6216015502324677,
+      "learning_rate": 1.5670598638627707e-05,
+      "loss": 0.3465,
+      "step": 498
+    },
+    {
+      "epoch": 1.5104048429814605,
+      "grad_norm": 0.5454476639824847,
+      "learning_rate": 1.564879721796258e-05,
+      "loss": 0.3703,
+      "step": 499
+    },
+    {
+      "epoch": 1.513431706394249,
+      "grad_norm": 0.6370971586589214,
+      "learning_rate": 1.5626956294978103e-05,
+      "loss": 0.385,
+      "step": 500
+    },
+    {
+      "epoch": 1.5164585698070374,
+      "grad_norm": 0.5796293191409025,
+      "learning_rate": 1.560507602240894e-05,
+      "loss": 0.3463,
+      "step": 501
+    },
+    {
+      "epoch": 1.519485433219826,
+      "grad_norm": 0.5656402082947922,
+      "learning_rate": 1.5583156553264923e-05,
+      "loss": 0.3678,
+      "step": 502
+    },
+    {
+      "epoch": 1.5225122966326143,
+      "grad_norm": 0.5915558460301508,
+      "learning_rate": 1.5561198040829994e-05,
+      "loss": 0.3711,
+      "step": 503
+    },
+    {
+      "epoch": 1.525539160045403,
+      "grad_norm": 0.610120684885826,
+      "learning_rate": 1.5539200638661106e-05,
+      "loss": 0.3732,
+      "step": 504
+    },
+    {
+      "epoch": 1.5285660234581915,
+      "grad_norm": 0.5842280589370393,
+      "learning_rate": 1.551716450058719e-05,
+      "loss": 0.3626,
+      "step": 505
+    },
+    {
+      "epoch": 1.53159288687098,
+      "grad_norm": 0.5444300039442264,
+      "learning_rate": 1.5495089780708062e-05,
+      "loss": 0.3421,
+      "step": 506
+    },
+    {
+      "epoch": 1.5346197502837684,
+      "grad_norm": 0.6040514512021319,
+      "learning_rate": 1.5472976633393325e-05,
+      "loss": 0.3601,
+      "step": 507
+    },
+    {
+      "epoch": 1.537646613696557,
+      "grad_norm": 0.5695317326407557,
+      "learning_rate": 1.5450825213281317e-05,
+      "loss": 0.3842,
+      "step": 508
+    },
+    {
+      "epoch": 1.5406734771093453,
+      "grad_norm": 0.5765182370904195,
+      "learning_rate": 1.5428635675278022e-05,
+      "loss": 0.3534,
+      "step": 509
+    },
+    {
+      "epoch": 1.543700340522134,
+      "grad_norm": 0.5772064028372753,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.4041,
+      "step": 510
+    },
+    {
+      "epoch": 1.5467272039349225,
+      "grad_norm": 0.5489124829043441,
+      "learning_rate": 1.53841428665532e-05,
+      "loss": 0.3607,
+      "step": 511
+    },
+    {
+      "epoch": 1.549754067347711,
+      "grad_norm": 0.5401307766516238,
+      "learning_rate": 1.5361839906972095e-05,
+      "loss": 0.366,
+      "step": 512
+    },
+    {
+      "epoch": 1.5527809307604994,
+      "grad_norm": 0.5442447913339895,
+      "learning_rate": 1.5339499451778363e-05,
+      "loss": 0.3766,
+      "step": 513
+    },
+    {
+      "epoch": 1.555807794173288,
+      "grad_norm": 0.5474992403959636,
+      "learning_rate": 1.531712165719992e-05,
+      "loss": 0.3192,
+      "step": 514
+    },
+    {
+      "epoch": 1.5588346575860763,
+      "grad_norm": 0.5869683234577026,
+      "learning_rate": 1.529470667972579e-05,
+      "loss": 0.3679,
+      "step": 515
+    },
+    {
+      "epoch": 1.5618615209988649,
+      "grad_norm": 0.5690417479292975,
+      "learning_rate": 1.5272254676105026e-05,
+      "loss": 0.3699,
+      "step": 516
+    },
+    {
+      "epoch": 1.5648883844116535,
+      "grad_norm": 0.5856953644947631,
+      "learning_rate": 1.5249765803345602e-05,
+      "loss": 0.3797,
+      "step": 517
+    },
+    {
+      "epoch": 1.567915247824442,
+      "grad_norm": 0.5872630482496503,
+      "learning_rate": 1.5227240218713326e-05,
+      "loss": 0.3779,
+      "step": 518
+    },
+    {
+      "epoch": 1.5709421112372304,
+      "grad_norm": 0.5437501734610226,
+      "learning_rate": 1.5204678079730724e-05,
+      "loss": 0.3481,
+      "step": 519
+    },
+    {
+      "epoch": 1.573968974650019,
+      "grad_norm": 0.5462994239072939,
+      "learning_rate": 1.5182079544175957e-05,
+      "loss": 0.3708,
+      "step": 520
+    },
+    {
+      "epoch": 1.5769958380628073,
+      "grad_norm": 0.5511038615742283,
+      "learning_rate": 1.5159444770081704e-05,
+      "loss": 0.3518,
+      "step": 521
+    },
+    {
+      "epoch": 1.5800227014755959,
+      "grad_norm": 0.5982991851085945,
+      "learning_rate": 1.5136773915734067e-05,
+      "loss": 0.3506,
+      "step": 522
+    },
+    {
+      "epoch": 1.5830495648883844,
+      "grad_norm": 0.5674239162199687,
+      "learning_rate": 1.5114067139671452e-05,
+      "loss": 0.359,
+      "step": 523
+    },
+    {
+      "epoch": 1.586076428301173,
+      "grad_norm": 0.544974999672973,
+      "learning_rate": 1.5091324600683472e-05,
+      "loss": 0.343,
+      "step": 524
+    },
+    {
+      "epoch": 1.5891032917139614,
+      "grad_norm": 0.5912765576426214,
+      "learning_rate": 1.5068546457809831e-05,
+      "loss": 0.3574,
+      "step": 525
+    },
+    {
+      "epoch": 1.59213015512675,
+      "grad_norm": 0.5886124541165613,
+      "learning_rate": 1.5045732870339213e-05,
+      "loss": 0.3698,
+      "step": 526
+    },
+    {
+      "epoch": 1.5951570185395383,
+      "grad_norm": 0.5416372890300947,
+      "learning_rate": 1.5022883997808171e-05,
+      "loss": 0.3459,
+      "step": 527
+    },
+    {
+      "epoch": 1.5981838819523269,
+      "grad_norm": 0.5189154096760485,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.3314,
+      "step": 528
+    },
+    {
+      "epoch": 1.6012107453651154,
+      "grad_norm": 0.5465686426814911,
+      "learning_rate": 1.4977081036943637e-05,
+      "loss": 0.3553,
+      "step": 529
+    },
+    {
+      "epoch": 1.604237608777904,
+      "grad_norm": 0.5886852800535238,
+      "learning_rate": 1.4954127268912525e-05,
+      "loss": 0.3921,
+      "step": 530
+    },
+    {
+      "epoch": 1.6072644721906924,
+      "grad_norm": 0.561129556992196,
+      "learning_rate": 1.4931138856423504e-05,
+      "loss": 0.3608,
+      "step": 531
+    },
+    {
+      "epoch": 1.610291335603481,
+      "grad_norm": 0.5348224260286456,
+      "learning_rate": 1.4908115960235683e-05,
+      "loss": 0.3747,
+      "step": 532
+    },
+    {
+      "epoch": 1.6133181990162693,
+      "grad_norm": 0.5740320127552693,
+      "learning_rate": 1.4885058741349322e-05,
+      "loss": 0.3522,
+      "step": 533
+    },
+    {
+      "epoch": 1.6163450624290578,
+      "grad_norm": 0.516633693944608,
+      "learning_rate": 1.4861967361004687e-05,
+      "loss": 0.3784,
+      "step": 534
+    },
+    {
+      "epoch": 1.6193719258418464,
+      "grad_norm": 0.5570420052633861,
+      "learning_rate": 1.483884198068096e-05,
+      "loss": 0.3694,
+      "step": 535
+    },
+    {
+      "epoch": 1.622398789254635,
+      "grad_norm": 0.5421203867144168,
+      "learning_rate": 1.4815682762095065e-05,
+      "loss": 0.3695,
+      "step": 536
+    },
+    {
+      "epoch": 1.6254256526674233,
+      "grad_norm": 0.5470071986064853,
+      "learning_rate": 1.479248986720057e-05,
+      "loss": 0.3647,
+      "step": 537
+    },
+    {
+      "epoch": 1.628452516080212,
+      "grad_norm": 0.5687150651429135,
+      "learning_rate": 1.476926345818654e-05,
+      "loss": 0.3593,
+      "step": 538
+    },
+    {
+      "epoch": 1.6314793794930003,
+      "grad_norm": 0.5623895371829716,
+      "learning_rate": 1.4746003697476406e-05,
+      "loss": 0.3836,
+      "step": 539
+    },
+    {
+      "epoch": 1.6345062429057888,
+      "grad_norm": 0.6345666632955639,
+      "learning_rate": 1.472271074772683e-05,
+      "loss": 0.3547,
+      "step": 540
+    },
+    {
+      "epoch": 1.6375331063185774,
+      "grad_norm": 0.5303872334612526,
+      "learning_rate": 1.4699384771826562e-05,
+      "loss": 0.3379,
+      "step": 541
+    },
+    {
+      "epoch": 1.640559969731366,
+      "grad_norm": 0.5916889584005367,
+      "learning_rate": 1.4676025932895315e-05,
+      "loss": 0.3554,
+      "step": 542
+    },
+    {
+      "epoch": 1.6435868331441543,
+      "grad_norm": 0.5578742203619609,
+      "learning_rate": 1.4652634394282608e-05,
+      "loss": 0.3668,
+      "step": 543
+    },
+    {
+      "epoch": 1.646613696556943,
+      "grad_norm": 0.5706287003024477,
+      "learning_rate": 1.4629210319566626e-05,
+      "loss": 0.3589,
+      "step": 544
+    },
+    {
+      "epoch": 1.6496405599697312,
+      "grad_norm": 0.5883459050791259,
+      "learning_rate": 1.4605753872553094e-05,
+      "loss": 0.3658,
+      "step": 545
+    },
+    {
+      "epoch": 1.6526674233825198,
+      "grad_norm": 0.5749125915015197,
+      "learning_rate": 1.4582265217274105e-05,
+      "loss": 0.3701,
+      "step": 546
+    },
+    {
+      "epoch": 1.6556942867953084,
+      "grad_norm": 0.5796901883762888,
+      "learning_rate": 1.4558744517986995e-05,
+      "loss": 0.3399,
+      "step": 547
+    },
+    {
+      "epoch": 1.658721150208097,
+      "grad_norm": 0.5937574416050302,
+      "learning_rate": 1.4535191939173179e-05,
+      "loss": 0.3741,
+      "step": 548
+    },
+    {
+      "epoch": 1.6617480136208853,
+      "grad_norm": 0.5894641094726274,
+      "learning_rate": 1.4511607645537009e-05,
+      "loss": 0.3695,
+      "step": 549
+    },
+    {
+      "epoch": 1.6647748770336739,
+      "grad_norm": 0.5513144788911749,
+      "learning_rate": 1.4487991802004625e-05,
+      "loss": 0.3487,
+      "step": 550
+    },
+    {
+      "epoch": 1.6678017404464622,
+      "grad_norm": 0.6062380971088138,
+      "learning_rate": 1.4464344573722788e-05,
+      "loss": 0.3493,
+      "step": 551
+    },
+    {
+      "epoch": 1.6708286038592508,
+      "grad_norm": 0.5248882186460455,
+      "learning_rate": 1.4440666126057743e-05,
+      "loss": 0.3793,
+      "step": 552
+    },
+    {
+      "epoch": 1.6738554672720394,
+      "grad_norm": 0.5698030748222763,
+      "learning_rate": 1.4416956624594047e-05,
+      "loss": 0.3752,
+      "step": 553
+    },
+    {
+      "epoch": 1.676882330684828,
+      "grad_norm": 0.5824067211439345,
+      "learning_rate": 1.4393216235133427e-05,
+      "loss": 0.349,
+      "step": 554
+    },
+    {
+      "epoch": 1.6799091940976163,
+      "grad_norm": 0.565829366148147,
+      "learning_rate": 1.4369445123693595e-05,
+      "loss": 0.3655,
+      "step": 555
+    },
+    {
+      "epoch": 1.6829360575104049,
+      "grad_norm": 0.5268457257034886,
+      "learning_rate": 1.4345643456507126e-05,
+      "loss": 0.3747,
+      "step": 556
+    },
+    {
+      "epoch": 1.6859629209231932,
+      "grad_norm": 0.5379301632259317,
+      "learning_rate": 1.4321811400020245e-05,
+      "loss": 0.3518,
+      "step": 557
+    },
+    {
+      "epoch": 1.6889897843359818,
+      "grad_norm": 0.5451257517376008,
+      "learning_rate": 1.4297949120891718e-05,
+      "loss": 0.362,
+      "step": 558
+    },
+    {
+      "epoch": 1.6920166477487704,
+      "grad_norm": 0.5774796506875564,
+      "learning_rate": 1.427405678599164e-05,
+      "loss": 0.3527,
+      "step": 559
+    },
+    {
+      "epoch": 1.695043511161559,
+      "grad_norm": 0.5589313872026416,
+      "learning_rate": 1.4250134562400301e-05,
+      "loss": 0.3393,
+      "step": 560
+    },
+    {
+      "epoch": 1.6980703745743473,
+      "grad_norm": 0.602449588932781,
+      "learning_rate": 1.4226182617406996e-05,
+      "loss": 0.3545,
+      "step": 561
+    },
+    {
+      "epoch": 1.7010972379871359,
+      "grad_norm": 0.562742649180259,
+      "learning_rate": 1.4202201118508863e-05,
+      "loss": 0.3653,
+      "step": 562
+    },
+    {
+      "epoch": 1.7041241013999242,
+      "grad_norm": 0.5577933221351792,
+      "learning_rate": 1.4178190233409717e-05,
+      "loss": 0.3536,
+      "step": 563
+    },
+    {
+      "epoch": 1.7071509648127128,
+      "grad_norm": 0.5400978551242521,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.3512,
+      "step": 564
+    },
+    {
+      "epoch": 1.7101778282255014,
+      "grad_norm": 0.537954419668241,
+      "learning_rate": 1.413008097644995e-05,
+      "loss": 0.3557,
+      "step": 565
+    },
+    {
+      "epoch": 1.71320469163829,
+      "grad_norm": 0.5237209880691893,
+      "learning_rate": 1.4105982941019751e-05,
+      "loss": 0.3592,
+      "step": 566
+    },
+    {
+      "epoch": 1.7162315550510783,
+      "grad_norm": 0.5483245176497303,
+      "learning_rate": 1.4081856192247032e-05,
+      "loss": 0.3835,
+      "step": 567
+    },
+    {
+      "epoch": 1.7192584184638668,
+      "grad_norm": 0.504606326543202,
+      "learning_rate": 1.405770089885134e-05,
+      "loss": 0.3431,
+      "step": 568
+    },
+    {
+      "epoch": 1.7222852818766552,
+      "grad_norm": 0.5198895416151927,
+      "learning_rate": 1.4033517229751843e-05,
+      "loss": 0.351,
+      "step": 569
+    },
+    {
+      "epoch": 1.7253121452894438,
+      "grad_norm": 0.5810656350330754,
+      "learning_rate": 1.4009305354066138e-05,
+      "loss": 0.3787,
+      "step": 570
+    },
+    {
+      "epoch": 1.7283390087022323,
+      "grad_norm": 0.5146397531139112,
+      "learning_rate": 1.3985065441109075e-05,
+      "loss": 0.3505,
+      "step": 571
+    },
+    {
+      "epoch": 1.731365872115021,
+      "grad_norm": 0.559962641056386,
+      "learning_rate": 1.396079766039157e-05,
+      "loss": 0.3673,
+      "step": 572
+    },
+    {
+      "epoch": 1.7343927355278093,
+      "grad_norm": 0.5196852177966552,
+      "learning_rate": 1.3936502181619415e-05,
+      "loss": 0.3695,
+      "step": 573
+    },
+    {
+      "epoch": 1.7374195989405978,
+      "grad_norm": 0.5801575160019491,
+      "learning_rate": 1.39121791746921e-05,
+      "loss": 0.3642,
+      "step": 574
+    },
+    {
+      "epoch": 1.7404464623533862,
+      "grad_norm": 0.560082970104919,
+      "learning_rate": 1.3887828809701623e-05,
+      "loss": 0.3493,
+      "step": 575
+    },
+    {
+      "epoch": 1.7434733257661748,
+      "grad_norm": 0.5364206961809506,
+      "learning_rate": 1.3863451256931286e-05,
+      "loss": 0.3654,
+      "step": 576
+    },
+    {
+      "epoch": 1.7465001891789633,
+      "grad_norm": 0.5887689320499502,
+      "learning_rate": 1.3839046686854538e-05,
+      "loss": 0.3751,
+      "step": 577
+    },
+    {
+      "epoch": 1.749527052591752,
+      "grad_norm": 0.5269594428000105,
+      "learning_rate": 1.381461527013374e-05,
+      "loss": 0.3552,
+      "step": 578
+    },
+    {
+      "epoch": 1.7525539160045402,
+      "grad_norm": 0.5542219938157557,
+      "learning_rate": 1.3790157177619005e-05,
+      "loss": 0.3669,
+      "step": 579
+    },
+    {
+      "epoch": 1.7555807794173288,
+      "grad_norm": 0.5681663817375245,
+      "learning_rate": 1.3765672580346986e-05,
+      "loss": 0.3613,
+      "step": 580
+    },
+    {
+      "epoch": 1.7586076428301172,
+      "grad_norm": 0.5394894735420986,
+      "learning_rate": 1.3741161649539695e-05,
+      "loss": 0.3511,
+      "step": 581
+    },
+    {
+      "epoch": 1.7616345062429057,
+      "grad_norm": 0.5855273371961216,
+      "learning_rate": 1.3716624556603275e-05,
+      "loss": 0.369,
+      "step": 582
+    },
+    {
+      "epoch": 1.7646613696556943,
+      "grad_norm": 0.5688954616840206,
+      "learning_rate": 1.3692061473126845e-05,
+      "loss": 0.3346,
+      "step": 583
+    },
+    {
+      "epoch": 1.7676882330684829,
+      "grad_norm": 0.525902938604159,
+      "learning_rate": 1.3667472570881264e-05,
+      "loss": 0.3229,
+      "step": 584
+    },
+    {
+      "epoch": 1.7707150964812712,
+      "grad_norm": 0.5574968601299607,
+      "learning_rate": 1.3642858021817944e-05,
+      "loss": 0.3594,
+      "step": 585
+    },
+    {
+      "epoch": 1.7737419598940598,
+      "grad_norm": 0.5531340236532221,
+      "learning_rate": 1.361821799806765e-05,
+      "loss": 0.3591,
+      "step": 586
+    },
+    {
+      "epoch": 1.7767688233068482,
+      "grad_norm": 0.5481878462749159,
+      "learning_rate": 1.3593552671939292e-05,
+      "loss": 0.3342,
+      "step": 587
+    },
+    {
+      "epoch": 1.7797956867196367,
+      "grad_norm": 0.5534693487111089,
+      "learning_rate": 1.356886221591872e-05,
+      "loss": 0.3566,
+      "step": 588
+    },
+    {
+      "epoch": 1.7828225501324253,
+      "grad_norm": 0.5871746863618214,
+      "learning_rate": 1.3544146802667517e-05,
+      "loss": 0.3699,
+      "step": 589
+    },
+    {
+      "epoch": 1.7858494135452139,
+      "grad_norm": 0.5422828637115944,
+      "learning_rate": 1.3519406605021797e-05,
+      "loss": 0.3337,
+      "step": 590
+    },
+    {
+      "epoch": 1.7888762769580022,
+      "grad_norm": 0.5509607560972102,
+      "learning_rate": 1.3494641795990986e-05,
+      "loss": 0.3622,
+      "step": 591
+    },
+    {
+      "epoch": 1.7919031403707908,
+      "grad_norm": 0.5673643427306578,
+      "learning_rate": 1.3469852548756626e-05,
+      "loss": 0.3486,
+      "step": 592
+    },
+    {
+      "epoch": 1.7949300037835791,
+      "grad_norm": 0.5367488345947885,
+      "learning_rate": 1.3445039036671152e-05,
+      "loss": 0.3788,
+      "step": 593
+    },
+    {
+      "epoch": 1.7979568671963677,
+      "grad_norm": 0.4947022963500923,
+      "learning_rate": 1.342020143325669e-05,
+      "loss": 0.334,
+      "step": 594
+    },
+    {
+      "epoch": 1.8009837306091563,
+      "grad_norm": 0.5869945444476703,
+      "learning_rate": 1.3395339912203829e-05,
+      "loss": 0.3333,
+      "step": 595
+    },
+    {
+      "epoch": 1.8040105940219449,
+      "grad_norm": 0.5523256147360104,
+      "learning_rate": 1.3370454647370418e-05,
+      "loss": 0.3434,
+      "step": 596
+    },
+    {
+      "epoch": 1.8070374574347332,
+      "grad_norm": 0.5477503696317212,
+      "learning_rate": 1.3345545812780354e-05,
+      "loss": 0.364,
+      "step": 597
+    },
+    {
+      "epoch": 1.8100643208475218,
+      "grad_norm": 0.5627603496232576,
+      "learning_rate": 1.3320613582622354e-05,
+      "loss": 0.345,
+      "step": 598
+    },
+    {
+      "epoch": 1.8130911842603101,
+      "grad_norm": 0.5365668286676434,
+      "learning_rate": 1.3295658131248737e-05,
+      "loss": 0.3425,
+      "step": 599
+    },
+    {
+      "epoch": 1.8161180476730987,
+      "grad_norm": 0.5672066665757356,
+      "learning_rate": 1.3270679633174219e-05,
+      "loss": 0.3721,
+      "step": 600
+    },
+    {
+      "epoch": 1.8191449110858873,
+      "grad_norm": 0.5587104092301314,
+      "learning_rate": 1.3245678263074668e-05,
+      "loss": 0.3752,
+      "step": 601
+    },
+    {
+      "epoch": 1.8221717744986758,
+      "grad_norm": 0.5427828991636513,
+      "learning_rate": 1.3220654195785917e-05,
+      "loss": 0.3619,
+      "step": 602
+    },
+    {
+      "epoch": 1.8251986379114642,
+      "grad_norm": 0.5926680708544364,
+      "learning_rate": 1.3195607606302501e-05,
+      "loss": 0.3669,
+      "step": 603
+    },
+    {
+      "epoch": 1.8282255013242528,
+      "grad_norm": 0.5583834828460088,
+      "learning_rate": 1.3170538669776469e-05,
+      "loss": 0.351,
+      "step": 604
+    },
+    {
+      "epoch": 1.8312523647370411,
+      "grad_norm": 0.6042641791122267,
+      "learning_rate": 1.3145447561516138e-05,
+      "loss": 0.3783,
+      "step": 605
+    },
+    {
+      "epoch": 1.8342792281498297,
+      "grad_norm": 0.5483791322843222,
+      "learning_rate": 1.3120334456984871e-05,
+      "loss": 0.3564,
+      "step": 606
+    },
+    {
+      "epoch": 1.8373060915626183,
+      "grad_norm": 0.5644563916406111,
+      "learning_rate": 1.3095199531799858e-05,
+      "loss": 0.3512,
+      "step": 607
+    },
+    {
+      "epoch": 1.8403329549754068,
+      "grad_norm": 0.5401983457525021,
+      "learning_rate": 1.3070042961730878e-05,
+      "loss": 0.3476,
+      "step": 608
+    },
+    {
+      "epoch": 1.8433598183881952,
+      "grad_norm": 0.5535106881078694,
+      "learning_rate": 1.3044864922699072e-05,
+      "loss": 0.3426,
+      "step": 609
+    },
+    {
+      "epoch": 1.8463866818009838,
+      "grad_norm": 0.5425844969006997,
+      "learning_rate": 1.3019665590775717e-05,
+      "loss": 0.3663,
+      "step": 610
+    },
+    {
+      "epoch": 1.849413545213772,
+      "grad_norm": 0.554778966858946,
+      "learning_rate": 1.2994445142180994e-05,
+      "loss": 0.3422,
+      "step": 611
+    },
+    {
+      "epoch": 1.8524404086265607,
+      "grad_norm": 0.5672637652870383,
+      "learning_rate": 1.296920375328275e-05,
+      "loss": 0.3476,
+      "step": 612
+    },
+    {
+      "epoch": 1.8554672720393492,
+      "grad_norm": 0.5425083320655277,
+      "learning_rate": 1.2943941600595271e-05,
+      "loss": 0.358,
+      "step": 613
+    },
+    {
+      "epoch": 1.8584941354521378,
+      "grad_norm": 0.5897125776599859,
+      "learning_rate": 1.2918658860778046e-05,
+      "loss": 0.3577,
+      "step": 614
+    },
+    {
+      "epoch": 1.8615209988649262,
+      "grad_norm": 0.594655229875934,
+      "learning_rate": 1.289335571063453e-05,
+      "loss": 0.3719,
+      "step": 615
+    },
+    {
+      "epoch": 1.8645478622777147,
+      "grad_norm": 0.5193747791720533,
+      "learning_rate": 1.2868032327110904e-05,
+      "loss": 0.3329,
+      "step": 616
+    },
+    {
+      "epoch": 1.867574725690503,
+      "grad_norm": 0.5938203877570072,
+      "learning_rate": 1.284268888729485e-05,
+      "loss": 0.3474,
+      "step": 617
+    },
+    {
+      "epoch": 1.8706015891032917,
+      "grad_norm": 0.5519595612997348,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.3393,
+      "step": 618
+    },
+    {
+      "epoch": 1.8736284525160802,
+      "grad_norm": 0.5298931896363447,
+      "learning_rate": 1.27919425478362e-05,
+      "loss": 0.3803,
+      "step": 619
+    },
+    {
+      "epoch": 1.8766553159288688,
+      "grad_norm": 0.5880690022665852,
+      "learning_rate": 1.2766540003065272e-05,
+      "loss": 0.387,
+      "step": 620
+    },
+    {
+      "epoch": 1.8796821793416572,
+      "grad_norm": 0.5426952990442959,
+      "learning_rate": 1.2741118111742778e-05,
+      "loss": 0.3527,
+      "step": 621
+    },
+    {
+      "epoch": 1.8827090427544457,
+      "grad_norm": 0.5326160357905958,
+      "learning_rate": 1.2715677051645259e-05,
+      "loss": 0.3633,
+      "step": 622
+    },
+    {
+      "epoch": 1.885735906167234,
+      "grad_norm": 0.5501142788597894,
+      "learning_rate": 1.2690217000683321e-05,
+      "loss": 0.3115,
+      "step": 623
+    },
+    {
+      "epoch": 1.8887627695800226,
+      "grad_norm": 0.5296356398431183,
+      "learning_rate": 1.266473813690035e-05,
+      "loss": 0.3526,
+      "step": 624
+    },
+    {
+      "epoch": 1.8917896329928112,
+      "grad_norm": 0.5610313935163528,
+      "learning_rate": 1.2639240638471317e-05,
+      "loss": 0.3521,
+      "step": 625
+    },
+    {
+      "epoch": 1.8948164964055998,
+      "grad_norm": 0.5562895305058339,
+      "learning_rate": 1.2613724683701491e-05,
+      "loss": 0.3469,
+      "step": 626
+    },
+    {
+      "epoch": 1.8978433598183881,
+      "grad_norm": 0.5126278056308082,
+      "learning_rate": 1.2588190451025209e-05,
+      "loss": 0.3349,
+      "step": 627
+    },
+    {
+      "epoch": 1.9008702232311767,
+      "grad_norm": 0.5269548739825417,
+      "learning_rate": 1.2562638119004627e-05,
+      "loss": 0.3631,
+      "step": 628
+    },
+    {
+      "epoch": 1.903897086643965,
+      "grad_norm": 0.5131891961710551,
+      "learning_rate": 1.2537067866328473e-05,
+      "loss": 0.3569,
+      "step": 629
+    },
+    {
+      "epoch": 1.9069239500567536,
+      "grad_norm": 0.5149892973056653,
+      "learning_rate": 1.2511479871810792e-05,
+      "loss": 0.3328,
+      "step": 630
+    },
+    {
+      "epoch": 1.9099508134695422,
+      "grad_norm": 0.5108906078949854,
+      "learning_rate": 1.2485874314389701e-05,
+      "loss": 0.3602,
+      "step": 631
+    },
+    {
+      "epoch": 1.9129776768823308,
+      "grad_norm": 0.5534490746393853,
+      "learning_rate": 1.2460251373126136e-05,
+      "loss": 0.3798,
+      "step": 632
+    },
+    {
+      "epoch": 1.9160045402951191,
+      "grad_norm": 0.5531637583636305,
+      "learning_rate": 1.2434611227202591e-05,
+      "loss": 0.3696,
+      "step": 633
+    },
+    {
+      "epoch": 1.9190314037079077,
+      "grad_norm": 0.5354578566389409,
+      "learning_rate": 1.2408954055921884e-05,
+      "loss": 0.3795,
+      "step": 634
+    },
+    {
+      "epoch": 1.922058267120696,
+      "grad_norm": 0.5115276756933597,
+      "learning_rate": 1.2383280038705885e-05,
+      "loss": 0.3286,
+      "step": 635
+    },
+    {
+      "epoch": 1.9250851305334846,
+      "grad_norm": 0.5377927715271165,
+      "learning_rate": 1.2357589355094275e-05,
+      "loss": 0.3579,
+      "step": 636
+    },
+    {
+      "epoch": 1.9281119939462732,
+      "grad_norm": 0.5578527103755802,
+      "learning_rate": 1.2331882184743273e-05,
+      "loss": 0.3616,
+      "step": 637
+    },
+    {
+      "epoch": 1.9311388573590618,
+      "grad_norm": 0.5474622331847911,
+      "learning_rate": 1.2306158707424402e-05,
+      "loss": 0.3357,
+      "step": 638
+    },
+    {
+      "epoch": 1.9341657207718501,
+      "grad_norm": 0.5025358294807459,
+      "learning_rate": 1.2280419103023219e-05,
+      "loss": 0.3476,
+      "step": 639
+    },
+    {
+      "epoch": 1.9371925841846387,
+      "grad_norm": 0.5473473724494916,
+      "learning_rate": 1.2254663551538047e-05,
+      "loss": 0.3348,
+      "step": 640
+    },
+    {
+      "epoch": 1.940219447597427,
+      "grad_norm": 0.5465379000334961,
+      "learning_rate": 1.2228892233078738e-05,
+      "loss": 0.347,
+      "step": 641
+    },
+    {
+      "epoch": 1.9432463110102156,
+      "grad_norm": 0.5352788134461759,
+      "learning_rate": 1.2203105327865407e-05,
+      "loss": 0.3748,
+      "step": 642
+    },
+    {
+      "epoch": 1.9462731744230042,
+      "grad_norm": 0.525633414900016,
+      "learning_rate": 1.2177303016227158e-05,
+      "loss": 0.3447,
+      "step": 643
+    },
+    {
+      "epoch": 1.9493000378357928,
+      "grad_norm": 0.5223985747950463,
+      "learning_rate": 1.215148547860084e-05,
+      "loss": 0.3377,
+      "step": 644
+    },
+    {
+      "epoch": 1.952326901248581,
+      "grad_norm": 0.5287177207739253,
+      "learning_rate": 1.2125652895529766e-05,
+      "loss": 0.3597,
+      "step": 645
+    },
+    {
+      "epoch": 1.9553537646613697,
+      "grad_norm": 0.5337120855691064,
+      "learning_rate": 1.2099805447662485e-05,
+      "loss": 0.3381,
+      "step": 646
+    },
+    {
+      "epoch": 1.958380628074158,
+      "grad_norm": 0.5569034169274,
+      "learning_rate": 1.2073943315751473e-05,
+      "loss": 0.3413,
+      "step": 647
+    },
+    {
+      "epoch": 1.9614074914869466,
+      "grad_norm": 0.5523868050821549,
+      "learning_rate": 1.2048066680651908e-05,
+      "loss": 0.3502,
+      "step": 648
+    },
+    {
+      "epoch": 1.9644343548997352,
+      "grad_norm": 0.5509328816855648,
+      "learning_rate": 1.2022175723320382e-05,
+      "loss": 0.3794,
+      "step": 649
+    },
+    {
+      "epoch": 1.9674612183125237,
+      "grad_norm": 0.541806199844615,
+      "learning_rate": 1.1996270624813642e-05,
+      "loss": 0.3479,
+      "step": 650
+    },
+    {
+      "epoch": 1.970488081725312,
+      "grad_norm": 0.5382748986625098,
+      "learning_rate": 1.1970351566287332e-05,
+      "loss": 0.3792,
+      "step": 651
+    },
+    {
+      "epoch": 1.9735149451381007,
+      "grad_norm": 0.5588634307219937,
+      "learning_rate": 1.194441872899471e-05,
+      "loss": 0.3487,
+      "step": 652
+    },
+    {
+      "epoch": 1.976541808550889,
+      "grad_norm": 0.548371003399634,
+      "learning_rate": 1.1918472294285402e-05,
+      "loss": 0.3635,
+      "step": 653
+    },
+    {
+      "epoch": 1.9795686719636776,
+      "grad_norm": 0.5919931177059518,
+      "learning_rate": 1.1892512443604103e-05,
+      "loss": 0.3659,
+      "step": 654
+    },
+    {
+      "epoch": 1.9825955353764662,
+      "grad_norm": 0.5794515147958162,
+      "learning_rate": 1.1866539358489347e-05,
+      "loss": 0.3478,
+      "step": 655
+    },
+    {
+      "epoch": 1.9856223987892547,
+      "grad_norm": 0.512653411733838,
+      "learning_rate": 1.1840553220572204e-05,
+      "loss": 0.372,
+      "step": 656
+    },
+    {
+      "epoch": 1.988649262202043,
+      "grad_norm": 0.5746638468912015,
+      "learning_rate": 1.1814554211575026e-05,
+      "loss": 0.3601,
+      "step": 657
+    },
+    {
+      "epoch": 1.9916761256148316,
+      "grad_norm": 0.6071984014948465,
+      "learning_rate": 1.1788542513310178e-05,
+      "loss": 0.3429,
+      "step": 658
+    },
+    {
+      "epoch": 1.99470298902762,
+      "grad_norm": 0.5505791784140044,
+      "learning_rate": 1.1762518307678747e-05,
+      "loss": 0.3562,
+      "step": 659
+    },
+    {
+      "epoch": 1.9977298524404086,
+      "grad_norm": 0.5682928157267071,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 0.3361,
+      "step": 660
+    },
+    {
+      "epoch": 2.000756715853197,
+      "grad_norm": 0.5743804531537003,
+      "learning_rate": 1.17104331023566e-05,
+      "loss": 0.3061,
+      "step": 661
+    },
+    {
+      "epoch": 2.0037835792659857,
+      "grad_norm": 0.9411175521819087,
+      "learning_rate": 1.1684372466900306e-05,
+      "loss": 0.2705,
+      "step": 662
+    },
+    {
+      "epoch": 2.0068104426787743,
+      "grad_norm": 0.7393149745242747,
+      "learning_rate": 1.1658300052543742e-05,
+      "loss": 0.2643,
+      "step": 663
+    },
+    {
+      "epoch": 2.0098373060915624,
+      "grad_norm": 0.8739081262162334,
+      "learning_rate": 1.1632216041612595e-05,
+      "loss": 0.239,
+      "step": 664
+    },
+    {
+      "epoch": 2.012864169504351,
+      "grad_norm": 0.83690011937627,
+      "learning_rate": 1.1606120616513648e-05,
+      "loss": 0.241,
+      "step": 665
+    },
+    {
+      "epoch": 2.0158910329171396,
+      "grad_norm": 0.7161168271338201,
+      "learning_rate": 1.15800139597335e-05,
+      "loss": 0.2318,
+      "step": 666
+    },
+    {
+      "epoch": 2.018917896329928,
+      "grad_norm": 0.9743027494790029,
+      "learning_rate": 1.1553896253837304e-05,
+      "loss": 0.2442,
+      "step": 667
+    },
+    {
+      "epoch": 2.0219447597427167,
+      "grad_norm": 0.8704398868867053,
+      "learning_rate": 1.1527767681467472e-05,
+      "loss": 0.2704,
+      "step": 668
+    },
+    {
+      "epoch": 2.0249716231555053,
+      "grad_norm": 0.9400400942149604,
+      "learning_rate": 1.1501628425342404e-05,
+      "loss": 0.2331,
+      "step": 669
+    },
+    {
+      "epoch": 2.0279984865682934,
+      "grad_norm": 0.8085127043741395,
+      "learning_rate": 1.1475478668255223e-05,
+      "loss": 0.2214,
+      "step": 670
+    },
+    {
+      "epoch": 2.031025349981082,
+      "grad_norm": 0.7935206156867401,
+      "learning_rate": 1.1449318593072468e-05,
+      "loss": 0.2617,
+      "step": 671
+    },
+    {
+      "epoch": 2.0340522133938705,
+      "grad_norm": 0.7693189199180338,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.231,
+      "step": 672
+    },
+    {
+      "epoch": 2.037079076806659,
+      "grad_norm": 0.6921768189726245,
+      "learning_rate": 1.1396968220245953e-05,
+      "loss": 0.2275,
+      "step": 673
+    },
+    {
+      "epoch": 2.0401059402194477,
+      "grad_norm": 0.6873376181245889,
+      "learning_rate": 1.1370778288690947e-05,
+      "loss": 0.2402,
+      "step": 674
+    },
+    {
+      "epoch": 2.0431328036322363,
+      "grad_norm": 0.6317927272457231,
+      "learning_rate": 1.1344578771215319e-05,
+      "loss": 0.2289,
+      "step": 675
+    },
+    {
+      "epoch": 2.0461596670450244,
+      "grad_norm": 0.690846595640899,
+      "learning_rate": 1.1318369851033604e-05,
+      "loss": 0.2437,
+      "step": 676
+    },
+    {
+      "epoch": 2.049186530457813,
+      "grad_norm": 0.7161875840319474,
+      "learning_rate": 1.1292151711426079e-05,
+      "loss": 0.2442,
+      "step": 677
+    },
+    {
+      "epoch": 2.0522133938706015,
+      "grad_norm": 0.7287484041742311,
+      "learning_rate": 1.1265924535737494e-05,
+      "loss": 0.2362,
+      "step": 678
+    },
+    {
+      "epoch": 2.05524025728339,
+      "grad_norm": 0.6146136158095936,
+      "learning_rate": 1.1239688507375792e-05,
+      "loss": 0.2322,
+      "step": 679
+    },
+    {
+      "epoch": 2.0582671206961787,
+      "grad_norm": 0.7293317218333972,
+      "learning_rate": 1.121344380981082e-05,
+      "loss": 0.2203,
+      "step": 680
+    },
+    {
+      "epoch": 2.0612939841089672,
+      "grad_norm": 0.7806299992827529,
+      "learning_rate": 1.1187190626573052e-05,
+      "loss": 0.2533,
+      "step": 681
+    },
+    {
+      "epoch": 2.0643208475217554,
+      "grad_norm": 0.7206366890445451,
+      "learning_rate": 1.1160929141252303e-05,
+      "loss": 0.2484,
+      "step": 682
+    },
+    {
+      "epoch": 2.067347710934544,
+      "grad_norm": 0.7137496714424608,
+      "learning_rate": 1.1134659537496442e-05,
+      "loss": 0.2589,
+      "step": 683
+    },
+    {
+      "epoch": 2.0703745743473325,
+      "grad_norm": 0.6815888975730414,
+      "learning_rate": 1.1108381999010111e-05,
+      "loss": 0.2239,
+      "step": 684
+    },
+    {
+      "epoch": 2.073401437760121,
+      "grad_norm": 0.8117150308340428,
+      "learning_rate": 1.1082096709553443e-05,
+      "loss": 0.24,
+      "step": 685
+    },
+    {
+      "epoch": 2.0764283011729097,
+      "grad_norm": 0.6780304137554947,
+      "learning_rate": 1.1055803852940772e-05,
+      "loss": 0.2389,
+      "step": 686
+    },
+    {
+      "epoch": 2.0794551645856982,
+      "grad_norm": 0.7152806331473901,
+      "learning_rate": 1.1029503613039347e-05,
+      "loss": 0.2345,
+      "step": 687
+    },
+    {
+      "epoch": 2.0824820279984864,
+      "grad_norm": 0.7096265182902401,
+      "learning_rate": 1.1003196173768051e-05,
+      "loss": 0.2361,
+      "step": 688
+    },
+    {
+      "epoch": 2.085508891411275,
+      "grad_norm": 0.6452200254272257,
+      "learning_rate": 1.0976881719096118e-05,
+      "loss": 0.2178,
+      "step": 689
+    },
+    {
+      "epoch": 2.0885357548240635,
+      "grad_norm": 0.6883827345412773,
+      "learning_rate": 1.0950560433041825e-05,
+      "loss": 0.2536,
+      "step": 690
+    },
+    {
+      "epoch": 2.091562618236852,
+      "grad_norm": 0.6922008351400569,
+      "learning_rate": 1.0924232499671248e-05,
+      "loss": 0.2523,
+      "step": 691
+    },
+    {
+      "epoch": 2.0945894816496406,
+      "grad_norm": 0.683022218826114,
+      "learning_rate": 1.0897898103096917e-05,
+      "loss": 0.2564,
+      "step": 692
+    },
+    {
+      "epoch": 2.097616345062429,
+      "grad_norm": 0.6566532647982883,
+      "learning_rate": 1.0871557427476585e-05,
+      "loss": 0.2333,
+      "step": 693
+    },
+    {
+      "epoch": 2.1006432084752173,
+      "grad_norm": 0.6377470954365588,
+      "learning_rate": 1.0845210657011893e-05,
+      "loss": 0.2332,
+      "step": 694
+    },
+    {
+      "epoch": 2.103670071888006,
+      "grad_norm": 0.6648773864559259,
+      "learning_rate": 1.0818857975947128e-05,
+      "loss": 0.2425,
+      "step": 695
+    },
+    {
+      "epoch": 2.1066969353007945,
+      "grad_norm": 0.7024793142813248,
+      "learning_rate": 1.0792499568567885e-05,
+      "loss": 0.2329,
+      "step": 696
+    },
+    {
+      "epoch": 2.109723798713583,
+      "grad_norm": 0.6185118652650472,
+      "learning_rate": 1.0766135619199825e-05,
+      "loss": 0.2147,
+      "step": 697
+    },
+    {
+      "epoch": 2.1127506621263716,
+      "grad_norm": 0.6587459506063513,
+      "learning_rate": 1.0739766312207344e-05,
+      "loss": 0.2342,
+      "step": 698
+    },
+    {
+      "epoch": 2.11577752553916,
+      "grad_norm": 0.6730890718539143,
+      "learning_rate": 1.0713391831992324e-05,
+      "loss": 0.2429,
+      "step": 699
+    },
+    {
+      "epoch": 2.1188043889519483,
+      "grad_norm": 0.6100826095967132,
+      "learning_rate": 1.068701236299281e-05,
+      "loss": 0.2336,
+      "step": 700
+    },
+    {
+      "epoch": 2.121831252364737,
+      "grad_norm": 0.6506458622519254,
+      "learning_rate": 1.0660628089681742e-05,
+      "loss": 0.2173,
+      "step": 701
+    },
+    {
+      "epoch": 2.1248581157775255,
+      "grad_norm": 0.6738275135205595,
+      "learning_rate": 1.0634239196565646e-05,
+      "loss": 0.2603,
+      "step": 702
+    },
+    {
+      "epoch": 2.127884979190314,
+      "grad_norm": 0.670430192080833,
+      "learning_rate": 1.0607845868183368e-05,
+      "loss": 0.2294,
+      "step": 703
+    },
+    {
+      "epoch": 2.1309118426031026,
+      "grad_norm": 0.6632733981548405,
+      "learning_rate": 1.0581448289104759e-05,
+      "loss": 0.2444,
+      "step": 704
+    },
+    {
+      "epoch": 2.133938706015891,
+      "grad_norm": 0.626553035386687,
+      "learning_rate": 1.0555046643929402e-05,
+      "loss": 0.2311,
+      "step": 705
+    },
+    {
+      "epoch": 2.1369655694286793,
+      "grad_norm": 0.6847474693533315,
+      "learning_rate": 1.0528641117285315e-05,
+      "loss": 0.2449,
+      "step": 706
+    },
+    {
+      "epoch": 2.139992432841468,
+      "grad_norm": 0.6390707124891285,
+      "learning_rate": 1.0502231893827652e-05,
+      "loss": 0.2296,
+      "step": 707
+    },
+    {
+      "epoch": 2.1430192962542565,
+      "grad_norm": 0.7037522034764115,
+      "learning_rate": 1.0475819158237426e-05,
+      "loss": 0.2447,
+      "step": 708
+    },
+    {
+      "epoch": 2.146046159667045,
+      "grad_norm": 0.61452309933875,
+      "learning_rate": 1.0449403095220205e-05,
+      "loss": 0.2155,
+      "step": 709
+    },
+    {
+      "epoch": 2.1490730230798336,
+      "grad_norm": 0.6485736468339458,
+      "learning_rate": 1.0422983889504831e-05,
+      "loss": 0.245,
+      "step": 710
+    },
+    {
+      "epoch": 2.152099886492622,
+      "grad_norm": 0.6314496923055156,
+      "learning_rate": 1.0396561725842124e-05,
+      "loss": 0.2261,
+      "step": 711
+    },
+    {
+      "epoch": 2.1551267499054103,
+      "grad_norm": 0.6512071371719522,
+      "learning_rate": 1.0370136789003582e-05,
+      "loss": 0.2521,
+      "step": 712
+    },
+    {
+      "epoch": 2.158153613318199,
+      "grad_norm": 0.6681180332880192,
+      "learning_rate": 1.0343709263780099e-05,
+      "loss": 0.2423,
+      "step": 713
+    },
+    {
+      "epoch": 2.1611804767309875,
+      "grad_norm": 0.6223035706838734,
+      "learning_rate": 1.031727933498068e-05,
+      "loss": 0.2248,
+      "step": 714
+    },
+    {
+      "epoch": 2.164207340143776,
+      "grad_norm": 0.6339006443573365,
+      "learning_rate": 1.0290847187431115e-05,
+      "loss": 0.2371,
+      "step": 715
+    },
+    {
+      "epoch": 2.1672342035565646,
+      "grad_norm": 0.6618895059728317,
+      "learning_rate": 1.0264413005972736e-05,
+      "loss": 0.2379,
+      "step": 716
+    },
+    {
+      "epoch": 2.170261066969353,
+      "grad_norm": 0.6312966657056531,
+      "learning_rate": 1.0237976975461074e-05,
+      "loss": 0.2472,
+      "step": 717
+    },
+    {
+      "epoch": 2.1732879303821413,
+      "grad_norm": 0.6653647067075577,
+      "learning_rate": 1.0211539280764617e-05,
+      "loss": 0.2358,
+      "step": 718
+    },
+    {
+      "epoch": 2.17631479379493,
+      "grad_norm": 0.6458555380195727,
+      "learning_rate": 1.0185100106763467e-05,
+      "loss": 0.2295,
+      "step": 719
+    },
+    {
+      "epoch": 2.1793416572077184,
+      "grad_norm": 0.6181439151350354,
+      "learning_rate": 1.015865963834808e-05,
+      "loss": 0.2337,
+      "step": 720
+    },
+    {
+      "epoch": 2.182368520620507,
+      "grad_norm": 0.6628645040509464,
+      "learning_rate": 1.013221806041797e-05,
+      "loss": 0.244,
+      "step": 721
+    },
+    {
+      "epoch": 2.1853953840332956,
+      "grad_norm": 0.6253484293076698,
+      "learning_rate": 1.0105775557880398e-05,
+      "loss": 0.2442,
+      "step": 722
+    },
+    {
+      "epoch": 2.188422247446084,
+      "grad_norm": 0.6125457489995962,
+      "learning_rate": 1.0079332315649097e-05,
+      "loss": 0.2354,
+      "step": 723
+    },
+    {
+      "epoch": 2.1914491108588727,
+      "grad_norm": 0.6255416116212591,
+      "learning_rate": 1.0052888518642978e-05,
+      "loss": 0.2374,
+      "step": 724
+    },
+    {
+      "epoch": 2.194475974271661,
+      "grad_norm": 0.6540579724296892,
+      "learning_rate": 1.0026444351784823e-05,
+      "loss": 0.2395,
+      "step": 725
+    },
+    {
+      "epoch": 2.1975028376844494,
+      "grad_norm": 0.5990636510131816,
+      "learning_rate": 1e-05,
+      "loss": 0.2268,
+      "step": 726
+    },
+    {
+      "epoch": 2.200529701097238,
+      "grad_norm": 0.649146717831707,
+      "learning_rate": 9.97355564821518e-06,
+      "loss": 0.2361,
+      "step": 727
+    },
+    {
+      "epoch": 2.2035565645100266,
+      "grad_norm": 0.679449603268564,
+      "learning_rate": 9.947111481357023e-06,
+      "loss": 0.2448,
+      "step": 728
+    },
+    {
+      "epoch": 2.206583427922815,
+      "grad_norm": 0.6794228110603918,
+      "learning_rate": 9.920667684350906e-06,
+      "loss": 0.2542,
+      "step": 729
+    },
+    {
+      "epoch": 2.2096102913356033,
+      "grad_norm": 0.6141101765615307,
+      "learning_rate": 9.894224442119606e-06,
+      "loss": 0.2355,
+      "step": 730
+    },
+    {
+      "epoch": 2.212637154748392,
+      "grad_norm": 0.6276375406303295,
+      "learning_rate": 9.867781939582036e-06,
+      "loss": 0.2378,
+      "step": 731
+    },
+    {
+      "epoch": 2.2156640181611804,
+      "grad_norm": 0.6436260440535428,
+      "learning_rate": 9.841340361651921e-06,
+      "loss": 0.2418,
+      "step": 732
+    },
+    {
+      "epoch": 2.218690881573969,
+      "grad_norm": 0.6763482840571772,
+      "learning_rate": 9.814899893236537e-06,
+      "loss": 0.2532,
+      "step": 733
+    },
+    {
+      "epoch": 2.2217177449867576,
+      "grad_norm": 0.6255680080401044,
+      "learning_rate": 9.788460719235386e-06,
+      "loss": 0.2272,
+      "step": 734
+    },
+    {
+      "epoch": 2.224744608399546,
+      "grad_norm": 0.6327076863954073,
+      "learning_rate": 9.762023024538928e-06,
+      "loss": 0.2364,
+      "step": 735
+    },
+    {
+      "epoch": 2.2277714718123347,
+      "grad_norm": 0.6692427118258562,
+      "learning_rate": 9.735586994027267e-06,
+      "loss": 0.2567,
+      "step": 736
+    },
+    {
+      "epoch": 2.230798335225123,
+      "grad_norm": 0.6893260287286201,
+      "learning_rate": 9.709152812568886e-06,
+      "loss": 0.2572,
+      "step": 737
+    },
+    {
+      "epoch": 2.2338251986379114,
+      "grad_norm": 0.6279345549187322,
+      "learning_rate": 9.682720665019325e-06,
+      "loss": 0.24,
+      "step": 738
+    },
+    {
+      "epoch": 2.2368520620507,
+      "grad_norm": 0.6562075891192821,
+      "learning_rate": 9.656290736219903e-06,
+      "loss": 0.2458,
+      "step": 739
+    },
+    {
+      "epoch": 2.2398789254634885,
+      "grad_norm": 0.6331349597994242,
+      "learning_rate": 9.62986321099642e-06,
+      "loss": 0.2195,
+      "step": 740
+    },
+    {
+      "epoch": 2.242905788876277,
+      "grad_norm": 0.6635481559030685,
+      "learning_rate": 9.603438274157878e-06,
+      "loss": 0.2605,
+      "step": 741
+    },
+    {
+      "epoch": 2.2459326522890652,
+      "grad_norm": 0.6189478967049777,
+      "learning_rate": 9.57701611049517e-06,
+      "loss": 0.2367,
+      "step": 742
+    },
+    {
+      "epoch": 2.248959515701854,
+      "grad_norm": 0.6428736919771693,
+      "learning_rate": 9.5505969047798e-06,
+      "loss": 0.2277,
+      "step": 743
+    },
+    {
+      "epoch": 2.2519863791146424,
+      "grad_norm": 0.658527849353508,
+      "learning_rate": 9.524180841762577e-06,
+      "loss": 0.2391,
+      "step": 744
+    },
+    {
+      "epoch": 2.255013242527431,
+      "grad_norm": 0.6448817914710812,
+      "learning_rate": 9.497768106172351e-06,
+      "loss": 0.2272,
+      "step": 745
+    },
+    {
+      "epoch": 2.2580401059402195,
+      "grad_norm": 0.6205919206206887,
+      "learning_rate": 9.471358882714687e-06,
+      "loss": 0.2444,
+      "step": 746
+    },
+    {
+      "epoch": 2.261066969353008,
+      "grad_norm": 0.6307846510682182,
+      "learning_rate": 9.444953356070601e-06,
+      "loss": 0.2346,
+      "step": 747
+    },
+    {
+      "epoch": 2.2640938327657967,
+      "grad_norm": 0.6655015244817076,
+      "learning_rate": 9.418551710895243e-06,
+      "loss": 0.2274,
+      "step": 748
+    },
+    {
+      "epoch": 2.267120696178585,
+      "grad_norm": 0.6524824921764198,
+      "learning_rate": 9.392154131816635e-06,
+      "loss": 0.255,
+      "step": 749
+    },
+    {
+      "epoch": 2.2701475595913734,
+      "grad_norm": 0.6765374248909245,
+      "learning_rate": 9.365760803434356e-06,
+      "loss": 0.2242,
+      "step": 750
+    },
+    {
+      "epoch": 2.273174423004162,
+      "grad_norm": 0.6465302078802556,
+      "learning_rate": 9.33937191031826e-06,
+      "loss": 0.2344,
+      "step": 751
+    },
+    {
+      "epoch": 2.2762012864169505,
+      "grad_norm": 0.650881719023694,
+      "learning_rate": 9.312987637007191e-06,
+      "loss": 0.2118,
+      "step": 752
+    },
+    {
+      "epoch": 2.279228149829739,
+      "grad_norm": 0.6701754451710951,
+      "learning_rate": 9.286608168007678e-06,
+      "loss": 0.2539,
+      "step": 753
+    },
+    {
+      "epoch": 2.282255013242527,
+      "grad_norm": 0.6854639201894199,
+      "learning_rate": 9.260233687792657e-06,
+      "loss": 0.2475,
+      "step": 754
+    },
+    {
+      "epoch": 2.285281876655316,
+      "grad_norm": 0.6537314852340039,
+      "learning_rate": 9.233864380800178e-06,
+      "loss": 0.2414,
+      "step": 755
+    },
+    {
+      "epoch": 2.2883087400681044,
+      "grad_norm": 0.653445818171961,
+      "learning_rate": 9.207500431432115e-06,
+      "loss": 0.2362,
+      "step": 756
+    },
+    {
+      "epoch": 2.291335603480893,
+      "grad_norm": 0.6272508106645552,
+      "learning_rate": 9.181142024052877e-06,
+      "loss": 0.2484,
+      "step": 757
+    },
+    {
+      "epoch": 2.2943624668936815,
+      "grad_norm": 0.6365666938013747,
+      "learning_rate": 9.154789342988108e-06,
+      "loss": 0.2466,
+      "step": 758
+    },
+    {
+      "epoch": 2.29738933030647,
+      "grad_norm": 0.6796164394092712,
+      "learning_rate": 9.128442572523418e-06,
+      "loss": 0.2637,
+      "step": 759
+    },
+    {
+      "epoch": 2.3004161937192587,
+      "grad_norm": 0.6205016338374488,
+      "learning_rate": 9.102101896903084e-06,
+      "loss": 0.2453,
+      "step": 760
+    },
+    {
+      "epoch": 2.303443057132047,
+      "grad_norm": 0.6593880809527846,
+      "learning_rate": 9.075767500328757e-06,
+      "loss": 0.2342,
+      "step": 761
+    },
+    {
+      "epoch": 2.3064699205448354,
+      "grad_norm": 0.6407629469733517,
+      "learning_rate": 9.049439566958176e-06,
+      "loss": 0.2243,
+      "step": 762
+    },
+    {
+      "epoch": 2.309496783957624,
+      "grad_norm": 0.650140811885611,
+      "learning_rate": 9.023118280903887e-06,
+      "loss": 0.2271,
+      "step": 763
+    },
+    {
+      "epoch": 2.3125236473704125,
+      "grad_norm": 0.6668000817258457,
+      "learning_rate": 8.99680382623195e-06,
+      "loss": 0.2443,
+      "step": 764
+    },
+    {
+      "epoch": 2.315550510783201,
+      "grad_norm": 0.6465208637832673,
+      "learning_rate": 8.970496386960657e-06,
+      "loss": 0.2583,
+      "step": 765
+    },
+    {
+      "epoch": 2.318577374195989,
+      "grad_norm": 0.6448469184946334,
+      "learning_rate": 8.944196147059233e-06,
+      "loss": 0.2358,
+      "step": 766
+    },
+    {
+      "epoch": 2.3216042376087778,
+      "grad_norm": 0.6316325553577151,
+      "learning_rate": 8.917903290446559e-06,
+      "loss": 0.2372,
+      "step": 767
+    },
+    {
+      "epoch": 2.3246311010215663,
+      "grad_norm": 0.6429537666576937,
+      "learning_rate": 8.89161800098989e-06,
+      "loss": 0.2498,
+      "step": 768
+    },
+    {
+      "epoch": 2.327657964434355,
+      "grad_norm": 0.600883646853335,
+      "learning_rate": 8.86534046250356e-06,
+      "loss": 0.2206,
+      "step": 769
+    },
+    {
+      "epoch": 2.3306848278471435,
+      "grad_norm": 0.6558361713764058,
+      "learning_rate": 8.839070858747697e-06,
+      "loss": 0.2531,
+      "step": 770
+    },
+    {
+      "epoch": 2.333711691259932,
+      "grad_norm": 0.6608328800932824,
+      "learning_rate": 8.812809373426951e-06,
+      "loss": 0.252,
+      "step": 771
+    },
+    {
+      "epoch": 2.3367385546727206,
+      "grad_norm": 0.6360350229289197,
+      "learning_rate": 8.786556190189183e-06,
+      "loss": 0.252,
+      "step": 772
+    },
+    {
+      "epoch": 2.3397654180855088,
+      "grad_norm": 0.639352478570481,
+      "learning_rate": 8.760311492624211e-06,
+      "loss": 0.2442,
+      "step": 773
+    },
+    {
+      "epoch": 2.3427922814982973,
+      "grad_norm": 0.6619991399521448,
+      "learning_rate": 8.734075464262507e-06,
+      "loss": 0.2395,
+      "step": 774
+    },
+    {
+      "epoch": 2.345819144911086,
+      "grad_norm": 0.6536339191489544,
+      "learning_rate": 8.707848288573926e-06,
+      "loss": 0.2403,
+      "step": 775
+    },
+    {
+      "epoch": 2.3488460083238745,
+      "grad_norm": 0.6852721506040604,
+      "learning_rate": 8.681630148966397e-06,
+      "loss": 0.2576,
+      "step": 776
+    },
+    {
+      "epoch": 2.351872871736663,
+      "grad_norm": 0.6215955912372084,
+      "learning_rate": 8.655421228784683e-06,
+      "loss": 0.2229,
+      "step": 777
+    },
+    {
+      "epoch": 2.354899735149451,
+      "grad_norm": 0.6630198995718284,
+      "learning_rate": 8.629221711309056e-06,
+      "loss": 0.235,
+      "step": 778
+    },
+    {
+      "epoch": 2.3579265985622397,
+      "grad_norm": 0.6522114837436219,
+      "learning_rate": 8.603031779754049e-06,
+      "loss": 0.2247,
+      "step": 779
+    },
+    {
+      "epoch": 2.3609534619750283,
+      "grad_norm": 0.6807144487342814,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.2528,
+      "step": 780
+    },
+    {
+      "epoch": 2.363980325387817,
+      "grad_norm": 0.6355229999539997,
+      "learning_rate": 8.550681406927534e-06,
+      "loss": 0.2221,
+      "step": 781
+    },
+    {
+      "epoch": 2.3670071888006055,
+      "grad_norm": 0.6435379721934988,
+      "learning_rate": 8.52452133174478e-06,
+      "loss": 0.2465,
+      "step": 782
+    },
+    {
+      "epoch": 2.370034052213394,
+      "grad_norm": 0.6185100233630717,
+      "learning_rate": 8.498371574657596e-06,
+      "loss": 0.2361,
+      "step": 783
+    },
+    {
+      "epoch": 2.3730609156261826,
+      "grad_norm": 0.6551922592424928,
+      "learning_rate": 8.472232318532531e-06,
+      "loss": 0.2549,
+      "step": 784
+    },
+    {
+      "epoch": 2.3760877790389707,
+      "grad_norm": 0.601612721178037,
+      "learning_rate": 8.446103746162699e-06,
+      "loss": 0.2241,
+      "step": 785
+    },
+    {
+      "epoch": 2.3791146424517593,
+      "grad_norm": 0.631422508704952,
+      "learning_rate": 8.419986040266502e-06,
+      "loss": 0.2296,
+      "step": 786
+    },
+    {
+      "epoch": 2.382141505864548,
+      "grad_norm": 0.6083958707625841,
+      "learning_rate": 8.393879383486356e-06,
+      "loss": 0.2328,
+      "step": 787
+    },
+    {
+      "epoch": 2.3851683692773364,
+      "grad_norm": 0.6393785501244605,
+      "learning_rate": 8.367783958387407e-06,
+      "loss": 0.2464,
+      "step": 788
+    },
+    {
+      "epoch": 2.388195232690125,
+      "grad_norm": 0.7241279614691993,
+      "learning_rate": 8.341699947456261e-06,
+      "loss": 0.2555,
+      "step": 789
+    },
+    {
+      "epoch": 2.391222096102913,
+      "grad_norm": 0.6237675599002793,
+      "learning_rate": 8.315627533099697e-06,
+      "loss": 0.2421,
+      "step": 790
+    },
+    {
+      "epoch": 2.3942489595157017,
+      "grad_norm": 0.6321237492689261,
+      "learning_rate": 8.289566897643401e-06,
+      "loss": 0.2292,
+      "step": 791
+    },
+    {
+      "epoch": 2.3972758229284903,
+      "grad_norm": 0.6260475127096129,
+      "learning_rate": 8.263518223330698e-06,
+      "loss": 0.2321,
+      "step": 792
+    },
+    {
+      "epoch": 2.400302686341279,
+      "grad_norm": 0.6326980834143849,
+      "learning_rate": 8.237481692321251e-06,
+      "loss": 0.2381,
+      "step": 793
+    },
+    {
+      "epoch": 2.4033295497540674,
+      "grad_norm": 0.6502888415096226,
+      "learning_rate": 8.211457486689829e-06,
+      "loss": 0.2342,
+      "step": 794
+    },
+    {
+      "epoch": 2.406356413166856,
+      "grad_norm": 0.6563466205537725,
+      "learning_rate": 8.185445788424975e-06,
+      "loss": 0.2457,
+      "step": 795
+    },
+    {
+      "epoch": 2.4093832765796446,
+      "grad_norm": 0.6680432566905454,
+      "learning_rate": 8.159446779427798e-06,
+      "loss": 0.2571,
+      "step": 796
+    },
+    {
+      "epoch": 2.4124101399924327,
+      "grad_norm": 0.6701093458322849,
+      "learning_rate": 8.133460641510653e-06,
+      "loss": 0.2625,
+      "step": 797
+    },
+    {
+      "epoch": 2.4154370034052213,
+      "grad_norm": 0.6405284951273623,
+      "learning_rate": 8.107487556395902e-06,
+      "loss": 0.2299,
+      "step": 798
+    },
+    {
+      "epoch": 2.41846386681801,
+      "grad_norm": 0.6711971777120053,
+      "learning_rate": 8.081527705714605e-06,
+      "loss": 0.2407,
+      "step": 799
+    },
+    {
+      "epoch": 2.4214907302307984,
+      "grad_norm": 0.6514817432917603,
+      "learning_rate": 8.055581271005292e-06,
+      "loss": 0.226,
+      "step": 800
+    },
+    {
+      "epoch": 2.424517593643587,
+      "grad_norm": 0.631830441888797,
+      "learning_rate": 8.029648433712671e-06,
+      "loss": 0.2307,
+      "step": 801
+    },
+    {
+      "epoch": 2.427544457056375,
+      "grad_norm": 0.6507549595548531,
+      "learning_rate": 8.00372937518636e-06,
+      "loss": 0.2586,
+      "step": 802
+    },
+    {
+      "epoch": 2.4305713204691637,
+      "grad_norm": 0.6659250181752053,
+      "learning_rate": 7.977824276679623e-06,
+      "loss": 0.2418,
+      "step": 803
+    },
+    {
+      "epoch": 2.4335981838819523,
+      "grad_norm": 0.6680353178668894,
+      "learning_rate": 7.951933319348095e-06,
+      "loss": 0.2455,
+      "step": 804
+    },
+    {
+      "epoch": 2.436625047294741,
+      "grad_norm": 0.6424936528722962,
+      "learning_rate": 7.926056684248529e-06,
+      "loss": 0.2511,
+      "step": 805
+    },
+    {
+      "epoch": 2.4396519107075294,
+      "grad_norm": 0.703102660703569,
+      "learning_rate": 7.900194552337516e-06,
+      "loss": 0.2569,
+      "step": 806
+    },
+    {
+      "epoch": 2.442678774120318,
+      "grad_norm": 0.6760699663645013,
+      "learning_rate": 7.874347104470234e-06,
+      "loss": 0.2389,
+      "step": 807
+    },
+    {
+      "epoch": 2.4457056375331065,
+      "grad_norm": 0.6676577290533771,
+      "learning_rate": 7.848514521399167e-06,
+      "loss": 0.2146,
+      "step": 808
+    },
+    {
+      "epoch": 2.4487325009458947,
+      "grad_norm": 0.6634017613335853,
+      "learning_rate": 7.822696983772847e-06,
+      "loss": 0.2528,
+      "step": 809
+    },
+    {
+      "epoch": 2.4517593643586832,
+      "grad_norm": 0.70509109874294,
+      "learning_rate": 7.796894672134594e-06,
+      "loss": 0.251,
+      "step": 810
+    },
+    {
+      "epoch": 2.454786227771472,
+      "grad_norm": 0.6544884329920871,
+      "learning_rate": 7.771107766921262e-06,
+      "loss": 0.2432,
+      "step": 811
+    },
+    {
+      "epoch": 2.4578130911842604,
+      "grad_norm": 0.6735519761433937,
+      "learning_rate": 7.745336448461958e-06,
+      "loss": 0.2463,
+      "step": 812
+    },
+    {
+      "epoch": 2.460839954597049,
+      "grad_norm": 0.7109222437659978,
+      "learning_rate": 7.719580896976788e-06,
+      "loss": 0.2405,
+      "step": 813
+    },
+    {
+      "epoch": 2.463866818009837,
+      "grad_norm": 0.6853531225505147,
+      "learning_rate": 7.6938412925756e-06,
+      "loss": 0.2572,
+      "step": 814
+    },
+    {
+      "epoch": 2.4668936814226257,
+      "grad_norm": 0.6457294700034135,
+      "learning_rate": 7.66811781525673e-06,
+      "loss": 0.2198,
+      "step": 815
+    },
+    {
+      "epoch": 2.4699205448354142,
+      "grad_norm": 0.618582025994861,
+      "learning_rate": 7.642410644905726e-06,
+      "loss": 0.2173,
+      "step": 816
+    },
+    {
+      "epoch": 2.472947408248203,
+      "grad_norm": 0.6875851497493306,
+      "learning_rate": 7.616719961294119e-06,
+      "loss": 0.2504,
+      "step": 817
+    },
+    {
+      "epoch": 2.4759742716609914,
+      "grad_norm": 0.6825978460833699,
+      "learning_rate": 7.591045944078119e-06,
+      "loss": 0.2362,
+      "step": 818
+    },
+    {
+      "epoch": 2.47900113507378,
+      "grad_norm": 0.6711914562196986,
+      "learning_rate": 7.565388772797412e-06,
+      "loss": 0.2294,
+      "step": 819
+    },
+    {
+      "epoch": 2.4820279984865685,
+      "grad_norm": 0.6641547858048515,
+      "learning_rate": 7.539748626873866e-06,
+      "loss": 0.2396,
+      "step": 820
+    },
+    {
+      "epoch": 2.4850548618993566,
+      "grad_norm": 0.6516934782304565,
+      "learning_rate": 7.514125685610299e-06,
+      "loss": 0.2308,
+      "step": 821
+    },
+    {
+      "epoch": 2.488081725312145,
+      "grad_norm": 0.6406103426204676,
+      "learning_rate": 7.488520128189209e-06,
+      "loss": 0.2388,
+      "step": 822
+    },
+    {
+      "epoch": 2.491108588724934,
+      "grad_norm": 0.6438079121431102,
+      "learning_rate": 7.46293213367153e-06,
+      "loss": 0.2488,
+      "step": 823
+    },
+    {
+      "epoch": 2.4941354521377224,
+      "grad_norm": 0.6547452848567821,
+      "learning_rate": 7.4373618809953755e-06,
+      "loss": 0.2489,
+      "step": 824
+    },
+    {
+      "epoch": 2.497162315550511,
+      "grad_norm": 0.709826447730048,
+      "learning_rate": 7.411809548974792e-06,
+      "loss": 0.2437,
+      "step": 825
+    },
+    {
+      "epoch": 2.500189178963299,
+      "grad_norm": 0.6181719706753194,
+      "learning_rate": 7.386275316298513e-06,
+      "loss": 0.2263,
+      "step": 826
+    },
+    {
+      "epoch": 2.5032160423760876,
+      "grad_norm": 0.64657832203985,
+      "learning_rate": 7.360759361528685e-06,
+      "loss": 0.2427,
+      "step": 827
+    },
+    {
+      "epoch": 2.506242905788876,
+      "grad_norm": 0.6518342978452752,
+      "learning_rate": 7.335261863099652e-06,
+      "loss": 0.2426,
+      "step": 828
+    },
+    {
+      "epoch": 2.509269769201665,
+      "grad_norm": 0.635063625858039,
+      "learning_rate": 7.309782999316683e-06,
+      "loss": 0.2485,
+      "step": 829
+    },
+    {
+      "epoch": 2.5122966326144534,
+      "grad_norm": 0.6281345561858895,
+      "learning_rate": 7.2843229483547405e-06,
+      "loss": 0.2341,
+      "step": 830
+    },
+    {
+      "epoch": 2.515323496027242,
+      "grad_norm": 0.6462616882643718,
+      "learning_rate": 7.2588818882572266e-06,
+      "loss": 0.247,
+      "step": 831
+    },
+    {
+      "epoch": 2.5183503594400305,
+      "grad_norm": 0.621677017805431,
+      "learning_rate": 7.233459996934731e-06,
+      "loss": 0.2326,
+      "step": 832
+    },
+    {
+      "epoch": 2.5213772228528186,
+      "grad_norm": 0.6332111913693412,
+      "learning_rate": 7.208057452163803e-06,
+      "loss": 0.2353,
+      "step": 833
+    },
+    {
+      "epoch": 2.524404086265607,
+      "grad_norm": 0.6475729400157463,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.2506,
+      "step": 834
+    },
+    {
+      "epoch": 2.5274309496783958,
+      "grad_norm": 0.609927001106603,
+      "learning_rate": 7.1573111127051496e-06,
+      "loss": 0.2297,
+      "step": 835
+    },
+    {
+      "epoch": 2.5304578130911843,
+      "grad_norm": 0.6112497940593078,
+      "learning_rate": 7.131967672889101e-06,
+      "loss": 0.213,
+      "step": 836
+    },
+    {
+      "epoch": 2.533484676503973,
+      "grad_norm": 0.6178618031525237,
+      "learning_rate": 7.106644289365474e-06,
+      "loss": 0.2384,
+      "step": 837
+    },
+    {
+      "epoch": 2.536511539916761,
+      "grad_norm": 0.6389023170460563,
+      "learning_rate": 7.081341139221955e-06,
+      "loss": 0.2337,
+      "step": 838
+    },
+    {
+      "epoch": 2.5395384033295496,
+      "grad_norm": 0.6704759389174486,
+      "learning_rate": 7.056058399404731e-06,
+      "loss": 0.2296,
+      "step": 839
+    },
+    {
+      "epoch": 2.542565266742338,
+      "grad_norm": 0.6206264860172712,
+      "learning_rate": 7.0307962467172555e-06,
+      "loss": 0.2267,
+      "step": 840
+    },
+    {
+      "epoch": 2.5455921301551268,
+      "grad_norm": 0.6090990958586081,
+      "learning_rate": 7.005554857819009e-06,
+      "loss": 0.2281,
+      "step": 841
+    },
+    {
+      "epoch": 2.5486189935679153,
+      "grad_norm": 0.6637436947813579,
+      "learning_rate": 6.9803344092242855e-06,
+      "loss": 0.2301,
+      "step": 842
+    },
+    {
+      "epoch": 2.551645856980704,
+      "grad_norm": 0.6404558814613964,
+      "learning_rate": 6.955135077300932e-06,
+      "loss": 0.2187,
+      "step": 843
+    },
+    {
+      "epoch": 2.5546727203934925,
+      "grad_norm": 0.614536565780932,
+      "learning_rate": 6.929957038269123e-06,
+      "loss": 0.2431,
+      "step": 844
+    },
+    {
+      "epoch": 2.5576995838062806,
+      "grad_norm": 0.6095415948401826,
+      "learning_rate": 6.904800468200144e-06,
+      "loss": 0.2268,
+      "step": 845
+    },
+    {
+      "epoch": 2.560726447219069,
+      "grad_norm": 0.6694887693150934,
+      "learning_rate": 6.87966554301513e-06,
+      "loss": 0.2425,
+      "step": 846
+    },
+    {
+      "epoch": 2.5637533106318577,
+      "grad_norm": 0.6669864733949095,
+      "learning_rate": 6.854552438483866e-06,
+      "loss": 0.2496,
+      "step": 847
+    },
+    {
+      "epoch": 2.5667801740446463,
+      "grad_norm": 0.6279168812729935,
+      "learning_rate": 6.8294613302235325e-06,
+      "loss": 0.2203,
+      "step": 848
+    },
+    {
+      "epoch": 2.569807037457435,
+      "grad_norm": 0.6311296567354054,
+      "learning_rate": 6.804392393697502e-06,
+      "loss": 0.238,
+      "step": 849
+    },
+    {
+      "epoch": 2.572833900870223,
+      "grad_norm": 0.6120565852476493,
+      "learning_rate": 6.779345804214088e-06,
+      "loss": 0.2442,
+      "step": 850
+    },
+    {
+      "epoch": 2.5758607642830116,
+      "grad_norm": 0.6292262512158885,
+      "learning_rate": 6.7543217369253335e-06,
+      "loss": 0.24,
+      "step": 851
+    },
+    {
+      "epoch": 2.5788876276958,
+      "grad_norm": 0.6352971617247055,
+      "learning_rate": 6.729320366825785e-06,
+      "loss": 0.2372,
+      "step": 852
+    },
+    {
+      "epoch": 2.5819144911085887,
+      "grad_norm": 0.6429796326224673,
+      "learning_rate": 6.704341868751263e-06,
+      "loss": 0.2415,
+      "step": 853
+    },
+    {
+      "epoch": 2.5849413545213773,
+      "grad_norm": 0.6551586278791376,
+      "learning_rate": 6.679386417377649e-06,
+      "loss": 0.2573,
+      "step": 854
+    },
+    {
+      "epoch": 2.587968217934166,
+      "grad_norm": 0.6404911827584128,
+      "learning_rate": 6.654454187219649e-06,
+      "loss": 0.2383,
+      "step": 855
+    },
+    {
+      "epoch": 2.5909950813469544,
+      "grad_norm": 0.6721124810836703,
+      "learning_rate": 6.629545352629583e-06,
+      "loss": 0.2573,
+      "step": 856
+    },
+    {
+      "epoch": 2.5940219447597426,
+      "grad_norm": 0.682230817315984,
+      "learning_rate": 6.604660087796176e-06,
+      "loss": 0.2419,
+      "step": 857
+    },
+    {
+      "epoch": 2.597048808172531,
+      "grad_norm": 0.6535426021235742,
+      "learning_rate": 6.579798566743314e-06,
+      "loss": 0.2554,
+      "step": 858
+    },
+    {
+      "epoch": 2.6000756715853197,
+      "grad_norm": 0.6609341282439698,
+      "learning_rate": 6.554960963328851e-06,
+      "loss": 0.2508,
+      "step": 859
+    },
+    {
+      "epoch": 2.6031025349981083,
+      "grad_norm": 0.6376460315662437,
+      "learning_rate": 6.530147451243377e-06,
+      "loss": 0.242,
+      "step": 860
+    },
+    {
+      "epoch": 2.606129398410897,
+      "grad_norm": 0.6370583599812931,
+      "learning_rate": 6.505358204009018e-06,
+      "loss": 0.2194,
+      "step": 861
+    },
+    {
+      "epoch": 2.609156261823685,
+      "grad_norm": 0.6248957668901398,
+      "learning_rate": 6.480593394978208e-06,
+      "loss": 0.2253,
+      "step": 862
+    },
+    {
+      "epoch": 2.6121831252364736,
+      "grad_norm": 0.621671163093181,
+      "learning_rate": 6.455853197332488e-06,
+      "loss": 0.2408,
+      "step": 863
+    },
+    {
+      "epoch": 2.615209988649262,
+      "grad_norm": 0.6601319517578017,
+      "learning_rate": 6.431137784081283e-06,
+      "loss": 0.2558,
+      "step": 864
+    },
+    {
+      "epoch": 2.6182368520620507,
+      "grad_norm": 0.6164011742333828,
+      "learning_rate": 6.40644732806071e-06,
+      "loss": 0.2236,
+      "step": 865
+    },
+    {
+      "epoch": 2.6212637154748393,
+      "grad_norm": 0.6127372360391244,
+      "learning_rate": 6.381782001932352e-06,
+      "loss": 0.2254,
+      "step": 866
+    },
+    {
+      "epoch": 2.624290578887628,
+      "grad_norm": 0.6500913547154086,
+      "learning_rate": 6.357141978182056e-06,
+      "loss": 0.2334,
+      "step": 867
+    },
+    {
+      "epoch": 2.6273174423004164,
+      "grad_norm": 0.6735641978204768,
+      "learning_rate": 6.33252742911874e-06,
+      "loss": 0.2703,
+      "step": 868
+    },
+    {
+      "epoch": 2.6303443057132045,
+      "grad_norm": 0.6484590910425324,
+      "learning_rate": 6.3079385268731575e-06,
+      "loss": 0.2299,
+      "step": 869
+    },
+    {
+      "epoch": 2.633371169125993,
+      "grad_norm": 0.646310899358486,
+      "learning_rate": 6.283375443396726e-06,
+      "loss": 0.2438,
+      "step": 870
+    },
+    {
+      "epoch": 2.6363980325387817,
+      "grad_norm": 0.6305023418691454,
+      "learning_rate": 6.258838350460308e-06,
+      "loss": 0.25,
+      "step": 871
+    },
+    {
+      "epoch": 2.6394248959515703,
+      "grad_norm": 0.6758220383206582,
+      "learning_rate": 6.234327419653013e-06,
+      "loss": 0.252,
+      "step": 872
+    },
+    {
+      "epoch": 2.642451759364359,
+      "grad_norm": 0.6251830324544638,
+      "learning_rate": 6.209842822380998e-06,
+      "loss": 0.2378,
+      "step": 873
+    },
+    {
+      "epoch": 2.645478622777147,
+      "grad_norm": 0.6214316336026975,
+      "learning_rate": 6.185384729866264e-06,
+      "loss": 0.2214,
+      "step": 874
+    },
+    {
+      "epoch": 2.6485054861899355,
+      "grad_norm": 0.6453318243694905,
+      "learning_rate": 6.160953313145464e-06,
+      "loss": 0.2507,
+      "step": 875
+    },
+    {
+      "epoch": 2.651532349602724,
+      "grad_norm": 0.6695280797809467,
+      "learning_rate": 6.136548743068713e-06,
+      "loss": 0.2441,
+      "step": 876
+    },
+    {
+      "epoch": 2.6545592130155127,
+      "grad_norm": 0.6223867171863179,
+      "learning_rate": 6.112171190298384e-06,
+      "loss": 0.2437,
+      "step": 877
+    },
+    {
+      "epoch": 2.6575860764283012,
+      "grad_norm": 0.6438835631949237,
+      "learning_rate": 6.087820825307904e-06,
+      "loss": 0.2399,
+      "step": 878
+    },
+    {
+      "epoch": 2.66061293984109,
+      "grad_norm": 0.7052524613550574,
+      "learning_rate": 6.063497818380587e-06,
+      "loss": 0.2356,
+      "step": 879
+    },
+    {
+      "epoch": 2.6636398032538784,
+      "grad_norm": 0.5930605819961416,
+      "learning_rate": 6.039202339608432e-06,
+      "loss": 0.2335,
+      "step": 880
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.6208233623252428,
+      "learning_rate": 6.014934558890926e-06,
+      "loss": 0.2379,
+      "step": 881
+    },
+    {
+      "epoch": 2.669693530079455,
+      "grad_norm": 0.6449351146575381,
+      "learning_rate": 5.990694645933866e-06,
+      "loss": 0.2284,
+      "step": 882
+    },
+    {
+      "epoch": 2.6727203934922437,
+      "grad_norm": 0.6611623611719284,
+      "learning_rate": 5.966482770248161e-06,
+      "loss": 0.2522,
+      "step": 883
+    },
+    {
+      "epoch": 2.6757472569050322,
+      "grad_norm": 0.6284599606738535,
+      "learning_rate": 5.9422991011486635e-06,
+      "loss": 0.2273,
+      "step": 884
+    },
+    {
+      "epoch": 2.678774120317821,
+      "grad_norm": 0.6393586494786687,
+      "learning_rate": 5.918143807752972e-06,
+      "loss": 0.2374,
+      "step": 885
+    },
+    {
+      "epoch": 2.681800983730609,
+      "grad_norm": 0.643678369680256,
+      "learning_rate": 5.894017058980249e-06,
+      "loss": 0.2417,
+      "step": 886
+    },
+    {
+      "epoch": 2.6848278471433975,
+      "grad_norm": 0.6444605265379367,
+      "learning_rate": 5.8699190235500534e-06,
+      "loss": 0.2369,
+      "step": 887
+    },
+    {
+      "epoch": 2.687854710556186,
+      "grad_norm": 0.6808655042137965,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.2466,
+      "step": 888
+    },
+    {
+      "epoch": 2.6908815739689746,
+      "grad_norm": 0.718182843703189,
+      "learning_rate": 5.821809766590287e-06,
+      "loss": 0.2452,
+      "step": 889
+    },
+    {
+      "epoch": 2.693908437381763,
+      "grad_norm": 0.6946850669089457,
+      "learning_rate": 5.797798881491138e-06,
+      "loss": 0.2545,
+      "step": 890
+    },
+    {
+      "epoch": 2.696935300794552,
+      "grad_norm": 0.6796389675615006,
+      "learning_rate": 5.773817382593008e-06,
+      "loss": 0.2505,
+      "step": 891
+    },
+    {
+      "epoch": 2.6999621642073404,
+      "grad_norm": 0.7130263407201116,
+      "learning_rate": 5.749865437599703e-06,
+      "loss": 0.261,
+      "step": 892
+    },
+    {
+      "epoch": 2.7029890276201285,
+      "grad_norm": 0.6291079921910409,
+      "learning_rate": 5.725943214008362e-06,
+      "loss": 0.2334,
+      "step": 893
+    },
+    {
+      "epoch": 2.706015891032917,
+      "grad_norm": 0.6213433389309921,
+      "learning_rate": 5.702050879108284e-06,
+      "loss": 0.2285,
+      "step": 894
+    },
+    {
+      "epoch": 2.7090427544457056,
+      "grad_norm": 0.6277622671352314,
+      "learning_rate": 5.678188599979754e-06,
+      "loss": 0.2465,
+      "step": 895
+    },
+    {
+      "epoch": 2.712069617858494,
+      "grad_norm": 0.65856654471035,
+      "learning_rate": 5.654356543492883e-06,
+      "loss": 0.2324,
+      "step": 896
+    },
+    {
+      "epoch": 2.715096481271283,
+      "grad_norm": 0.6212836211135754,
+      "learning_rate": 5.630554876306407e-06,
+      "loss": 0.2364,
+      "step": 897
+    },
+    {
+      "epoch": 2.718123344684071,
+      "grad_norm": 0.6479524202436808,
+      "learning_rate": 5.606783764866576e-06,
+      "loss": 0.2351,
+      "step": 898
+    },
+    {
+      "epoch": 2.7211502080968595,
+      "grad_norm": 0.6558744863755732,
+      "learning_rate": 5.583043375405951e-06,
+      "loss": 0.2404,
+      "step": 899
+    },
+    {
+      "epoch": 2.724177071509648,
+      "grad_norm": 0.6600445324551689,
+      "learning_rate": 5.559333873942259e-06,
+      "loss": 0.2453,
+      "step": 900
+    },
+    {
+      "epoch": 2.7272039349224366,
+      "grad_norm": 0.66888442042798,
+      "learning_rate": 5.535655426277217e-06,
+      "loss": 0.249,
+      "step": 901
+    },
+    {
+      "epoch": 2.730230798335225,
+      "grad_norm": 0.6475883269482804,
+      "learning_rate": 5.512008197995379e-06,
+      "loss": 0.2451,
+      "step": 902
+    },
+    {
+      "epoch": 2.7332576617480138,
+      "grad_norm": 0.6356900492061184,
+      "learning_rate": 5.488392354462996e-06,
+      "loss": 0.2205,
+      "step": 903
+    },
+    {
+      "epoch": 2.7362845251608023,
+      "grad_norm": 0.6506613513302171,
+      "learning_rate": 5.464808060826825e-06,
+      "loss": 0.2393,
+      "step": 904
+    },
+    {
+      "epoch": 2.7393113885735905,
+      "grad_norm": 0.6148322685714376,
+      "learning_rate": 5.44125548201301e-06,
+      "loss": 0.2193,
+      "step": 905
+    },
+    {
+      "epoch": 2.742338251986379,
+      "grad_norm": 0.6245591630791876,
+      "learning_rate": 5.417734782725896e-06,
+      "loss": 0.2348,
+      "step": 906
+    },
+    {
+      "epoch": 2.7453651153991676,
+      "grad_norm": 0.6366519884721765,
+      "learning_rate": 5.394246127446911e-06,
+      "loss": 0.2243,
+      "step": 907
+    },
+    {
+      "epoch": 2.748391978811956,
+      "grad_norm": 0.681834936873786,
+      "learning_rate": 5.370789680433376e-06,
+      "loss": 0.2518,
+      "step": 908
+    },
+    {
+      "epoch": 2.7514188422247448,
+      "grad_norm": 0.7010973300923488,
+      "learning_rate": 5.347365605717394e-06,
+      "loss": 0.2455,
+      "step": 909
+    },
+    {
+      "epoch": 2.754445705637533,
+      "grad_norm": 0.6424452214012089,
+      "learning_rate": 5.323974067104687e-06,
+      "loss": 0.2397,
+      "step": 910
+    },
+    {
+      "epoch": 2.7574725690503215,
+      "grad_norm": 0.636108404480733,
+      "learning_rate": 5.300615228173442e-06,
+      "loss": 0.2335,
+      "step": 911
+    },
+    {
+      "epoch": 2.76049943246311,
+      "grad_norm": 0.6136517204954844,
+      "learning_rate": 5.277289252273175e-06,
+      "loss": 0.2268,
+      "step": 912
+    },
+    {
+      "epoch": 2.7635262958758986,
+      "grad_norm": 0.646774719907475,
+      "learning_rate": 5.253996302523596e-06,
+      "loss": 0.2338,
+      "step": 913
+    },
+    {
+      "epoch": 2.766553159288687,
+      "grad_norm": 0.6287318989877378,
+      "learning_rate": 5.230736541813463e-06,
+      "loss": 0.2249,
+      "step": 914
+    },
+    {
+      "epoch": 2.7695800227014757,
+      "grad_norm": 0.6125130769372704,
+      "learning_rate": 5.207510132799436e-06,
+      "loss": 0.2201,
+      "step": 915
+    },
+    {
+      "epoch": 2.7726068861142643,
+      "grad_norm": 0.624872684515624,
+      "learning_rate": 5.184317237904939e-06,
+      "loss": 0.2352,
+      "step": 916
+    },
+    {
+      "epoch": 2.7756337495270524,
+      "grad_norm": 0.6351049288807447,
+      "learning_rate": 5.161158019319041e-06,
+      "loss": 0.231,
+      "step": 917
+    },
+    {
+      "epoch": 2.778660612939841,
+      "grad_norm": 0.614144378782359,
+      "learning_rate": 5.138032638995315e-06,
+      "loss": 0.2359,
+      "step": 918
+    },
+    {
+      "epoch": 2.7816874763526296,
+      "grad_norm": 0.6525482913759885,
+      "learning_rate": 5.114941258650686e-06,
+      "loss": 0.2538,
+      "step": 919
+    },
+    {
+      "epoch": 2.784714339765418,
+      "grad_norm": 0.5873765262324834,
+      "learning_rate": 5.091884039764321e-06,
+      "loss": 0.2168,
+      "step": 920
+    },
+    {
+      "epoch": 2.7877412031782067,
+      "grad_norm": 0.6224246447433406,
+      "learning_rate": 5.0688611435764975e-06,
+      "loss": 0.2434,
+      "step": 921
+    },
+    {
+      "epoch": 2.790768066590995,
+      "grad_norm": 0.6241390073773446,
+      "learning_rate": 5.045872731087479e-06,
+      "loss": 0.2402,
+      "step": 922
+    },
+    {
+      "epoch": 2.7937949300037834,
+      "grad_norm": 0.666649003419413,
+      "learning_rate": 5.022918963056364e-06,
+      "loss": 0.2319,
+      "step": 923
+    },
+    {
+      "epoch": 2.796821793416572,
+      "grad_norm": 0.647061885845558,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.2464,
+      "step": 924
+    },
+    {
+      "epoch": 2.7998486568293606,
+      "grad_norm": 0.6192359605570994,
+      "learning_rate": 4.9771160021918305e-06,
+      "loss": 0.2278,
+      "step": 925
+    },
+    {
+      "epoch": 2.802875520242149,
+      "grad_norm": 0.6426965954317514,
+      "learning_rate": 4.954267129660789e-06,
+      "loss": 0.2315,
+      "step": 926
+    },
+    {
+      "epoch": 2.8059023836549377,
+      "grad_norm": 0.6173264673790622,
+      "learning_rate": 4.931453542190172e-06,
+      "loss": 0.2123,
+      "step": 927
+    },
+    {
+      "epoch": 2.8089292470677263,
+      "grad_norm": 0.6181796966517056,
+      "learning_rate": 4.908675399316534e-06,
+      "loss": 0.2176,
+      "step": 928
+    },
+    {
+      "epoch": 2.8119561104805144,
+      "grad_norm": 0.63376529970278,
+      "learning_rate": 4.8859328603285525e-06,
+      "loss": 0.2138,
+      "step": 929
+    },
+    {
+      "epoch": 2.814982973893303,
+      "grad_norm": 0.6447576363160566,
+      "learning_rate": 4.863226084265939e-06,
+      "loss": 0.2326,
+      "step": 930
+    },
+    {
+      "epoch": 2.8180098373060916,
+      "grad_norm": 0.6222659383575014,
+      "learning_rate": 4.840555229918299e-06,
+      "loss": 0.2336,
+      "step": 931
+    },
+    {
+      "epoch": 2.82103670071888,
+      "grad_norm": 0.6434875828203607,
+      "learning_rate": 4.817920455824045e-06,
+      "loss": 0.2531,
+      "step": 932
+    },
+    {
+      "epoch": 2.8240635641316687,
+      "grad_norm": 0.6788952626073534,
+      "learning_rate": 4.795321920269279e-06,
+      "loss": 0.2478,
+      "step": 933
+    },
+    {
+      "epoch": 2.827090427544457,
+      "grad_norm": 0.6152386565088812,
+      "learning_rate": 4.772759781286679e-06,
+      "loss": 0.2194,
+      "step": 934
+    },
+    {
+      "epoch": 2.8301172909572454,
+      "grad_norm": 0.6384832901787484,
+      "learning_rate": 4.7502341966544e-06,
+      "loss": 0.2339,
+      "step": 935
+    },
+    {
+      "epoch": 2.833144154370034,
+      "grad_norm": 0.6327881274551513,
+      "learning_rate": 4.727745323894976e-06,
+      "loss": 0.2239,
+      "step": 936
+    },
+    {
+      "epoch": 2.8361710177828225,
+      "grad_norm": 0.692955173621687,
+      "learning_rate": 4.705293320274213e-06,
+      "loss": 0.2503,
+      "step": 937
+    },
+    {
+      "epoch": 2.839197881195611,
+      "grad_norm": 0.629275219626828,
+      "learning_rate": 4.682878342800087e-06,
+      "loss": 0.2399,
+      "step": 938
+    },
+    {
+      "epoch": 2.8422247446083997,
+      "grad_norm": 0.6138824304496313,
+      "learning_rate": 4.66050054822164e-06,
+      "loss": 0.2198,
+      "step": 939
+    },
+    {
+      "epoch": 2.8452516080211883,
+      "grad_norm": 0.6546920006888297,
+      "learning_rate": 4.638160093027908e-06,
+      "loss": 0.2242,
+      "step": 940
+    },
+    {
+      "epoch": 2.8482784714339764,
+      "grad_norm": 0.6151299786126675,
+      "learning_rate": 4.6158571334468035e-06,
+      "loss": 0.2211,
+      "step": 941
+    },
+    {
+      "epoch": 2.851305334846765,
+      "grad_norm": 0.5848622142238439,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.2229,
+      "step": 942
+    },
+    {
+      "epoch": 2.8543321982595535,
+      "grad_norm": 0.6212626396667569,
+      "learning_rate": 4.571364324721982e-06,
+      "loss": 0.2483,
+      "step": 943
+    },
+    {
+      "epoch": 2.857359061672342,
+      "grad_norm": 0.6340582944913485,
+      "learning_rate": 4.549174786718684e-06,
+      "loss": 0.2377,
+      "step": 944
+    },
+    {
+      "epoch": 2.8603859250851307,
+      "grad_norm": 0.6134606334821089,
+      "learning_rate": 4.527023366606678e-06,
+      "loss": 0.2163,
+      "step": 945
+    },
+    {
+      "epoch": 2.863412788497919,
+      "grad_norm": 0.605903089816217,
+      "learning_rate": 4.504910219291941e-06,
+      "loss": 0.2268,
+      "step": 946
+    },
+    {
+      "epoch": 2.8664396519107074,
+      "grad_norm": 0.6460468920942336,
+      "learning_rate": 4.482835499412812e-06,
+      "loss": 0.2602,
+      "step": 947
+    },
+    {
+      "epoch": 2.869466515323496,
+      "grad_norm": 0.6288730620426358,
+      "learning_rate": 4.460799361338898e-06,
+      "loss": 0.2438,
+      "step": 948
+    },
+    {
+      "epoch": 2.8724933787362845,
+      "grad_norm": 0.6355417780345397,
+      "learning_rate": 4.438801959170014e-06,
+      "loss": 0.2316,
+      "step": 949
+    },
+    {
+      "epoch": 2.875520242149073,
+      "grad_norm": 0.6656356282601408,
+      "learning_rate": 4.416843446735077e-06,
+      "loss": 0.2449,
+      "step": 950
+    },
+    {
+      "epoch": 2.8785471055618617,
+      "grad_norm": 0.6465586548085274,
+      "learning_rate": 4.394923977591059e-06,
+      "loss": 0.231,
+      "step": 951
+    },
+    {
+      "epoch": 2.8815739689746502,
+      "grad_norm": 0.6746916385953997,
+      "learning_rate": 4.373043705021899e-06,
+      "loss": 0.2477,
+      "step": 952
+    },
+    {
+      "epoch": 2.8846008323874384,
+      "grad_norm": 0.6887152235292493,
+      "learning_rate": 4.351202782037426e-06,
+      "loss": 0.2444,
+      "step": 953
+    },
+    {
+      "epoch": 2.887627695800227,
+      "grad_norm": 0.6689963610841018,
+      "learning_rate": 4.3294013613722944e-06,
+      "loss": 0.2509,
+      "step": 954
+    },
+    {
+      "epoch": 2.8906545592130155,
+      "grad_norm": 0.6246818932919441,
+      "learning_rate": 4.307639595484924e-06,
+      "loss": 0.2275,
+      "step": 955
+    },
+    {
+      "epoch": 2.893681422625804,
+      "grad_norm": 0.6397094814740238,
+      "learning_rate": 4.2859176365564294e-06,
+      "loss": 0.228,
+      "step": 956
+    },
+    {
+      "epoch": 2.8967082860385927,
+      "grad_norm": 0.6258018157817,
+      "learning_rate": 4.264235636489542e-06,
+      "loss": 0.2413,
+      "step": 957
+    },
+    {
+      "epoch": 2.8997351494513808,
+      "grad_norm": 0.5854171067560046,
+      "learning_rate": 4.2425937469075626e-06,
+      "loss": 0.2289,
+      "step": 958
+    },
+    {
+      "epoch": 2.9027620128641693,
+      "grad_norm": 0.6407565966520492,
+      "learning_rate": 4.220992119153302e-06,
+      "loss": 0.218,
+      "step": 959
+    },
+    {
+      "epoch": 2.905788876276958,
+      "grad_norm": 0.6372100488677686,
+      "learning_rate": 4.19943090428802e-06,
+      "loss": 0.2378,
+      "step": 960
+    },
+    {
+      "epoch": 2.9088157396897465,
+      "grad_norm": 0.6458399642344448,
+      "learning_rate": 4.177910253090357e-06,
+      "loss": 0.2337,
+      "step": 961
+    },
+    {
+      "epoch": 2.911842603102535,
+      "grad_norm": 0.6218169458370939,
+      "learning_rate": 4.1564303160552935e-06,
+      "loss": 0.2353,
+      "step": 962
+    },
+    {
+      "epoch": 2.9148694665153236,
+      "grad_norm": 0.6313931066353392,
+      "learning_rate": 4.134991243393097e-06,
+      "loss": 0.2303,
+      "step": 963
+    },
+    {
+      "epoch": 2.917896329928112,
+      "grad_norm": 0.6531301330971137,
+      "learning_rate": 4.113593185028273e-06,
+      "loss": 0.2221,
+      "step": 964
+    },
+    {
+      "epoch": 2.9209231933409003,
+      "grad_norm": 0.6465885281980449,
+      "learning_rate": 4.092236290598499e-06,
+      "loss": 0.2395,
+      "step": 965
+    },
+    {
+      "epoch": 2.923950056753689,
+      "grad_norm": 0.6218936913420795,
+      "learning_rate": 4.070920709453597e-06,
+      "loss": 0.2295,
+      "step": 966
+    },
+    {
+      "epoch": 2.9269769201664775,
+      "grad_norm": 0.6566040758668049,
+      "learning_rate": 4.049646590654487e-06,
+      "loss": 0.2422,
+      "step": 967
+    },
+    {
+      "epoch": 2.930003783579266,
+      "grad_norm": 0.6532140222840412,
+      "learning_rate": 4.028414082972141e-06,
+      "loss": 0.2392,
+      "step": 968
+    },
+    {
+      "epoch": 2.9330306469920546,
+      "grad_norm": 0.6236470383405065,
+      "learning_rate": 4.007223334886531e-06,
+      "loss": 0.2425,
+      "step": 969
+    },
+    {
+      "epoch": 2.9360575104048428,
+      "grad_norm": 0.6286348945011191,
+      "learning_rate": 3.986074494585619e-06,
+      "loss": 0.2179,
+      "step": 970
+    },
+    {
+      "epoch": 2.9390843738176313,
+      "grad_norm": 0.62836289543254,
+      "learning_rate": 3.964967709964287e-06,
+      "loss": 0.263,
+      "step": 971
+    },
+    {
+      "epoch": 2.94211123723042,
+      "grad_norm": 0.6217994199682595,
+      "learning_rate": 3.943903128623336e-06,
+      "loss": 0.2415,
+      "step": 972
+    },
+    {
+      "epoch": 2.9451381006432085,
+      "grad_norm": 0.6202733122587974,
+      "learning_rate": 3.9228808978684195e-06,
+      "loss": 0.231,
+      "step": 973
+    },
+    {
+      "epoch": 2.948164964055997,
+      "grad_norm": 0.6470888805197212,
+      "learning_rate": 3.9019011647090465e-06,
+      "loss": 0.2578,
+      "step": 974
+    },
+    {
+      "epoch": 2.9511918274687856,
+      "grad_norm": 0.6137197146582352,
+      "learning_rate": 3.880964075857535e-06,
+      "loss": 0.2215,
+      "step": 975
+    },
+    {
+      "epoch": 2.954218690881574,
+      "grad_norm": 0.6297820686028612,
+      "learning_rate": 3.860069777727983e-06,
+      "loss": 0.2462,
+      "step": 976
+    },
+    {
+      "epoch": 2.9572455542943623,
+      "grad_norm": 0.6225626907922844,
+      "learning_rate": 3.839218416435251e-06,
+      "loss": 0.2437,
+      "step": 977
+    },
+    {
+      "epoch": 2.960272417707151,
+      "grad_norm": 0.6033446218389062,
+      "learning_rate": 3.818410137793947e-06,
+      "loss": 0.2201,
+      "step": 978
+    },
+    {
+      "epoch": 2.9632992811199395,
+      "grad_norm": 0.6007327594780554,
+      "learning_rate": 3.797645087317401e-06,
+      "loss": 0.218,
+      "step": 979
+    },
+    {
+      "epoch": 2.966326144532728,
+      "grad_norm": 0.6478133910517775,
+      "learning_rate": 3.7769234102166365e-06,
+      "loss": 0.2537,
+      "step": 980
+    },
+    {
+      "epoch": 2.9693530079455166,
+      "grad_norm": 0.6295219821596777,
+      "learning_rate": 3.7562452513993676e-06,
+      "loss": 0.2338,
+      "step": 981
+    },
+    {
+      "epoch": 2.9723798713583047,
+      "grad_norm": 0.6290140099552729,
+      "learning_rate": 3.735610755468988e-06,
+      "loss": 0.2418,
+      "step": 982
+    },
+    {
+      "epoch": 2.9754067347710933,
+      "grad_norm": 0.6261606967373853,
+      "learning_rate": 3.715020066723557e-06,
+      "loss": 0.2249,
+      "step": 983
+    },
+    {
+      "epoch": 2.978433598183882,
+      "grad_norm": 0.6222642646596266,
+      "learning_rate": 3.6944733291547784e-06,
+      "loss": 0.2394,
+      "step": 984
+    },
+    {
+      "epoch": 2.9814604615966704,
+      "grad_norm": 0.6215887182125999,
+      "learning_rate": 3.6739706864470048e-06,
+      "loss": 0.2168,
+      "step": 985
+    },
+    {
+      "epoch": 2.984487325009459,
+      "grad_norm": 0.6110422715113881,
+      "learning_rate": 3.653512281976238e-06,
+      "loss": 0.232,
+      "step": 986
+    },
+    {
+      "epoch": 2.9875141884222476,
+      "grad_norm": 0.627418160404746,
+      "learning_rate": 3.633098258809119e-06,
+      "loss": 0.2385,
+      "step": 987
+    },
+    {
+      "epoch": 2.990541051835036,
+      "grad_norm": 0.6154835559908926,
+      "learning_rate": 3.612728759701919e-06,
+      "loss": 0.2161,
+      "step": 988
+    },
+    {
+      "epoch": 2.9935679152478243,
+      "grad_norm": 0.6155081147736654,
+      "learning_rate": 3.5924039270995624e-06,
+      "loss": 0.2218,
+      "step": 989
+    },
+    {
+      "epoch": 2.996594778660613,
+      "grad_norm": 0.6160818598297803,
+      "learning_rate": 3.5721239031346067e-06,
+      "loss": 0.2305,
+      "step": 990
+    },
+    {
+      "epoch": 2.9996216420734014,
+      "grad_norm": 0.646732740367869,
+      "learning_rate": 3.5518888296262723e-06,
+      "loss": 0.2302,
+      "step": 991
+    },
+    {
+      "epoch": 3.00264850548619,
+      "grad_norm": 1.1987960503250197,
+      "learning_rate": 3.5316988480794255e-06,
+      "loss": 0.1585,
+      "step": 992
+    },
+    {
+      "epoch": 3.0056753688989786,
+      "grad_norm": 1.1253803564643572,
+      "learning_rate": 3.5115540996836174e-06,
+      "loss": 0.1375,
+      "step": 993
+    },
+    {
+      "epoch": 3.008702232311767,
+      "grad_norm": 0.9626281928550182,
+      "learning_rate": 3.4914547253120655e-06,
+      "loss": 0.149,
+      "step": 994
+    },
+    {
+      "epoch": 3.0117290957245553,
+      "grad_norm": 0.8623583122705603,
+      "learning_rate": 3.4714008655207e-06,
+      "loss": 0.1419,
+      "step": 995
+    },
+    {
+      "epoch": 3.014755959137344,
+      "grad_norm": 0.9894354006959183,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.147,
+      "step": 996
+    },
+    {
+      "epoch": 3.0177828225501324,
+      "grad_norm": 1.1570853923341926,
+      "learning_rate": 3.4314302503097906e-06,
+      "loss": 0.1461,
+      "step": 997
+    },
+    {
+      "epoch": 3.020809685962921,
+      "grad_norm": 1.0876737712601052,
+      "learning_rate": 3.4115137744067516e-06,
+      "loss": 0.1431,
+      "step": 998
+    },
+    {
+      "epoch": 3.0238365493757096,
+      "grad_norm": 1.0021111160155307,
+      "learning_rate": 3.3916433721149323e-06,
+      "loss": 0.1314,
+      "step": 999
+    },
+    {
+      "epoch": 3.026863412788498,
+      "grad_norm": 1.1261417676826264,
+      "learning_rate": 3.37181918238904e-06,
+      "loss": 0.138,
+      "step": 1000
+    },
+    {
+      "epoch": 3.0298902762012863,
+      "grad_norm": 0.9436331362664303,
+      "learning_rate": 3.3520413438606215e-06,
+      "loss": 0.1264,
+      "step": 1001
+    },
+    {
+      "epoch": 3.032917139614075,
+      "grad_norm": 0.9443798105228549,
+      "learning_rate": 3.3323099948370853e-06,
+      "loss": 0.1355,
+      "step": 1002
+    },
+    {
+      "epoch": 3.0359440030268634,
+      "grad_norm": 0.9337976919362775,
+      "learning_rate": 3.3126252733007314e-06,
+      "loss": 0.1327,
+      "step": 1003
+    },
+    {
+      "epoch": 3.038970866439652,
+      "grad_norm": 0.7669690884046846,
+      "learning_rate": 3.292987316907792e-06,
+      "loss": 0.1273,
+      "step": 1004
+    },
+    {
+      "epoch": 3.0419977298524405,
+      "grad_norm": 0.7729773646336908,
+      "learning_rate": 3.273396262987475e-06,
+      "loss": 0.1227,
+      "step": 1005
+    },
+    {
+      "epoch": 3.045024593265229,
+      "grad_norm": 0.79288820954428,
+      "learning_rate": 3.253852248540994e-06,
+      "loss": 0.1235,
+      "step": 1006
+    },
+    {
+      "epoch": 3.0480514566780172,
+      "grad_norm": 0.8189545941934355,
+      "learning_rate": 3.2343554102406104e-06,
+      "loss": 0.1406,
+      "step": 1007
+    },
+    {
+      "epoch": 3.051078320090806,
+      "grad_norm": 0.8570458258161203,
+      "learning_rate": 3.2149058844286796e-06,
+      "loss": 0.1387,
+      "step": 1008
+    },
+    {
+      "epoch": 3.0541051835035944,
+      "grad_norm": 0.7608079915326709,
+      "learning_rate": 3.1955038071167034e-06,
+      "loss": 0.1323,
+      "step": 1009
+    },
+    {
+      "epoch": 3.057132046916383,
+      "grad_norm": 0.8187176386278655,
+      "learning_rate": 3.1761493139843734e-06,
+      "loss": 0.1445,
+      "step": 1010
+    },
+    {
+      "epoch": 3.0601589103291715,
+      "grad_norm": 0.7569608417633935,
+      "learning_rate": 3.1568425403786175e-06,
+      "loss": 0.1385,
+      "step": 1011
+    },
+    {
+      "epoch": 3.06318577374196,
+      "grad_norm": 0.7378892420628202,
+      "learning_rate": 3.1375836213126653e-06,
+      "loss": 0.1302,
+      "step": 1012
+    },
+    {
+      "epoch": 3.0662126371547482,
+      "grad_norm": 0.7826963636856568,
+      "learning_rate": 3.1183726914650903e-06,
+      "loss": 0.1337,
+      "step": 1013
+    },
+    {
+      "epoch": 3.069239500567537,
+      "grad_norm": 0.7170056098167819,
+      "learning_rate": 3.099209885178882e-06,
+      "loss": 0.1291,
+      "step": 1014
+    },
+    {
+      "epoch": 3.0722663639803254,
+      "grad_norm": 0.7653284336044542,
+      "learning_rate": 3.080095336460491e-06,
+      "loss": 0.1296,
+      "step": 1015
+    },
+    {
+      "epoch": 3.075293227393114,
+      "grad_norm": 0.7595939801367457,
+      "learning_rate": 3.0610291789789094e-06,
+      "loss": 0.138,
+      "step": 1016
+    },
+    {
+      "epoch": 3.0783200908059025,
+      "grad_norm": 0.726571136574179,
+      "learning_rate": 3.042011546064724e-06,
+      "loss": 0.1268,
+      "step": 1017
+    },
+    {
+      "epoch": 3.081346954218691,
+      "grad_norm": 0.7514071429515331,
+      "learning_rate": 3.023042570709185e-06,
+      "loss": 0.1346,
+      "step": 1018
+    },
+    {
+      "epoch": 3.084373817631479,
+      "grad_norm": 0.7096311910136123,
+      "learning_rate": 3.004122385563273e-06,
+      "loss": 0.134,
+      "step": 1019
+    },
+    {
+      "epoch": 3.087400681044268,
+      "grad_norm": 0.7983729742124769,
+      "learning_rate": 2.9852511229367862e-06,
+      "loss": 0.1368,
+      "step": 1020
+    },
+    {
+      "epoch": 3.0904275444570564,
+      "grad_norm": 0.7875613167994405,
+      "learning_rate": 2.9664289147974046e-06,
+      "loss": 0.1417,
+      "step": 1021
+    },
+    {
+      "epoch": 3.093454407869845,
+      "grad_norm": 0.7289218998215642,
+      "learning_rate": 2.9476558927697605e-06,
+      "loss": 0.14,
+      "step": 1022
+    },
+    {
+      "epoch": 3.0964812712826335,
+      "grad_norm": 0.7258863060060438,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.132,
+      "step": 1023
+    },
+    {
+      "epoch": 3.099508134695422,
+      "grad_norm": 0.7399726942901795,
+      "learning_rate": 2.9102579318274994e-06,
+      "loss": 0.1377,
+      "step": 1024
+    },
+    {
+      "epoch": 3.10253499810821,
+      "grad_norm": 0.7040002645890767,
+      "learning_rate": 2.891633254438685e-06,
+      "loss": 0.1346,
+      "step": 1025
+    },
+    {
+      "epoch": 3.105561861520999,
+      "grad_norm": 0.7182540886750363,
+      "learning_rate": 2.8730582862113743e-06,
+      "loss": 0.136,
+      "step": 1026
+    },
+    {
+      "epoch": 3.1085887249337874,
+      "grad_norm": 0.713496893300805,
+      "learning_rate": 2.854533157041238e-06,
+      "loss": 0.1303,
+      "step": 1027
+    },
+    {
+      "epoch": 3.111615588346576,
+      "grad_norm": 0.6704086662426908,
+      "learning_rate": 2.8360579964754277e-06,
+      "loss": 0.1228,
+      "step": 1028
+    },
+    {
+      "epoch": 3.1146424517593645,
+      "grad_norm": 0.7589911724643795,
+      "learning_rate": 2.8176329337116604e-06,
+      "loss": 0.141,
+      "step": 1029
+    },
+    {
+      "epoch": 3.117669315172153,
+      "grad_norm": 0.6838499097432651,
+      "learning_rate": 2.7992580975973136e-06,
+      "loss": 0.1393,
+      "step": 1030
+    },
+    {
+      "epoch": 3.120696178584941,
+      "grad_norm": 0.7071723704684244,
+      "learning_rate": 2.780933616628523e-06,
+      "loss": 0.1325,
+      "step": 1031
+    },
+    {
+      "epoch": 3.1237230419977298,
+      "grad_norm": 0.6866170951568361,
+      "learning_rate": 2.7626596189492983e-06,
+      "loss": 0.1232,
+      "step": 1032
+    },
+    {
+      "epoch": 3.1267499054105183,
+      "grad_norm": 0.6650653067695044,
+      "learning_rate": 2.7444362323506145e-06,
+      "loss": 0.1316,
+      "step": 1033
+    },
+    {
+      "epoch": 3.129776768823307,
+      "grad_norm": 0.6979919650401635,
+      "learning_rate": 2.726263584269513e-06,
+      "loss": 0.1318,
+      "step": 1034
+    },
+    {
+      "epoch": 3.1328036322360955,
+      "grad_norm": 0.7384824318579697,
+      "learning_rate": 2.70814180178823e-06,
+      "loss": 0.1463,
+      "step": 1035
+    },
+    {
+      "epoch": 3.135830495648884,
+      "grad_norm": 0.7019749733032239,
+      "learning_rate": 2.690071011633284e-06,
+      "loss": 0.1296,
+      "step": 1036
+    },
+    {
+      "epoch": 3.138857359061672,
+      "grad_norm": 0.6889435070120475,
+      "learning_rate": 2.6720513401746128e-06,
+      "loss": 0.129,
+      "step": 1037
+    },
+    {
+      "epoch": 3.1418842224744608,
+      "grad_norm": 0.6702693102926474,
+      "learning_rate": 2.6540829134246683e-06,
+      "loss": 0.1319,
+      "step": 1038
+    },
+    {
+      "epoch": 3.1449110858872493,
+      "grad_norm": 0.7207122610103165,
+      "learning_rate": 2.6361658570375528e-06,
+      "loss": 0.1452,
+      "step": 1039
+    },
+    {
+      "epoch": 3.147937949300038,
+      "grad_norm": 0.6997034563627192,
+      "learning_rate": 2.618300296308135e-06,
+      "loss": 0.1227,
+      "step": 1040
+    },
+    {
+      "epoch": 3.1509648127128265,
+      "grad_norm": 0.7412929552840333,
+      "learning_rate": 2.6004863561711633e-06,
+      "loss": 0.1398,
+      "step": 1041
+    },
+    {
+      "epoch": 3.1539916761256146,
+      "grad_norm": 0.6781034165258759,
+      "learning_rate": 2.582724161200405e-06,
+      "loss": 0.1287,
+      "step": 1042
+    },
+    {
+      "epoch": 3.157018539538403,
+      "grad_norm": 0.6877510416472407,
+      "learning_rate": 2.5650138356077715e-06,
+      "loss": 0.144,
+      "step": 1043
+    },
+    {
+      "epoch": 3.1600454029511917,
+      "grad_norm": 0.681427471305092,
+      "learning_rate": 2.5473555032424534e-06,
+      "loss": 0.1344,
+      "step": 1044
+    },
+    {
+      "epoch": 3.1630722663639803,
+      "grad_norm": 0.7554482416072129,
+      "learning_rate": 2.529749287590042e-06,
+      "loss": 0.1535,
+      "step": 1045
+    },
+    {
+      "epoch": 3.166099129776769,
+      "grad_norm": 0.7172279573573774,
+      "learning_rate": 2.5121953117716744e-06,
+      "loss": 0.1305,
+      "step": 1046
+    },
+    {
+      "epoch": 3.1691259931895575,
+      "grad_norm": 0.6896133475440172,
+      "learning_rate": 2.494693698543179e-06,
+      "loss": 0.1241,
+      "step": 1047
+    },
+    {
+      "epoch": 3.172152856602346,
+      "grad_norm": 0.700532657812981,
+      "learning_rate": 2.477244570294206e-06,
+      "loss": 0.1269,
+      "step": 1048
+    },
+    {
+      "epoch": 3.175179720015134,
+      "grad_norm": 0.7074025641832723,
+      "learning_rate": 2.4598480490473753e-06,
+      "loss": 0.1355,
+      "step": 1049
+    },
+    {
+      "epoch": 3.1782065834279227,
+      "grad_norm": 0.6641524502652301,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.1217,
+      "step": 1050
+    },
+    {
+      "epoch": 3.1812334468407113,
+      "grad_norm": 0.6923462795977636,
+      "learning_rate": 2.4252133138103416e-06,
+      "loss": 0.1248,
+      "step": 1051
+    },
+    {
+      "epoch": 3.1842603102535,
+      "grad_norm": 0.7193244924473611,
+      "learning_rate": 2.4079753420225694e-06,
+      "loss": 0.1284,
+      "step": 1052
+    },
+    {
+      "epoch": 3.1872871736662884,
+      "grad_norm": 0.7064168388675774,
+      "learning_rate": 2.3907904616400855e-06,
+      "loss": 0.126,
+      "step": 1053
+    },
+    {
+      "epoch": 3.190314037079077,
+      "grad_norm": 0.6619152799905513,
+      "learning_rate": 2.3736587928376197e-06,
+      "loss": 0.1229,
+      "step": 1054
+    },
+    {
+      "epoch": 3.193340900491865,
+      "grad_norm": 0.6963273119905737,
+      "learning_rate": 2.356580455417776e-06,
+      "loss": 0.128,
+      "step": 1055
+    },
+    {
+      "epoch": 3.1963677639046537,
+      "grad_norm": 0.6791778130529682,
+      "learning_rate": 2.339555568810221e-06,
+      "loss": 0.1214,
+      "step": 1056
+    },
+    {
+      "epoch": 3.1993946273174423,
+      "grad_norm": 0.7208411523148566,
+      "learning_rate": 2.3225842520708264e-06,
+      "loss": 0.1253,
+      "step": 1057
+    },
+    {
+      "epoch": 3.202421490730231,
+      "grad_norm": 0.6539864225667831,
+      "learning_rate": 2.305666623880858e-06,
+      "loss": 0.1312,
+      "step": 1058
+    },
+    {
+      "epoch": 3.2054483541430194,
+      "grad_norm": 0.7150524464982211,
+      "learning_rate": 2.288802802546124e-06,
+      "loss": 0.1343,
+      "step": 1059
+    },
+    {
+      "epoch": 3.208475217555808,
+      "grad_norm": 0.7615230150916095,
+      "learning_rate": 2.27199290599617e-06,
+      "loss": 0.1506,
+      "step": 1060
+    },
+    {
+      "epoch": 3.211502080968596,
+      "grad_norm": 0.684214433147453,
+      "learning_rate": 2.255237051783431e-06,
+      "loss": 0.1252,
+      "step": 1061
+    },
+    {
+      "epoch": 3.2145289443813847,
+      "grad_norm": 0.6432864037385053,
+      "learning_rate": 2.2385353570824308e-06,
+      "loss": 0.12,
+      "step": 1062
+    },
+    {
+      "epoch": 3.2175558077941733,
+      "grad_norm": 0.7066441155398018,
+      "learning_rate": 2.2218879386889515e-06,
+      "loss": 0.1294,
+      "step": 1063
+    },
+    {
+      "epoch": 3.220582671206962,
+      "grad_norm": 0.7329111720042054,
+      "learning_rate": 2.2052949130192136e-06,
+      "loss": 0.1399,
+      "step": 1064
+    },
+    {
+      "epoch": 3.2236095346197504,
+      "grad_norm": 0.7027878381750445,
+      "learning_rate": 2.1887563961090664e-06,
+      "loss": 0.1398,
+      "step": 1065
+    },
+    {
+      "epoch": 3.226636398032539,
+      "grad_norm": 0.7585370272197892,
+      "learning_rate": 2.172272503613183e-06,
+      "loss": 0.1372,
+      "step": 1066
+    },
+    {
+      "epoch": 3.229663261445327,
+      "grad_norm": 0.7409677520336692,
+      "learning_rate": 2.155843350804243e-06,
+      "loss": 0.1405,
+      "step": 1067
+    },
+    {
+      "epoch": 3.2326901248581157,
+      "grad_norm": 0.7553198525028305,
+      "learning_rate": 2.1394690525721275e-06,
+      "loss": 0.1462,
+      "step": 1068
+    },
+    {
+      "epoch": 3.2357169882709043,
+      "grad_norm": 0.708750842437752,
+      "learning_rate": 2.1231497234231113e-06,
+      "loss": 0.1187,
+      "step": 1069
+    },
+    {
+      "epoch": 3.238743851683693,
+      "grad_norm": 0.6941581239936369,
+      "learning_rate": 2.1068854774790783e-06,
+      "loss": 0.1311,
+      "step": 1070
+    },
+    {
+      "epoch": 3.2417707150964814,
+      "grad_norm": 0.6671294120863492,
+      "learning_rate": 2.090676428476709e-06,
+      "loss": 0.1207,
+      "step": 1071
+    },
+    {
+      "epoch": 3.24479757850927,
+      "grad_norm": 0.7217993681413384,
+      "learning_rate": 2.0745226897666858e-06,
+      "loss": 0.1416,
+      "step": 1072
+    },
+    {
+      "epoch": 3.247824441922058,
+      "grad_norm": 0.6882344407588535,
+      "learning_rate": 2.0584243743129018e-06,
+      "loss": 0.133,
+      "step": 1073
+    },
+    {
+      "epoch": 3.2508513053348467,
+      "grad_norm": 0.6954093990832545,
+      "learning_rate": 2.0423815946916783e-06,
+      "loss": 0.1248,
+      "step": 1074
+    },
+    {
+      "epoch": 3.2538781687476352,
+      "grad_norm": 0.719536165897936,
+      "learning_rate": 2.0263944630909736e-06,
+      "loss": 0.1399,
+      "step": 1075
+    },
+    {
+      "epoch": 3.256905032160424,
+      "grad_norm": 0.6847641378879086,
+      "learning_rate": 2.010463091309587e-06,
+      "loss": 0.1305,
+      "step": 1076
+    },
+    {
+      "epoch": 3.2599318955732124,
+      "grad_norm": 0.6758901512982688,
+      "learning_rate": 1.994587590756397e-06,
+      "loss": 0.1273,
+      "step": 1077
+    },
+    {
+      "epoch": 3.2629587589860005,
+      "grad_norm": 0.7118619655347667,
+      "learning_rate": 1.9787680724495617e-06,
+      "loss": 0.1327,
+      "step": 1078
+    },
+    {
+      "epoch": 3.265985622398789,
+      "grad_norm": 0.7111414543406603,
+      "learning_rate": 1.963004647015764e-06,
+      "loss": 0.1346,
+      "step": 1079
+    },
+    {
+      "epoch": 3.2690124858115777,
+      "grad_norm": 0.7057007427956181,
+      "learning_rate": 1.947297424689414e-06,
+      "loss": 0.1362,
+      "step": 1080
+    },
+    {
+      "epoch": 3.2720393492243662,
+      "grad_norm": 0.6860575535672581,
+      "learning_rate": 1.9316465153119e-06,
+      "loss": 0.1389,
+      "step": 1081
+    },
+    {
+      "epoch": 3.275066212637155,
+      "grad_norm": 0.7517463231416913,
+      "learning_rate": 1.9160520283308115e-06,
+      "loss": 0.137,
+      "step": 1082
+    },
+    {
+      "epoch": 3.2780930760499434,
+      "grad_norm": 0.7337007723861229,
+      "learning_rate": 1.9005140727991678e-06,
+      "loss": 0.1284,
+      "step": 1083
+    },
+    {
+      "epoch": 3.281119939462732,
+      "grad_norm": 0.691804920741409,
+      "learning_rate": 1.8850327573746584e-06,
+      "loss": 0.1214,
+      "step": 1084
+    },
+    {
+      "epoch": 3.28414680287552,
+      "grad_norm": 0.6922140599803238,
+      "learning_rate": 1.8696081903188955e-06,
+      "loss": 0.1251,
+      "step": 1085
+    },
+    {
+      "epoch": 3.2871736662883086,
+      "grad_norm": 0.6683978165640897,
+      "learning_rate": 1.854240479496643e-06,
+      "loss": 0.1211,
+      "step": 1086
+    },
+    {
+      "epoch": 3.290200529701097,
+      "grad_norm": 0.6917755904967634,
+      "learning_rate": 1.8389297323750644e-06,
+      "loss": 0.1282,
+      "step": 1087
+    },
+    {
+      "epoch": 3.293227393113886,
+      "grad_norm": 0.7129993709978478,
+      "learning_rate": 1.8236760560229715e-06,
+      "loss": 0.1337,
+      "step": 1088
+    },
+    {
+      "epoch": 3.2962542565266744,
+      "grad_norm": 0.7072452605471916,
+      "learning_rate": 1.808479557110081e-06,
+      "loss": 0.135,
+      "step": 1089
+    },
+    {
+      "epoch": 3.2992811199394625,
+      "grad_norm": 0.7028787211547175,
+      "learning_rate": 1.7933403419062689e-06,
+      "loss": 0.1333,
+      "step": 1090
+    },
+    {
+      "epoch": 3.302307983352251,
+      "grad_norm": 0.6726252942081367,
+      "learning_rate": 1.7782585162808152e-06,
+      "loss": 0.1128,
+      "step": 1091
+    },
+    {
+      "epoch": 3.3053348467650396,
+      "grad_norm": 0.7073230696100601,
+      "learning_rate": 1.7632341857016733e-06,
+      "loss": 0.1284,
+      "step": 1092
+    },
+    {
+      "epoch": 3.308361710177828,
+      "grad_norm": 0.7129885702256057,
+      "learning_rate": 1.7482674552347334e-06,
+      "loss": 0.1372,
+      "step": 1093
+    },
+    {
+      "epoch": 3.311388573590617,
+      "grad_norm": 0.7152965471090009,
+      "learning_rate": 1.7333584295430894e-06,
+      "loss": 0.1295,
+      "step": 1094
+    },
+    {
+      "epoch": 3.3144154370034054,
+      "grad_norm": 0.7221761816544192,
+      "learning_rate": 1.7185072128862934e-06,
+      "loss": 0.1479,
+      "step": 1095
+    },
+    {
+      "epoch": 3.317442300416194,
+      "grad_norm": 0.7177358325502864,
+      "learning_rate": 1.7037139091196396e-06,
+      "loss": 0.1374,
+      "step": 1096
+    },
+    {
+      "epoch": 3.320469163828982,
+      "grad_norm": 0.7055040388485216,
+      "learning_rate": 1.6889786216934389e-06,
+      "loss": 0.1368,
+      "step": 1097
+    },
+    {
+      "epoch": 3.3234960272417706,
+      "grad_norm": 0.691545205540142,
+      "learning_rate": 1.6743014536522872e-06,
+      "loss": 0.1202,
+      "step": 1098
+    },
+    {
+      "epoch": 3.326522890654559,
+      "grad_norm": 0.727070397923638,
+      "learning_rate": 1.659682507634347e-06,
+      "loss": 0.1576,
+      "step": 1099
+    },
+    {
+      "epoch": 3.3295497540673478,
+      "grad_norm": 0.6806119484155511,
+      "learning_rate": 1.6451218858706374e-06,
+      "loss": 0.1211,
+      "step": 1100
+    },
+    {
+      "epoch": 3.3325766174801363,
+      "grad_norm": 0.6982463842495003,
+      "learning_rate": 1.630619690184303e-06,
+      "loss": 0.1372,
+      "step": 1101
+    },
+    {
+      "epoch": 3.3356034808929245,
+      "grad_norm": 0.7545625835221444,
+      "learning_rate": 1.616176021989926e-06,
+      "loss": 0.1329,
+      "step": 1102
+    },
+    {
+      "epoch": 3.338630344305713,
+      "grad_norm": 0.7610618090642491,
+      "learning_rate": 1.6017909822927869e-06,
+      "loss": 0.1291,
+      "step": 1103
+    },
+    {
+      "epoch": 3.3416572077185016,
+      "grad_norm": 0.7622321197115256,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.1245,
+      "step": 1104
+    },
+    {
+      "epoch": 3.34468407113129,
+      "grad_norm": 0.7043551378987364,
+      "learning_rate": 1.573197190360729e-06,
+      "loss": 0.1242,
+      "step": 1105
+    },
+    {
+      "epoch": 3.3477109345440788,
+      "grad_norm": 0.7496001191085345,
+      "learning_rate": 1.558988638083616e-06,
+      "loss": 0.1327,
+      "step": 1106
+    },
+    {
+      "epoch": 3.3507377979568673,
+      "grad_norm": 0.7107031225432933,
+      "learning_rate": 1.5448391142179575e-06,
+      "loss": 0.1313,
+      "step": 1107
+    },
+    {
+      "epoch": 3.353764661369656,
+      "grad_norm": 0.7263777697515049,
+      "learning_rate": 1.5307487177120773e-06,
+      "loss": 0.1365,
+      "step": 1108
+    },
+    {
+      "epoch": 3.356791524782444,
+      "grad_norm": 0.6588279002084336,
+      "learning_rate": 1.516717547100821e-06,
+      "loss": 0.1241,
+      "step": 1109
+    },
+    {
+      "epoch": 3.3598183881952326,
+      "grad_norm": 0.7273962819018849,
+      "learning_rate": 1.5027457005048573e-06,
+      "loss": 0.1288,
+      "step": 1110
+    },
+    {
+      "epoch": 3.362845251608021,
+      "grad_norm": 0.714221212605477,
+      "learning_rate": 1.4888332756300027e-06,
+      "loss": 0.1347,
+      "step": 1111
+    },
+    {
+      "epoch": 3.3658721150208097,
+      "grad_norm": 0.6927134286324752,
+      "learning_rate": 1.4749803697665366e-06,
+      "loss": 0.1183,
+      "step": 1112
+    },
+    {
+      "epoch": 3.3688989784335983,
+      "grad_norm": 0.7446791932301232,
+      "learning_rate": 1.4611870797885196e-06,
+      "loss": 0.1379,
+      "step": 1113
+    },
+    {
+      "epoch": 3.3719258418463864,
+      "grad_norm": 0.6654536801330815,
+      "learning_rate": 1.4474535021531099e-06,
+      "loss": 0.1118,
+      "step": 1114
+    },
+    {
+      "epoch": 3.374952705259175,
+      "grad_norm": 0.6807919500279922,
+      "learning_rate": 1.433779732899897e-06,
+      "loss": 0.1177,
+      "step": 1115
+    },
+    {
+      "epoch": 3.3779795686719636,
+      "grad_norm": 0.7444310754124427,
+      "learning_rate": 1.4201658676502294e-06,
+      "loss": 0.1333,
+      "step": 1116
+    },
+    {
+      "epoch": 3.381006432084752,
+      "grad_norm": 0.6962882397345854,
+      "learning_rate": 1.4066120016065442e-06,
+      "loss": 0.1318,
+      "step": 1117
+    },
+    {
+      "epoch": 3.3840332954975407,
+      "grad_norm": 0.6948618025041309,
+      "learning_rate": 1.3931182295516965e-06,
+      "loss": 0.1327,
+      "step": 1118
+    },
+    {
+      "epoch": 3.3870601589103293,
+      "grad_norm": 0.7394121924782194,
+      "learning_rate": 1.379684645848307e-06,
+      "loss": 0.1293,
+      "step": 1119
+    },
+    {
+      "epoch": 3.390087022323118,
+      "grad_norm": 0.6983619826923315,
+      "learning_rate": 1.3663113444380905e-06,
+      "loss": 0.1304,
+      "step": 1120
+    },
+    {
+      "epoch": 3.393113885735906,
+      "grad_norm": 0.6895595049063865,
+      "learning_rate": 1.3529984188412104e-06,
+      "loss": 0.1215,
+      "step": 1121
+    },
+    {
+      "epoch": 3.3961407491486946,
+      "grad_norm": 0.7327502465388068,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.1354,
+      "step": 1122
+    },
+    {
+      "epoch": 3.399167612561483,
+      "grad_norm": 0.678650006161856,
+      "learning_rate": 1.3265540670563904e-06,
+      "loss": 0.1392,
+      "step": 1123
+    },
+    {
+      "epoch": 3.4021944759742717,
+      "grad_norm": 0.7253805186808874,
+      "learning_rate": 1.3134228257951142e-06,
+      "loss": 0.1406,
+      "step": 1124
+    },
+    {
+      "epoch": 3.4052213393870603,
+      "grad_norm": 0.7360447965554704,
+      "learning_rate": 1.3003523301992105e-06,
+      "loss": 0.1344,
+      "step": 1125
+    },
+    {
+      "epoch": 3.4082482027998484,
+      "grad_norm": 0.675159073384005,
+      "learning_rate": 1.2873426716713012e-06,
+      "loss": 0.1183,
+      "step": 1126
+    },
+    {
+      "epoch": 3.411275066212637,
+      "grad_norm": 0.7516003685268435,
+      "learning_rate": 1.2743939411885753e-06,
+      "loss": 0.1346,
+      "step": 1127
+    },
+    {
+      "epoch": 3.4143019296254256,
+      "grad_norm": 0.6708007386151681,
+      "learning_rate": 1.2615062293021508e-06,
+      "loss": 0.1233,
+      "step": 1128
+    },
+    {
+      "epoch": 3.417328793038214,
+      "grad_norm": 0.6917755624861373,
+      "learning_rate": 1.248679626136433e-06,
+      "loss": 0.1354,
+      "step": 1129
+    },
+    {
+      "epoch": 3.4203556564510027,
+      "grad_norm": 0.6641796315568148,
+      "learning_rate": 1.2359142213884933e-06,
+      "loss": 0.118,
+      "step": 1130
+    },
+    {
+      "epoch": 3.4233825198637913,
+      "grad_norm": 0.7454728522905301,
+      "learning_rate": 1.2232101043274437e-06,
+      "loss": 0.1425,
+      "step": 1131
+    },
+    {
+      "epoch": 3.42640938327658,
+      "grad_norm": 0.671480697668419,
+      "learning_rate": 1.2105673637938054e-06,
+      "loss": 0.1192,
+      "step": 1132
+    },
+    {
+      "epoch": 3.429436246689368,
+      "grad_norm": 0.6856852236399914,
+      "learning_rate": 1.1979860881988903e-06,
+      "loss": 0.1205,
+      "step": 1133
+    },
+    {
+      "epoch": 3.4324631101021565,
+      "grad_norm": 0.6780700381737955,
+      "learning_rate": 1.1854663655241804e-06,
+      "loss": 0.125,
+      "step": 1134
+    },
+    {
+      "epoch": 3.435489973514945,
+      "grad_norm": 0.6983799739717494,
+      "learning_rate": 1.17300828332072e-06,
+      "loss": 0.1211,
+      "step": 1135
+    },
+    {
+      "epoch": 3.4385168369277337,
+      "grad_norm": 0.6596359832397064,
+      "learning_rate": 1.1606119287084982e-06,
+      "loss": 0.1166,
+      "step": 1136
+    },
+    {
+      "epoch": 3.4415437003405223,
+      "grad_norm": 0.7176409015382039,
+      "learning_rate": 1.1482773883758357e-06,
+      "loss": 0.1282,
+      "step": 1137
+    },
+    {
+      "epoch": 3.4445705637533104,
+      "grad_norm": 0.6532518376120142,
+      "learning_rate": 1.136004748578785e-06,
+      "loss": 0.1306,
+      "step": 1138
+    },
+    {
+      "epoch": 3.447597427166099,
+      "grad_norm": 0.7137818903014344,
+      "learning_rate": 1.1237940951405257e-06,
+      "loss": 0.1341,
+      "step": 1139
+    },
+    {
+      "epoch": 3.4506242905788875,
+      "grad_norm": 0.7102285086887318,
+      "learning_rate": 1.1116455134507665e-06,
+      "loss": 0.1475,
+      "step": 1140
+    },
+    {
+      "epoch": 3.453651153991676,
+      "grad_norm": 0.6784512424611988,
+      "learning_rate": 1.0995590884651363e-06,
+      "loss": 0.1187,
+      "step": 1141
+    },
+    {
+      "epoch": 3.4566780174044647,
+      "grad_norm": 0.674923928786375,
+      "learning_rate": 1.0875349047046113e-06,
+      "loss": 0.1194,
+      "step": 1142
+    },
+    {
+      "epoch": 3.4597048808172532,
+      "grad_norm": 0.6548391790535211,
+      "learning_rate": 1.0755730462549008e-06,
+      "loss": 0.1325,
+      "step": 1143
+    },
+    {
+      "epoch": 3.462731744230042,
+      "grad_norm": 0.6879578506372566,
+      "learning_rate": 1.0636735967658785e-06,
+      "loss": 0.1313,
+      "step": 1144
+    },
+    {
+      "epoch": 3.46575860764283,
+      "grad_norm": 0.6925919569049426,
+      "learning_rate": 1.0518366394509805e-06,
+      "loss": 0.1252,
+      "step": 1145
+    },
+    {
+      "epoch": 3.4687854710556185,
+      "grad_norm": 0.6765941537945104,
+      "learning_rate": 1.0400622570866426e-06,
+      "loss": 0.1238,
+      "step": 1146
+    },
+    {
+      "epoch": 3.471812334468407,
+      "grad_norm": 0.6685733963210995,
+      "learning_rate": 1.0283505320117016e-06,
+      "loss": 0.1149,
+      "step": 1147
+    },
+    {
+      "epoch": 3.4748391978811957,
+      "grad_norm": 0.7217463157090274,
+      "learning_rate": 1.0167015461268303e-06,
+      "loss": 0.1237,
+      "step": 1148
+    },
+    {
+      "epoch": 3.4778660612939842,
+      "grad_norm": 0.664259814009184,
+      "learning_rate": 1.0051153808939683e-06,
+      "loss": 0.1127,
+      "step": 1149
+    },
+    {
+      "epoch": 3.4808929247067724,
+      "grad_norm": 0.678461934078288,
+      "learning_rate": 9.935921173357444e-07,
+      "loss": 0.1201,
+      "step": 1150
+    },
+    {
+      "epoch": 3.483919788119561,
+      "grad_norm": 0.7604970886503312,
+      "learning_rate": 9.821318360349142e-07,
+      "loss": 0.1324,
+      "step": 1151
+    },
+    {
+      "epoch": 3.4869466515323495,
+      "grad_norm": 0.7219300283511431,
+      "learning_rate": 9.707346171337895e-07,
+      "loss": 0.128,
+      "step": 1152
+    },
+    {
+      "epoch": 3.489973514945138,
+      "grad_norm": 0.6887986669710066,
+      "learning_rate": 9.594005403336881e-07,
+      "loss": 0.1165,
+      "step": 1153
+    },
+    {
+      "epoch": 3.4930003783579266,
+      "grad_norm": 0.7447354305445412,
+      "learning_rate": 9.481296848943744e-07,
+      "loss": 0.1343,
+      "step": 1154
+    },
+    {
+      "epoch": 3.496027241770715,
+      "grad_norm": 0.6922543258473028,
+      "learning_rate": 9.369221296335007e-07,
+      "loss": 0.1285,
+      "step": 1155
+    },
+    {
+      "epoch": 3.499054105183504,
+      "grad_norm": 0.7076794148162799,
+      "learning_rate": 9.257779529260558e-07,
+      "loss": 0.1339,
+      "step": 1156
+    },
+    {
+      "epoch": 3.5020809685962924,
+      "grad_norm": 0.7141576656640798,
+      "learning_rate": 9.146972327038173e-07,
+      "loss": 0.1311,
+      "step": 1157
+    },
+    {
+      "epoch": 3.5051078320090805,
+      "grad_norm": 0.6706391598789441,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.1254,
+      "step": 1158
+    },
+    {
+      "epoch": 3.508134695421869,
+      "grad_norm": 0.7333883290180401,
+      "learning_rate": 8.927264712227823e-07,
+      "loss": 0.1441,
+      "step": 1159
+    },
+    {
+      "epoch": 3.5111615588346576,
+      "grad_norm": 0.6776599387487536,
+      "learning_rate": 8.818365836066101e-07,
+      "loss": 0.1199,
+      "step": 1160
+    },
+    {
+      "epoch": 3.514188422247446,
+      "grad_norm": 0.7077267273359388,
+      "learning_rate": 8.710104597598224e-07,
+      "loss": 0.1382,
+      "step": 1161
+    },
+    {
+      "epoch": 3.5172152856602343,
+      "grad_norm": 0.7051350116746629,
+      "learning_rate": 8.602481753900427e-07,
+      "loss": 0.1387,
+      "step": 1162
+    },
+    {
+      "epoch": 3.520242149073023,
+      "grad_norm": 0.6730491202178467,
+      "learning_rate": 8.495498057584605e-07,
+      "loss": 0.1101,
+      "step": 1163
+    },
+    {
+      "epoch": 3.5232690124858115,
+      "grad_norm": 0.6772212751972969,
+      "learning_rate": 8.389154256793042e-07,
+      "loss": 0.1216,
+      "step": 1164
+    },
+    {
+      "epoch": 3.5262958758986,
+      "grad_norm": 0.7304812166365322,
+      "learning_rate": 8.283451095193229e-07,
+      "loss": 0.1202,
+      "step": 1165
+    },
+    {
+      "epoch": 3.5293227393113886,
+      "grad_norm": 0.7146483333864209,
+      "learning_rate": 8.178389311972612e-07,
+      "loss": 0.1337,
+      "step": 1166
+    },
+    {
+      "epoch": 3.532349602724177,
+      "grad_norm": 0.6645856709962106,
+      "learning_rate": 8.073969641833446e-07,
+      "loss": 0.1164,
+      "step": 1167
+    },
+    {
+      "epoch": 3.5353764661369658,
+      "grad_norm": 0.6760752778184539,
+      "learning_rate": 7.970192814987676e-07,
+      "loss": 0.1251,
+      "step": 1168
+    },
+    {
+      "epoch": 3.5384033295497543,
+      "grad_norm": 0.7527393030886043,
+      "learning_rate": 7.867059557151813e-07,
+      "loss": 0.1426,
+      "step": 1169
+    },
+    {
+      "epoch": 3.5414301929625425,
+      "grad_norm": 0.7207363038063721,
+      "learning_rate": 7.764570589541876e-07,
+      "loss": 0.1346,
+      "step": 1170
+    },
+    {
+      "epoch": 3.544457056375331,
+      "grad_norm": 0.7204748281848526,
+      "learning_rate": 7.662726628868233e-07,
+      "loss": 0.1343,
+      "step": 1171
+    },
+    {
+      "epoch": 3.5474839197881196,
+      "grad_norm": 0.7066006850526803,
+      "learning_rate": 7.561528387330797e-07,
+      "loss": 0.1337,
+      "step": 1172
+    },
+    {
+      "epoch": 3.550510783200908,
+      "grad_norm": 0.7150262910587308,
+      "learning_rate": 7.460976572613888e-07,
+      "loss": 0.1312,
+      "step": 1173
+    },
+    {
+      "epoch": 3.5535376466136963,
+      "grad_norm": 0.6733928663905246,
+      "learning_rate": 7.361071887881376e-07,
+      "loss": 0.1224,
+      "step": 1174
+    },
+    {
+      "epoch": 3.556564510026485,
+      "grad_norm": 0.7233736712903273,
+      "learning_rate": 7.261815031771602e-07,
+      "loss": 0.1339,
+      "step": 1175
+    },
+    {
+      "epoch": 3.5595913734392735,
+      "grad_norm": 0.7095539582286534,
+      "learning_rate": 7.163206698392744e-07,
+      "loss": 0.1274,
+      "step": 1176
+    },
+    {
+      "epoch": 3.562618236852062,
+      "grad_norm": 0.701663403630143,
+      "learning_rate": 7.065247577317747e-07,
+      "loss": 0.1268,
+      "step": 1177
+    },
+    {
+      "epoch": 3.5656451002648506,
+      "grad_norm": 0.6729500468983306,
+      "learning_rate": 6.96793835357964e-07,
+      "loss": 0.1265,
+      "step": 1178
+    },
+    {
+      "epoch": 3.568671963677639,
+      "grad_norm": 0.7300300387198451,
+      "learning_rate": 6.871279707666634e-07,
+      "loss": 0.1315,
+      "step": 1179
+    },
+    {
+      "epoch": 3.5716988270904277,
+      "grad_norm": 0.6998485103967874,
+      "learning_rate": 6.775272315517423e-07,
+      "loss": 0.129,
+      "step": 1180
+    },
+    {
+      "epoch": 3.5747256905032163,
+      "grad_norm": 0.6973002866414547,
+      "learning_rate": 6.679916848516499e-07,
+      "loss": 0.1156,
+      "step": 1181
+    },
+    {
+      "epoch": 3.5777525539160044,
+      "grad_norm": 0.695421080732386,
+      "learning_rate": 6.585213973489335e-07,
+      "loss": 0.1296,
+      "step": 1182
+    },
+    {
+      "epoch": 3.580779417328793,
+      "grad_norm": 0.7121299790759412,
+      "learning_rate": 6.491164352697854e-07,
+      "loss": 0.1357,
+      "step": 1183
+    },
+    {
+      "epoch": 3.5838062807415816,
+      "grad_norm": 0.7072003167411328,
+      "learning_rate": 6.397768643835755e-07,
+      "loss": 0.1216,
+      "step": 1184
+    },
+    {
+      "epoch": 3.58683314415437,
+      "grad_norm": 0.7210231866303369,
+      "learning_rate": 6.305027500023841e-07,
+      "loss": 0.1306,
+      "step": 1185
+    },
+    {
+      "epoch": 3.5898600075671583,
+      "grad_norm": 0.7045626383182585,
+      "learning_rate": 6.212941569805508e-07,
+      "loss": 0.1309,
+      "step": 1186
+    },
+    {
+      "epoch": 3.592886870979947,
+      "grad_norm": 0.6692839497335036,
+      "learning_rate": 6.121511497142274e-07,
+      "loss": 0.1292,
+      "step": 1187
+    },
+    {
+      "epoch": 3.5959137343927354,
+      "grad_norm": 0.6701418480948146,
+      "learning_rate": 6.030737921409169e-07,
+      "loss": 0.1235,
+      "step": 1188
+    },
+    {
+      "epoch": 3.598940597805524,
+      "grad_norm": 0.6626458839972716,
+      "learning_rate": 5.940621477390307e-07,
+      "loss": 0.1175,
+      "step": 1189
+    },
+    {
+      "epoch": 3.6019674612183126,
+      "grad_norm": 0.7221584729392889,
+      "learning_rate": 5.851162795274445e-07,
+      "loss": 0.1295,
+      "step": 1190
+    },
+    {
+      "epoch": 3.604994324631101,
+      "grad_norm": 0.6697223908366807,
+      "learning_rate": 5.762362500650598e-07,
+      "loss": 0.1152,
+      "step": 1191
+    },
+    {
+      "epoch": 3.6080211880438897,
+      "grad_norm": 0.6995439832015345,
+      "learning_rate": 5.674221214503639e-07,
+      "loss": 0.138,
+      "step": 1192
+    },
+    {
+      "epoch": 3.6110480514566783,
+      "grad_norm": 0.6936269390474235,
+      "learning_rate": 5.586739553209974e-07,
+      "loss": 0.1244,
+      "step": 1193
+    },
+    {
+      "epoch": 3.6140749148694664,
+      "grad_norm": 0.6812490921353546,
+      "learning_rate": 5.499918128533155e-07,
+      "loss": 0.1265,
+      "step": 1194
+    },
+    {
+      "epoch": 3.617101778282255,
+      "grad_norm": 0.691180727263234,
+      "learning_rate": 5.413757547619747e-07,
+      "loss": 0.1192,
+      "step": 1195
+    },
+    {
+      "epoch": 3.6201286416950436,
+      "grad_norm": 0.689479163146746,
+      "learning_rate": 5.328258412994958e-07,
+      "loss": 0.1356,
+      "step": 1196
+    },
+    {
+      "epoch": 3.623155505107832,
+      "grad_norm": 0.6948324067295191,
+      "learning_rate": 5.243421322558506e-07,
+      "loss": 0.131,
+      "step": 1197
+    },
+    {
+      "epoch": 3.6261823685206203,
+      "grad_norm": 0.7004617508454012,
+      "learning_rate": 5.159246869580348e-07,
+      "loss": 0.1285,
+      "step": 1198
+    },
+    {
+      "epoch": 3.629209231933409,
+      "grad_norm": 0.6914234592584946,
+      "learning_rate": 5.075735642696611e-07,
+      "loss": 0.1344,
+      "step": 1199
+    },
+    {
+      "epoch": 3.6322360953461974,
+      "grad_norm": 0.7002086761448016,
+      "learning_rate": 4.992888225905467e-07,
+      "loss": 0.136,
+      "step": 1200
+    },
+    {
+      "epoch": 3.635262958758986,
+      "grad_norm": 0.6779479398042831,
+      "learning_rate": 4.910705198563004e-07,
+      "loss": 0.1332,
+      "step": 1201
+    },
+    {
+      "epoch": 3.6382898221717745,
+      "grad_norm": 0.6979911164624134,
+      "learning_rate": 4.829187135379221e-07,
+      "loss": 0.1245,
+      "step": 1202
+    },
+    {
+      "epoch": 3.641316685584563,
+      "grad_norm": 0.7024585290248009,
+      "learning_rate": 4.7483346064139513e-07,
+      "loss": 0.1278,
+      "step": 1203
+    },
+    {
+      "epoch": 3.6443435489973517,
+      "grad_norm": 0.7030212148031239,
+      "learning_rate": 4.6681481770729844e-07,
+      "loss": 0.1285,
+      "step": 1204
+    },
+    {
+      "epoch": 3.6473704124101403,
+      "grad_norm": 0.7080012748382167,
+      "learning_rate": 4.5886284081039676e-07,
+      "loss": 0.1296,
+      "step": 1205
+    },
+    {
+      "epoch": 3.6503972758229284,
+      "grad_norm": 0.6840363669875889,
+      "learning_rate": 4.509775855592613e-07,
+      "loss": 0.1319,
+      "step": 1206
+    },
+    {
+      "epoch": 3.653424139235717,
+      "grad_norm": 0.6943919302210196,
+      "learning_rate": 4.4315910709587296e-07,
+      "loss": 0.1208,
+      "step": 1207
+    },
+    {
+      "epoch": 3.6564510026485055,
+      "grad_norm": 0.6847007168574144,
+      "learning_rate": 4.354074600952407e-07,
+      "loss": 0.1287,
+      "step": 1208
+    },
+    {
+      "epoch": 3.659477866061294,
+      "grad_norm": 0.7084360755865498,
+      "learning_rate": 4.277226987650129e-07,
+      "loss": 0.1264,
+      "step": 1209
+    },
+    {
+      "epoch": 3.6625047294740822,
+      "grad_norm": 0.697096946222449,
+      "learning_rate": 4.2010487684511105e-07,
+      "loss": 0.1242,
+      "step": 1210
+    },
+    {
+      "epoch": 3.665531592886871,
+      "grad_norm": 0.6968109877950773,
+      "learning_rate": 4.12554047607342e-07,
+      "loss": 0.1221,
+      "step": 1211
+    },
+    {
+      "epoch": 3.6685584562996594,
+      "grad_norm": 0.7006549746845336,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.1203,
+      "step": 1212
+    },
+    {
+      "epoch": 3.671585319712448,
+      "grad_norm": 0.6835729220374582,
+      "learning_rate": 3.976535779226376e-07,
+      "loss": 0.1296,
+      "step": 1213
+    },
+    {
+      "epoch": 3.6746121831252365,
+      "grad_norm": 0.6665712963178916,
+      "learning_rate": 3.9030404167542777e-07,
+      "loss": 0.1269,
+      "step": 1214
+    },
+    {
+      "epoch": 3.677639046538025,
+      "grad_norm": 0.7397584046060942,
+      "learning_rate": 3.8302170650907023e-07,
+      "loss": 0.1396,
+      "step": 1215
+    },
+    {
+      "epoch": 3.6806659099508137,
+      "grad_norm": 0.6631190793528081,
+      "learning_rate": 3.7580662334929517e-07,
+      "loss": 0.1232,
+      "step": 1216
+    },
+    {
+      "epoch": 3.6836927733636022,
+      "grad_norm": 0.6718844245014493,
+      "learning_rate": 3.6865884265153584e-07,
+      "loss": 0.1156,
+      "step": 1217
+    },
+    {
+      "epoch": 3.6867196367763904,
+      "grad_norm": 0.7311208892451181,
+      "learning_rate": 3.615784144005796e-07,
+      "loss": 0.1344,
+      "step": 1218
+    },
+    {
+      "epoch": 3.689746500189179,
+      "grad_norm": 0.6987175634646997,
+      "learning_rate": 3.545653881102118e-07,
+      "loss": 0.1303,
+      "step": 1219
+    },
+    {
+      "epoch": 3.6927733636019675,
+      "grad_norm": 0.7153006202782871,
+      "learning_rate": 3.476198128228736e-07,
+      "loss": 0.1463,
+      "step": 1220
+    },
+    {
+      "epoch": 3.695800227014756,
+      "grad_norm": 0.6935939116814561,
+      "learning_rate": 3.4074173710931804e-07,
+      "loss": 0.1346,
+      "step": 1221
+    },
+    {
+      "epoch": 3.698827090427544,
+      "grad_norm": 0.7005633234416655,
+      "learning_rate": 3.339312090682689e-07,
+      "loss": 0.1252,
+      "step": 1222
+    },
+    {
+      "epoch": 3.701853953840333,
+      "grad_norm": 0.691280819973838,
+      "learning_rate": 3.271882763260881e-07,
+      "loss": 0.1269,
+      "step": 1223
+    },
+    {
+      "epoch": 3.7048808172531214,
+      "grad_norm": 0.6934986059256726,
+      "learning_rate": 3.2051298603643754e-07,
+      "loss": 0.124,
+      "step": 1224
+    },
+    {
+      "epoch": 3.70790768066591,
+      "grad_norm": 0.6991815650369816,
+      "learning_rate": 3.139053848799556e-07,
+      "loss": 0.1333,
+      "step": 1225
+    },
+    {
+      "epoch": 3.7109345440786985,
+      "grad_norm": 0.697169013017705,
+      "learning_rate": 3.0736551906392354e-07,
+      "loss": 0.1221,
+      "step": 1226
+    },
+    {
+      "epoch": 3.713961407491487,
+      "grad_norm": 0.6803348079769721,
+      "learning_rate": 3.008934343219483e-07,
+      "loss": 0.1186,
+      "step": 1227
+    },
+    {
+      "epoch": 3.7169882709042756,
+      "grad_norm": 0.6518105969070985,
+      "learning_rate": 2.9448917591363923e-07,
+      "loss": 0.1249,
+      "step": 1228
+    },
+    {
+      "epoch": 3.720015134317064,
+      "grad_norm": 0.6947437759148959,
+      "learning_rate": 2.881527886242941e-07,
+      "loss": 0.123,
+      "step": 1229
+    },
+    {
+      "epoch": 3.7230419977298523,
+      "grad_norm": 0.6881948016278483,
+      "learning_rate": 2.818843167645835e-07,
+      "loss": 0.135,
+      "step": 1230
+    },
+    {
+      "epoch": 3.726068861142641,
+      "grad_norm": 0.706281938540367,
+      "learning_rate": 2.7568380417024253e-07,
+      "loss": 0.1324,
+      "step": 1231
+    },
+    {
+      "epoch": 3.7290957245554295,
+      "grad_norm": 0.7124993483737403,
+      "learning_rate": 2.6955129420176193e-07,
+      "loss": 0.1282,
+      "step": 1232
+    },
+    {
+      "epoch": 3.732122587968218,
+      "grad_norm": 0.6918620687913427,
+      "learning_rate": 2.6348682974408956e-07,
+      "loss": 0.1204,
+      "step": 1233
+    },
+    {
+      "epoch": 3.735149451381006,
+      "grad_norm": 0.7158247149190025,
+      "learning_rate": 2.5749045320632824e-07,
+      "loss": 0.1378,
+      "step": 1234
+    },
+    {
+      "epoch": 3.7381763147937948,
+      "grad_norm": 0.6636782903816454,
+      "learning_rate": 2.5156220652143403e-07,
+      "loss": 0.124,
+      "step": 1235
+    },
+    {
+      "epoch": 3.7412031782065833,
+      "grad_norm": 0.7132188182481783,
+      "learning_rate": 2.4570213114592957e-07,
+      "loss": 0.1244,
+      "step": 1236
+    },
+    {
+      "epoch": 3.744230041619372,
+      "grad_norm": 0.734404942113102,
+      "learning_rate": 2.399102680596155e-07,
+      "loss": 0.1225,
+      "step": 1237
+    },
+    {
+      "epoch": 3.7472569050321605,
+      "grad_norm": 0.6755342834582262,
+      "learning_rate": 2.3418665776527738e-07,
+      "loss": 0.1275,
+      "step": 1238
+    },
+    {
+      "epoch": 3.750283768444949,
+      "grad_norm": 0.7119729210783879,
+      "learning_rate": 2.2853134028840594e-07,
+      "loss": 0.1257,
+      "step": 1239
+    },
+    {
+      "epoch": 3.7533106318577376,
+      "grad_norm": 0.7084026415905846,
+      "learning_rate": 2.2294435517691504e-07,
+      "loss": 0.1322,
+      "step": 1240
+    },
+    {
+      "epoch": 3.756337495270526,
+      "grad_norm": 0.7047552900351938,
+      "learning_rate": 2.174257415008707e-07,
+      "loss": 0.1278,
+      "step": 1241
+    },
+    {
+      "epoch": 3.7593643586833143,
+      "grad_norm": 0.6609926792808695,
+      "learning_rate": 2.119755378522137e-07,
+      "loss": 0.121,
+      "step": 1242
+    },
+    {
+      "epoch": 3.762391222096103,
+      "grad_norm": 0.7298387914261356,
+      "learning_rate": 2.0659378234448524e-07,
+      "loss": 0.131,
+      "step": 1243
+    },
+    {
+      "epoch": 3.7654180855088915,
+      "grad_norm": 0.7040276209683777,
+      "learning_rate": 2.0128051261257165e-07,
+      "loss": 0.1393,
+      "step": 1244
+    },
+    {
+      "epoch": 3.76844494892168,
+      "grad_norm": 0.6671357350048638,
+      "learning_rate": 1.960357658124301e-07,
+      "loss": 0.1227,
+      "step": 1245
+    },
+    {
+      "epoch": 3.771471812334468,
+      "grad_norm": 0.6998952211870678,
+      "learning_rate": 1.908595786208367e-07,
+      "loss": 0.1305,
+      "step": 1246
+    },
+    {
+      "epoch": 3.7744986757472567,
+      "grad_norm": 0.6968141847590695,
+      "learning_rate": 1.8575198723512434e-07,
+      "loss": 0.128,
+      "step": 1247
+    },
+    {
+      "epoch": 3.7775255391600453,
+      "grad_norm": 0.6554169644835662,
+      "learning_rate": 1.8071302737293294e-07,
+      "loss": 0.1213,
+      "step": 1248
+    },
+    {
+      "epoch": 3.780552402572834,
+      "grad_norm": 0.674414467410239,
+      "learning_rate": 1.7574273427196087e-07,
+      "loss": 0.1222,
+      "step": 1249
+    },
+    {
+      "epoch": 3.7835792659856224,
+      "grad_norm": 0.705701920196972,
+      "learning_rate": 1.7084114268971275e-07,
+      "loss": 0.1384,
+      "step": 1250
+    },
+    {
+      "epoch": 3.786606129398411,
+      "grad_norm": 0.6817425582860241,
+      "learning_rate": 1.6600828690326087e-07,
+      "loss": 0.1146,
+      "step": 1251
+    },
+    {
+      "epoch": 3.7896329928111996,
+      "grad_norm": 0.7284615095770618,
+      "learning_rate": 1.612442007090076e-07,
+      "loss": 0.1386,
+      "step": 1252
+    },
+    {
+      "epoch": 3.792659856223988,
+      "grad_norm": 0.7196194029267956,
+      "learning_rate": 1.565489174224444e-07,
+      "loss": 0.1371,
+      "step": 1253
+    },
+    {
+      "epoch": 3.7956867196367763,
+      "grad_norm": 0.6846954987344251,
+      "learning_rate": 1.519224698779198e-07,
+      "loss": 0.1241,
+      "step": 1254
+    },
+    {
+      "epoch": 3.798713583049565,
+      "grad_norm": 0.6890229747930346,
+      "learning_rate": 1.4736489042840972e-07,
+      "loss": 0.1212,
+      "step": 1255
+    },
+    {
+      "epoch": 3.8017404464623534,
+      "grad_norm": 0.6972438248152127,
+      "learning_rate": 1.4287621094529524e-07,
+      "loss": 0.137,
+      "step": 1256
+    },
+    {
+      "epoch": 3.804767309875142,
+      "grad_norm": 0.6984966971624574,
+      "learning_rate": 1.3845646281813508e-07,
+      "loss": 0.1172,
+      "step": 1257
+    },
+    {
+      "epoch": 3.80779417328793,
+      "grad_norm": 0.6727174279817845,
+      "learning_rate": 1.3410567695444576e-07,
+      "loss": 0.1221,
+      "step": 1258
+    },
+    {
+      "epoch": 3.8108210367007187,
+      "grad_norm": 0.7109263753614155,
+      "learning_rate": 1.298238837794885e-07,
+      "loss": 0.122,
+      "step": 1259
+    },
+    {
+      "epoch": 3.8138479001135073,
+      "grad_norm": 0.687809623281965,
+      "learning_rate": 1.2561111323605714e-07,
+      "loss": 0.1314,
+      "step": 1260
+    },
+    {
+      "epoch": 3.816874763526296,
+      "grad_norm": 0.6816762291167833,
+      "learning_rate": 1.2146739478426483e-07,
+      "loss": 0.1308,
+      "step": 1261
+    },
+    {
+      "epoch": 3.8199016269390844,
+      "grad_norm": 0.6726006653927287,
+      "learning_rate": 1.1739275740134004e-07,
+      "loss": 0.122,
+      "step": 1262
+    },
+    {
+      "epoch": 3.822928490351873,
+      "grad_norm": 0.6791801576748797,
+      "learning_rate": 1.1338722958142311e-07,
+      "loss": 0.1297,
+      "step": 1263
+    },
+    {
+      "epoch": 3.8259553537646616,
+      "grad_norm": 0.6878564739342548,
+      "learning_rate": 1.0945083933537104e-07,
+      "loss": 0.1246,
+      "step": 1264
+    },
+    {
+      "epoch": 3.82898221717745,
+      "grad_norm": 0.6753574392027365,
+      "learning_rate": 1.055836141905553e-07,
+      "loss": 0.1265,
+      "step": 1265
+    },
+    {
+      "epoch": 3.8320090805902383,
+      "grad_norm": 0.664681025214523,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.1163,
+      "step": 1266
+    },
+    {
+      "epoch": 3.835035944003027,
+      "grad_norm": 0.672531544994734,
+      "learning_rate": 9.805676689555788e-08,
+      "loss": 0.1316,
+      "step": 1267
+    },
+    {
+      "epoch": 3.8380628074158154,
+      "grad_norm": 0.7101366030816906,
+      "learning_rate": 9.439719738099318e-08,
+      "loss": 0.1319,
+      "step": 1268
+    },
+    {
+      "epoch": 3.841089670828604,
+      "grad_norm": 0.7001804738441401,
+      "learning_rate": 9.080689823853017e-08,
+      "loss": 0.1317,
+      "step": 1269
+    },
+    {
+      "epoch": 3.844116534241392,
+      "grad_norm": 0.6675902371100275,
+      "learning_rate": 8.728589457530857e-08,
+      "loss": 0.1219,
+      "step": 1270
+    },
+    {
+      "epoch": 3.8471433976541807,
+      "grad_norm": 0.6742505534092966,
+      "learning_rate": 8.383421101388344e-08,
+      "loss": 0.1219,
+      "step": 1271
+    },
+    {
+      "epoch": 3.8501702610669692,
+      "grad_norm": 0.7267450253473334,
+      "learning_rate": 8.04518716920466e-08,
+      "loss": 0.1272,
+      "step": 1272
+    },
+    {
+      "epoch": 3.853197124479758,
+      "grad_norm": 0.689741535617822,
+      "learning_rate": 7.713890026266768e-08,
+      "loss": 0.1289,
+      "step": 1273
+    },
+    {
+      "epoch": 3.8562239878925464,
+      "grad_norm": 0.6747671544887421,
+      "learning_rate": 7.389531989351773e-08,
+      "loss": 0.1267,
+      "step": 1274
+    },
+    {
+      "epoch": 3.859250851305335,
+      "grad_norm": 0.7150920576402213,
+      "learning_rate": 7.072115326711704e-08,
+      "loss": 0.1287,
+      "step": 1275
+    },
+    {
+      "epoch": 3.8622777147181235,
+      "grad_norm": 0.6977338707243536,
+      "learning_rate": 6.761642258056977e-08,
+      "loss": 0.127,
+      "step": 1276
+    },
+    {
+      "epoch": 3.865304578130912,
+      "grad_norm": 0.7052835376146881,
+      "learning_rate": 6.458114954541294e-08,
+      "loss": 0.1341,
+      "step": 1277
+    },
+    {
+      "epoch": 3.8683314415437002,
+      "grad_norm": 0.6834053197443895,
+      "learning_rate": 6.161535538745877e-08,
+      "loss": 0.1181,
+      "step": 1278
+    },
+    {
+      "epoch": 3.871358304956489,
+      "grad_norm": 0.6890253688841128,
+      "learning_rate": 5.87190608466548e-08,
+      "loss": 0.1298,
+      "step": 1279
+    },
+    {
+      "epoch": 3.8743851683692774,
+      "grad_norm": 0.6782102205215697,
+      "learning_rate": 5.5892286176932875e-08,
+      "loss": 0.123,
+      "step": 1280
+    },
+    {
+      "epoch": 3.877412031782066,
+      "grad_norm": 0.7288394685509304,
+      "learning_rate": 5.3135051146068203e-08,
+      "loss": 0.123,
+      "step": 1281
+    },
+    {
+      "epoch": 3.880438895194854,
+      "grad_norm": 0.6631250485241632,
+      "learning_rate": 5.044737503554165e-08,
+      "loss": 0.1229,
+      "step": 1282
+    },
+    {
+      "epoch": 3.8834657586076426,
+      "grad_norm": 0.7170895171742782,
+      "learning_rate": 4.7829276640406486e-08,
+      "loss": 0.1338,
+      "step": 1283
+    },
+    {
+      "epoch": 3.886492622020431,
+      "grad_norm": 0.6965922940890412,
+      "learning_rate": 4.528077426915412e-08,
+      "loss": 0.1388,
+      "step": 1284
+    },
+    {
+      "epoch": 3.88951948543322,
+      "grad_norm": 0.6837658892318492,
+      "learning_rate": 4.280188574358857e-08,
+      "loss": 0.1244,
+      "step": 1285
+    },
+    {
+      "epoch": 3.8925463488460084,
+      "grad_norm": 0.6919432603705619,
+      "learning_rate": 4.0392628398699954e-08,
+      "loss": 0.1288,
+      "step": 1286
+    },
+    {
+      "epoch": 3.895573212258797,
+      "grad_norm": 0.7273152136114581,
+      "learning_rate": 3.805301908254455e-08,
+      "loss": 0.1325,
+      "step": 1287
+    },
+    {
+      "epoch": 3.8986000756715855,
+      "grad_norm": 0.6811200746816398,
+      "learning_rate": 3.578307415612714e-08,
+      "loss": 0.1341,
+      "step": 1288
+    },
+    {
+      "epoch": 3.901626939084374,
+      "grad_norm": 0.7216384710769473,
+      "learning_rate": 3.3582809493284406e-08,
+      "loss": 0.1338,
+      "step": 1289
+    },
+    {
+      "epoch": 3.904653802497162,
+      "grad_norm": 0.7136409779528067,
+      "learning_rate": 3.1452240480577265e-08,
+      "loss": 0.1298,
+      "step": 1290
+    },
+    {
+      "epoch": 3.907680665909951,
+      "grad_norm": 0.702151719093669,
+      "learning_rate": 2.9391382017178726e-08,
+      "loss": 0.1276,
+      "step": 1291
+    },
+    {
+      "epoch": 3.9107075293227394,
+      "grad_norm": 0.6829617843839416,
+      "learning_rate": 2.7400248514776184e-08,
+      "loss": 0.1279,
+      "step": 1292
+    },
+    {
+      "epoch": 3.913734392735528,
+      "grad_norm": 0.6741347057276731,
+      "learning_rate": 2.547885389746485e-08,
+      "loss": 0.1166,
+      "step": 1293
+    },
+    {
+      "epoch": 3.916761256148316,
+      "grad_norm": 0.7122123149860387,
+      "learning_rate": 2.3627211601651157e-08,
+      "loss": 0.1377,
+      "step": 1294
+    },
+    {
+      "epoch": 3.9197881195611046,
+      "grad_norm": 0.6957680589032047,
+      "learning_rate": 2.184533457596394e-08,
+      "loss": 0.1295,
+      "step": 1295
+    },
+    {
+      "epoch": 3.922814982973893,
+      "grad_norm": 0.7298755483154954,
+      "learning_rate": 2.013323528115674e-08,
+      "loss": 0.1348,
+      "step": 1296
+    },
+    {
+      "epoch": 3.9258418463866818,
+      "grad_norm": 0.7224320571619296,
+      "learning_rate": 1.849092569002342e-08,
+      "loss": 0.1284,
+      "step": 1297
+    },
+    {
+      "epoch": 3.9288687097994703,
+      "grad_norm": 0.7002282673912389,
+      "learning_rate": 1.6918417287318245e-08,
+      "loss": 0.1328,
+      "step": 1298
+    },
+    {
+      "epoch": 3.931895573212259,
+      "grad_norm": 0.676577261771961,
+      "learning_rate": 1.5415721069669265e-08,
+      "loss": 0.1224,
+      "step": 1299
+    },
+    {
+      "epoch": 3.9349224366250475,
+      "grad_norm": 0.7058486708776619,
+      "learning_rate": 1.3982847545507271e-08,
+      "loss": 0.1325,
+      "step": 1300
+    },
+    {
+      "epoch": 3.937949300037836,
+      "grad_norm": 0.677288513497866,
+      "learning_rate": 1.2619806734986972e-08,
+      "loss": 0.1166,
+      "step": 1301
+    },
+    {
+      "epoch": 3.940976163450624,
+      "grad_norm": 0.7210974689630478,
+      "learning_rate": 1.1326608169920373e-08,
+      "loss": 0.1268,
+      "step": 1302
+    },
+    {
+      "epoch": 3.9440030268634128,
+      "grad_norm": 0.6847309293945656,
+      "learning_rate": 1.0103260893709055e-08,
+      "loss": 0.121,
+      "step": 1303
+    },
+    {
+      "epoch": 3.9470298902762013,
+      "grad_norm": 0.6951200764811203,
+      "learning_rate": 8.949773461282008e-09,
+      "loss": 0.1157,
+      "step": 1304
+    },
+    {
+      "epoch": 3.95005675368899,
+      "grad_norm": 0.6954503552838828,
+      "learning_rate": 7.866153939033449e-09,
+      "loss": 0.1281,
+      "step": 1305
+    },
+    {
+      "epoch": 3.953083617101778,
+      "grad_norm": 0.6949580652054457,
+      "learning_rate": 6.8524099047695415e-09,
+      "loss": 0.1379,
+      "step": 1306
+    },
+    {
+      "epoch": 3.9561104805145666,
+      "grad_norm": 0.6693386351560809,
+      "learning_rate": 5.9085484476506525e-09,
+      "loss": 0.1265,
+      "step": 1307
+    },
+    {
+      "epoch": 3.959137343927355,
+      "grad_norm": 0.7019396513566264,
+      "learning_rate": 5.034576168149175e-09,
+      "loss": 0.1345,
+      "step": 1308
+    },
+    {
+      "epoch": 3.9621642073401437,
+      "grad_norm": 0.7133284977295001,
+      "learning_rate": 4.230499177994007e-09,
+      "loss": 0.1316,
+      "step": 1309
+    },
+    {
+      "epoch": 3.9651910707529323,
+      "grad_norm": 0.7280955821151646,
+      "learning_rate": 3.4963231001383657e-09,
+      "loss": 0.1399,
+      "step": 1310
+    },
+    {
+      "epoch": 3.968217934165721,
+      "grad_norm": 0.6446170261348445,
+      "learning_rate": 2.832053068709817e-09,
+      "loss": 0.1139,
+      "step": 1311
+    },
+    {
+      "epoch": 3.9712447975785095,
+      "grad_norm": 0.6414736152914798,
+      "learning_rate": 2.237693728981416e-09,
+      "loss": 0.1131,
+      "step": 1312
+    },
+    {
+      "epoch": 3.974271660991298,
+      "grad_norm": 0.6513493495818518,
+      "learning_rate": 1.7132492373395094e-09,
+      "loss": 0.1106,
+      "step": 1313
+    },
+    {
+      "epoch": 3.977298524404086,
+      "grad_norm": 0.6991553023785897,
+      "learning_rate": 1.2587232612493172e-09,
+      "loss": 0.136,
+      "step": 1314
+    },
+    {
+      "epoch": 3.9803253878168747,
+      "grad_norm": 0.7179032549001348,
+      "learning_rate": 8.741189792338401e-10,
+      "loss": 0.1299,
+      "step": 1315
+    },
+    {
+      "epoch": 3.9833522512296633,
+      "grad_norm": 0.698262796397405,
+      "learning_rate": 5.594390808494332e-10,
+      "loss": 0.1202,
+      "step": 1316
+    },
+    {
+      "epoch": 3.986379114642452,
+      "grad_norm": 0.7113457921287619,
+      "learning_rate": 3.1468576666915383e-10,
+      "loss": 0.1191,
+      "step": 1317
+    },
+    {
+      "epoch": 3.98940597805524,
+      "grad_norm": 0.6913352130523229,
+      "learning_rate": 1.3986074826388697e-10,
+      "loss": 0.1312,
+      "step": 1318
+    },
+    {
+      "epoch": 3.9924328414680286,
+      "grad_norm": 0.69065108323836,
+      "learning_rate": 3.496524819457392e-11,
+      "loss": 0.1409,
+      "step": 1319
+    },
+    {
+      "epoch": 3.995459704880817,
+      "grad_norm": 0.686219267302617,
+      "learning_rate": 0.0,
+      "loss": 0.128,
+      "step": 1320
+    },
+    {
+      "epoch": 3.995459704880817,
+      "step": 1320,
+      "total_flos": 5.509516886812918e+17,
+      "train_loss": 0.30280588573465744,
+      "train_runtime": 6040.8082,
+      "train_samples_per_second": 27.992,
+      "train_steps_per_second": 0.219
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1320,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.509516886812918e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}